European Migration Law (Oxford European Union Law Library) 0192894277, 9780192894274

This title provides a comprehensive overview of European migration law. More than three dozen directives and regulations

310 38 8MB

English Pages 720 [721] Year 2023

Report DMCA / Copyright

DOWNLOAD FILE

Polecaj historie

European Migration Law (Oxford European Union Law Library) 9780192894274, 0192894277

This title provides a comprehensive overview of European migration law. More than three dozen directives and regulations

128 50 Read more

EU Citizenship Law (Oxford European Union Law Library) 9780198795315, 0198795319

European Union citizenship is a novel and complex legal status. Since its formal conception in the Maastricht Treaty, EU

138 82 Read more

EU Anti-Discrimination Law (Oxford European Union Law Library) 9780199698462, 0199698465

EU Anti-Discrimination Law provides a detailed and critical analysis of the corpus of European Union law prohibiting dis

123 113 Read more

EU Justice and Home Affairs Law: Volume 1: EU Immigration and Asylum Law (Oxford European Union Law Library) 9780198890232, 0198890230

EU Asylum and Immigration Law sets out the institutional framework and offers comprehensive coverage and analysis of EU

107 78 Read more

Oxford Principles of European Union Law: Volume 1: The European Union Legal Order 9780191058806, 0191058807

Since the 1957 Rome Treaty, the European Union has changed dramatically - in terms of its composition, scope and depth.

103 99 80MB Read more

European Union Corporate Tax Law 9781107347861, 9781107018990

How does EU law affect Member State corporate tax systems and the cross-border activities of companies? This unique stud

272 45 4MB Read more

European Union Law [3 ed.] 9780993336591

1,806 164 4MB Read more

European Union Law and Defence Integration 9781472563477, 9781841134406

This monograph examines the legal dimension of European defence integration from the Second World War to the Treaty Esta

218 27 4MB Read more

Internet and Electronic Commerce Law in the European Union 9781472562005

This book outlines and analyses the legislative activity of the Union in an area which is currently experiencing exponen

278 11 8MB Read more

The Outer Limits of European Union Law 9781472564702, 9781841138602

A commonly expressed view is that the citizens and the Member States are destined to be overcome by the European Union.

273 126 2MB Read more

European Migration Law (Oxford European Union Law Library)
0192894277, 9780192894274

Author / Uploaded
Prof Daniel Thym

Citation preview

OXFORD EU LAW LIBRARY General Editors ROBERT SCHÜTZE Professor of European and Global Law, Durham Law School and Co-Director, Global Policy Institute, Durham Law School PIET EECKHOUT Professor of EU Law and Dean of the Faculty of Laws, UCL, and Academic Director of the European Institute

European Migration Law

OX F O R D E U L AW L I B R A RY The aim of the series is to publish important and original studies of the various branches of EU law. Each work provides a clear, concise, and critical exposition of the law in its social, economic, and political context, at a level which will interest the advanced student, the practitioner, the academic, and government officials. OTHER TITLES IN THIS SERIES EU Securities and Financial Markets Regulation Fourth Edition Niamh Moloney

EU Procedural Law Koen Lenaerts, Ignace Maselis, Kathleen Gutman, Janek Tomasz Nowak

EU Diplomatic Law Sanderijn Duquet

The EU Common Security and Defence Policy Panos Koutrakos

EU Constitutional Law Koen Lenaerts, Piet Van Nuffel, Tim Corthaut

EU Anti-Discrimination Law Second Edition Evelyn Ellis and Philippa Watson

EU Customs Law Third Edition Timothy Lyons

EU Employment Law Fourth Edition Catherine Barnard

Principles and Practice in EU Sports Law Stephen Weatherill

EU External Relations Law Second Edition Piet Eeckhout

EU Justice and Home Affairs Law Fourth Edition Steve Peers

The EU Common Fisheries Policy Robin Churchill and Daniel Owen

European Migration Law DA N I E L T H YM

Great Clarendon Street, Oxford, OX2 6DP, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © Daniel Thym 2023 The moral rights of the author have been asserted First Edition published in 2023 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Public sector information reproduced under Open Government Licence v3.0 (http://www.nationalarchives.gov.uk/doc/open-government-licence/open-government-licence.htm) Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2023936578 ISBN 978–0–19–289427–4 DOI: 10.1093/oso/9780192894274.001.0001 Printed and bound in the UK by TJ Books Limited Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.

Acknowledgements This book builds upon more than two decades of experience with European migration law and policy. Numerous friends and colleagues have informed my thinking on this complex and controversial topic ever since I started writing my PhD at the Walter Hallstein Institute for European Constitutional Law at Humboldt University in Berlin in the early 2000s. An important change came when I was appointed full Professor of Public, European, and International Law at the University of Konstanz ten years later, as the successor to Kay Hailbronner who was a prominent participant in the foundational period of European migration law. To be not primarily a German academic, writing in the national language for a domestic audience, had always been my aspiration as well. The dynamic and interdisciplinary atmosphere in Konstanz proved a perfect academic base to do so. Membership in the transnational Odysseus Network, coordinated by Philippe De Bruycker, equally supports a pan-European outlook. I am indebted to friends and colleagues with whom I have discussed migration law throughout all these years. Anyone reading the book will realise that my analysis moves beyond questions of legal interpretation by considering both the policy concept and the theoretical positions that lie beneath. Doing so is not self-evident for someone who studied law on the continent. I benefited greatly from my involvement in the interdisciplinary Clusters of Excellence ‘Cultural Foundations of Cultural Integration’ and ‘The Politics of Inequality’ at the University of Konstanz. Similarly, I learnt new perspectives during the seven years I served as a member and vice-chairperson of the German Expert Council on Integration and Migration, an independent advisory body in Berlin. Our annual reports covered diverse issues such as asylum, labour migration, integration, diversity, and relations with Africa. Equally instructive were ongoing discussions with journalists and regular appearances as an expert witness in the home affairs committee of the German parliament in the hectic years following the events of 2015/16. Over the years, I became acquainted with a rich arsenal of research questions, practical challenges, and methodological approaches. Few people have the luxury of being able to consider these multiple experiences at an intermediate level of abstraction. Writing this book allowed me to do precisely this. My home university generously supported this venture with a supplementary sabbatical, which I spent in the inspiring and charming environment of the European University Institute in Florence, just before another round of lockdowns during the Covid-19 pandemic. They interrupted the usual travel circuit and brought back the intimate atmosphere of the academic study room, where I spent endless days and evenings writing the manuscript. My partner deserves praise for having accepted my repeated assurances of ‘just one more hour’ again and again. Generations of student assistants have accumulated countless articles, book chapters, and other documents as print copies or as digital files. They go unnamed, but their relentless efforts are much appreciated indeed. Our university library must have one of the most comprehensive collections of European migration law by now. The research assistants at my university chair invested time and energy in adapting the references to the OSCOLA

vi Acknowledgements style guide. Sabine Gerber did a wonderful job in coordinating the different activities and supporting me on the administrative front. Without her valuable work, this book could not have been written besides all my other obligations. Finally, I am grateful to Oxford University Press for having accepted this contribution to the EU Law Library Series and for having accompanied the production process diligently. Once my literary ‘baby’ is born, it will be up to readers to receive it as a new player in the world of migration law and policy. Any kind of feedback is welcome (daniel.thym@uni-konstanz.de). The highly dynamic nature of this object of analysis calls for a reasonably swift second edition for sure. Konstanz February 2023

Summary Contents Table of Cases Table of EU Legislation List of Abbreviations Introduction

xxi xxxi xliii 1

PA RT I OV E R A R C H I N G T H E M E S 1. Building an Area of Freedom, Security, and Justice

13

2. Institutional Prerogatives and Decision-making

42

3. Court of Justice: Achievements and Limitations

69

4. Interdisciplinary Perspectives and Methodology

96

5. Human Rights and State Sovereignty

122

6. Doctrinal Foundations of the Case Law

150

7. Administrative Dimension

172

8. Agencies (Frontex and Asylum Agency)

197

9. Databases

223

PA RT I I SE C T O R A L L E G I SL AT IO N A N D P O L IC I E S 10. General Features of EU Legislation

253

11. Visa Policy

280

12. Border Controls

301

13. Common European Asylum System

337

14. Legal Migration

427

15. Integration and Settlement

469

16. Irregular Presence and Return

505

17. Association Agreements with Neighbours

550

18. International Cooperation with Third States

570

Bibliography Index

597 635

Contents Table of Cases Table of EU Legislation List of Abbreviations Introduction: European Migration Law as a Field of Inquiry

xxi xxxi xliii 1

PA RT I OV E R A R C H I N G T H E M E S 1. Building an Area of Freedom, Security, and Justice 1.1 From International to EU Migration Law

13 14

1.2 Towards Distinct Policies for Third Country Nationals

24

1.3 Distinguishing Union Citizenship from Third Country Nationals

33

1.4 Summary

40

1.1.1 Parallelism of emigration, imperialism, and tourism 1.1.2 Migration law as a by-product of state formation 1.1.3 Towards international protection for refugees 1.1.4 Prehistory of EU rules on labour migration 1.1.5 Free movement within the single market 1.1.6 Colonial migration: the downside of free movement 1.1.7 Theoretical ambiguity of Union citizenship

1.2.1 ‘Schengen’: compensatory logic 1.2.2 Treaty of Maastricht: fragile intergovernmental cooperation 1.2.3 Treaty of Amsterdam: gradual supranationalisation 1.2.4 Treaty of Lisbon: autonomy of migration law 1.2.5 European Council: political programming 1.2.6 Public discourse between ‘fortress Europe’ and ‘common values’

1.3.1 Primary law: open-ended Treaty objectives 1.3.2 Legislative leeway within the confines of human rights 1.3.3 Overlap with Union citizenship 1.3.4 Tipping points in the legislative process

14 16 17 18 20 20 22 24 26 27 29 29 31 34 35 37 39

2. Institutional Prerogatives and Decision-making 2.1 Driving Forces behind Europeanisation

42 42

49

2.1.1 The choice for Schengen as an example 2.1.2 Continuity of mixed results in migration policy 2.1.3 From ‘permissive consensus’ to ‘constraining dissensus’ 2.1.4 The danger of institutional blockage

2.2 Influence on the Policy Output

2.2.1 Commission 2.2.1.1 Civil servants between expertise and politics 2.2.1.2 Technocratic posture on migration law 2.2.2 European Council 2.2.3 Council

43 44 46 47 49 49 50 51 52

x Contents 2.2.3.1 The ‘backroom’: working parties of national civil servants 2.2.3.2 Consensus culture in the shadow of majority voting 2.2.4 European Parliament 2.2.4.1 Committees as the ‘engines’ of parliamentary impact 2.2.4.2 Shifting positions on migration law

52 53 54 55 55

2.3 Interinstitutional Practices

56

2.4 Differentiated Integration: Opt-outs after Brexit

63

2.5 Summary

67

2.3.1 Treaty change: limited relevance 2.3.2 Secondary legislation: prevalence of informal ‘trilogues’ 2.3.3 Curtailment of delegated and implementing acts 2.3.4 Ancillary role of ‘soft law’ 2.3.5 International cooperation: side-effects of informalisation 2.3.6 Private actors: a democratic virtue for migration law?

2.4.1 Denmark: the long shadow of the referenda 2.4.2 Ireland: relic of British reticence

56 57 58 59 60 61 64 65

3. Court of Justice: Achievements and Limitations 3.1 Influence of the Court Architecture

69 69

3.2 Constitutional Authority of the Supranational Judiciary

74

3.3 Access to and Output in the Area of Migration

79

3.4 Alternative Accountability Mechanisms

90

3.5 Summary

94

3.1.1 Collective and multi-lingual adjudication 3.1.2 Horizontal outlook beyond migration law 3.1.3 Frontex, Asylum Agency, and the pitfalls of specialised tribunals 3.1.4 Non-governmental organisations and the limits of third party intervention

3.2.1 Reputation of ‘constitutional imagination’ 3.2.2 Migration law: ‘administrative mindset’ 3.2.3 Interaction with the legislature

3.3.1 Validity disputes 3.3.1.1 Direct actions by individuals on migratory matters 3.3.1.2 Privileged access by the EU institutions 3.3.2 Infringement proceedings 3.3.2.1 Importance of the pre-litigation stage 3.3.2.2 Changing dynamics in migration law 3.3.3 Preliminary references 3.3.3.1 Stark discrepancies between the Member States 3.3.3.2 Carrots and sticks for loyal cooperation 3.3.3.3 Between micromanagement and vagueness 3.3.4 ‘Demand’ side 3.3.4.1 Thematic asymmetries in migration law 3.3.4.2 Strategic litigation to the benefit of migrants

3.4.1 Accountability for wrongdoing beyond courts 3.4.2 Ombudspersons as vehicles of administrative control 3.4.3 Financial and managerial accountability 3.4.4 Political oversight by parliaments 3.4.5 Access to documents as a source of public debates

70 71 72 73 75 76 77 79 79 80 80 81 81 83 83 85 86 87 87 89 90 91 92 93 94

Contents xi

4. Interdisciplinary Perspectives and Methodology 4.1 Drivers of Migratory Movements

4.1.1 Numerous ‘push’ and ‘pull’ factors 4.1.2 Migration as a process over time 4.1.3 Intermediate level: networks, regimes, and infrastructures 4.1.4 Continuum of ‘voluntary’ and ‘forced’ migration

4.2.1 Normative background: the open borders debate 4.2.2 Explaining the ‘control gap’: the liberal paradox 4.2.3 Public opinion and intergroup threat perceptions 4.2.4 Multiple actors of migration governance 4.2.5 Individual migrant agency

4.3.1 Starting point: Foucault and constructivism 4.3.2 ‘Labels’ as a means of government 4.3.3 Language between ‘abuse’, ‘crisis’, and ‘normality’ 4.3.4 Securitisation versus rights-based approach

4.4.1 Academic discourse: from enthusiasm to scepticism 4.4.2 Between ‘black letter’ and doctrinal constructivism 4.4.3 Interdisciplinary and critical approaches

96 96

97 99 101 102

4.2 Significance of State Measures

103

4.3 Law as an Instrument of Government

110

4.4 Legal Methodology between Doctrine and Critique

115

4.5 Summary

120

5. Human Rights and State Sovereignty 5.1 ‘Aliens Law’ as Protection qua Nationality

5.1.1 Historic minimum standards as inter-state obligations 5.1.2 Limited impact of the conventions of the Council of Europe 5.1.3 Reduction of statelessness as a legal obligation

5.2.1 Individual guarantees for ‘everyone’ 5.2.2 Hanna Arendt and the ‘right to have rights’ 5.2.3 Normative counterweight to state sovereignty

5.3.1 Judicial dynamism in the field of migration 5.3.2 Focus on general principles and the Grand Chamber 5.3.3 Interaction with EU law

5.4.1 More generous protection 5.4.2 Rights of the child as an example 5.4.3 EU not a ‘human rights organisation’ 5.4.4 Limited scope of application of the Charter

5.5.1 Plethora of human rights treaty bodies 5.5.2 Potential source of dynamism in the field of migration 5.5.3 Special rapporteurs and diplomatic initiatives 5.5.4 Global Compacts for Migration and on Refugees

5.6.1 Interaction with Union law 5.6.2 Supporting role in the judicial practice 5.6.3 Practical priority of human rights

104 105 106 107 109 110 111 112 114 115 116 118

122 123

123 124 126

5.2 Human Rights qua Personhood

127

5.3 Lead Function of the European Convention

130

5.4 Added Value of the Charter

134

5.5 Complementary Role of International Bodies and Political Fora

139

5.6 Refugee Convention: Distant Lodestar

143

5.7 Summary

148

127 128 129 130 132 133 134 135 136 137 139 140 141 142 143 145 147

xii Contents

6. Doctrinal Foundations of the Case Law 6.1 ‘Legal Order’ as Doctrinal Self-sufficiency 6.2 The Promise and Limits of Coherence

150 150 152

6.3 Constitutional Essentials in Migration Law

155

6.4 Pitfalls of the Interpretative Exercise

158

6.5 Individual Rights of Migrants

163

6.6 Practical Tips for Dealing with the Case Law 6.7 Summary

169 170

7. Administrative Dimension 7.1 European Migration Law as a Composite System

172 173

7.2 National Procedural Autonomy and Its Limits

179

7.3 Transnational Effects of Horizontal Cooperation

187

7.4 Money as a Means of Governance 7.5 Summary

191 195

6.2.1 Legislative fragmentation 6.2.2 Vision of an ‘immigration code’ 6.2.3 Interpretative approximation

6.3.1 Direct effect and primacy as an everyday practice 6.3.2 No categorical distinction between directives and regulations 6.3.3 Subsidiarity as political preference

6.4.1 Multilingualism of EU migration law 6.4.2 Revival of the drafting history 6.4.3 Indeterminacy of teleological interpretation 6.4.4 Autonomous concepts and their limits 6.4.5 Effet utile as a useful doctrinal tool

6.5.1 Individual rights enshrined in legislation 6.5.2 Added valued in comparison to human rights 6.5.3 Ambiguous Court practice on the Charter 6.5.3.1 Interpretation in conformity with human rights 6.5.3.2 Silence or vagueness on the Charter 6.5.4 Scope of the right to an effective remedy

7.1.1 Conceptual vagueness of ‘areas’ and ‘systems’ 7.1.2 Multiplicity of actors at different levels 7.1.3 Operational character of asylum and border controls 7.1.4 Diverse avenues of inter-state cooperation

7.2.1 The many faces of procedural harmonisation 7.2.2 Effectiveness as a corrective vehicle 7.2.2.1 Court practice on migration law 7.2.2.2 Time limits as a test case 7.2.3 Fundamental right to an effective remedy 7.2.3.1 Availability of legal remedies 7.2.3.2 Scope of judicial oversight 7.2.3.3 Suspensive effect and legal assistance 7.2.4 Good administration 7.2.4.1 Giving reasons and right to be heard 7.2.4.2 Consequences of procedural deficits

7.3.1 Visas and databases: rules in secondary legislation 7.3.2 Entry bans: failure of inter-state consultation 7.3.3 Intra-EU mobility: ‘golden visas’ as an abusive practice?

152 153 154 155 156 157 158 159 160 162 163 164 164 166 166 166 168

173 175 176 177 179 180 181 182 182 183 183 184 185 185 186 188 189 191

Contents xiii

8. Agencies (Frontex and Asylum Agency) 8.1 Theory and Policy Design

197 198

8.2 Constitutional Foundations

203

8.3 Frontex: Institutional Muscle Formation

207

8.4 Asylum Agency: Latecomer with Subtle Influence

211

8.5 Accountability for Wrongdoing

214

8.6 Summary

221

8.1.1 Elusive vision of administrative centralisation 8.1.2 The choice for Frontex and the Asylum Office 8.1.3 Formal expansion of the mandates 8.1.4 Exponential growth in practice

8.2.1 Involvement in administrative decision-making 8.2.2 ‘Meroni’ doctrine: no decisive hurdle 8.2.3 Territorial scope

8.3.1 Operational powers 8.3.2 Supervision of the Member States 8.3.3 Cooperation with third states 8.3.4 Eurosur and procurement of equipment

8.4.1 Information gathering and practical guidance 8.4.2 Operational support in ‘hotpots’ and beyond 8.4.3 Supervision and international cooperation

8.5.1 Legislative provisions on agency involvement 8.5.2 Institutional governance 8.5.3 Legal remedies for composite procedures

198 200 201 202 203 205 206 207 208 209 210 211 212 214 215 217 218

9. Databases 9.1 Theory and Policy Design

223 224

9.2 Constitutional Foundations

227

9.3 An Overview of Migration Databases

230

9.4 Prospect of Interoperability and Artificial Intelligence

236

9.5 Respect for Data Protection Standards

243

9.6 Summary

248

9.1.1 Information technology as ‘invisible’ infrastructure 9.1.2 ‘Function creep’ over three decades

9.2.1 Supranational competences 9.2.2 Decentralised network structure 9.2.3 Territorial scope

9.3.1 Schengen Information System (SIS) 9.3.2 Eurodac 9.3.3 Visa Information System (VIS) 9.3.4 Travel authorisation (ETIAS) 9.3.5 Entry/Exit System (EES)

9.4.1 From support to automated processing 9.4.2 Quality concerns and implementation deficits 9.4.3 The spectre of ‘interoperability’

9.5.1 Principles of data protection 9.5.2 Core elements of the human rights assessment 9.5.3 Preventive and reactive supervision

224 225 227 228 228 230 232 233 234 236 237 239 241 243 244 247

xiv Contents

PA RT I I SE C T O R A L L E G I SL AT IO N A N D P O L IC I E S 10. General Features of EU Legislation 10.1 Acquisition of Residence Permits

253 254

10.2 Sufficient Resources Requirement 10.3 Public Policy Exception

257 258

10.4 Equal Treatment as a Constitutional Guarantee

262

10.5 Proportionality

267

10.6 More Favourable Domestic Rules

270

10.1.1 Of ‘residence permits’ and ‘visas’ 10.1.2 Constitutive state authorisation 10.1.3 Fee levels 10.3.1 Overlap with Union citizenship 10.3.2 Sector-specific outcomes

10.4.1 Human rights instead of Union citizenship 10.4.2 Different scrutiny levels in the judicial practice 10.4.3 Distinctions based on migration status 10.4.4 Racial and ethnic discrimination 10.5.1 Context-specific outcomes 10.5.2 Abstract or individual assessment? 10.6.1 Identification of mandatory provisions 10.6.2 Different patterns in secondary legislation

10.7 Exceptional Non-Compliance (Article 72 TFEU) 10.8 Abusive Practices 10.9 Practical Tips for Dealing with Legislation 10.10 Summary

254 255 256 259 260 262 264 265 266 267 269 270 271

273 275 276 278

11. Visa Policy 11.1 Theory and Policy Design 11.2 Constitutional Foundations

280 281 282

11.3 Visa Requirements

283

11.4 Visa Procedure

288

11.5 Special Case: Humanitarian Visas 11.6 Other Pre-arrival Measures

294 295

11.7 Summary

299

11.2.1 Supranational competences 11.2.2 Territorial scope 11.3.1 Rationale behind ‘black’ and ‘white’ lists 11.3.2 Visa liberalisation as a bargaining chip 11.3.3 Reciprocity in relations with third states 11.4.1 Admission criteria 11.4.2 Application procedure 11.4.3 Decision and legal remedies 11.4.4 Visa format 11.4.5 Inter-state cooperation 11.6.1 Carrier sanctions 11.6.2 Export of visa policy to transit countries 11.6.3 Liaison officers abroad

282 282 283 285 287 288 290 291 292 292 295 297 298

Contents xv

12. Border Controls 12.1 Theory and Policy Design

301 302

12.2 Constitutional Foundations

305

12.3 Schengen Governance

311

12.4 Internal Movements within the Schengen Area

316

12.5 External Controls at Crossing Points

324

12.6 Surveillance of the External Borders

328

12.7 Document Security 12.8 Summary

334 335

12.1.1 Symbolic weight of borders 12.1.2 Incomplete ‘territorialisation’ of the EU 12.1.3 Towards a new ‘culture’ of border controls

12.2.1 Supranational competences 12.2.2 Extraterritorial reach of human rights 12.2.3 Search and rescue at sea 12.2.4 Rejection at the border 12.2.4.1 Right to asylum and prohibition of refoulement 12.2.4.2 Prohibition of collective expulsion 12.2.5 Territorial scope

12.3.1 Membership and phased implementation 12.3.2 Crises and reform efforts 12.3.3 Scope of the ‘Schengen acquis’

12.4.1 Circulation of third country nationals 12.4.1.1 Free travel for short stays 12.4.1.2 Secondary movements of asylum applicants 12.4.2 Internal border controls 12.4.2.1 Requirements for temporary reintroduction 12.4.2.2 Extensive and illegal state practice 12.4.3 Police checks in border areas

12.5.1 Entry conditions 12.5.2 Checks on persons 12.5.3 Refusal of entry and legal remedies 12.5.4 Local border traffic

12.6.1 Land borders 12.6.2 Interaction with asylum law 12.6.3 Sea borders 12.6.3.1 Sea Borders Regulation 12.6.3.2 Basic procedural safeguards

302 303 304

305 306 308 309 309 310 311 312 313 315 317 317 318 319 319 320 322 324 325 327 328 329 330 331 332 333

13. Common European Asylum System 13.1 Theory and Policy Design

337 338

349

13.1.1 A history of half-hearted commitment 13.1.2 Early years of intergovernmental cooperation 13.1.3 First and second phase of harmonisation 13.1.4 Policy crisis of 2015/16 13.1.5 Elusive reform efforts 13.1.6 Geopolitics enter the political equation 13.1.7 Structural deficits and their fallout

13.2.1 Supranational competences

13.2 Constitutional Foundations

338 339 341 342 344 345 347 349

xvi Contents 13.2.2 Prohibition of refoulement 13.2.3 Right to asylum 13.2.3.1 Article 18 CFR 13.2.3.2 Protocol on Union citizens 13.2.4 Solidarity (Article 80 TFEU) 13.2.5 Territorial scope

351 353 353 354 355 357

13.3 Asylum Jurisdiction

357

13.4 Asylum Procedures

372

13.5 Refugee Status

386

13.6 Subsidiary and Complementary Protection

400

13.3.1 Origin of the first entry rule 13.3.2 Futile quest for solidarity 13.3.3 Hierarchy of the substantive criteria 13.3.4 Scope and procedure 13.3.5 Take charge/back requests 13.3.6 Mutual trust and its limits 13.3.7 Secondary movements and the transfer of jurisdiction

13.4.1 Access to the procedure 13.4.2 Personal interview and individualised assessment 13.4.3 Sixfold procedural differentiation 13.4.3.1 Asylum jurisdiction and admissibility 13.4.3.2 Regular and accelerated examination 13.4.3.3 Subsequent applications and withdrawal 13.4.4 Border procedures 13.4.5 Safe countries 13.4.5.1 Safe countries of origin and common lists 13.4.5.2 Rebuttal of the presumption of safety 13.4.5.3 Controversies about safe third countries 13.4.6 Legal remedies 13.4.7 Recurring ‘vision’: external processing 13.5.1 Preventing disparate recognition quotas 13.5.2 Standards and burden of proof 13.5.3 Notion of persecution 13.5.3.1 Severe violation of basic human rights 13.5.3.2 Refugees sur place and family members 13.5.3.3 Poverty 13.5.3.4 Climate change 13.5.4 Actors of protection 13.5.4.1 Home state and international organisations 13.5.4.2 Internal protection alternative 13.5.5 Reasons for persecution 13.5.5.1 Membership of a particular social group 13.5.5.2 Refusal to perform military service 13.5.6 Exclusion and cessation 13.5.6.1 Palestinian refugees 13.5.6.2 Terrorism and other serious crime 13.5.6.3 Change of circumstance in the home state

13.6.1 Added value of Europeanisation 13.6.2 Protection against indiscriminate violence 13.6.3 Other scenarios: divergence from human rights

358 360 362 364 365 368 370 373 374 376 376 377 378 379 381 381 382 383 384 385 387 388 390 390 392 392 393 394 394 395 396 396 397 398 398 399 400 400 401 403

Contents xvii

13.6.4 ‘Complementary’ protection under national laws

403

13.7 Temporary Protection

405

13.8 Reception Conditions for Asylum Applicants

410

13.9 Rights of Beneficiaries of International Protection

415

13.7.1 Rationale behind blanket recognition 13.7.2 Conditions and contents of temporary protection 13.7.3 Interaction with EU migration law 13.8.1 Protracted differences between the Member States 13.8.2 Living conditions and other guarantees 13.8.3 Restrictions and human rights compliance 13.8.4 Right to remain and ‘fiction of non-entry’ 13.9.1 Limits of equal treatment 13.9.2 Specific guarantees 13.9.3 Free movement and option of ‘uniform status’ 13.9.4 Refugees without protection status

405 406 408 410 411 412 414 415 416 417 418

13.10 Resettlement and Other Legal Pathways

420

13.11 Summary

424

13.10.1 Soft side of asylum policy 13.10.2 EU resettlement framework 13.10.3 Other protracted entry procedures

420 421 423

14. Legal Migration 14.1 Theory and Policy Design

427 428

14.2 Constitutional Foundations

435

14.3 Family Reunification

440

14.4 Skilled Labour Migration

448

14.1.1 ‘Guest workers’ outside the reach of EU institutions 14.1.2 Family bonds as a critical gateway 14.1.3 Globalisation and the ‘battle’ over labour migration 14.1.4 Sectoral approach to economic migration 14.1.5 Limits of Europeanisation 14.1.6 Effects on the welfare state 14.2.1 Supranational competences 14.2.2 National prerogatives for labour migration 14.2.3 Human rights, family life, and rights of the child 14.2.4 Territorial scope

14.3.1 EU citizens and their family members 14.3.2 Scope of the Family Reunification Directive 14.3.2.1 Limitation to the nuclear family 14.3.2.2 Optional clauses 14.3.3 Conditions under the Directive 14.3.4 Procedure and rights of family members 14.3.5 Beneficiaries of international protection 14.4.1 Regulatory toolbox and driving forces beyond the law 14.4.2 Blue Card Directive 14.4.2.1 Personal and thematic scope 14.4.2.2 Procedure and rights of blue card holders 14.4.3 Students and Researchers Directive 14.4.3.1 Students as ‘ideal immigrants’ 14.4.3.2 Delegation of responsibility to research organisations

428 429 431 432 433 434 436 437 438 439 440 443 443 444 445 445 447

448 450 450 452 453 453 454

xviii Contents

14.4.4 Single Permit Directive 14.4.5 Vexed issue of intra-European mobility

14.5.1 Posted workers in the single market 14.5.2 GATS and other trade agreements 14.5.3 Inter-Corporate Transfers Directive 14.5.4 Seasonal Workers Directive 14.5.5 Legal pathways for economic purposes

455 456

14.5 Temporary Economic Activities

458

14.6 Summary

466

459 460 461 463 464

15. Integration and Settlement 15.1 Theory and Policy Design

469 470

15.2 Constitutional Foundations

476

15.3 Equality Provisions in Secondary Legislation

479

15.4 Mandatory Integration Requirements

487

15.5 Other Integration Measures 15.6 Long-Term Residents Directive

492 494

15.7 Acquisition of Nationality

498

15.8 Summary

503

15.1.1 From ‘denizenship’ to citizenship 15.1.2 Models of migrant integration 15.1.3 Disputes over the direction of EU migration law 15.1.4 Status change in multiple directions 15.1.5 Mixed output in migration law and beyond

15.2.1 Scope of supranational competences 15.2.2 Human rights protection against expulsion 15.2.3 Territorial scope

15.3.1 Complexity of the legislative framework 15.3.2 Overarching doctrinal characteristics 15.3.3 Social security other benefits 15.3.3.1 Exclusion from social assistance and advantages 15.3.3.2 Human rights compliance 15.3.4 Disparate effects of equal treatment 15.3.5 Distinctions between refugees and subsidiary protection

15.4.1 Proliferation and significance of domestic laws 15.4.2 Supranational legal framework 15.4.3 Judicial endorsement, subject to a caveat

15.6.1 Independence from the original purpose 15.6.2 Scope and conditions 15.6.3 Procedure and loss

15.7.1 Conceptual lacuna of EU migration law 15.7.2 Member State prerogatives and their limits under EU law 15.7.3 ‘Golden passport’ schemes

470 472 473 474 475 477 477 478 479 481 482 482 484 484 486 488 489 491 494 496 497 499 500 502

16. Irregular Presence and Return 16.1 Theory and Policy Design

505 505

509

16.1.1 Beyond binary conceptions of (il)legality 16.1.2 Relentless search for effectiveness 16.1.3 Contrasting policy dynamics

16.2 Constitutional Foundations

506 507 508

Contents xix

16.2.1 Supranational competences 16.2.2 Detention in conformity with human rights 16.2.3 Human rights of those irregularly present 16.2.4 Territorial scope

16.3.1 Prosecution of ‘smuggling’ 16.3.2 Intimidation of search and rescue 16.3.3 Limits for criminal sanctions 16.3.4 Victims of trafficking 16.3.5 Employer sanctions

16.4.1 Refusal for different reasons 16.4.2 Entry bans under the Return Directive

16.5.1 Reasons for illegal stay 16.5.2 Exceptions from the scope 16.5.3 Return decision 16.5.4 Voluntary departure and removal 16.5.5 Limbo of ‘non-removable’ returnees

16.6.1 Mutual recognition of expulsion decisions 16.6.2 Cooperation between the Member States 16.6.3 Return to other Member States 16.6.4 Readmission agreements with third states

510 511 513 514

16.3 Criminalisation of Illegal Entry and Stay

514

16.4 Bans on Entry and Stay

524

16.5 Structure of the Return Directive

526

16.6 Inter-state Cooperation

533

16.7 Detention

539

16.8 Regularisation 16.9 Summary

546 547

16.7.1 Delimitation of EU legislation 16.7.2 Grounds for detention 16.7.2.1 Risk of absconding 16.7.2.2 Other criteria for asylum seekers 16.7.2.3 Additional grounds for returnees 16.7.3 Alternatives and length 16.7.4 Conditions of detention

515 517 518 520 522 524 525 527 528 528 530 532 534 535 536 537 539 540 541 542 543 544 545

17. Association Agreements with Neighbours 17.1 Constitutional Foundations

550 551

17.2 ‘Best friends’: European Economic Area and Switzerland

555

17.3 United Kingdom: Reversed Dynamics

559

17.4 Turkey: ‘Consolation Prize’ with Much Practical Effects

561

17.1.1 Supranational competences 17.1.2 Status in the EU legal order 17.1.3 Interpretative parallelism and its limits

17.2.1 Free movement of persons 17.2.2 Membership in Schengen and Dublin 17.2.3 Micro states and Gibraltar

17.3.1 Acquired rights of (former) Union citizens 17.3.2 Sovereign control over future movements

17.4.1 Turkish nationals residing in the Member States 17.4.2 ‘Standstill’ as dynamism for first admission

551 552 553 555 557 558 559 560 561 563

xx Contents

17.5 Western Balkans: Pre-accession Guarantees

564

17.6 Neighbourhood: Novel Focus on Migration Control

566

17.7 Summary

568

17.5.1 Former agreements and transitional periods 17.5.2 Limited impact of contemporary agreements

17.6.1 Eastern partnership: pragmatic cooperation 17.6.2 Union for the Mediterranean: hotbed of control practices

564 565 566 567

18. International Cooperation with Third States 18.1 Theory and Policy Design

570 571

18.2 Constitutional Foundations

577

18.3 Externalisation of Control Practices

581

18.4 Development Cooperation and Legal Pathways

588

18.5 Summary

595

18.1.1 ‘Migration management’ as an overarching narrative 18.1.2 A short history of the EU’s strategic vision 18.1.3 Trial and error of external migration policy 18.1.4 Give and take in the mutual interest

18.2.1 Supranational competences 18.2.2 Spread of informal cooperation frameworks

18.3.1 A recurring theme in the policy discourse 18.3.2 Conditionality: ‘carrots and sticks’ for cooperation 18.3.3 Capacity building and operational cooperation 18.3.4 EU involvement via Frontex and CSDP missions

18.4.1 Contextual complexity of the migration-development nexus 18.4.2 Financial support for multiple purposes 18.4.3 Refugee protection in the Global South 18.4.4 From ‘mobility’ to ‘talent’ partnerships

Bibliography Index

571 573 575 576 578 579 581 583 586 587 589 590 591 593

597 635

Table of Cases COURT OF JUSTICE 1/03 [Opinion] Lugano Convention EU:C:2006:81�� 574n.25 1/13 [Opinion] Accession of Third States to the Hague Convention EU:C:2014:2303�� 578n.44 1/15 [Opinion] Draft Agreement between Canada and the EU EU:C:2017:592 �� 239n.84, 239n.87, 244n.121, 245–46 1/17 [Opinion] EU-Canada CETA-Agreement EU:C:2019:341 �� 262n.69 1/19 [Opinion] Istanbul Convention EU:C:2021:832�� 579n.51 1/91 [Opinion] EEA I EU:C:1991:490�� 553n.18 2/13 [Opinion] Accession to the ECHR EU:C:2014:2454�� 133n.56 2/15 [Opinion] EU-Singapore FTA EU:C:2017:376�� 579n.50 2/94 [Opinion] Accession to the ECHR EU:C:1996:140�� 83n.66, 86n.81 C-4/11 Puid EU:C:2013:740 �� 164n.74 6/64 Costa v ENEL EU:C:1964:66�� 150n.1 C-8/15 P to C-10/15 P Ledra Advertising v Commission and ECB EU:C:2016:701��220n.142 C-8/20 LR EU:C:2021:404�� 558n.52 C-9/16 A EU:C:2017:483��268n.105, 323, 323n.151 9/56 Meroni EU:C:1958:7 �� 56n.83, 203, 205–6 12/86 Demirel EU:C:1987:400��550n.1, 551–52, 563n.90 C-13/01 Safalero EU:C:2003:447 ��219n.129 C-14/09 Genc EU:C:2010:57�� 554n.22 C-14/21 and C-15/21 Sea Watch EU:C:2022:604��308n.51, 309n.57, 517–18 C-15/11 Sommer EU:C:2012:371��454n.174 C-16/05 Tum and Dari EU:C:2007:530�� 563n.92 C-17/16 El Dakkak and Intercontinental EU:C:2017:341��327n.180 C-18/16 K EU:C:2017:680�� 512n.37, 542n.268, 544n.283 C-18/19 Stadt Frankfurt am Main EU:C:2020:511�� 159n.44, 163n.70, 259n.42, 273n.146, 546n.312 C-18/20 Bundesamt für Fremdenwesen und Asyl EU:C:2021:710�� 379n.306, n.308 C-19/08 Petrosian EU:C:2009:41 ��367n.205 C-19/21 Staatssecretaris van Justitie en Veiligheid EU:C:2022:605��169 C-20/12 Giersch and others EU:C:2013:411�� 518n.80 C-22/08 and C-23/08 Vatsouras and Koupatantze EU:C:2009:344�� 262n.69 C-23/12 Zakaria EU:C:2013:24 �� 138n.93, 327n.183 25/62 Plaumann EU:C:1963:17�� 79–80 26/62 van Gend en Loos EU:C:1963:1��18n.26, 122n.1, 150n.1, 356 C-27/04 Commission v Council EU:C:2004:436�� 80n.45 30/77 Bouchereau EU:C:1977:172 �� 260n.50 C-31/09 Bolbol EU:C:2010:351��398n.434 33/76 Rewe-Zentralfinanz EU:C:1976:188�� 181n.35 C-36/17 Ahmed EU:C:2017:273�� 364n.180, 377n.279 C-38/14 Zaizoune EU:C:2015:260 ��519 C-40/11 Iida EU:C:2012:2405�� 37n.157, 496n.189 41/74 Van Duyn EU:C:1974:133�� 260n.51 C-44/14 Spain v Parliament and Council EU:C:2015:554 �� 65n.137 C-45/07 Commission v Greece EU:C:2009:81 ��144n.121 C-47/15 Affum EU:C:2016:408��518n.79, 519n.90, 528n.159, 536n.220 C-47/17 and C-48/17 X and X EU:C:2018:900�� 161n.61, 367n.202 48/75 Royer EU:C:1976:57�� 255n.11 C-51/03 Georgescu EU:C:2004:200�� 83n.68 C-56/17 Fathi EU:C:2018:803�� 364n.182, 389–90, 391n.390

xxii Table of Cases C-57/09 and 101/09 B and D EU:C:2010:661 �� 271n.125, 272n.139, 399n.447, 399n.449 C-60/16 Khir Amayry EU:C:2017:675�� 540n.249, 544–45 C-61/11 PPU El Dridi EU:C:2011:268��161n.63, 519, 519n.82, 530n.175, 530n.177, 544n.284 C-63/09 Walz EU:C:2010:251��144n.127 C-63/15 Ghezelbash EU:C:2016:409�� 161n.61, 164n.75 C-64/96 and C-65/96 Uecker and Jacquet EU:C:1997:285�� 442n.87 C-66/21 Staatssecretaris van Justitie en Veiligheid EU:C:2022:809��521n.105, 522 C-67/14 Alimanovic EU:C:2015:597 ��269n.115 C-68/89 Commission v Netherlands EU:C:1991:226��326n.173 C-69/10 Samba Diouf EU:C:2011:524�� 182n.45, 219n.129, 269n.112, 374n.260 C-70/09 Hengartner and Gasser EU:C:2010:430�� 556n.39 C-70/18 A and others EU:C:2019:823��245, 245n.124, 246n.134 72/83 Campus Oil EU:C:1984:256 �� 260n.53 C-77/05 United Kingdom v Council EU:C:2007:803��66 C-79/13 Saciri and others EU:C:2014:103�� 412n.532, 413n.542 C-81/13 United Kingdom v Council EU:C:2014:2449�� 436n.53 C-82/16, KA and others EU:C:2018:308��37n.157, 259n.42, 442n.91, 526n.145 C-83/12 PPU Minh Khao Vo EU:C:2012:202 �� 188n.83 C-83/14 CHEZ Razpredelenie Bulgaria EU:C:2015:480��266n.100 C-84/12 Koushkaki EU:C:2013:862 ��162n.67, 163n.69, 255n.10, 261n.65, 271n.127, 289n.59, 290n.61, 292n.84 C-85/96 Martínez Sala EU:C:1998:217 �� 155n.21 C-86/12 Alopka and Moudoulou EU:C:2013:645�� 442n.93 C-88/14 Commission v Parliament and Council EU:C:2015:499�� 59n.95 C-89/18 A EU:C:2019:580�� 564n.97 90/63 and 91/63 Commission v Luxembourg and Belgium EU:C:1964:80�� 156n.26 C-91/05 Commission v Council EU:C:2008:288�� 579n.49 C-91/20 Bundesrepublik Deutschland EU:C:2021:898�� 146n.137, 160n.49, 271n.130, 391n.385 C-94/20 Land Oberösterreich EU:C:2021:477��138n.91, 266n.101, 483n.94, 484n.100 C-95/99–C-98/99 and C-180/99 Khalil EU:C:2001:532 �� 262n.69 C-97/91 Oleificio Borelli v Commission EU:C:1992:491��220n.135 98/80 Romano EU:C:1981:104 �� 205n.39 C-101/13 U EU:C:2014:2249�� 111n.81, 245n.124, 334n.220 C-105/15 P to C-109/15 P Mallis and Malli v Commission and ECB EU:C:2016:702�� 218n.124, 220n.137 106/77 Simmenthal EU:C:1978:49�� 155n.22 C-109/01 Akrich EU:C:2003:491�� 441n.83 C-112/20 État belge EU:C:2021:197��533n.197 C-113/89 Rush Portuguesa EU:C:1990:142��459n.206 C-118/20 Wiener Landesregierung EU:C:2022:34��501 C-123/17 Yön EU:C:2018:632�� 563n.93, 564n.97 C-127/08 Metock EU:C:2008:449�� 441n.83 C-129/18 SM EU:C:2019:248 �� 136n.76, 443n.99 C-130/08 Commission v Greece EU:C:2008:854��365n.194 C-131/12 Google Spain SL and Google Inc EU:C:2014:317 ��243n.115 C-133/06 Parliament v Council EU:C:2008:257 �� 58n.89, 382n.324 C-133/19, C-136/19 and C-137/19 État belge EU:C:2020:577 �� 161n.59, 162n.68, 167n.91, 448n.130 C-135/08 Rottmann EU:C:2010:104��269–70, 500–1 C-137/05 United Kingdom v Council EU:C:2007:805�� 66n.148 C-138/13 Dogan EU:C:2014:2066 �� 563n.95 139/85 Kempf EU:C:1986:223�� 23n.59 C-140/12 Brey EU:C:2013:565 �� 155n.21 C-141/08 P Foshan Shunde Yongjian Housewares and Hardware v Council EU:C:2009:598��169n.95 C-145/09 Tsakouridis EU:C:2010:708 �� 260n.54 C-146/14 Mahdi EU:C:2014:1320�� 161n.65, 167, 185n.69, 532n.193, 541n.263, 543n.282, 545, 546n.316 C-148/02 Garcia Avello EU:C:2003:539�� 442n.88

Table of Cases xxiii C-148/13–C-150/13 A, B and C EU:C:2014:2406�� 389–90, 389nn.374–75 C-153/14 K and A EU:C:2015:453 ��266n.94, 490–91, 490n.144, 492n.153 C-155/11 PPU Mohammad Imran EU:C:2011:387 ��490n.145 157/79 Pieck EU:C:1980:179��326n.173 C-158/07 Förster EU:C:2008:630 ��269n.115 C-159/21 Országos Idegenrendeszeti Főigazgatóság and others EU:C:2022:708 �� 185n.70, 261n.64, 273n.150, 379n.312, 399n.446 C-161/15 Bensada Benallal EU:C:2016:175�� 181n.41 C-162/00 Pokrzeptowiez-Meyer EU:C:2002:57��565n.107 C-163/17 Jawo EU:C:2019:218��161n.61, 352n.91, 367n.209, 368n.217 C-165/14 Rendón Marín EU:C:2016:675�� 259n.45 C-165/16 Lounes EU:C:2017:862�� 442n.88, 499n.206 C-166/13 Mukarubega EU:C:2014:2336�� 185n.67, 186n.74, 377n.290 C-168/91 Konstantinidis EU:C:1993:115��334n.220 C-171/13 Demirci and others EU:C:2015:8�� 476n.49, 562n.84 C-171/95 Tetik EU:C:1997:31��562 C-175/08, C-176/08, C-178/08 and C-179/08 Abdulla EU:C:2010:105�� 144n.126, 388n.369, 390n.381, 395n.414, 396n.420, 400 C-175/11 D and A EU:C:2013:45�� 266n.94, 378n.294 C-175/17 Belastingdienst/Toeslagen EU:C:2018:776 �� 181n.37, 184n.64 C-179/11 Cimade and GISTI EU:C:2012:594��411, 413n.542, 414n.545 C-181/16 Gnandi EU:C:2018:465��71n.7, 184n.62, 377n.290, 411n.529, 527 181/73 Haegeman EU:C:1974:41 �� 552n.9 C-184/16 Petrea EU:C:2017:684�� 84n.71, 528n.167 C-184/99 Grzelczyk EU:C:2001:458�� 23n.64 C-186/21 Republika Slovenija EU:C:2021:447 ��542n.273 C-188/10 and C-189/10 Melki and Abdeli EU:C:2010:363 �� 322–23 C-192/89 Sevince EU:C:1990:322�� 553n.12, 562n.81 C-192/99 Kaur EU:C:2001:106��500n.217 C-193/19 Migrationsverket EU:C:2021:168�� 190n.99 C-194/01 Commission v Austria EU:C:2004:248 �� 156n.29 C-194/19 État belge EU:C:2021:270��168n.93, 183–84, 363n.169, 367n.203 C-199/12–C-201/12 X and others EU:C:2013:720��134n.67, 263n.72, 389n.371, 391, 396–97, 396n.423 C-200/02 Zhu and Chen EU:C:2004:639�� 442n.85 C-201/16 Shiri EU:C:2017:805��168n.93, 367 C-202/13 McCarthy and others EU:C:2014:2450��276n.170 C-202/18 Rimšēvičs v Latvia EU:C:2019:139��219n.126 C-203/15 and C-698/15 Tele2 Sverige, Watson and others EU:C:2016:970 ��246n.133 C-205/06 Commission v Austria EU:C:2009:118 �� 553n.16 C-205/15 Toma and Biroul Executorului Judecătoresc Horaţiu-Vasile Cruduleci EU:C:2016:499�� 183n.54 C-208/09 Sayn-Wittgenstein EU:C:2010:806 �� 260n.52, 274n.154 C-208–210/17 P NF and others v European Council EU:C:2018:705�� 79n.44 C-215/03 Oulane EU:C:2005:95�� 255n.13 C-216/18 PPU Minister for Justice and Equality EU:C:2018:589�� 353n.96 C-219/17 Berlusconi and Fininvest EU:C:2018:1023 ��219n.131 C-220/17 Planta Tabak EU:C:2019:76�� 239n.85 C-221/11 Demirkan EU:C:2013:583�� 285n.29, 563n.96 C-221/17 Tjebbes and others EU:C:2019:189�� 501–2 C-225/16 Ouhrami EU:C:2017:590 ��526 C-225/19 and C-226/19 Minister van Buitenlandse Zaken EU:C:2020:951�� 185n.68, 188n.89, 220n.135, 255n.10, 291n.77 C-228/06 Soysal and Savatli EU:C:2009:101�� 563n.94 C-231/21 Bundesamt für Fremdenwesen und Asyl EU:C:2022:237�� 159n.41, 367n.210 C-233/18 Haqbin EU:C:2019:956�� 136n.77, 352n.91, 413 C-233/19 CPAS de Liège EU:C:2020:757��166n.83, 184 C-235/95 Dumon and Froment EU:C:1998:365 �� 83n.67, 268n.108

xxiv Table of Cases C-236/09 Association belge des Consommateurs Test-Achats ASBL and others EU:C:2011:100�� 239n.85, 484n.97 C-237/15 PPU Lanigan EU:C:2015:474�� 190n.98 C-237/91 Kus EU:C:1992:527�� 563n.90 C-238/19 Bundesamt für Migration und Flüchtlinge EU:C:2020:945��86n.84, 389n.371, 397–98 C-239/14 Tall EU:C:2015:824��134n.64, 184n.62, 307n.46 C-240/17 E EU:C:2018:8�� 190n.101, 190n.102, 259n.38, 259n.42, 261n.62, 525n.136, 536n.219 C-241/05 Bot EU:C:2006:634 �� 289n.57, 317n.108 C-241/21 Politsei-ja Piirivalveamet EU:C:2022:753 ��512, 543n.278 C-244/10 and C-245/10 Mesopotamia Broadcast EU:C:2011:607 �� 187n.82 C-244/17 Commission v Council EU:C:2018:662�� 579n.52 C-245/11 K EU:C:2012:685�� 363n.171, 365n.184 C-245/21 and C-248/21 Bundesrepublik Deutschland EU:C:2022:709�� 161n.61, 367n.206 C-246/17 Diallo EU:C:2018:499�� 182n.47 C-247/20 Commissioners for Her Majesty’s Revenue and Customs EU:C:2022:177�� 559n.65 C-249/13 Boudjlida EU:C:2014:2431�� 184n.66, 186n.74 C-249/96 Grant EU:C:1998:63 ��140n.101 C-254/11 Shomodi EU:C:2013:182��328n.186 C-255/19 Secretary of State for the Home Department EU:C:2021:36�� 392n.399, 393n.400, 395n.415, 396n.420, 400 C-256/11 Dereci and others EU:C:2011:734 ��37n.157, 302n.7, 442n.91 C-257/01 Commission v Council EU:C:2005:25�� 58n.92 C-257/99 Barkoci and Malik EU:C:2001:491��565n.103 C-261/08 and C-348/08 Zurita García and Choque Cabrera EU:C:2009:648�� 159n.40, 327n.179 C-262/96 Sürül EU:C:1999:228�� 563n.88 C-265/03 Simutenkov EU:C:2005:213��566n.111 C-268/99 Jany EU:C:2001:616 ��565n.103 C-269/18 C and others EU:C:2018:544��540n.251 C-270/12 United Kingdom v Council and Parliament EU:C:2014:18�� 204n.31, 205n.39 270/80 Polydor EU:C:1982:43�� 554n.19 C-273/20 and C-355/20 Bundesrepublik Deutschland EU:C:2022:617��448n.130 C-275/02 Ayaz EU:C:2004:570�� 554n.23 C-277/11 MM EU:C:2012:2479 �� 169n.96 C-278/12 Adil EU:C:2012:508��323, 323n.148 C-279/09 DEB EU:C:2010:811�� 184n.66 C-279/20 Bundesrepublik Deutschland EU:C:2022:618�� 447–48 281, 283, 284, 285 and 287/85 Germany and others v Commission EU:C:1987:351 �� 25n.74, 437n.58 283/81 CILFIT EU:C:1982:335�� 85n.80 C-285/12 Diakité EU:C:2014:39��160n.49, 402 C-290/14 Celaj EU:C:2015:640��161n.63, 518n.80, 519n.82, 519n.91 C-291/12 Schwarz EU:C:2013:670�� 282n.13, 334n.216 C-293/12 and C-594/12 Digital Rights Ireland EU:C:2014:238��243n.115, 245 C-294/06 Payir EU:C:2008:36�� 562n.83 294/83 Les Verts EU:C:1986:166��218n.123, 356 C-297/12 Filev and Osmani EU:C:2013:569 �� 526n.146, 528n.162 C-297/17, C-318/17, C-319/17 and C-438/17 Ibrahim and others EU:C:2019:219�� 354n.105, 364n.182, 369n.226, 370n.231 C-300/09 and C-301/09 Toprak and Oguz EU:C:2010:756�� 563n.93 C-300/11 ZZ EU:C:2013:363�� 185n.70 C-300/98 and C-392/98 Christian Dior EU:C:2000:688 �� 552n.6 C-302/18 X EU:C:2019:830��159n.42, 257n.31, 258n.34 C-302/19 Istituto Nazionale della Previdenza Sociale EU:C:2020:957�� 480n.71, 486n.113 C-303/08 Bozkurt EU:C:2010:800 �� 38n.159 C-303/19 Istituto Nazionale della Previdenza Sociale EU:C:2020:958�� 476n.48, 481, 482n.79 C-308/06 Intertanko EU:C:2008:312�� 553n.11, 553n.14 C-309/14 CGIL and INCA EU:C:2015:523��256 C-310/08, Ibrahim EU:C:2010:80�� 441n.82

Table of Cases xxv C-312/12 Ajdini EU:C:2013:103�� 138n.88, 272n.137 C-322/19 and C-385/19 The International Protection Appeals Tribunal and others EU:C:2021:11��155n.20, 167n.92, 276n.171, 319n.117, 411n.528, 412n.536, 413n.542 C-323/21, C-324/21 and C-325/21 Staatssecretaris van Justitie en Veiligheid EU:C:2023:4 ��161, 168, 365n.191, 366n.198, 367n.203 C-327/02 Panayotova and others EU:C:2004:718��565n.104 C-329/06 and C-343/06 Wiedemann and Funk EU:C:2008:366�� 187n.82 C-329/11 Achughbabian EU:C:2011:807��519, 527n.154 C-336/05 Eckhouik EU:C:2006:394 ��567n.118 C-338/13 Noorzia EU:C:2014:2092 ��161n.55, 269, 444 C-34/09 Ruiz Zambrano EU:C:2011:124��37, 442 C-340/97 Nazli EU:C:2000:77�� 554n.24 C-341/18 Staatssecretaris van Justitie en Veiligheid EU:C:2020:76�� 161n.66, 327n.179 C-348/09 I EU:C:2012:300�� 260n.54 C-348/16 Sacko EU:C:2017:591 �� 186n.76 C-349/20 Secretary of State for the Home Department EU:C:2022:151�� 389n.370, 398–99 C-350/20 INPS EU:C:2021:659�� 481n.75, 482n.83, 484n.100 C-351/95 Kadiman EU:C:1997:205�� 255n.14, 562n.80 C-353/16 MP EU:C:2018:276��393n.402, 403 C-355/10 Parliament v Council EU:C:2012:516 �� 58n.90, 332n.203 C-356/11 and C-357/11 O and S EU:C:2012:776��37n.157, 136n.77, 439n.74, 443n.97 C-357/09 PPU Kadzoev EU:C:2009:741 ��543, 544n.291, 545n.295 C-359/16 Altun and others EU:C:2018:63��191n.106 C-36/20 PPU Ministerio Fiscal EU:C:2020:495�� 170n.104, 331n.198, 354n.107, 373–74, 540n.247, 542n.267 C-360/16 Hasan EU:C:2018:35��366n.198 C-364/11 El Karem El Kott and others EU:C:2012:286�� 159n.45, 398, 399n.442, 487n.125 C-365/02 Lindfors EU:C:2004:449�� 167n.88 C-368/20 and C-369/20 Landespolizeidirektion Steiermark and Bezirkshauptmannschaft Leibnitz EU:C:2022:298�� 274n.155, 319n.119, 320–21 C-369/17 Ahmed EU:C:2018:713�� 145n.133, 399n.445, 399n.448, 487n.118 C-369/90 Micheletti EU:C:1992:295��500n.216 C-370/17 CRPNPAC EU:C:2020:260�� 191n.106, 460n.216 C-371/08 Ziebell EU:C:2011:809�� 23n.60, 259–60, 554 C-373/13 T EU:C:2015:413��255n.9, 259n.39, 259n.41, 260n.51, 261n.63, 419 C-377/12 Commission v Council EU:C:2014:1903�� 579n.53 C-378/97 Wijsenbeek EU:C:1999:439�� 26n.79, 227n.22 C-379/20 Udlændingenævnet EU:C:2021:660�� 564n.97 C-380/17 K and B EU:C:2018:877 ��269n.114, 447 C-380/18 EP EU:C:2019:1071��259–60, 261n.59, 261n.65 C-381/18 and C-382/18 GS and VG EU:C:2019:1072�� 160n.50, 260n.48, 261n.59, 262n.66, 272n.134 C-383/03 Dogan EU:C:2005:436�� 562n.82 C-383/13 PPU G and R EU:C:2013:533�� 161n.65, 186n.74 C-391/16, C-77/17 and C-78/17 M and others EU:C:2019:403�� 144n.125, 166n.81, 354n.108, 419 C-392/95 Parliament v Council EU:C:1997:289 �� 284n.17 C-394/12 Abdullahi EU:C:2013:813�� 164n.74, 368n.218 C-397/01–C-403/01 Pfeiffer EU:C:2004:584�� 156n.28 C-399/12 Germany v Council EU:C:2014:2258��144n.122 C-400/10 PPU McB EU:C:2010:582�� 134n.65 C-401/11 Soukupová EU:C:2013:223�� 138n.94 C-402/19 CPAS de Seraing EU:C:2020:759�� 532n.192, 532n.195 C-403/05 Parliament v Commission EU:C:2007:624 �� 579n.54 C-403/16 El Hassani EU:C:2017:960�� 85n.75, 169n.94, 183n.50, 183n.53, 292n.78 C-404/17 A EU:C:2018:588��381n.323 C-406/18 Bevándorlási és Menekültügyi Hivatal EU:C:2020:216��156n.25, 167n.87, 182n.48 C-409/20 Subdelegación del Gobierno en Pontevedra EU:C:2022:148 ��530n.176 C-411/06 Commission v Council EU:C:2009:518�� 305n.25

xxvi Table of Cases C-411/10 and C-493/10 NS and others EU:C:2011:865��134n.63, 161n.60, 177n.20, 368n.217, 369n.225, 369n.229 C-413/99 Baumbast and R EU:C:2002:493�� 441n.82 C-416/96 El-Yassini EU:C:1999:107��456n.191 C-422/21 Ministero dell’Interno EU:C:2022:616��413 C-429/15 Danqua EU:C:2016:789�� 182n.45 C-43/93 Vander Elst EU:C:1994:310 ��459 C-430/11 Sagor EU:C:2012:777 �� 519n.85, 519n.89 C-431/11 United Kingdom v Council EU:C:2013:589�� 436n.51, 552n.7 C-432/20 Landeshauptmann von Wien EU:C:2022:39��475n.38, 498 C-434/09 McCarthy EU:C:2011:277�� 302n.6 C-434/93 Bozkurt EU:C:1995:168 �� 554n.21 C-441/02 Commission v Germany EU:C:2006:253�� 83n.63 C-441/19 Staatssecretaris van Justitie en Veiligheid EU:C:2021:9�� 154n.19, 363n.166, 533n.196 C-443/14 and C-444/14 Alo and Osso EU:C:2016:127�� 264n.79, 416n.555, 417, 481n.76, 487n.125 C-444/17 Arib and others EU:C:2019:220�� 319n.117, 528n.161 C-445/03 Commission v Luxemburg EU:C:2004:655��460n.215 C-446/12–C-449/12 Willems and others EU:C:2015:238��334n.218 C-448/19 Subdelegación del Gobierno en Guadalajara EU:C:2020:467�� 85n.77 C-449/16 Martinez Silva EU:C:2017:485�� 482n.83 C-456/12 O and B EU:C:2014:135 �� 276n.169, 442n.86 C-457/17 Maniero EU:C:2018:912��266n.101 C-459/99 MRAX EU:C:2002:461��326n.174 C-462/20 ASGI and others EU:C:2021:894 �� 482n.79 C-465/07 Elgafaji EU:C:2009:94�� 86, 401–3 C-467/02 Cetinkaya EU:C:2004:708�� 554n.24 C-469/13 Tahir EU:C:2014:2094�� 271n.125, 272n.137, 496n.181, 497n.190 C-472/13 Shepherd EU:C:2015:117 �� 389n.373, 391n.387, 397n.428 C-473/13 and C-514/13 Bero and Bouzalmate EU:C:2014:2095��546n.310 C-473/16 F EU:C:2018:36�� 167n.90, 389n.372, 389n.378 C-474/13 Pham EU:C:2014:2096��546n.311 C-479/21 PPU Governor of Cloverhill Prison and others EU:C:2021:929�� 552n.7 C-481/13 Qurbani EU:C:2014:2101�� 144n.120, 518n.79 C-482/08 United Kingdom v Council EU:C:2010:631�� 66n.148, 316n.101 C-483/20 Commissaire général aux réfugiés et aux apatrides EU:C:2022:103�� 377n.284, 447n.125 C-484/22 Bundesrepublik Deutschland EU:C:2023:122��533n.196 C-490/16 AS EU:C:2017:585�� 164n.75, 367n.205 C-491/13 Ben Alaya EU:C:2014:2187 �� 162n.67, 255n.7, 454 C-497/21 Bundesrepublik Deutschland, EU:C:2022:721 �� 65n.139, 379n.303 C-502/04 Torun EU:C:2006:112�� 562n.85 C-502/10 Singh EU:C:2012:3076 ��496, 497n.191 C-503/03 Commission v Spain EU:C:2006:74�� 246n.136, 326n.174 C-507/19 Bundesrepublik Deutschland EU:C:2021:3�� 146n.135, 398n.437 C-508/10 Commission v Netherlands EU:C:2012:243 ��255n.8, 256, 256n.19, 476n.46 C-513/12 Ayalti EU:C:2013:210��490n.145 C-517/17 Addis EU:C:2020:579 �� 186n.72, 374n.263 C-519/18 Bevándorlási és Menekültügyi Hivatal EU:C:2019:1070�� 257n.32, 258n.35, 272n.138, 272n.139, 447 C-519/20 Landkreis Gifhorn EU:C:2022:178��545n.304, 546 C-524/06 Huber EU:C:2008:724�� 245n.124, 245n.127, 246n.134, 246n.137 C-528/11 Halaf EU:C:2013:342 ��145n.133, 365 C-528/15 Al Chodor EU:C:2017:213 ��541, 541n.263, 544n.283 C-534/11 Arslan EU:C:2013:343�� 527n.157, 542n.266, 543n.278 C-540/03 Parliament v Council EU:C:2006:429 ��36n.147, 138n.90, 141n.104, 160n.53, 165, 167n.86, 255n.6, 438n.63, 439n.74, 444n.101, 444n.105, 491n.150, 491n.151 C-540/17 and C-541/17 Hamed and Omar EU:C:2019:964 �� 370n.230, 377n.283

Table of Cases xxvii C-542/13 M’Bodj EU:C:2014:2452��138n.88, 272n.140, 350n.71, 392–93, 403n.479 C-544/15 Fahimian EU:C:2017:255�� 162n.67, 163n.69, 259, 259n.38, 260n.50, 261n.60, 261n.65 C-546/19 Westerwaldkreis EU:C:2021:432 �� 525n.139, 526n.151, 529n.172, 532n.192, 533n.198, 546n.317 C-550/16 A and S EU:C:2018:248��448n.130 C-554/13 Zh and O EU:C:2015:377�� 259n.38, 525n.142 C-554/19 Staatsanwaltschaft Offenburg EU:C:2020:439��323n.153 C-556/17 Torubarov EU:C:2019:626 �� 166n.83, 186n.78 C-557/17 YZ and others EU:C:2019:203 ��255n.16, 498 C-558/14 Khachab EU:C:2016:285��161n.56, 167n.87, 257n.30, 258n.33 C-560/14 M EU:C:2017:101 ��185n.71, 186n.73, 375n.266, 376n.274, 389n.374 C-561/14 Genc EU:C:2016:247�� 564n.97 C-562/13 Abdida EU:C:2014:2453�� 184n.62, 532, 532n.190 C-564/18 Bevándorlási és Menekültügyi Hivatal (Tompa) EU:C:2020:218 �� 182n.45, 377n.280, 382n.330, 383n.339 C-564/22 Bundesrepublik Deutschland EU:C:2022:951�� 185n.70 C-565/19 P Carvalho and others v Parliament and Council EU:C:2021:252��218n.125 C-568/19 Subdelegación del Gobierno en Toledo EU:C:2020�� 520n.92 C-571/10 Kamberaj EU:C:2012:233��165n.79, 266n.96, 481n.74, 483n.87, 483n.92, 484n.97, 484n.100 C-573/14 Lounani EU:C:2017:71��399n.450 C-575/12 Air Baltic Corporation EU:C:2014:2155�� 161n.66, 271n.127, 324n.157, 325n.166 C-578/08 Chakroun EU:C:2010:117��160n.53, 165n.79, 257n.31, 257n.32 C-578/16 PPU CK and others EU:C:2017:127�� 369n.221, 369n.229 C-579/13 P and S EU:C:2015:369��264n.79, 476n.48, 490–91 C-582/17 and 583/17 H and R EU:C:2019:280�� 161n.60, 367n.202 C-583/11 P Inuit Tapiriit Kanatami and others v Parliament and Council EU:C:2013:625�� 79n.42, 220n.134 C-584/18 Blue Air –Airline Management Solutions EU:C:2020:324��185n.68, 296 C-585/16 Alheto EU:C:2018:584�� 85n.75, 157n.33, 186n.78, 384n.346, 399n.441, 399n.443 C-600/14 Germany v Council EU:C:2017:935�� 578n.42 C-601/15 PPU N EU:C:2016:84 �� 259n.38, 512, 540n.248, 542n.265, 543n.275 C-604/12 HN EU:C:2014:302�� 185n.67 C-604/12 N EU:C:2014:302�� 376n.274, 377n.287 C-606/10 ANAFE EU:C:2012:348 �� 319n.117, 324n.156 C-609/17 and C-610/17 TSN and AKT EU:C:2019:981�� 138n.89 C-616/19 Minister for Justice and Equality EU:C:2020:1010�� 66n.153, 161n.60, 377n.282 C-617/10 Åkerberg Fransson EU:C:2013:280 �� 137n.84, 138n.87 C-620/18 and C-626/18 Hungary and Poland v Parliament and Council EU:C:2020:1001�� 477n.54 C-635/17 E EU:C:2019:192��60n.99, 136n.76, 439n.75, 446, 447n.119 C-636/16 López Pastuzano EU:C:2017:949�� 259n.43, 262n.66 C-638/16 PPU X and X EU:C:2017:173�� 74, 77, 294–95 C-643/15 and C-647/15 Slovakia and Hungary v Council EU:C:2017:631 �� 54n.74, 156n.23, 268n.104, 350–51, 354n.106, 361n.152, 407n.502, 472n.17 C-646/16 Jafari EU:C:2017:586��77, 153n.9, 274, 325n.168, 356n.124, 363n.173, 369n.221 C-647/16 Hassan EU:C:2018:368�� 77n.31, 160n.48, 160n.49, 161n.61, 367n.202 C-648/11 MA and others EU:C:2013:367�� 136n.77, 363nn.165–66 C-651/19 Commissaire général aux réfugiés and aux apatrides EU:C:2020:681 ��181, 182n.46 C-652/16 Ahmedbekova EU:C:2018:801�� 272n.133, 392, 396n.423, 397n.427 C-656/11 United Kingdom v Council EU:C:2014:97�� 436n.51 C-660/13 Council v Commission EU:C:2016:616�� 61n.109 C-661/17 MA and others EU:C:2019:53��365 C-662/17 E G EU:C:2018:847�� 166n.83, 377n.288 C-668/15 Jyske Finans EU:C:2017:278�� 266n.98, 266n.101 C-670/16 Mengesteab EU:C:2017:587 ��160n.49, 161n.61, 164n.75, 367n.201 C-673/16 Coman and others EU:C:2018:385�� 443n.98 C-673/19 M and others EU:C:2021:127�� 532n.195, 536n.221, 536n.222, 546n.316

xxviii Table of Cases C-673/20 Préfet du Gers and Institut National de la Statistique et des Études Économiques EU:C:2022:449 ��500n.214 C-680/17 Vethanayagam EU:C:2019:627 �� 189n.94, 556n.42 C-682/15 Berlioz Investment Fund EU:C:2017:373��219, 220n.134 C-69/21 Staatssecretaris van Justitie en Veiligheid EU:C:2022:913�� 352–53, 529n.172 C-695/15 PPU Mirza EU:C:2016:188�� 161n.60, 364n.183 C-706/18 Belgische Staat EU:C:2019:993�� 161n.58, 182n.47 C-71/11 and C-99/11 Y and Z EU:C:2012:518 ��388n.369, 391, 391n.393 C-713/17 Ayubi EU:C:2018:929�� 156n.27, 481n.77, 487n.125 C-715/17, C-718/17 and C-719/17 Commission v Poland and others EU:C:2020:257��82n.61, 261n.61, 273, 274n.156, 356n.124, 361n.153 C-718/19 Ordre des barreaux francophones et germanophone and others EU:C:2021:505�� 35n.143, 264n.79, 269n.111 C-72/22 PPU Valstybės sienos apsaugos tarnyba EU:C:2022:505�� 275n.163, 331n.199, 354n.109, 374n.256, 542n.267, 543n.274 C-720/17 Bilali EU:C:2019:448�� 145n.133, 379n.311, 446n.117, 498n.199 C-720/19 Stadt Duisburg EU:C:2020:847 �� 562n.86 C-720/20 Bundesrepublik Deutschland EU:C:2022:603��363n.167 C-754/18 Ryanair Designated Activity Company EU:C:2020:478�� 326n.171, 326n.174 C-755/19 THC EU:C:2021:108�� 182n.46 C-768/19 Bundesrepublik Deutschland EU:C:2021:709��448n.130 C-806/18 JZ EU:C:2020:723 �� 519n.90, 526n.150 C-808/18 Commission v Hungary EU:C:2020:493 ��82n.59, 255n.5, 273n.144, 274n.153, 274n.157, 330n.197, 374n.256, 385n.350, 539n.244, 542 C-817/19 Ligue des droits humains EU:C:2022:491 �� 238n.82, 245n.124, 245n.126 C-821/19 Commission v Hungary EU:C:2021:930 �� 276n.172, 354n.107, 373n.250, 375 C-825/21 Centre public d’action sociale de Liège EU:C:2022:810��546n.316 C-848/19 P Germany v Poland EU:C:2021:598��356 C-897/19 PPU Ruska Federacija EU:C:2020:262 �� 556n.40 C-901/19 Bundesrepublik Deutschland EU:C:2021:472�� 402–3 C-921/19 Staatssecretaris van Justitie en Veiligheid EU:C:2021:478�� 378–79, 379n.305, 389n.374 C-924/19 PPU and C-925/19 PPU Országos Idegenrendészeti Főigazgatóság Dél-alföldi Regionális Igazgatóság EU:C:2020:367��73n.15, 181n.42, 354n.107, 377n.281, 378n.300, 379n.310, 414n.549, 511–12, 531n.186, 540n.246, 542–43, 544–45 C-930/19 État belge EU:C:2021:657��35n.143, 262n.70, 264n.79 C-949/19 Konsul Rzeczypospolitej Polskiej w N EU:C:2021:186�� 183n.53 GENERAL COURT T-31/18 Izuzquiza and Semsrott v Frontex EU:T:2019:815 �� 79n.43 T-192/16 NF v European Council EU:T:2017:128�� 79n.44 T-229/05 PKK v Council EU:T:2008:87��220n.138 T-326/99 Olivieri v Commission and EMEA EU:T:2003:351��219n.130 T-429/18 BRF and SHB Comercio e Industria de Alimentos v Commission EU:T:2020:322�� 169n.95 T-528/20 Kočner v Europol EU:T:2021:631��220n.143 T-600/21 WS and others v Frontex EU:T:2022:474�� 74n.21 EUROPEAN COURT OF HUMAN RIGHTS AA and others v North Macedonia App no 55798/16 (ECtHR, 5 April 2022) �� 311n.72 AM v Netherlands App no 29094/09 (ECtHR, 5 July 2016)�� 351n.85 Abdulaziz and others v United Kingdom App nos 9214/80, 9473/81 and 9474/81 (ECtHR, 28 May 1985) ��22n.47, 129n.34, 266n.96, 438n.62 Amuur v France App no 19776/92 (ECtHR, 15 June 1996)�� 511n.31, 512n.40 Animal Defenders International v the United Kingdom App no 48876/08 (ECtHR [GC], 22 April 2013)��270n.120 Antwi and others v Norway App no 26940/10 (ECtHR, 14 February 2012) �� 478n.59

Table of Cases xxix Asady and others v Slovakia App no 24917/15 (ECtHR, 24 March 2020) �� 311n.73 Bah v United Kingdom App no 56328/07 (ECtHR, 27 November 2011) �� 265n.90 Berrehab v Netherlands App no 10730/84 (ECtHR, 21 June 1988)�� 477n.55 Biao v Denmark App no 38590/10 (ECtHR [GC], 24 May 2016)��267n.102 Biao v Denmark App no 38590/10 (ECtHR, 25 March 2013)�� 265n.92 Bosphorus Airways v Ireland App no 45036/98 (ECtHR [GC], 30 June 2005)�� 133n.57 Boultif v Switzerland App no 54273/00 (ECtHR, 2 August 2001)�� 478n.56 Carson and others v United Kingdom App no 42184/05 (ECtHR [GC], 16 March 2010)�� 266n.93 Čonka v Belgium App no 51564/99 (ECtHR, 5 February 2002) �� 184n.61, 310n.68 DH and others v Czech Republic App no 57325/00 (ECtHR [GC], 13 November 2007) ��266n.100 DNM v Sweden App no 28379/11 (ECtHR, 27 June 2013)�� 395–96 Dalea v France App no 964/07 (ECtHR, 2 February 2010) ��246n.138 Daytbegova and Magomedova v Austria App no 6198/12 (ECtHR, 4 June 2013)��369n.222 De Souza Ribeiro v France App no 22689/07 (ECtHR [GC], 13 December 2012)�� 184n.57 Dhabi v Italy App no 17120/09 (ECtHR, 8 April 2013) �� 264n.84 El Ghatet v Switzerland App no 56971/10 (ECtHR, 8 November 2016)�� 439n.68 FG v Sweden App no 43611/11 (ECtHR [GC], 23 March 2016) ��388n.369 Gaygusuz v Austria App no 17371/90 (ECtHR, 16 September 1996)��264 HLR v France App no 24573/94 (ECtHR [GC], 29 April 1997)�� 351n.82 Hirsi Jamaa and others v Italy App no 27765/09 (ECtHR [GC], 23 February 2012)��306, 306n.27 Hode and Abdi v United Kingdom App no 22341/09 (ECtHR, 6 November 2012)��487 Hoti v Croatia App no 63311/14 (ECtHR, 26 April 2018)�� 126n.20, 513n.50 IAA and others v United Kingdom App no 25960/13 (ECtHR, 8 March 2016) �� 135n.72, 438n.66, 439 Ilias and Ahmed v Hungary App no 47287/15 (ECtHR [GC], 21 November 2019)�� 310n.63, 511, 545n.305 JK v Sweden App no 59166/12 (ECtHR [GC], 23 August 2016) ��389, 394n.411 JR and others v Greece App no 22696/16 (ECtHR, 25 January 2018) �� 511n.31 Jeunesse v Netherlands App no 12738/10 (ECtHR [GC], 3 October 2014)�� 514n.52 KI v France App no 5560/19 (ECtHR, 15 April 2021)��399n.444 KN v United Kingdom App no 28774/22 (ECtHR, 14 June 2022)��385n.357 Károly Nagy v Hungary App no 56665/09 (ECtHR [GC], 14 September 2016)�� 169n.98 Khlaifia and others v Italy App no 16483/12 (ECtHR [GC], 12 December 2016) �� 184n.59, 311n.73, 331n.198 Konstatinov v Netherlands App no 16351/03 (ECtHR, 26 April 2013)�� 257n.28 Koua Poirrez v France App no 40892/98 (ECtHR, 30 September 2003)�� 265n.88 Kudła v Poland App no 30210/96 (ECtHR [GC], 26 October 2000) �� 73n.14 LM and others v Russia App nos 40081/14, 40088/14 and 40127/14 (ECtHR, 15 October 2015)��146n.140 MA v Denmark App no 6697/18 (ECtHR [GC], 9 July 2021)�� 270n.119, 438–39, 438n.64, 487n.125 MA v Lithuania App no 59793/17 (ECtHR, 11 December 2018)��331n.198 MB and RA v Spain App no 20351/17 (ECtHR, 5 July 2022)�� 307n.40, 310n.66 MH and others v Croatia App nos 15670/18 and 43115/18 (ECtHR, 18 November 2021)�� 311n.72, 512n.41, 545n.307 MK and others v Poland App nos 40503/17, 42902/17 and 43643/17 (ECtHR, 23 July 2020)�� 310, 311n.72, 330n.197 MN and others v Belgium App no 3599/18 (ECtHR [GC], 5 May 2020)�� 131n.45, 183n.51, 306, 306n.26 MSS v Belgium and Greece App no 30696/09 (ECtHR [GC], 21 January 2011)��133n.59, 146n.140, 183n.52, 184n.65, 352, 368n.212, 369n.225, 545n.305 MT and others v Sweden App no 22105/18 (ECtHR, 20 October 2022)��487n.124 MT v Netherlands App no 46595/19 (ECtHR, 23 March 2021)��369n.223 MYH and others v Sweden App no 50859/10 (ECtHR, 27 June 2013) �� 352n.86 Maaouia v France App no 39652/98 (ECtHR [GC], 5 October 2000) �� 183n.51 Maslov v Austria App no 1638/03 (ECtHR [GC], 23 June 2008) �� 478n.57, 478n.60 Mohammed Hussein and others v Netherlands and Italy App no 27725/10 (ECtHR, 2 April 2013)��369n.222 Moustaquim v Belgium App no 12313/86 (ECtHR, 18 February 1991)�� 264n.81

xxx Table of Cases Muskhadzhiyeva and others v Belgium App no 41442/07 (ECtHR, 19 January 2010) �� 512n.41 NA v Finland App no 25244/18 (ECtHR, 14 November 2019)��531n.183 ND and NT v Spain App nos 8675/15 and 8697/15 (ECtHR [GC], 13 February 2020)�� 130n.38, 140n.100, 307n.38, 310–11, 353n.97, 586n.97 NH and others v France App nos 28820/13, 75547/13 and 13114/15 (ECtHR, 2 July 2020)�� 352n.90, 410n.521 NA v United Kingdom App no 25904/07 (ECtHR, 17 July 2008)�� 351–52 Nunez v Norway App no 55597/09 (ECtHR, 28 June 2011)��135n.71, 439n.71, 513n.49 Omwenyeke v Germany App no 44294/04 (ECtHR, 20 November 2008) ��419n.570 Othman (Abu Qatada) v United Kingdom App no 8139/09 (ECtHR, 17 January 2012)�� 351n.81, 353n.95 Paposhvili v Belgium App no 41738/10 (ECtHR [GC], 13 December 2016)�� 352–53 Ponomaryovi and others v Bulgaria App no 5335/05 (ECtHR, 21 June 2011)��264n.81, 264n.85, 265n.88 Pormes v Netherlands App no 25402/14 (ECtHR, 28 July 2020)�� 514n.51 S and Marper v United Kingdom App nos 30562 and 30566/04 (ECtHR [GC], 4 December 2008) �� 244n.121, 245n.126 SK v Russia App no 52722/15 (ECtHR, 14 February 2017)�� 352n.87 SM v Croatia App no 60561/14 (ECtHR [GC], 25 June 2020)�� 521n.98 Saadi v Italy App no 37201/06 (ECtHR [GC], 28 February 2008) �� 351n.80 Saadi v United Kingdom App no 13229/03 (ECtHR [GC], 29 January 2008) ��146n.139, 511 Salah Sheekh v Netherlands App no 1948/04 (ECtHR, 11 January 2007)��146n.139, 351n.83, 395n.417 Savickis and others v Latvia App no 49270/11 (ECtHR [GC], 9 June 2022)�� 265n.86 Savran v Denmark App no 57467/15 (ECtHR [GC], 7 December 2021) �� 352n.92, 478n.58, 526n.144 Shahzad v Hungary App no 12625/17 (ECtHR, 8 July 2021)�� 311n.72 Sisojeva v Latvia App no 60654/00 (ECtHR [GC], 15 January 2007)��126n.20, 478n.63, 513n.50 Slivenko and others v Latvia App no 48321/99 (ECtHR [GC], 9 October 2003)�� 438n.65 Sufi and Elmi v United Kingdom App nos 8319/07 and 11449/07 (ECtHR, 28 June 2011) �� 352n.86, 395n.416 TI v United Kingdom App no 43844/98 (ECtHR, 7 March 2000)��368n.215 Tanda-Muzinga v France App no 2260/10 (ECtHR, 10 July 2014)�� 439n.69, 487n.122 Tarakhel v Switzerland App no 29217/12 (ECtHR [GC], 4 November 2014) �� 352n.90, 369n.223, 369n.227 Tyrer v United Kingdom App no 5856/72 (ECtHR, 25 April 1978)�� 129n.32 Üner v Netherlands App no 46410/99 (ECtHR [GC], 18 October 2006�� 475n.38, 476n.45, 478n.56, 478n.60 Xhavara and others v Italy and Albania App no 39473/98 (ECtHR, 11 January 2001)�� 309n.60, 583n.78 Yeshtla v Netherlands App no 37115/11 (ECtHR, 15 January 2019)�� 265n.89 INTERNATIONAL COURT OF JUSTICE Application of the International Convention on the Elimination of All Forms of Racial Discrimination (Qatar v United Arab Emirates) (ICJ Judgment No 172, 4 February 2021)�� 266n.97 Nottebohm (Liechtenstein v Guatemala) (Judgment, Second Phase) [1955] ICJ Rep 4, 23��501n.227

Table of EU Legislation PROTOCOLS Protocol [No 2] on the application of the principles of subsidiarity and proportionality [2004] OJ C310/207��157n.35 Protocol [No 3] on the Statute of the Court of Justice [2008] OJ C115/210�� 74n.20 Protocol [No 6] on the location of the seats of the institutions [2008] OJ C115/265�� 72n.10 Protocol [No 7] on the privileges and immunities of the European Union [2008] OJ L115/266 ��219n.127 Protocol [No 19] integrating the Schengen acquis into the framework of the European Union [2008] OJ C115/290��28, 65–66, 315n.99, 557n.43 Protocol [No 21] on the position of the United Kingdom and Ireland in respect of the area of freedom, security, and justice [2008] OJ C115/295��65–67 Protocol [No 22] on the position of Denmark [2008] OJ C115/299�� 64–65 Protocol [No 23] on external relations of the Member States with regard to the crossing of external borders [2008] OJ C115/304��328, 578–79 Protocol [No 24] on asylum for nationals of Member States of the European Union [2008] OJ C115/305�� 354–55 Protocol [No 35] on Article 67 [2006] OJ C321E/317��28 REGULATIONS Regulation No 1 determining the languages to be used by the European Economic Community [1958] OJ Spec Ed Vol I/59��159n.43 Regulation (EEC) No 1612/68 on freedom of movement for workers within the Community [1968] OJ Spec Ed I-1968(II)/475�� 20n.37 Regulation (EC) No 1683/95 laying down a uniform format for visas [1995] OJ L164/1�� 283t, 292 Regulation (EC) 2317/95 determining the third countries whose nationals must be in possession of visas [1995] OJ L234/1 ��284n.16

Regulation (EC) No 574/1999 determining the third countries whose nationals must be in possession of visas [1999] OJ L72/2 ��284n.17 Regulation (EC) No 2725/2000 concerning the establishment of ‘Eurodac’ [2000] OJ L316/1 �� 226n.12, 229n.b Regulation (EC) No 539/2001 listing the third countries whose nationals must be in possession of visas [2001] OJ L81/1��284n.21, 287, 563 Regulation (EC) No 1049/2001 regarding public access to documents [2001] OJ L145/43 ��94 Regulation (EC) No 1030/2002 laying down a uniform format for residence permits [2002] OJ L157/1 �� 254n.2, 254n.3, 312t, 334n.223 Regulation (EC) No 343/2003 establishing the criteria and mechanisms for determining the Member State responsible for examining an asylum application [2003] OJ L50/1 �� 53–54, 65n.136, 164, 341–42, 360, 366, 368–69 Regulation (EC) No 453/2003 amending Regulation (EC) No 539/2001 listing the third countries whose nationals must be in possession of visas [2003] OJ L69/10��284n.23 Regulation (EC) No 261/2004 establishing common rules on compensation and assistance to passengers in the event of denied boarding and of cancellation or long delay of flights [2004] OJ L46/1 ��296n.110 Regulation (EC) No 377/2004 on the creation of an immigration liaison officers network [2004] OJ L64/1��178n.27 Regulation (EC) No 883/2004 on the coordination of social security system [2004] OJ L166/1 �� 18–19, 437, 459n.212, 479t, 482–83, 485–86, 503–4, 555–56, 560–61, 562–63 Regulation (EC) No 2007/2004 establishing a European Agency for the Management of Operational Cooperation at the External Borders [2004] OJ L349/1 ��200–2, 203–4

xxxii Table of EU Legislation Regulation (EC) No 2252/2004 on standards for security features and biometrics in passports and travel documents issued by Member States [2004] OJ L385/1 ��66, 312t, 334–35 Regulation (EC) No 562/2006 establishing a Community Code on the rules governing the movement of persons across borders (Schengen Borders Code) [2006] OJ L105/1 �� 58n.90 Regulation (EC) No 765/2006 concerning restrictive measures in respect of Belarus [2006] OJ L134/1 ��586n.93 Regulation (EC) No 1907/2006 concerning the registration, evaluation, authorisation and restriction of chemicals (REACH) [2006] OJ L396/1�� 73n.12 Regulation (EC) No 1931/2006 laying down rules on local border traffic at the external land borders [2006] OJ L405/1�� 312t, 328 Regulation (EC) No 1932/2006 amending Regulation (EC) No 539/2001 listing the third countries whose nationals must be in possession of visas [2007] OJ L405/23 ��284n.23 Regulation (EC) No 1987/2006 on the establishment, operation and use of the second generation SIS [2006] OJ L381/4�� 188n.86, 231, 239–41 Regulation (EC) No 863/2007 establishing a mechanism for the creation of Rapid Border Intervention Teams [2008] OJ L199/30 �� 201–2 Regulation (EC) No 380/2008 amending Regulation (EC) No 1030/2002 laying down a uniform format for residence permits [2008] OJ L114/88�� 334–35 Regulation (EC) No 767/2008 concerning the Visa Information System (VIS) [2008] OJ L218/60 �� 188–89, 225–27, 229t, 233–34, 238–39, 242n.107, 243n.118, 247n.140 Regulation (EC) No 390/2009 amending the Common Consular Instructions [2009] OJ L131/1 �� 285n.32, 288–89, 291n.74, 293n.91 Regulation (EC) No 444/2009 amending Regulation (EC) No 2252/2004 on standards for security features and biometrics in passports and travel documents [2009] OJ L142/1 �� 334–35 Regulation (EC) No 810/2009 establishing a Community Code on Visas (Visa Code) [2009] OJ L243/1 �� 7–8, 28, 58–59, 154, 161–62, 168, 183n.53, 188–89,

233n.50, 255–57, 258, 271, 277, 278, 280–83, 283t, 284–85, 287, 288–95, 300, 442–43, 475n.40, 576–77, 584–85 Regulation (EU) No 265/2010 as regards movement of persons with a long-stay visa [2010] OJ L85/1��317n.110 Regulation (EU) No 439/2010 establishing a European Asylum Support Office [2010] OJ L132/11 �� 202, 203–4, 212–13 Regulation (EU) No 1231/2010 extending Regulation (EC) No 883/2004 to nationals of third countries [2010] OJ L344/1��436n.52, 479t, 485–86 Regulation (EU) No 492/2011 on freedom of movement for workers within the Union [2011] OJ L141/1�� 441n.80, 482–83 Regulation (EU) No 1168/2011 amending the initial Frontex Regulation (EC) No 2007/2004 [2011] OJ L304/1�� 201–2 Regulation (EU) No 610/2013 amending Regulation (EC) No 562/2006 establishing a Community Code on the rules governing the movement of persons across borders [2013] OJ L182/1��277, 317n.108 Regulation (EU) No 1053/2013 establishing an evaluation and monitoring mechanism to verify the application of the Schengen acquis [2013] OJ L295/27 ��314 Regulation (EU) No 1289/2013 amending the former Visa List Regulation [2013] OJ L347/74 �� 286n.38, 287n.44, 584n.84 Regulation (EU) No 603/2013 on the establishment of ‘Eurodac’ for the comparison of fingerprints [2013] OJ L180/1��65, 226n.13, 229t, 232–33, 243n.118, 247n.140, 248n.150 Regulation (EU) No 604/2013 establishing the criteria and mechanisms for determining the Member State responsible for examining an application for international protection lodged in one of the Member States by a third-country national or a stateless person [2013] OJ L180/31 �� 7, 47–48, 65, 77, 125, 138, 152–53, 156, 157, 164, 168–69, 171, 183–84, 189–91, 214, 244–45, 276, 325, 342, 356, 357–72, 373–75, 376–77, 409, 424, 447, 522, 533, 534–35, 536, 540, 544–45, 561 Regulation (EU) No 515/2014 establishing, as part of the Internal Security Fund, the instrument for financial support

Table of EU Legislation xxxiii for external borders and visa [2014] OJ L150/143 ��194n.121 Regulation (EU) No 516/2014 establishing the Asylum, Migration and Integration Fund [2014] OJ L150/168��194n.121 Regulation (EU) No 656/2014 establishing rules for the surveillance of the external sea borders in the context of operational cooperation coordinated by the European Agency for the Management of Operational Cooperation at the External Borders [2014] OJ L189/93 �� 58, 305, 312t, 331–34, 336 Regulation (EU) 2016/399 on a Union Code on the rules governing the movement of persons across borders (Schengen Borders Code) [2016] OJ L77/1 ��8, 17, 28, 60, 73, 138, 152–53, 154, 158–59, 161–62, 178n.29, 183n.53, 184n.63, 200n.12, 235n.63, 236, 254n.4, 255–56, 258, 259–60, 261, 268, 271, 278, 281n.6, 284n.27, 288–90, 300, 301, 312t, 313–15, 317n.104, 319–32, 334n.219, 336, 363–64, 371n.240, 442n.95, 518, 524n.134, 558n.55, 578–79 Regulation (EU) 2016/679 on the protection of natural persons with regard to the processing of personal data and the free movement of such data [2016] OJ L119/1 �� 243–44 Regulation (EU) 2016/1624 on the European Border and Coast Guard [2016] OJ L251/1��201–2, 210n.71, 320n.127 Regulation (EU) 2016/1953 on the establishment of a European travel document for the return of illegally staying third-country nationals [2016] OJ L311/13 ��537n.228 Regulation (EU) 2017/371 amending Council Regulation (EC) No 539/2001 listing the third countries whose nationals must be in possession of visas [2017] OJ L61/1 �� 286n.38, 584n.84 Regulation (EU) 2017/458 as regards the reinforcement of checks against relevant databases at external borders [2017] OJ L74/1 ��324, 326n.175 Regulation (EU) 2017/1001 on the European Union trade mark [2017] OJ L154/1 �� 73n.12 Regulation (EU) 2017/1954 amending Council Regulation (EC) No 1030/2002 laying down a uniform format for residence permits [2017] OJ L286/9 ��335n.224

Regulation (EU) 2017/2225 as regards the use of the Entry/Exit System [2017] OJ L327/28 ��236n.71, 281, 324, 325–27, 558 Regulation (EU) 2017/2226 establishing an Entry/Exit System (EES) [2017] OJ L327/28 �� 225–27, 229t, 230, 236, 243–44, 247–48, 317n.109 Regulation (EU) 2018/1240 establishing a European Travel Information and Authorisation System (ETIAS) [2018] OJ L236/1 �� 58–59, 229t, 234–35, 237–39, 241–42, 243–44, 247 Regulation (EU) 2018/1725 on the protection of natural persons with regard to the processing of personal data by the Union institutions [2018] OJ L295/39 ��248n.149 Regulation (EU) 2018/1726 on the European Union Agency for the Operational Management of Large-Scale IT Systems in the Area of Freedom, Security and Justice (eu-LISA) [2018] OJ L295/99 ��228 Regulation (EU) 2018/1806 listing the third countries whose nationals must be in possession of visas [2018] OJ L303/39 ��189, 234–35, 283–88, 299–300, 506, 584n.84 Regulation (EU) 2018/1860 on the use of the Schengen Information System (SIS) for the return of illegally staying third-country nationals [2018] OJ L312/1�� 190, 229t, 230–31, 247n.140, 524–25 Regulation (EU) 2018/1861 on the establishment, operation and use of the Schengen Information System (SIS) in the field of border checks [2018] OJ L312/14 �� 188, 189–91, 229t, 230–31, 237n.73, 241n.100, 243n.118, 244n.123, 247, 524–25 Regulation (EU) 2018/1862 on the establishment, operation and use of the SIS in criminal matters [2018] OJ L312/56 ��229t, 230–31 Regulation (EU) 2019/592 as regards the withdrawal of the United Kingdom from the Union [2019] OJ L103I/1��288n.46 Regulation (EU) 2019/816 establishing a centralised system for the identification of Member States holding conviction information on third-country nationals (ECRIS-TCN) [2019] OJ L135/1 ��230

xxxiv Table of EU Legislation Regulation (EU) 2019/817 on establishing a framework for interoperability between EU information systems in the fields of borders and visa and amending [2019] OJ L135/27�� 229t, 241–42, 247n.140 Regulation (EU) 2019/818 on establishing a framework for interoperability between EU information systems in the fields of police and judicial cooperation, asylum and migration [2019] OJ L135/27 �� 229t, 230, 241–42, 247n.140 Regulation (EU) 2019/1155 amending Regulation (EC) No 810/2009 establishing a Community Code on Visas [2019] OJ L188/25�� 58n.94, 189n.92, 281n.7, 584–85 Regulation (EU) 2019/1240 on the creation of a European network of immigration liaison officers [2019] OJ L198/88 �� 178n.27, 299n.127 Regulation (EU) 2019/1896 on the European Border and Coast Guard [2019] OJ L295/1 �� 92, 93, 154, 199, 201–2, 205–6, 206t, 207–11, 213, 215–21, 253, 298–99, 535–36, 557 Regulation (EU) 2021/695 establishing Horizon Europe [2021] OJ L170/1��193 Regulation (EU) 2021/947 establishing the Neighbourhood, Development and International Cooperation Instrument –Global Europe [2021] OJ L209/1�� 193–94, 585n.91, 591 Regulation (EU) 2021/1134 for the purpose of reforming the Visa Information System [2021] OJ L248/11 ��233–34, 237–39, 242n.107 Regulation (EU) 2021/1147 establishing the Asylum, Migration and Integration Fund [2021] OJ L251/1�� 194–95 Regulation (EU) 2021/1148 establishing the Instrument for Financial Support for Border Management and Visa Policy [2021] OJ L251/48�� 194–95 Regulation (EU) 2021/1151 as regards the establishment of the conditions for accessing other EU information systems for the purposes of the ETIAS [2021] OJ L249/7 �� 234–35 Regulation (EU) 2021/1152 as regards the establishment of the conditions for accessing other EU information systems for the purposes of the ETIAS [2021] OJ L249/15 �� 234–35, 242n.107

Regulation (EU) 2021/1529 establishing the Instrument for Pre-Accession assistance (IPA III) [2021] OJ L330/1�� 193–94 Regulation (EU) 2021/1985 amending Regulation (EC) No 765/2006 concerning restrictive measures in respect of Belarus [2021] OJ L405/1��586n.93 Regulation (EU) 2021/2303 on the European Union Agency for Asylum [2021] OJ L468/1 ��202, 205–6, 206t, 211–14, 216–21 Regulation (EU) 2022/562 as regards Cohesion’s Action for Refugees in Europe (CARE) [2022] OJ L109/1��193n.111 DIRECTIVES Directive 64/221/EEC on the co-ordination of special measures concerning the movement and residence of foreign nationals [1963/64] OJ Special Ed I/117�� 261n.56 Directive 77/486/EEC on the education of the children of migrant workers [1977] OJ L199/32 ��473n.22 Directive 96/71/EC concerning the posting of workers in the framework of the provision of services [1997] OJ L18/1�� 459–60 Directive 2000/43/EC implementing the principle of equal treatment between persons irrespective of racial or ethnic origin [2000] OJ L180/22 ��266–67, 479t, 493–94 Directive 2001/40/EC on the mutual recognition of expulsion decisions [2001] OJ L149/34 �� 534–35 Directive 2001/51/EC supplementing the provisions of Article 26 of the Convention Implementing the Schengen Agreement [2001] OJ L187/45 �� 283t, 295–97, 522 Directive 2001/55/EC on minimum standards for giving temporary protection in the event of a mass influx of displaced persons [2001] OJ L212/12 �� 346–47, 349–50, 357t, 359–60, 405–9, 425, 447n.128, 479n.64, 483n.91, 566–67 Directive 2002/90/EC defining the facilitation of unauthorised entry, transit and residence [2002] OJ L328/17 ��515–17, 515t

Table of EU Legislation xxxv Directive 2003/9/EC laying down minimum standards for the reception of asylum seekers [2003] OJ L31/18 ��341–42, 357t, 411–12, 542–43 Directive 2003/86/EC on the right to family reunification [2003] OJ L251/12 �� 27–28, 39, 48, 64–65, 101, 124n.8, 153–54, 159–61, 165, 167, 253, 257–58, 268, 271–72, 427, 430, 440–48, 440t, 452–53, 467, 474, 475n.39, 479n.64, 487–92, 495n.178, 496 Directive 2003/109/EC concerning the status of third-country nationals who are long-term residents [2004] OJ L16/44�� 27–28, 39, 126, 152n.8, 154, 158–59, 191, 253, 257, 262, 272, 418, 452–53, 457, 474, 477n.54, 478–79, 479t, 482–84, 489n.137, 494–98, 504, 527n.152, 552 Directive 2003/110/EC on transit for the purposes of removal by air [2010] OJ L321/26 �� 255n.12, 258, 259–60, 261, 535–36 Directive 2004/38/EC on the right of citizens of the Union and their family members to move and reside freely within the territory of the Member States [2004] OJ L158/77��38, 39, 161n.57, 287n.42, 288, 440–43, 445, 554, 555–56, 559–60 Directive 2004/81/EC on the residence permit issued to third-country nationals who are victims of trafficking in human beings or who have been the subject of an action to facilitate illegal immigration, who cooperate with the competent authorities [2004] OJ L261/19��367, 515t, 520–22 Directive 2004/82/EC on the obligation of carriers to communicate passenger data [2004] OJ L261/24 �� 230n.27, 324n.160 Directive 2004/83/EC on minimum standards for the qualification and status of third country nationals or stateless persons as refugees or as persons who otherwise need international protection and the content of the protection granted [2004] OJ L304/12 ��341–42, 357n.c, 395n.416, 400–1 Directive 2004/114/EC on the conditions of admission of third-country nationals for the purposes of studies, pupil exchange, unremunerated training or voluntary service [2004] OJ L375/12�� 261n.60, 431, 453

Directive 2005/36/EC on the recognition of professional qualifications [2005] OJ L255/22 ��485, 555–56 Directive 2005/71/EC on a specific procedure for admitting third-country nationals for the purposes of scientific research [2005] OJ L289/15��431, 439–40, 440n.a, 453, 455n.184 Directive 2005/85/EC on minimum standards on procedures in Member States for granting and withdrawing refugee status [2005] OJ L326/13 �� 58, 66n.151, 341–42, 357n.c, 372, 376, 377n.289, 381–49 Directive 2008/115/EC on common standards and procedures in Member States for returning illegally staying third-country nationals [2008] OJ L348/98 ��73, 141, 152n.8, 184–85, 254n.1, 255–56, 270n.121, 271n.129, 319n.117, 329–30, 353n.94, 381, 506n.2, 512n.42, 513, 515t, 518–20, 521n.104, 524–33, 536–37, 538, 539–42, 543–47, 548 Directive 2009/16/EC on port State control [2009] OJ L131/62��517n.72 Directive 2009/50/EC on the conditions of entry and residence of third-country nationals for the purposes of highly qualified employment [2009] OJ L155/17 ��432–33, 450 Directive 2009/52/EC providing for minimum standards on sanctions and measures against employers of illegally staying third-country nationals [2009] OJ L168/24��271n.129, 506n.3, 515t, 522–24, 548 Directive 2011/36/EU on preventing and combating trafficking in human beings and protecting its victims [2011] OJ L101/1 �� 520–22 Directive 2011/51/EU amending Council Directive 2003/109/EC to extend its scope to beneficiaries of international protection [2011] OJ L32/1�� 494–98 Directive 2011/95/EU on standards for the qualification of third-country nationals or stateless persons as beneficiaries of international protection, for a uniform status for refugees or for persons eligible for subsidiary protection, and for the content of the protection granted [2011] OJ L337/9 �� 65, 66, 117, 124n.7, 138, 144–46, 147, 153, 156, 260, 271n.124, 272, 342, 351–52, 386–404, 408, 410, 415–19, 425, 447, 479–80, 481, 483n.90, 486–87, 489n.138

xxxvi Table of EU Legislation Directive 2011/98/EU on a single application procedure for a single permit for third-country nationals to reside and work in the territory of a Member State and on a common set of rights for third-country workers legally residing in a Member State [2011] OJ L343/1 �� 179–80, 256n.24, 272n.132, 432–33, 440t, 446, 454n.175, 454n.180, 455–57, 464–65, 467, 479–82, 483n.95, 484–86, 484n.98, 565, 566 Directive 2012/29/EU establishing minimum standards on the rights, support and protection of victims of crime [2012] OJ L315/57��484n.97 Directive 2013/32/EU on common procedures for granting and withdrawing international protection [2013] OJ L180/60 �� 65, 66, 73, 156, 157n.32, 178n.24, 179–80, 182n.44, 183–85, 256n.24, 276, 330, 342, 354, 357t, 364–65, 369–70, 372–85, 386, 389, 408–9, 412, 424–25, 527, 531, 536 Directive 2013/33/EU laying down standards for the reception of applicants for international protection [2013] OJ L180/96 �� 57–58, 167–68, 271n.124, 342, 352, 374–76, 408–9, 410–15, 425–26, 512, 527, 539, 545 Directive 2014/36/EU on the conditions of entry and stay of third-country nationals for the purpose of employment as seasonal workers [2014] OJ L94/375 �� 256n.24, 257n.26, 271n.131, 418n.566, 433, 440t, 463–64, 467–68, 475n.40, 479n.64, 482n.82, 486, 522, 593n.139 Directive 2014/66/EU on the conditions of entry and residence of third-country nationals in the framework of an intra-corporate transfer [2014] OJ L157/1�� 126, 255n.17, 257n.26, 271n.131, 418n.566, 432–33, 438n.64, 440t, 450n.142, 457–58, 461–63, 467–68, 484n.98, 489n.139, 566 Directive (EU) 2016/801 on the conditions of entry and residence of third-country nationals for the purposes of research, studies, training, voluntary service, pupil exchange schemes or educational projects and au pairing [2016] OJ L132/21��154, 178n.23, 253, 256n.24, 257–58, 272n.132, 277, 409n.517, 418n.565, 432n.33, 440t, 443, 450n.142, 453–55, 457–58, 467, 479–80, 489n.139 Directive (EU) 2018/957 amending Directive 96/71/EC concerning the

posting of workers in the framework of the provision of services [2018] OJ L173/16 ��460n.214 Directive (EU) 2021/1883 on the conditions of entry and residence of third-country nationals for the purpose of highly qualified employment [2021] OJ L382/1�� 124n.8, 152n.8, 256n.24, 271n.131, 272n.136, 409n.517, 418n.565, 433n.39, 437n.54, 450–53, 457–58, 467, 475n.39, 479n.64, 484n.98, 489n.139, 495n.178, 497, 593 DECISIONS Decision 85/381/EC setting up a prior communication and consultation procedure on migration policies [1985] OJ L217/25 �� 25n.74 Decision 94/262/ECSC, EC, Euratom governing the performance of the Ombudsman’s duties [1994] OJ L113/15 ��91n.100 Decision 96/198/JHA on an alert and emergency procedure for burden- sharing [1996] OJ L63/10 ��359n.139 Decision 97/340/JHA on the exchange of information concerning assistance for the voluntary repatriation [1997] OJ L147/3 ��531n.180 Decision SCH/Com-ex (98)10 of the Executive Committee on cooperation in returning foreign nationals by air [2000] OJ L239/193��535n.213 Decision SCH/Com-ex (98) 37 on the adoption of measures to fight illegal immigration [2000] OJ L293/203 ��298n.126 Decision SCH/Com-ex (99)13 on the definitive versions of the Common Manual and the Common Consular Instructions [2000] OJ L239/317�� 284n.20 Decision 1999/352/EC, ECSC, Euratom establishing the European Anti-Fraud Office [1999] OJ L136/20��92n.107 Decision 1999/437/EC on certain arrangements for the application of the Agreement concluded by the Council of the European Union and the Republic of Iceland and the Kingdom of Norway concerning the association of those two States with the Schengen acquis [1999] OJ L176/31��229n.d, 515n.b Decision 2000/365 concerning the request of the United Kingdom to take part in some of the provisions of the Schengen acquis [2000] OJ L131/43 �� 66n.145

Table of EU Legislation xxxvii Decision 2000/596/EC establishing a European Refugee Fund [2000] OJ L252/12 ��194n.119 Decision 2002/192 concerning Ireland’s request to take part in some of the provisions of the Schengen acquis [2002] OJ L64/20 �� 66n.145, 229n.f Decision 2002/463/EC adopting an ARGO programme [2002] OJ L161/11 ��194n.119 Framework Decision 2002/946/JHA on the strengthening of the penal framework to prevent the facilitation of unauthorised entry, transit and residence [2002] OJ L328/1��515–17, 515t Decision 2003/690/EC on the request by Ireland to accept Council Directive 2001/55/EC on minimum standards for giving temporary protection [2003] OJ L251/23 ��357t, 405–9 Decision 2004/191/EC setting out the criteria and practical arrangements for the compensation of the financial imbalances resulting from the application of Directive 2001/40/EC [2004] OJ L60/55 ��535n.212 Decision 2004/512/EC establishing the Visa Information System (VIS) [2004] OJ L213/5 ��233n.47 Decision 2004/573/EC on the organisation of joint flights for removals from the territory of two or more Member States [2004] OJ L261/28 �� 178n.29, 536n.218 Decision 2004/904/EG establishing the European Refugee Fund for the period 2005 to 2010 [2004] OJ L381/52 �� 194n.119, 522–24 Decision 2004/927/EC providing for certain areas covered by Title IV of Part Three of the EX Treaty to be governed by the procedure laid down in Article 251 of that Treaty [2004] OJ L396/45�� 28n.92 Decision 2006/188/EC on the conclusion of the Agreement extending to Denmark the provisions of the former Dublin II Regulation (EC) No 343/2003 [2006] OJ L66/37��65n.136 Decision 2006/688/EC on the establishment of a mutual information mechanism concerning Member States’ measures [2006] OJ L283/40 ��547n.322 Decision 2007/435/EC establishing the European Fund for the Integration of Third-Country Nationals [2007] OJ L168/18 ��194n.120

Decision 2007/573/EC establishing the European Refugee Fund [2007] OJ L144/1 ��194n.120 Decision 2007/574/EC establishing the External Borders Fund [2007] OJ L144/22 ��194n.120 Decision 2007/575/EC establishing the European Return Fund [2007] OJ L144/45 ��194n.120 Decision 2008/381/EC establishing a European Migration Network [2008] OJ L131/7��178n.25 Decision 2008/615/JHA on the stepping up of cross-border cooperation [2008] OJ L210/1 ��535n.216 Decision 2008/633/JHA concerning access for consultation of the VIS [2008] OJ L218/129 ��226n.15 Decision 2009/350/EC on the request by Ireland to accept Council Decision 2008/381/EC [2009] OJ L108/5 ��66n.145 Decision 2009/937/EU adopting the Council’s Rules of Procedure [2009] OJ L325/35 �� 53n.60 Decision 2010/365/EU on the application of the provisions relating to the SIS in the Republic of Bulgaria and Romania [2010] OJ L166/17 �� 229n.c Decision 2012/776/EU on the position to be taken within the Association Council with regard to the adoption of provisions on the coordination of social security systems [2012] OJ L340/19��563n.89 Decision 2013/233/CFSP on the European Union Integrated Border Management Assistance Mission in Libya (EUBAM Libya) [2013] OJ L138/15 ��587n.103 Decision 2014/668/EU on the signing and provisional application of the Association Agreement, as regards Title III [2014] OJ L278/1 �� 552n.8 Decision (CFSP) 2015/778 on a European Union military operation in the Southern Central Mediterranean (EUNAVFOR MED) [2015] OJ L122/31 ��588n.105 Decision (EU) 2015/1523 establishing provisional measures in the area of international protection for the benefit of Italy and Greece [2015] OJ L239/146 ��64–65, 82, 274, 344, 350–51, 357t, 360–61 Decision (EU) 2015/1601 establishing provisional measures in the area of international protection for the

xxxviii Table of EU Legislation benefit of Italy and Greece [2015] OJ L248/80 �� 54n.73, 64–65, 82, 274, 344, 350–51, 357t, 360–61 Decision (CFSP) 2016/993 amending Decision (CFSP) 2015/778 on a European Union military operation in the Southern Central Mediterranean [2016] OJ L162/18 ��588n.105 Decision (EU) 2017/733 on the application of the provisions of the Schengen acquis relating to the SIS in the Republic of Croatia [2018] OJ L165/37��229n.c Decision (EU) 2017/1908 on the putting into effect of certain provisions relating to the VIS in the Republic of Bulgaria and Romania [2017] OJ L269/39 ��229n.a Decision (EU) 2018/934 on the putting into effect of the remaining provisions of the SIS in the Republic of Bulgaria and Romania [2018] OJ L165/37��229n.e Decision (EU) 2018/1993 on the EU Integrated Political Crisis Response Arrangements [2018] OJ L320/28�� 576n.34 Decision (EU) 2019/267 concerning the conclusion of the Status Agreement between the EU and Albania [2019] OJ L46/3��210n.70 Decision (CFSP) 2020/472 on a European Union military operation in the Mediterranean (EUNAVFOR MED IRINI) [2020] OJ L101/4��588 Decision (EU) 2020/1745 on the putting into effect of the provisions of the Schengen acquis in Ireland [2020] OJ L393/3��66n.145 Decision (EU) 2021/1781 on the suspension of certain provisions of Regulation (EC) No 810/2009 with respect to The Gambia [2021] OJ L360/124 ��585n.89 Decision (EU) 2022/366 on the partial suspension of the application of the Agreement between the European Union and the Republic of Vanuatu on the short-stay visa waiver [2022] OJ 2022 L69/105��285n.31 Decision (EU) 2022/382 establishing the existence of a mass influx of displaced persons from Ukraine and having the effect of introducing temporary protection [2022] OJ L71/1�� 357t, 406, 408n.512 RECOMMENDATIONS Council Recommendation on harmonising means of combating

illegal immigration and illegal employment [1996] OJ C5/1��514n.55 Council Recommendation concerning a specimen bilateral readmission agreement [1996] OJ C274/20 �� 537–38 Council Recommendation on the guiding principles to be followed in drawing up protocols [1996] OJ C274/25 ��537n.229 Council Recommendation on combating the illegal employment [1996] OJ C304/1 ��522n.114 Council Resolution on measures to be adopted on the combating of marriages of convenience [1997] OJ C382/1 ��430n.14 Commission Recommendation (EU) 2015/914 on a European Resettlement Scheme [2015] OJ L148/32��421n.582 Commission Recommendation for a voluntary humanitarian admission scheme with Turkey, C(2015) 9490 final��421n.585 Commission Recommendation (EU) 2017/1803 on enhancing legal pathways [2017] OJ L259/21��421n.583 Commission Recommendation (EU) 2017/1804 on the implementation of the provisions of the Schengen Borders Code on temporary reintroduction of border control [2017] OJ L259/125 ��320n.131 Commission Recommendation (EU) 2017/432 on making returns more effective [2017] OJ L66/15��524n.135, 525n.137, 528n.168, 541n.260 Commission Recommendation (EU) 2017/ 820 on proportionate police checks and police cooperation in the Schengen area [2017] OJ C122/79��322n.145 Commission Recommendation (EU) 2017/2338 establishing a common ‘Return Handbook’ [2017] OJ L339/83 �� 60n.98, 528n.160 Commission Recommendation establishing a common ‘Practical Handbook for Border Guards’, C(2019) 7131 final �� 60n.98, 528n.160 Commission Recommendation (EU) 2020/1364 on legal pathways to protection in the EU [2020] OJ C317/13��421n.583, 423n.599 Commission Recommendation (EU) 2020/1365 concerning operations carried out by vessels owned or operated by private entities for the purpose of search and rescue activities [2020] OJ L317/23 ��518n.75

Table of EU Legislation xxxix Commission Recommendation (EU) 2020/ 1366 on a Migration Preparedness and Crisis Blueprint [2020] OJ C317/26�� 345n.46 Council Recommendation (EU) 2020/912 on the temporary restriction on non-essential travel into the EU [2020] OJ C208I/1 ��60n.103 Commission Recommendation on immediate steps in the context of the Russian invasion of Ukraine, C(2022) 2028 final ��502n.232 INTERNATIONAL AGREEMENTS Agreement between the European Community and its Member States, of the one part, and the Swiss Confederation, of the other, on the Free Movement of Persons (adopted 21 June 1999, entered into force 1 June 2002) [2002] OJ L114/6 �� 555–56 Agreement between the European Community and the Islamic Republic of Pakistan on the readmission of persons residing without authorisation (adopted 26 October 2010, entered into force 1 December 2010) [2010] OJ L287/52��537n.227 Agreement between the European Community and the Republic of Serbia on the facilitation of the issuance of visas (adopted 18 September 2007, entered into force 1 January 2008) [2007] OJ L334/137�� 286n.36 Agreement between the European Union and Georgia on the facilitation of the issuance of visas (adopted 22 November 2010, entered into force 1 March 2011) [2011] OJ L52/34�� 235n.66 Agreement between the European Union and the Republic of Armenia on the facilitation of the issuance of visas (adopted 17 December 2014, entered into force 1 January 2014) [2013] OJ L289/2��286n.36 Agreement concerning the criteria and mechanisms for establishing the State responsible for examining a request for asylum lodged in a Member State or in Iceland or Norway (adopted 19 January 2001, entered into force 1 April 2001) [2001] OJ L93/40�� 557–58 Agreement concerning the criteria and mechanisms for establishing the State responsible for examining a request for asylum lodged in a Member State or in Switzerland (adopted of

26 October 2004, entered into force 1 March 2008) [2008] OJ L53/5�� 557–58 Agreement concluded by the Council of the European Union and the Republic of Iceland and the Kingdom of Norway concerning the latters’ association with the Schengen acquis (adopted 18 May 1999, entered into force 26 June 2000) [1999] OJ L176/36��557–58 Agreement establishing an Association between the European Economic Community and Turkey (adopted 12 September 1963, entered into force 1 December 1964) [1977] OJ L361/29 �� 553–54, 561–64 ——Decision No 1/80 of the Association Council on the development of the association [not published in the OJ]��561–64 ——Decision No 3/80 of the Association Council on the application of the social security schemes to Turkish workers and members of their families [1983] OJ C110/60�� 561–64 Agreement on the European Economic Area (adopted 2 May 1992, entered into force 1 January 1994) [1994] OJ L1/3 �� 173–74, 325–26, 551–52, 555–56, 558, 559–60, 568–69 Agreement on the Gradual Abolition of Checks at their Common Borders (adopted and entered into force 14 June 1985) [2000] OJ L239/13�� 13, 24–26 Agreement on the Swiss Confederation’s association with the implementation, application and development of the Schengen acquis (adopted 26 October 2004, entered into force 1 March 2008) [2008] OJ L53/52�� 557–58 Agreement on the Withdrawal of the UK from the EU (adopted of 24 January 2020, entered into force 1 February 2020) [2020] OJ L29/7�� 9, 67, 495–96, 552–53, 554, 559–60, 569 Association Agreement between the European Union and its Member States, of the one part, and Ukraine, of the other part (adopted 27 June 2014, entered into force 1 September 2017) [2014] OJ L161/3 �� 566–67 Convention Determining the State Responsible for Examining Applications for Asylum Lodged in one of the Member States of the European Communities (adopted 15 June 1990, entered into force 1 September 1997) [1997] OJ C254/1 �� 13, 27, 232, 341, 357–60, 363n.169, 365–70, 582

xl Table of EU Legislation Convention for the Protection of Individuals with Regard to Automatic Processing of Personal Data (adopted 20 January 1981, entered into force 1 October 1985) ETS No 108��243n.117 Convention Implementing the Schengen Agreement of 14 June 1985 (adopted 19 June 1990, entered into force 1 September 1993) [2000] OJ L239/19 �� 24–26, 34–35, 64, 190, 230n.28, 284n.19, 288–89, 295–97, 298n.125, 311–19, 313n.76, 325n.169, 328n.185, 331n.202, 341, 358, 515–17, 518n.77, 524n.131, 527n.152, 536n.219 Convention on the Reduction of Cases of Multiple Nationality and Military Obligations in Cases of Multiple Nationality (adopted 6 May 2002, entered into force 28 March 1968) ETS No 43��126n.18 Convention on the Reduction of Statelessness (adopted 30 August 1961, entered into force 13 December 1975) 989 UNTS 175�� 126–27 Convention relating to the Status of Refugees (adopted 28 July 1951, entered into force 22 April 1954) 189 UNTS 137, as amended by the Protocol relating to the Status of Refugees (adopted 31 January 1967, entered into force 4 October 1967) 606 UNTS 267�� 6, 14, 17–18, 72, 102, 108, 112, 122–24, 137, 143–47, 148, 149, 159, 174, 296, 307, 338, 339–40, 341, 349–56, 372, 383–84, 390–91, 394–95, 398–401, 404, 405, 415–19, 420, 423, 425–26, 460, 481–82, 498, 518, 553, 570, 578, 591–93 Council of Europe Agreement on Transfer of Responsibility for Refugees (adopted 16 October 1980, entered into force 1 December 1982) ETS No 107�� 370–71, 417–18 Council of Europe Convention on Action against Trafficking in Human Beings (adopted 16 May 2005, entered into force 1 February 2008) ETS No 197��521n.97 Euro-Mediterranean Agreement establishing an association between the European Communities and their Member States, of the one part, and the Republic of Tunisia, of the other part (adopted 17 July 1995, entered into force 1 March 1998) [1998] OJ L97/2�� 567–68 Euro-Mediterranean Agreement establishing an Association between the European Communities and their

Member States, of the one part, and the Arab Republic of Egypt, of the other part (adopted 25 June 2001, entered into force 1 June 2004) [2004] OJ L304/39 �� 567–68 Europe Agreement establishing an association between the European Communities and their Member States, of the one part, and the Republic of Poland, of the other part (adopted 16 December 1991, entered into force 1 February 1994) [1993] OJ L348/2�� 564–65 European Convention for the Prevention of Torture (CPT) (adopted 26 November 1987, entered into force 1 February 1989) ETS No 126�� 139–40 European Convention on Establishment (adopted 13 December 1955, entered into force 23 February 1965) ETS No 19�� 19, 125 European Convention on Nationality (adopted 6 November 1997, entered into force 1 March 2000) ETS No 166�� 501 European Convention on Social and Medical Assistance (adopted 11 December 1953, entered into force 1 July 1954) ETS No 14 ��472 European Convention on the Legal Status of Migrant Workers (adopted 24 November 1977, entered into force 1 May 1983) ETS No 93��472 European Social Charter (adopted 18 October 1961, entered into force 26 February 1965) ETS No 35; and Revised European Social Charter (adopted 3 May 1996, entered into force 1 July 1999) ETS No 163��139, 392, 438, 513 General Agreement on Trade in Services (adopted 15 April 1994, entered into force 1 January 1995) [1994] OJ L336/191 �� 460–62, 467–68 [ILO] Migrant Workers (Supplementary Provisions) Convention [No 143] (adopted 24 June 1975, entered into force 9 December 1978) 1120 UNTS 323��522 [ILO] Migration for Employment Convention [No 97] (adopted 1 July 1949, entered into force 22 January 1952) 120 UNTS 1616�� 482–83 [ILO] Migration for Employment Convention [No 66] (adopted 28 June 1939, never entered into force)�� 19n.29 International Convention on Maritime Search and Rescue (SAR Convention) (adopted 27 April 1979, entered into force 22 June 1985) 1405 UNTS 118��308–9

Table of EU Legislation xli International Convention on the Protection of the Rights of All Migrant Workers and Members of their Families (adopted 18 December 1990, entered into force 1 July 2003) 2220 UNTS 3�� 513n.47 International Covenant on Civil and Political Rights (adopted 16 December 1966, entered into force 23 March 1976) 999 UNTS 171�� 139–40, 309n.58 International Covenant on Economic, Social and Cultural Rights (adopted 16 December 1966, entered into force 3 January 1976) 993 UNTS 3��307n.44 Partnership Agreement between the members of the African, Caribbean and Pacific Group of States and the European Community and its Member States (adopted 23 June 2000, entered into force 1 April 2003) [2000] OJ L317/6 �� 537–38, 551–52 Partnership and Cooperation Agreement between the European Communities and their Member States, of the one part, and the Russian Federation, of the other part (adopted 24 June 1996, entered into force 1 December 1997) [1997] OJ L327/3 ��566n.111 Partnership and Cooperation Agreement between the European Communities and their Member States, of the one part, and the Republic of Azerbaijan, of the other part (adopted 22 April 1996, entered into force 1 July 1999) [1999] OJ L246/3 �� 566–67 Protocol against the Smuggling of Migrants by Land, Sea, and Air (adopted 15 November 2000, entered into force 28 January 2004) [2006] OJ L262/34�� 332–33, 507–8, 515–17, 518n.77, 548 Protocol on Social Security Coordination [2021] OJ L149/2292, which is attached to the Trade and Cooperation Agreement (adopted 30 December 2020, provisionally applied 1 January 2021, entered into force 1 May 2021) [2021] OJ L149/10�� 552–54, 560–61 Protocol to Prevent, Suppress and Punish Trafficking in Persons, especially Women and Children (adopted 15 November 2000, entered into force 25 December 2003) [2006] OJ L262/51�� 520n.93 Prüm Convention on the stepping up of cross-border cooperation, particularly in combating terrorism, cross-border

crime and illegal migration (adopted 27 May 2005, entered into force 23 November 2006) [2006] Federal Law Gazette (Bundesgesetzblatt) II/626�� 535 Stabilisation and Association Agreement between the European Communities and their Member States of the one part, and the Republic of Serbia, of the other part (adopted 29 April 2008, entered into force 1 September 2013) [2013] OJ L278/16 ��565n.102 Stabilisation and Association Agreement between the European Union and the European Atomic Energy Community, of the one part, and Kosovo, of the other part (adopted 27 October 2015, entered into force 1 April 2016) [2016] OJ L71/3 �� 565–66 Status Agreement between the European Union and the Republic of Albania on actions carried out by the European Border and Coast Guard Agency (adopted 5 October 2018, entered into force 1 May 2019) [2019] OJ L46/3��209–10 Statute of the Council of Europe (adopted 5 May 1949, entered into force 3 August 1949) ETS No 1�� 124–25 Trade and Cooperation Agreement between the European Union and the European Atomic Energy Community, of the one part, and the United Kingdom of Great Britain and Northern Ireland, of the other part (adopted 30 December 2020, provisionally applied 1 January 2021, entered into force 1 May 2021) [2021] OJ L149/10 �� 551–53, 554, 560–61 United Nations Convention against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment (adopted 10 December 1984, entered into force 26 June 1987) 1465 UNTS 85��139–40 United Nations Convention on the Law of the Sea (adopted 10 December 1982, entered into force 16 November 1994) [1998] OJ L179/3 �� 308–9, 517–18 United Nations Convention on the Rights of the Child (adopted 20 November 1989, entered into force 2 September 1990) 1577 UNTS 3 �� 135–36 Vienna Convention on the Law of Treaties (adopted 23 May 1969, entered into force 27 January 1980) 1155 UNTS 331�� 130n.38, 145n.132, 159–60, 314n.93

List of Abbreviations AG AMIF Regulation (EU) 2021/1147 Anti-Trafficking Directive 2011/36/EU Asylum and Migration Interoperability Regulation (EU) 2019/ 818 Asylum Procedures Directive 2013/32/EU Blue Card Directive (EU) 2021/1883 BMVI Regulation (EU) 2021/1148 Borders and Visa Interoperability Regulation (EU) 2019/ 817 Carrier Sanctions Directive 2001/51/EC CEAS CFR CISA CJEU CML Rev Convention Implementing the Schengen Agreement CRC CUP Dublin II Regulation (EC) No 343/2003 Dublin III Regulation (EU) No 604/2013 EASO ECA

Advocate General Regulation (EU) 2021/1147 establishing the Asylum, Migration and Integration Fund [2021] OJ L251/1, with later amendments Directive 2011/36/EU on preventing and combating trafficking in human beings and protecting its victims [2011] OJ L101/1 Regulation (EU) 2019/818 on establishing a framework for interoperability between EU information systems in the fields of police and judicial cooperation, asylum and migration [2019] OJ L135/27, with later amendments Directive 2013/32/EU on common procedures for granting and withdrawing international protection [2013] OJ L180/60 Directive (EU) 2021/1883 on the conditions of entry and residence of third-country nationals for the purpose of highly qualified employment [2021] OJ C382/1 Regulation (EU) 2021/1148 establishing the Instrument for Financial Support for Border Management and Visa Policy [2021] OJ L251/48 Regulation (EU) 2019/817 on establishing a framework for interoperability between EU information systems in the fields of borders and visa and amending [2019] OJ L135/27, with later amendments Directive 2001/51/EC supplementing the provisions of Article 26 of the Convention Implementing the Schengen Agreement [2001] OJ L187/45 Common European Asylum System Charter of Fundamental Rights of the European Union see Convention Implementing the Schengen Agreement Court of Justice of the European Union Common Market Law Review Convention Implementing the Schengen Agreement of 14 June 1985 (adopted 19 June 1990, entered into force 1 September 1993) [2000] OJ L239/19 Convention on the Rights of the Child Cambridge University Press Regulation (EC) No 343/2003 establishing the criteria and mechanisms for determining the Member State responsible for examining an asylum application lodged in one of the Member States [2003] OJ L50/1 Regulation (EU) No 604/2013 establishing the criteria and mechanisms for determining the Member State responsible for examining an application for international protection lodged in one of the Member States [2013] OJ L180/31 European Asylum Support Office European Court of Auditors

xliv List of Abbreviations ECHR

ECRE ECtHR ed/eds EDPS EEA EES Regulation (EU) 2017/ 2226 EES EJIL EJML ELJ ELRev EMN Employers Sanctions Directive 2009/52/EC EP EPRS ETIAS Regulation (EU) 2018/1240 ETIAS ETS EUAA Regulation (EU) 2021/2303 EUAA EuConst Eurodac Regulation (EU) No 603/2013 Eurostat ExCom Facilitators Directive 2002/ 90/EC Facilitators Framework Decision 2002/946/JHA Family Reunification Directive 2003/86/EC former Asylum Procedures Directive 2005/85/EC former Blue Card Directive 2009/50/EC former EASO Regulation (EU) No 439/2010

Convention for the Protection of Human Rights and Fundamental Freedoms (adopted 4 November 1950, entered into force 3 September 1953) ETS No 5, as amended by Protocol No 15 (adopted of 24 June 2013, entered into force 1 August 2021) ETS No 213 European Council on Refugees and Exiles European Court of Human Rights editor/editors European Data Protection Supervisor European Economic Area Regulation (EU) 2017/2226 establishing an Entry/Exit System (EES) [2017] OJ L327/28, with later amendments Entry/Exit System European Journal of International Law European Journal of Migration and Law European Law Journal European Law Review European Migration Network Directive 2009/52/EC providing for minimum standards on sanctions and measures against employers of illegally staying third-country nationals [2009] OJ L168/24 European Parliament European Parliamentary Research Service Regulation (EU) 2018/1240 establishing a European Travel Information and Authorisation System (ETIAS) [2018] OJ L236/ 1, with later amendments European Travel Information and Authorisation System European Treaty Series Regulation (EU) 2021/2303 on the European Union Agency for Asylum [2021] OJ L468/1 European Union Asylum Agency European Constitutional Law Review Regulation (EU) No 603/2013 on the establishment of ‘Eurodac’ for the comparison of fingerprints [2013] OJ L180/1, with later amendments Statistical Office of the European Union Executive Committee Directive 2002/90/EC defining the facilitation of unauthorised entry, transit and residence [2002] OJ L328/17 Framework Decision 2002/946/JHA on the strengthening of the penal framework to prevent the facilitation of unauthorised entry, transit and residence [2002] OJ L328/1 Directive 2003/86/EC on the right to family reunification [2003] OJ L251/12 Directive 2005/85/EC on minimum standards on procedures in Member States for granting and withdrawing refugee status [2005] OJ L326/13 Directive 2009/50/EC on the conditions of entry and residence of third-country nationals for the purposes of highly qualified employment [2009] OJ L155/17 Regulation (EU) No 439/2010 establishing a European Asylum Support Office [2010] OJ L132/11

list of Abbreviations xlv former Frontex Regulation (EC) No 2007/2004 former Frontex Regulation (EU) 2016/1624 former Qualification Directive 2004/83/EC former Reception Directive 2003/9/EC former Researchers Directive 2005/71/EC former Schengen Borders Code Regulation (EC) No 562/2006 former Students Directive 2004/114/EC former Visa List Regulation (EC) No 539/2001 FRA Free Movement Directive 2004/38/EC Frontex Regulation (EU) 2019/1896 GC Geneva Convention GLJ HRC ibid ICCPR ICESCR ICLQ ICON ICT ICT Directive 2014/66/EU IJRL ILO IM Rev IOM JCMS

Regulation (EC) No 2007/2004 establishing a European Agency for the Management of Operational Cooperation at the External Borders [2004] OJ L349/1 Regulation (EU) 2016/1624 on the European Border and Coast Guard [2016] OJ L251/1 Directive 2004/83/EC on minimum standards for the qualification and status of third country nationals or stateless persons as refugees or as persons who otherwise need international protection and the content of the protection granted [2004] OJ L304/12 Directive 2003/9/EC laying down minimum standards for the reception of asylum seekers [2003] OJ L31/18 Directive 2005/71/EC on a specific procedure for admitting third-country nationals for the purposes of scientific research [2005] OJ L289/15 Regulation (EC) No 562/2006 establishing a Community Code on the rules governing the movement of persons across borders (Schengen Borders Code) [2006] OJ L105/1 Directive 2004/114/EC on the conditions of admission of third- country nationals for the purposes of studies, pupil exchange, unremunerated training or voluntary service [2004] OJ L375/12 Regulation (EC) No 539/2001 listing the third countries whose nationals must be in possession of visas when crossing the external borders [2001] OJ L81/1 EU Agency for Fundamental Rights Directive 2004/38/EC on the right of citizens of the Union and their family members to move and reside freely within the territory of the Member States [2004] OJ L158/77, with later amendments Regulation (EU) 2019/1896 on the European Border and Coast Guard [2019] OJ L295/1, with later amendments Grand Chamber See Refugee Convention German Law Journal Human Rights Committee ibidem/in the same place/the same International Covenant on Civil and Political Rights (adopted 16 December 1966, entered into force 23 March 1976) 999 UNTS 171 International Covenant on Economic, Social and Cultural Rights (adopted 16 December 1966, entered into force 3 January 1976) 993 UNTS 3 International and Comparative Law Quarterly International Journal of Constitutional Law intra-corporate transfers Directive 2014/66/EU on the conditions of entry and residence of third-country nationals in the framework of an intra- corporate transfer [2014] OJ L157/1 International Journal of Refugee Law International Labour Organization International Migration Review International Organization for Migration Journal of Common Market Studies

xlvi List of Abbreviations JEMS Local Border Traffic Regulation (EC) No 1931/2006 Long-Term Residents Directive 2003/109/EC MN NDICI Regulation (EU) 2021/947 OJ OUP PACE para/s Passports and Travel Documents Regulation (EC) No 2252/2004 Posted Workers Directive 96/71/EC Qualification Directive 2011/95/EU

Racial Equality Directive 2000/43/EC Reception Directive 2013/ 33/EU Recognition of Professional Qualifications Directive 2005/36/EC Refugee Convention

Relocation Decisions (EU) 2015/1523 & (EU) 2015/ 1601 Residence Permit Format Regulation (EC) No 1030/2002 Return Directive 2008/ 115/EC

Journal of Ethnic and Migration Studies Regulation (EC) No 1931/2006 laying down rules on local border traffic at the external land borders of the Member States [2006] OJ L405/1, with later amendments Directive 2003/109/EC concerning the status of third-country nationals who are long-term residents [2004] OJ L16/44, as amended by Directive 2011/51/EU to extend its scope to beneficiaries of international protection [2011] OJ L132/1 margin number/note Regulation (EU) 2021/947 establishing the Neighbourhood, Development and International Cooperation Instrument [2021] OJ L209/1 Official Journal of the European Union Oxford University Press Parliamentary Assembly of the Council of Europe paragraph/s Regulation (EC) No 2252/2004 on standards for security features and biometrics in passports and travel documents issued by Member States [2004] OJ L385/1, with later amendments Directive 96/71/EC concerning the posting of workers in the framework of the provision of services [1997] OJ L18/1, as amended by Directive (EU) 2018/957 [2018] OJ L173/16 Directive 2011/95/EU on standards for the qualification of third-country nationals or stateless persons as beneficiaries of international protection, for a uniform status for refugees or for persons eligible for subsidiary protection, and for the content of the protection granted [2011] OJ L337/9 Directive 2000/43/EC implementing the principle of equal treatment between persons irrespective of racial or ethnic origin [2000] OJ L180/22 Directive 2013/33/EU laying down standards for the reception of applicants for international protection [2013] OJ L180/96 Directive 2005/36/EC on the recognition of professional qualifications [2005] OJ L255/22, as amended by Directive 2013/55/EU [2013] OJ L158/368 Convention relating to the Status of Refugees (adopted 28 July 1951, entered into force 22 April 1954) 189 UNTS 137, as amended by the Protocol relating to the Status of Refugees (adopted 31 January 1967, entered into force 4 October 1967) 606 UNTS 267 Decision (EU) 2015/1523 establishing provisional measures in the area of international protection for the benefit of Italy and Greece [2015] OJ L239/146; and Decision (EU) 2015/1601 for the benefit of Italy and Greece [2015] OJ L248/80, with later amendments Regulation (EC) No 1030/2002 laying down a uniform format for residence permits for third-country nationals [2002] OJ L157/1, with later amendments Directive 2008/115/EC on common standards and procedures in Member States for returning illegally staying third-country nationals [2008] OJ L348/98

list of Abbreviations xlvii Schengen Borders Code Regulation (EU) 2016/ 399 Schengen Implementing Convention Sea Borders Regulation (EU) No 656/2014 Seasonal Workers Directive 2014/36/EU Single Permit Directive 2011/98/EU

SIS Border Checks Regulation (EU) 2018/ 1861 SIS Criminal Matters Regulation (EU) 2018/ 1862 SIS Return Regulation (EU) 2018/1860 SIS Social Security Coordination Regulation (EC) No 883/2004 Students and Researchers Directive (EU) 2016/801

Temporary Protection Directive 2001/55/EC Temporary Protection Implementing Decision (EU) 2022/382 TEU TFEU Treaty of Amsterdam

Regulation (EU) 2016/399 on a Union Code on the rules governing the movement of persons across borders (Schengen Borders Code) [2016] OJ L77/1, with later amendments see Convention Implementing the Schengen Agreement Regulation (EU) No 656/2014 establishing rules for the surveillance of the external sea borders in the context of operational cooperation coordinated by the European Agency for the Management of Operational Cooperation [2014] OJ L189/93) Directive 2014/36/EU on the conditions of entry and stay of third-country nationals for the purpose of employment as seasonal workers [2014] OJ L94/375) Directive 2011/98/EU on a single application procedure for a single permit for third-country nationals to reside and work in the territory of a Member State and on a common set of rights for third-country workers legally residing in a Member State [2011] OJ L343/1 Regulation (EU) 2018/1861 on the establishment, operation and use of the Schengen Information System (SIS) in the field of border checks [2018] OJ L312/14, with later amendments Regulation (EU) 2018/1862 on the establishment, operation and use of the SIS in criminal matters [2018] OJ L312/56, with later amendments Regulation (EU) 2018/1860 on the use of the Schengen Information System (SIS) for the return of illegally staying third- country nationals [2018] OJ L312/1, with later amendments Schengen Information System Regulation (EC) No 883/2004 on the coordination of social security system) [2004] OJ L166/1, with later amendments, as extended to legally resident third country nationals by Regulation (EU) No 1231/2010 [2010] OJ L344/1 Directive (EU) 2016/801 on the conditions of entry and residence of third-country nationals for the purposes of research, studies, training, voluntary service, pupil exchange schemes or educational projects and au pairing [2016] OJ L132/21, with later amendments Directive 2001/55/EC on minimum standards for giving temporary protection in the event of a mass influx of displaced persons [2001] OJ L212/12 Implementing Decision (EU) 2022/382 establishing the existence of a mass influx of displaced persons from Ukraine and having the effect of introducing temporary protection [2022] OJ L71/1 Treaty on European Union, as amended by the Treaty of Lisbon (adopted 13 December 2007, entered into force on 1 December 2009) [2008] OJ C115/13 Treaty on the Functioning of the European Union (adopted 13 December 2007, entered into force 1 December 2009) [2008] OJ C115/47 Treaty of Amsterdam (adopted 2 October 1997, entered into force 1 May 1999) [1997] OJ C340/173

xlviii List of Abbreviations Treaty of Maastricht UN UNGA UNHCR UNTS UP Victims of Trafficking Directive 2004/81/EC Vienna Convention on the Law of Treaties VIS Regulation (EC) No 767/2008 VIS Visa Code Regulation (EC) No 810/2009 Visa Format Regulation (EC) No 1683/95 Visa List Regulation (EU) 2018/1806

Treaty of Maastricht (adopted 7 February 1992, entered into force 1 November 1993) [1992] OJ C224/36 United Nations UN General Assembly United Nations High Commissioner for Refugees United National Treaty System University Press Directive 2004/81/EC on the residence permit issued to third- country nationals who are victims of trafficking in human beings who cooperate with the competent authorities [2004] OJ L261/19 Vienna Convention on the Law of Treaties (adopted 23 May 1969, entered into force 27 January 1980) 1155 UNTS 331 Regulation (EC) No 767/2008 concerning the Visa Information System (VIS) [2008] OJ L218/60, with later amendments Visa Information System Regulation (EC) No 810/2009 establishing a Community Code on Visas (Visa Code) [2009] OJ L243/1, with later amendments Regulation (EC) No 1683/95 laying down a uniform format for visas [1995] OJ L164/1, with later amendments Regulation (EU) 2018/1806 listing the third countries whose nationals must be in possession of visas when crossing the external borders [2018] OJ L303/39, with later amendments

Introduction European Migration Law as a Field of Inquiry

Migration law is politically contested, practically relevant, and full of legal conundrums. This book has been written to help readers navigate the often Byzantine European rulebook at a time where it has become increasingly difficult to keep an oversight. More than three dozen directives and regulations will be discussed throughout this volume, together with numerous court judgments, international treaties, reform proposals, and factual developments on the ground. Our objective is to provide readers with a mental map which empowers them to find solutions to pressing political, practical, and legal challenges. Eighteen chapters will present core features of visas and border controls, asylum and legal migration, integration and return, association agreements, and international cooperation. This comprehensive assessment of ‘European migration law’ introduces beginners to the state of the art and serves as a reliable inventory for experts. The need for an overarching treatise is apparent: the sheer complexity of the legislation, feedback loops between rules on related themes, and discrepancies between the law in the books and law in practice render it a formidable challenge to understand the rulebook and its implications. Questions of legal interpretation coalesce with the policy design and factual events. We will combine these aspects. Careful inspection of EU legislation and Court judgments shall be accompanied by due attention to domestic and international developments, as well as contextual factors influencing the real world of migratory movements. Whereas newcomers to European migration law will consult the book as a gate of entry into a dynamic and fascinating area of the law, experts will hopefully benefit from extra information and new perspectives. Political contestation and normative salience mean that there is no neutral standpoint from which we may describe migration law. Critical observers highlight that the very idea of nationality embodies a neo-feudal privilege of people born in the Global North, as does the distinction between ‘Union citizens’ and ‘third country nationals’. That is why comments made throughout this book make an effort to present the theoretical positions and conceptual choices that lie beneath the policy design. To do so allows the reader to reflect critically on the legal material which forms the basis of our analysis. The contents of positive law serves as the starting point of this venture. Chapter titles have been kept deliberately short and descriptive in an attempt to facilitate the identification of where information on specific themes can be found. At the same time, European migration law is more than the sum of legislation, court judgments, and practical measures on specific subject matters. There are overarching themes that a purely sectoral approach fails to recognise. Judges at the Court of Justice of the European Union (CJEU), for instance, will address legislation on a specific subject

European Migration Law. Daniel Thym, Oxford University Press. © Daniel Thym 2023. DOI: 10.1093/oso/9780192894274.003.0001

2 Introduction matter as an integral part of the supranational legal order; judgments on asylum procedure benefit from a horizontal comparison with what the Court says on the return or visa procedure. Political negotiations in Brussels are informed by entrenched patterns of behaviour, of which readers without a background in EU studies will often be unaware. Conversely, those who are familiar with the European project may wish to grasp the specificities of international human rights and refugee law. Interdisciplinary perspectives are a novelty for many legal experts; they reveal the dynamics behind the design and application of the legal material. Our holistic presentation will bring these different points of view together, both in the horizontal chapters in the first part and in the sectoral comments which can be found in the second part. We submit that such a holistic outlook adds value in terms of using scarce resources effectively. The constant flood of new legislation, court judgments, reform proposals, and changing contextual factors may lead to observers missing the proverbial forest for the trees. Anyone working on EU migration law knows the risk of ‘data overload’ when you have to digest highly complex information from multiple sources. The author readily admits that his mind was spinning repeatedly when dealing with what seemed like another impossible question that he struggled to grasp. Stepping back and trying to see the bigger picture helps in such scenarios. Technical problems of great practical or political importance will be better understood when the reader is aware of the framework in which they are embedded. The response to data overload should be the search for replicable patterns and horizontal cross-fertilization. Doing so may ideally allow us to see a mosaic instead of a convoluted patchwork of unrelated events. I have written this book with several audiences in view: policy experts, legal practitioners, and academic researchers. Policy experts working in ministries, parliaments, non- governmental organisations (NGOs), or think tanks discuss how to design new measures or to reform existing legislation. They will appreciate our comments on the policy concept, the constitutional foundations, the administrative practices, and the drivers of migratory movements. Having said this, new policies are rarely designed from scratch. More than three decades of European migration law have left us with a copious reservoir of rules and instruments. A good knowledge of the law is a precondition for success in the policy debate. That is particularly true at the EU level, where new initiatives often build upon previous choices in a path-dependent manner. Legal practitioners usually have a good knowledge of the law. They will consult this book to identify additional information, such as Court judgments they might be unaware of. This will happen frequently, considering that the CJEU has delivered more than 250 judgments on the directives and regulations discussed in this volume. They will be mentioned in the relevant passages, together with the case law of the European Court of Human Rights (ECtHR) and the position of the academic literature or international actors, such as UNHCR. Moreover, practitioners have to deal with legal questions, irrespective of whether courts have answered them already. Awareness of the doctrinal foundations of the case law and horizontal parallels between several pieces of legislation helps to identify solutions to new problems. This book presents practitioners with the nuances of the legislation and the arguments they may use. Academic researchers will have different outlooks depending on their area of expertise and level of seniority. Young researchers writing their PhD theses may struggle to get an overview and to identify a theme where they can add value. Being aware of the state of

Broad Understanding of ‘European Migration Law’ 3 the art is a precondition for a successful PhD project, which may seek to merge interdisciplinary viewpoints with classic arguments of legal interpretation. We are confident that young researchers will welcome this book as a gateway into a galaxy full of research puzzles and unresolved legal problems. Along similar lines, senior academics will benefit from our comments whenever they deal with new topics or are looking for additional information on specific legal areas.

Broad Understanding of ‘European Migration Law’ Neither the choice of terminology nor the thematic scope of ‘European migration law’ are self-explanatory. In essence, this book concentrates on the legal status of third country nationals entering or staying in the Member States of the European Union. EU legislation will be the starting point of our analysis, which, at the same time, will consider international developments, policy debates, and national laws. Our choice of title signals this broader outlook, which transcends the ‘nitty gritty’ of EU legislation and CJEU judgments. At the same time, the terminology reflects the interconnectivity of the subject matters discussed in this volume. The traditional dichotomy between ‘asylum’ and ‘immigration’ increasingly gives way to an awareness of overlap. This contribution employs the term ‘migration/migrant’ as an overarching category for the legal rules and everyday practices on both voluntary and forced migration. Such broad usage of the term does not negate the well-established distinction between refugee law and asylum legislation, on the one hand, and the rules on legal migration and return, on the other hand. The chapters in the second part will address the specificities of the legal material. Having said this, interdisciplinary studies emphasise that the binary juxtaposition of ‘voluntary’ and ‘forced’ fails to address the complexity of migratory movements. Interlinkages also define the lived experience of migrants: rules on visas, border controls, asylum, legal migration, integration, and return will often be inter-related in practice. Think of a beneficiary of international protection receiving long-term resident status, or of return decisions being issued once an asylum applicant has exhausted legal remedies. Visa requirements and cooperation with third states render it difficult to reach Union territory in the first place where an asylum application may be lodged. Policy papers, legal practice, and scholarly analyses have increasingly recognised this interconnectivity and employ the term ‘migration’ in an overarching manner, mirroring the title of this book. A broad terminology of ‘migration law and policy’ appears to be in the making, in Europe more than in North America. Nevertheless, the practice is far from uniform and the future is difficult to predict. As a matter of principle, the language of ‘immigration and asylum law’ and ‘migration law’ can be used interchangeably; both expressions are equally valid. For our purposes, however, the added value of a holistic understanding of ‘migration law’ stands out. It draws the attention of the reader to feedback loops between the different instruments and policies discussed throughout this book. Interdependencies concern the policy design and questions of legal interpretation. By way of example, procedural fundamental rights in the Charter apply to all instruments discussed in this volume, as do the best interests of the child. In comparison to ‘immigration’, the neutral descriptor ‘migration’ has the advantage of embracing temporary movements semantically. Numerous third country nationals leave

4 Introduction the Member States after having worked or stayed there for some time, often entirely voluntarily. The habitual meaning of ‘migration’ covers both temporary sojourn and permanent settlement, as well as the perspective of third country nationals who are residing abroad. Prevalent externalisation practices frustrate the aspiration of many migrants to reach European soil. To use the neutral descriptor ‘migration’ is better placed to cover the external dimension as well. Speaking of ‘European migration law’ does not exhibit quasi-imperial claims towards European countries which are not members of the European Union. Rather, reference to ‘Europe’, instead of the ‘EU’, in the main title reflects a broad thematic outlook which transcends the output of the supranational institutions in Brussels. International law and policy developments will be discussed throughout this book, with human rights and refugee law as the most prominent examples. Association agreements with neighbouring states will be considered in the same way as cooperation with countries further afield. Both the EU institutions and the Member States are particularly active in the field of international cooperation. Our semantic emphasis on ‘Europe’ in the main title indicates that this book is about much more than the idiosyncrasies of the supranational legislation and Court judgments. They will, of course, receive much attention, albeit as an integral part of the broader European response to cross-border movements. The European Union is a multi-level system. National governments play an active part in decision-making, and domestic authorities and courts are indispensable for effective implementation. Focusing on the supranational level runs the risk of missing half of the story. Our comments will highlight how legal rules and everyday practices at multiple levels inform the composite structure of European migration law. Member State action is particularly relevant in areas where the EU has no competences or prefers not to exercise them. Nationality law, legal pathways for refugees and economic purposes, regularisation of illegal stay, and operational cooperation with third states are prime examples of where little or no supranational activities exist. A holistic examination would be incomplete without due attention to the national contribution. Emphasising the significance of the national level has implications beyond questions of policy design and administrative implementation. Legal cultures differ across Europe. The Anglo-Saxon common law with its focus on courts and judges interacts with the continental civil law tradition of doctrinal hermeneutics. Positivist conceptions of legal interpretation coexist with normative and interdisciplinary approaches, both within and across national legal systems. For many practitioners, contributions written in English will not be the main or preferred source of information. The author is regularly confronted with these different methodological perspectives on the law and the potential of tensions between them. This book has been written with the intention of taking everybody on board. Legal hermeneutics have been combined with interdisciplinary comments to achieve this goal. We hope that every reader will find what they expect and will perceive alternative viewpoints as an invitation to question their own working method.

Outline of Part I on Overarching Themes Chapter 1 will explain that the EU started dealing with migration law in earnest thirty years ago. The abolition of border controls in the Schengen area was considered to require flanking

Outline of Part I on Overarching Themes 5 measures on visas and databases, amongst others. At the same time, states promoted intergovernmental cooperation in response to an increase in the number of asylum applications. National governments defined core features of the European rulebook on ‘third country nationals’, whose contents differs markedly from the freedom-enhancing rationale of the single market and ‘Union citizenship’. Successive Treaty amendments buttressed the relative autonomy of migration law within the area of freedom, security, and justice. EU institutions must decide on the degree of openness or closure within the confines of human rights, thus presenting the contemporary variant on how European states have dealt with migration ever since the nineteenth century. Inspection of the pre-history of European migration law exposes a fascinating combination of continuity and transformative disruption, which continues to define policy developments to this date. Institutional provisions have a direct impact on the policy outcome by determining the weight of the relevant actors. Chapter 2 will introduce readers to the prerogatives of the supranational institutions and their track record on migration. The informal trilogue format and the rise of soft law in the internal and external dimension demonstrate that everyday practices can be as relevant as legal rules in the Treaties. While the European Parliament often serves as a counterbalance to the preferences of national interior ministries, assembled in the Council, the Commission habitually responds to changing political priorities with technocratic pragmatism. Legislation on migration is full of compromise packages, and the salience of migratory matters in the public discourse complicates the ongoing search for a common ground. EU migration law is the product of state interests as much as it is a reflection of normative preferences. Chapter 3 will turn the spotlight on the CJEU in Luxembourg. Its judgments are a central object of analysis for anyone dealing with European migration law. Our comments will point out that the judgments do not come out of thin air. Inspection of procedural rules unveils the black box of judicial decision-making; rules of procedure prioritise the views of some actors over the conflicting views of others. Deference to the position of the legislature is one explanation why the judicial output fluctuates between dynamic constitutional rulings and an administrative mindset of doctrinal hermeneutics. A statistical survey of the case law unearths significant thematic asymmetries. Judgments on migration law reaffirm the centrality of preliminary references, notwithstanding divergences in the number of references between the Member States and according to subject matter. In subject areas where judges deliver few rulings, alternative accountability mechanisms are particularly important. Ombudspersons, financial and managerial control, and public attention can play a critical role in such situations. Migration law affects the lives of millions, raises formidable normative challenges, and does not always work well in practice. The rich interdisciplinary literature, presented in Chapter 4, sheds light on the underlying dynamics. Knowledge of the multiple economic, social, political, and cultural drivers of migration is essential for the design of reasonably well functioning policies, especially in the areas of asylum and return. Whereas legal experts concentrate on the interpretation of the law, other disciplines explore the making and the effects of migration policies. Sceptical public opinions support restrictions in times of increasing politicisation, and critical studies emphasise that perceptions of threat and normative preferences can be influenced by the use of language. Sociolegal research in the ethnographic tradition shows the impact of migrant agency and street-level bureaucracy. Policy experts can use these interdisciplinary insights to improve the policy output, whereas

6 Introduction legal scholars can develop sophisticated research designs; doing so combines doctrinal hermeneutics with a critical awareness of the law as an instrument of government. Contemporary forms of state control over the entry and stay of foreigners emerged during the nineteenth century, as a by-product of the formation of sovereign statehood. Despite being a cross-border phenomenon, migratory movements are not governed by public international law to a significant extent, with the Refugee Convention and human rights as the exceptions to the rule. The inherent tension between state sovereignty and human rights in the evolution of European migration law will take centre stage in Chapter 5. Our remarks will focus on the contribution of the European Convention on Human Rights (ECHR), the Charter of Fundamental Rights, the Refugee Convention, and international treaty bodies or diplomatic fora, including the Global Compacts. These general explanations will prepare the ground for sector-specific comments on human rights, refugee law, and international treaties in the substantive chapters. Chapter 6 will concentrate on the doctrinal foundations of the case law, which are essential for anyone trying to understand—or influence—the outcome of the interpretative exercise. Inspection of the supranational judicial practice unearths the predominance of seemingly technical arguments of doctrinal hermeneutics in the continental tradition, focusing on the wording, the general scheme, objectives, and the drafting history. To do so is easier said than done. While many judgments explore overarching features and cross- linkages between the different pieces of legislation, other rulings exhibit an almost bewildering lack of coherence, for instance with respect to teleological interpretation in light of objectives. Individual rights of migrants often have similar effects as human rights, under the conditions laid down in secondary legislation. Practitioners will appreciate that familiarity with doctrinal arguments increases legal certainty, supports interpretative coherence, and forms the basis for forward-looking proposals. Experts in EU law rightly pay much attention to the output of the supranational institutions. At the same time, the national level is equally important in terms of defining the conditions for the effective and uniform application of EU legislation. European migration law is a multi-level system, with national practices varying to a considerable extent. The vision of an ‘area’ of freedom, security, and justice and a common asylum ‘system’ should not be mistaken as shorthand for quasi-federal uniformity. Particularly important are differences in the administrative and court procedure. Chapter 7 will present the CJEU case law on unwritten general principles and fundamental rights in the Charter that limit the procedural autonomy of the Member States. Our horizontal assessment of judgments on different segments of migration law illustrates how rulings on asylum can be used in the context of return or legal migration, and vice versa. Chapter 8 will explore the fortification of Frontex and the Asylum Agency. Both agencies have received stronger mandates and have started hiring an abundance of new staff. While the Asylum Agency concentrates on information gathering and the quality and coherence of asylum procedures and reception conditions, Frontex increasingly assumes operational functions. Supranational staff and seconded national personnel assist Member States under pressure. In addition, cooperation with third states appears to become a focal point of future activities. ‘Pushback’ allegations and controversies about agency involvement in ‘hotspots’ underline that enhanced powers must go hand in hand with stronger accountability, both with regard to internal governance and external oversight. Challenging the behaviour of the agencies before the Court in Luxembourg is complicated due to the rules on standing, case

Outline of Part II on Sectoral Legislation 7 law on composite procedures, and the primary responsibility of host states of Frontex and EUAA missions under the mandate. Databases are less visible than the agencies, but they are equally important in practice. Recent years have witnessed a constant trend towards ‘function creep’, with the creation of new databases, the collection of ever more information, and the extension of access rights. The new legislative and technical infrastructure for the SIS, Eurodac, the Visa Information System, ETIAS, and the Entry/Exit System was scheduled to be put into effect at the time of writing. Doing so will mark a turning point for a highly vibrant area. The prospect of interoperability and the advent of automated processing will increase the practical, political, and conceptual significance of the databases. Data protection laws are the most important legal control standard, to be assessed by courts and other mechanisms of preventive or reactive supervision.

Outline of Part II on Sectoral Legislation As a reference work, this book will not usually be read from A to Z. Rather, readers will choose a specific theme from the table of contents. In order to facilitate orientation, the sectoral chapters will have the same structure: the policy concept and theoretical positions are presented first, followed by competences and human rights. The remainder of the analysis will be structured thematically: each section will consider a specific subject matter which can be identified easily from the heading. Often, one piece of legislation will be dealt with, for instance the Dublin III Regulation or the Seasonal Workers Directive. Our comments will contain cross-references to other chapters for overarching characteristics and whenever there are thematic interlinkages. Classic examples of such horizontal topics are supranational prescriptions for administrative procedures, the constitutive character of residence permits, or equal treatment provisions. The sheer complexity of European migration law entails that it is a matter of self-interest to identify general features which can then be studied jointly. Doing so saves time and resources and also allows for the identification of replicable patterns. Chapter 10 will highlight that the Court explores horizontal cross-fertilisation between the legislation on migration, as well as with the free movement of Union citizens. The public policy exception, fee levels, and the sufficient resources threshold are prime examples of the judicial search for coherence: judgments emphasising similarities interact with rulings accentuating differences, reflecting the specific contents and context of the different pieces of legislation. Of great practical significance are the principle of proportionality and the human right to equality before the law. Our analysis will adopt a bird’s eye view on the multi-faceted CJEU and ECtHR case law, stressing the persisting uncertainties. Judgments on Article 72 of the Treaty on the Functioning of the European Union (TFEU) and the principle of abuse demonstrate why states struggle to justify non-compliance with EU legislation in regular circumstances. Chapter 11 will examine instruments of pre-arrival control. Visa policy and carrier sanctions are essential elements of European migration law which sustain the profound stratification of contemporary entry policies: easy access for privileged travellers coexists with hurdles and restrictions for the ‘unwanted’. Visa requirements classify countries on the basis of an abstract risk profile. They are binding on all Member States and many neighbours

8 Introduction follow them as well; visa facilitation agreements serve as a bargaining chip to convince third states to cooperate. Procedural safeguards in the Visa Code Regulation apply to full members of the Schengen area only; they embody a high level of harmonisation in relation to documentation, admission criteria, and administrative procedure. Critical observers highlight the significance of carrier sanctions as a measure of privatised migration control, which is subject to limited legal oversight and hinders access to Union territory by persons coming from the Global South. Border controls have received much attention as a result of the controversies about ‘pushback’ allegations. Unfortunately, it can be difficult to identify the outer limits of the legal framework under human rights law and the provisions on border surveillance in the Schengen Borders Code Regulation and the Sea Borders Regulation. Chapter 12 will familiarise readers with these highly significant legal questions. By contrast, the rules on checks on persons at official border crossing points are comparatively straightforward, including for refusal of entry. The proliferation of prolonged ‘temporary’ internal border controls within the Schengen area is illustrative of profound governance deficits, which EU institutions struggle to overcome. Secondary movements of asylum applicants are a bone of contention, partly because of the complexity of the legal regime. Chapter 13 will focus on EU asylum legislation, which is the object of protracted policy debates and defines the work of many practitioners. Reiterating the idea behind a reference book, we will describe the political and theoretical framework, introduce newcomers to core features of the supranational legislation, and provide a reliable inventory of more than 100 Court judgments. Structural deficits of the Common European Asylum System will be discussed along with reform proposals and emerging themes, such as climate change. The evolution of European asylum law can be described as a history of half-hearted commitment. Our thematic comments will follow the structure of the legislation. Each section will discuss one instrument: asylum jurisdiction; asylum procedures; criteria for refugee status and subsidiary protection; temporary protection; reception conditions; the legal status of beneficiaries of international protection; resettlement and other legal pathways. We are confident that we managed to present both the broader legal and conceptual repercussions and the small print, which informs the legal practice. Legal migration receives less attention, although it concerns several million third country nationals. Chapter 14 will assess the contents and the effects of domestic and supranational rules for family members and labour migration. The need for common action on these matters was never straightforward, in the absence of enhanced transnational effects. Debates about economic migration were particularly divisive, before the EU settled on a sectoral approach. Inspection of the Blue Card Directive and the legislation on students and researchers, intra-corporate transferees, and seasonal workers will explain why these rules have a limited impact. States remain a laboratory for the design of innovative entry channels, both for the highly skilled and for migrants with fewer qualifications. While the ‘academic route’ supports the longer stay of students, other instruments are designed as temporary admission schemes. With regard to family reunification, optional derogations and exceptions have been criticised, but the Family Reunification Directive contains protective elements as well. Its standards go beyond the level of protection under human rights law. Chapter 15 will investigate integration policies. First admission sets in motion a process of status change, which culminates in the acquisition of long-term resident status or nationality. Many third country nationals benefit from residence security and equal

Outline of Part II on Sectoral Legislation 9 treatment, thus turning them into members of the societies they live in. A comprehensive analysis of integration policies must go beyond the explicit and implicit integration conditions enshrined in the instruments of migration law. Other state policies and contextual factors are equally important: think of the welfare state or schooling. In this overall context, the contribution of migration law is limited. Mandatory integration conditions have received much attention, especially language skills as a precondition for family reunification. Judges accepted their legality, subject to a hardship clause. Furthermore, many migration law instruments provide for equal treatment. We will demonstrate the legal characteristics and the practical implications of these provisions, with a special emphasis on the welfare state. The very idea of equal treatment with nationals and the limited inroads of EU law on the acquisition of nationality reaffirm our general conclusion that European migration law cannot be adequately understood without considering the national level. Return is the flipside of integration. EU Treaties call for ‘enhanced measures to combat illegal immigration’, and the institutions have taken that invitation seriously. Chapter 16 will describe the patchwork of domestic and supranational measures and how they interact, starting with a critical reflection on the use of the language. The EU approach to the criminalisation of irregular entry and stay is multi-faceted: instruments instructing Member States to adopt sanctions coexists with case law limiting the use of custodial sentences and legislation on victims of human trafficking. The Return Directive serves as the backbone of the supranational legal framework, establishing common standards for the return procedure, voluntary departure, and forced removal. Its provisions apply whenever third country nationals are ‘illegally staying’. EU legislation does not generally harmonise the conditions for unlawful stay and regularisation, which remain the domain of national laws and practices. Supranational legislation governs the administrative detention of returnees and asylum seekers, thus complementing human rights law. Chapter 17 will consider the position of neighbouring countries which prefer not to join the European Union or are waiting to be admitted. Association agreements and international treaties establish concentric circles with different degrees of participation. While Norway, Iceland, and Switzerland subscribe to the free movement of persons and are associated with the Dublin and Schengen rulebook, the UK stays outside. There is currently no enhanced cooperation with the UK on free movement or migration control, apart from the protection of acquired rights under the Withdrawal Agreement and transnational social security coordination. The status of Turkish nationals residing in the Member States has given rise to many Court judgments, which increased the level of protection significantly. Association agreements with other neighbours to the South and East contain few provisions on legal migration nowadays. The management paradigm of justice and home affairs has replaced the single market model as the focus of attention. International cooperation with third states is a controversial topic on which the EU institutions and national governments have spent considerable political capital. Chapter 18 will complete our survey with a portrayal of the external dimension. Its evolution can be described as a history of trial and error. While legal pathways and development feature prominently in policy documents, the output remains meagre. Mobility partnerships have not delivered, and it remains to be seen whether talent partnerships will prove more successful. Instead, the externalisation of migration control by means of international cooperation has become the hallmark of external action. Capacity building and operational

10 Introduction cooperation serve as sweeteners to convince third states to cooperate, combined with the—not so subtle—threat of sanctions in the domain of visas and beyond. Frontex positions itself to become an important player, in addition to national initiatives. EU legislation and Court judgments play a limited role, but this does not undo the practical and political significance of the external dimension for a holistic understanding of European migration law.

PART I

OV E R A RC HING T H E M E S

1

Building an Area of Freedom, Security, and Justice The European Union is a strange creature. It mirrors an international organisation in some respects and reproduces sovereign statehood in other regards. The perspective of migration law is a perfect illustration of this conceptual ambiguity. At the heart of the EU’s approach to cross-border movements of people lies what seems to be a paradox. For decades, European integration had been defined by the abolition of state borders. Swedes may move to the Netherlands and reside there indefinitely. However, internal mobility was not replicated externally. Treaty objectives call upon the institutions to ensure, amongst others, the ‘efficient monitoring of the crossing of external borders’.1 A Moroccan IT specialist cannot simply relocate to Spain for work; she will usually have to apply for a residence permit before entry, which can be refused for various reasons. To understand the supranational policy design on migration requires at least a basic awareness of the idiosyncrasies of the existing rulebook, as developed over time. To this date, the path- dependency of earlier developments informs debates about both the basic contours of European migration policy and seemingly technical legal questions that define the everyday work of most experts. Invocation of the ‘Schengen Agreement’ and the ‘Dublin Convention’ permeate the media discourse. Meanwhile, experts who are not familiar with EU law may be confused by the fundamental distinction between the free movement of ‘Union citizens’ and migration law towards ‘third country nationals’, or complaints about pervasive ‘intergovernmentalism’. The objective of this chapter is to familiarise readers with these idiosyncrasies in light of the historic evolution. Our analysis will combine the peculiarities of supranational migration law with a rough sketch of the evolution of migration law on the European continent. Awareness of the historic background similarly allows for the identification of overarching trends behind contemporary events and challenges. The sheer complexity of the legal material, the constant flood of new legislation and judgments, and changing contextual factors create the danger of observers not seeing the proverbial forest behind the trees. Our inspection of the history of migration law in Europe will concentrate on long- term transformation—in line with the historical concept of the ‘longue durée’,2 which accentuates slowly changing structures that the widespread focus on contemporary occurrences tends to lose sight of. Our exploration will unfold in three steps. It begins with the emergence of EU policies, which can be traced back to earlier inter-state cooperation on economic matters and refugee protection. These initiatives prepared the ground for the free movement of workers within

1 2

TFEU, art 77(1). See Fernand Braudel, ‘La longue durée’ (1958) 13 Annales: Économies, Sociétés, Civilisations 725.

European Migration Law. Daniel Thym, Oxford University Press. © Daniel Thym 2023. DOI: 10.1093/oso/9780192894274.003.0002

14 Building an Area of Freedom, Security and Justice the European Communities (1.1). Spillover effects on border controls, visas, and asylum resulted in the gradual emancipation of the rules on migration from third states, as opposed to mobility within the single market. The distinct features of today’s Treaty regime developed in a period of twenty-five years, spanning from the original Schengen agreements to the Treaty of Lisbon. The latter completed the ‘supranationalisation’ of the institutional framework and established a broad set of competences for the realisation of an area of freedom, security, and justice (1.2). An inspection of the Treaty rules for third country nationals will illustrate in how far the constitutional foundations differ from Union citizenship. EU Treaties leave the institutions much leeway to define the contents of secondary legislation within the confines of human rights (1.3).

1.1 From International to EU Migration Law European integration is commonly presented as an innovation that broke with previous attempts at promoting peace and prosperity on the continent. Once you pierce the veil of the self-description of a new beginning, the historic material on how states dealt with foreigners exposes a fascinating combination of underlying continuity and transformative disruption. An earlier phase of comparatively liberal migration regimes throughout the nineteenth century (1.1.1) coincided with the gradual emergence of the sovereignty-based migration law which persists until today (1.1.2). Important strands of EU legislation build upon earlier developments, especially with respect to the protection of refugees (1.1.3) and established patterns of labour migration on the European continent (1.1.4). The latter informed the choice for free movement within the European Communities (1.1.5), in contrast to the predominantly strict approach towards the former colonies (1.1.6). EU Treaties distinguish between ‘third country nationals’ and ‘citizens’, notwithstanding conceptual ambiguities about the rationale behind Union citizenship (1.1.7).

1.1.1 Parallelism of emigration, imperialism, and tourism To date, international law on cross-border movements is defined by a built-in discrepancy. A complex body of rules promoted the free exchange of goods after the Second World War, within the World Trade Organization and through free trade agreements. By contrast, Chapter 5 will demonstrate that the entry and stay of foreigners remains, by and large, the sovereign domain of states. With the exception of the Refugee Convention and human rights law, few agreements have tangible effects on migration. The Global Compact for Safe, Orderly and Regular Migration, adopted in 2018, did not fundamentally reverse this overall picture.3 Chapter 19 on the external dimension will illustrate the prevalence of informal cooperation frameworks underlying the dynamic evolution of relations with states beyond the European continent. Absence of far-fetched international rules need not be equated with restrictive policy outcomes. The second half of the nineteenth century is often remembered as a period of mass

3

See ch 5.5.4.

From International to EU Migration Law 15 migration, not least from a European perspective. Millions left their homes to start a new life in the ‘new world’, especially the United States and Latin America.4 Films and novels enshrined this period in the collective memory. ‘Little Italy’ in Manhattan and Scandinavian city names in the wider Great Lakes Region are testimony to a period of mass migration across the Atlantic. Intense movements also took place on the European continent, which technically mostly qualified as intra-state migration, even though the social and cultural change involved in the relocation from rural areas to urban centres was often dramatic in a period in which contemporary nation states were still in the making.5 By way of example, think of movements from Galicia to the industrial centres of Bohemia, Budapest, or Vienna within the Habsburg empire. Historical studies have put into perspective the occasionally one-sided memory of seemingly ‘free’ migration to the Americas. The US authorities rejected Europeans who suffered from bad health, were considered morally deviant, and, later on, could not read. They also regulated steam boat travel and introduced a historic precedent for contemporary carrier sanctions: shipping companies had to transport back home anyone refused entry.6 Most importantly, however, entry rules were asymmetric. The Chinese Exclusion Act, adopted by the US Congress in 1883, was the epitome of policy developments which effectively closed the borders to migrants from East Asia. This prepared the ground for overtly racist admission on the basis of country quotas, which willingly favoured Europeans (and among Europeans those from the Protestant North over people from the Catholic South and the poorer East). Other countries, such as Canada, Australia, and Brazil similarly pursued openly racist selection criteria.7 Drastic asymmetries also characterised the policies of European states. The nineteenth century saw both mass emigration to the Americas and intensified colonialism and imperial conquest. Colonial administrations supported semi-voluntary colonial migration patterns, for instance when Indians built railways or engaged in trade in Africa; forced labour was a frequent practice, including across longer distances.8 Doing so built upon centuries of experience with the slave trade, notably across the Atlantic, which defines the prehistory of European migration law in the same way as mass emigration to the United States.9 We may even draw an indirect line from colonial migration to the formation of the single market. Although the single market is occasionally presented as a model for global mobility, we shall see that its formation was consciously limited to nationals of the Member States, thus excluding those living in the (former) colonies. From a legal perspective, both policy restrictions and the facilitation of entry and stay were governed by domestic laws. In the long nineteenth century, few international initiatives dealt with migration law in today’s sense. Having said this, states actively cooperated in the domain of technical regulation. Some of the first international organisations pursued an administrative agenda on navigation, postal and telegraph services, or railway transportation.10 Such technical coordinated facilitated movements across borders. Faster and cheaper 4 See Vincent Chetail, International Migration Law (OUP 2019) 39–46. 5 See Jan Lucassen and Leo Lucassen, ‘The Mobility Transition Revisited, 1500–1900’ (2009) 4 J Global Hist 347. 6 See Aristide R Zolberg, A Nation by Design (Harvard UP 2006) chs 5–7. 7 See David Scott FitzGerald and David Cook-Martin, Culling the Masses (Harvard UP 2014) chs 3–4. 8 See Robert J Steinfeld, Coercion, Contract and Free Labor in the Nineteenth Century (CUP 2001). 9 See Seymour Drescher and Paul Finkelman, ‘Slavery’ in Bardo Fassbender and Anne Peters (eds), The Oxford Handbook of the History of International Law (OUP 2012) 890. 10 In conventional accounts of public international law, the Central Commission for Navigation on the Rhine became the first international organisation in 1831, to be followed, amongst others, by the International Telegraph

16 Building an Area of Freedom, Security and Justice transportation and communication provided the infrastructural bedrock for heightened mobility. Chapter 11.1 will remind us of the immense technical and regulatory efforts states invest in easing travel for bona fide travellers; this bifurcation of movements is often overlooked in studies of migration law. With regard to tourism, the English were among the first to spearhead tailor-made tours of the remnants of antiquity in the Mediterranean by means of rail-bound transportation.

1.1.2 Migration law as a by-product of state formation A diachronic comparison between the ‘migration law’ of today and previous legal regimes is complicated by the absence of a direct comparator. Most European countries developed the idea of nationality only gradually, thereby supplanting earlier forms of feudal limitations and local membership. Contemporary notions of nationality, entry control, and residence status are intricately linked to the formation of the sovereign nation state, which monopolised and harmonised membership and mobility regimes within and across state borders step by step.11 To describe statehood as the congruence of territory, population, and government became standard practice at the time, thus conceptualising cross-border movements as pertaining to core matters of sovereignty. States remained free to regulate the entry and stay of foreigners. In the nineteenth century, economic upheaval and the profound social, political, technical, and cultural renewal that characterised the ‘great transformation’12 created the migration law we know today. During that period, previous rules on sectoral limitations and local membership gradually lost relevance. The English ‘poor laws’ were emblematic of how public authorities had regulated migration before the introduction of contemporary notions of nationality and residence status. This was done on the basis of communal affiliation, which proved incapable of adequately responding to domestic population movements during the industrial revolution.13 On the continent, nationality laws in the French Code napoléon served as a model for state membership.14 They were imitated by countries across the continent.15 A sovereignty-based concept of nationality came into being; doing so necessitated the emergence of a ‘modern’ migration law regulating the entry and stay of foreigners. One important factor behind state legislation was an increase in the movements of Jews and other residents of the Habsburg Empire and Tsarist Russia to Western Europe.16 All in all, these

Union in 1865, the Universal Postal Union in 1874, and the less formal (continental European) Office central des transports internationaux par chemins de fer in 1893. 11 See the classic account by Stein Rokkan, State Formation, Nation-Building, and Mass Politics in Europe (OUP 1999)ch 2. 12 Karl Polanyi, The Great Transformation (1944) (2nd edn, Beacon Press 2001). 13 See David Feldman, ‘Was the Nineteenth Century a Golden Age for Immigrants?’ in Andreas Fahrmeir and others (eds), Migration Control in the North Atlantic World (Berghahn 2003) 167; and, for Germany, Ferdinand Weber, Staatsangehörigkeit und Status (Mohr Siebeck 2018)ch 1. 14 See Patrick Weil, How to be French (Duke UP 2008)ch 2. 15 Historic accounts have moved beyond the binary distinction between territorial ius soli and ethno-cultural ius sanguinis, instead emphasising how different conceptions overlapped and interacted across Europe; see Andreas Fahrmeir, Citizenship (Yale UP 2007)chs 1–4. 16 The Aliens Act 1905 served as a watershed in the UK, while, in Germany, labour mobility from today’s Poland resulted in fierce debates; in Southern Europe, other factors were more important.

From International to EU Migration Law 17 diverse developments fostered a continuous appearance of today’s migration law, which has been succinctly described as the ‘invention of the passport’17. That process turned full circle when most European countries introduced stringent border controls and visa requirements on the eve of the First World War. To be sure, control measures had existed previously, but it was not until the early twentieth century that a seemingly archetypical picture became a widespread phenomenon: border fences and border guards checking identity papers of people crossing the border. Earlier control systems had taken place on the territory and gave a greater emphasis on expulsion.18 Frontex and the Schengen Borders Code Regulation are the contemporary continuation of this historic realignment of control mechanisms. Somewhat paradoxically, the fortification of state borders was accompanied by enhanced international coordination, which laid the ground for the European migration law of today.

1.1.3 Towards international protection for refugees Inter-state cooperation started in earnest during the inter-war period within the broader framework of the League of Nations. While the League ultimately failed to secure peace, it spearheaded cooperation in various domains. Refugee protection was among them. Functional predecessors of asylum and refuge had existed since antiquity, but the contemporary refugee regime evolved in response to mass displacement before and after the first world. The first High Commissioner for Refugees, Fridtjof Nansen, established an identity document to facilitate cross-border movements for Russian exiles at a time of severe travel restrictions. Other initiatives were put forward, yet most of them were ad hoc measures which could not prevent moral failure when Jews and others fled Nazi Germany.19 When the Refugee Convention established a stable framework for displacement in 1951, it built upon these earlier initiatives. Interdisciplinary studies show that the principled distinction between ‘forced’ displacement and ‘voluntary’ migration was not self-evident. Sociological findings demonstrate that movements can rarely be classified as being predominantly voluntary or forced, as a result of the relative character and contextual fluidity of pertinent push and pull factors.20 Moreover, we should be cautious not to read history backwards. Debates in the 1920s and 1930 viewed refugees both as victims in need of protection and as economic actors building a new life. There were ample activities on refugees within the International Labour Organization (ILO),21 and even the Refugee Convention was concerned primarily with the social and economic integration of the ‘last million’ of displaced people who could not return home after the end of the Second World War.22 It was not a foregone conclusion that

17 John Torpey, The Invention of the Passport (2nd edn, CUP 2018)ch 4. 18 See Thomas Nail, Theory of the Border (OUP 2016) chs 2–6; and Andreas Fahrmeir, ‘Staatliche Abgrenzungen durch Passwesen und Visumzwang’ in Jochen Oltmer (ed), Handbuch Staat und Migration in Deutschland seit dem 17. Jahrhundert (De Gruyter 2015) 221. 19 See James C Hathaway, ‘The Evolution of International Refugee Law: 1920–1950’ (1984) 33 ICLQ 348; and Nevzat Soguk, States and Stranger (Minnesota UP 1999) chs 3–4. 20 See ch 4.1. 21 See Katy Long, ‘When Refugees Stopped Being Migrants’ (2013) 1 Migration St 4. 22 See Refugee Convention, arts 12–30, 34.

18 Building an Area of Freedom, Security and Justice the distinction between ‘migrants’ and ‘refugees’ would turn into the conceptual backbone of international and European migration law. Critical scholars have highlighted that the refugee regime replicates the concept of sovereign statehood, by means of correcting the failure of home states to protect their citizens. Refugee law is, in other words, firmly embedded in the Westphalian state system and the concept of state sovereignty.23 In the specific context of the inter-war period, it was much more than a simple humanitarian exercise. The newly created states in Central and Eastern Europe drew borders in areas where diverse ethno-cultural groupings had been living for centuries, thereby creating important minority populations (or dreadful practices of forced displacement and population exchange24). Pioneering ‘minority treaties’ were supposed to protect minorities in these states but turned out to be ineffective. The experience with the flagrant failure of the inter-state protection regime explains the initial scepticism with which Hannah Arendt approached the move towards international human rights.25 Things changed thereafter. Chapter 5 will explain that human rights law and the Refugee Convention have turned into essential pillars of migration law.

1.1.4 Prehistory of EU rules on labour migration The prevalent self-description of the European project as a ‘new legal order’26, distinct from classic international law, was motivated, in part at least, by the desire to overcome the weaknesses of the international regime of the inter-war period. This should not prevent us, however, from recognising the political and, to a certain extent, legal continuity of the approach to labour migration on the European continent from the early twentieth century until the creation of today’s single market. Italy was the most significant country of emigration, with millions leaving for the US, South America, and other European states, in addition to sizeable domestic movements from the Mezzogiorno to the industrial North. Within Europe, Italians went primarily to France, Switzerland, Luxembourg, and, later, Belgium and Germany; many returned to the Italian peninsula. Notwithstanding the image of ‘migration’ as a one-way street, labour migration, including later ‘guest worker’ programmes, was often a temporary project: many returned after having earned some money and acquired new skills. Return quotas differed between countries and over time, but it was always an important factor on the European continent and beyond. Italian labour migration is relevant for our topic, since the government actively promoted legal guarantees for Italians living abroad, in response to unsatisfactory practices of private companies which dominated the recruitment market at the time. Italy spearheaded the conclusion of a net of bilateral agreements, altogether negotiating more than one hundred treaties between the early twentieth century and the 1950s, starting with France. These bilateral treaties dealt with working conditions and safety at the workplace, financial remittances, accidence insurance, social security, and recruitment.27 Other countries concluded 23 See Nevzat (n 19) chs 1–2; and Emma Haddad, The Refugee in International Society (CUP 2008) ch 5. 24 They were widespread in south-eastern Europe, including the territories of today’s Greece and Turkey, in the context of the Balkan Wars of 1912/13 and the Treaties of Sèvres and Lausanne of 1920 and 1923. 25 See ch 5.2.2. 26 See Case 26/62 van Gend en Loos EU:C:1963:1. 27 See Christopher Rass, Institutionalisierunsprozesse auf einem internationalen Arbeitsmarkt (Schöningh 2010)ch 3.

From International to EU Migration Law 19 similar agreements. In the field of social security, bilateral treaties Germany negotiated with Poland and Czechoslovakia with respect to well-established regional migration patterns provided a template which continues to inform the contents of the Social Security Coordination Regulation (EC) No 883/2004.28 Migrant workers were an important field of activity for the ILO. The Geneva-based organisation promoted the rights of migrant workers, designed a model bilateral agreement in the late 1930s, and put forward a multilateral convention on the eve of the Second World War.29 Until today, the conventions and recommendations of the Geneva-based organisation are an important point of reference for the global debate about of migrant workers.30 Chapter 15.3.3 will explain why ILO Conventions do not, however, have much practical relevance for EU migration law, for the simple reason that secondary legislation habitually establishes a higher level of protection. ILO Conventions and bilateral treaties concentrated on the rights of workers and labour recruitment without embracing first admission. They complemented the entrenched body of international laws and customs on the treatment of aliens, which will be discussed in Chapter 5.1.1 on human rights. One step further, bilateral treaties on commerce, navigation, settlement, and friendship occasionally promised the admission of individuals for purposes of commerce and trade. Such abstract pledges seem not to have held much legal sway, however, at a time when entry and stay of foreigners was regarded as a matter of political preference and executive discretion.31 These bilateral treaties can be compared, at an abstract level, to the association agreements the EU has signed with neighbouring states; they will be presented in Chapter 17. Attempts to streamline international rules continued after the Second World War. Within Europe, the Organisation for European Economic Co-operation (OEEC), which later became the Organisation for Economic Co-operation and Development (OECD), and the Council of Europe were important players. Ideas developed in their framework provided the backdrop for discussions on the Treaties of Paris and Rome in 1951 and 1957 on what would gradually turn in the EU we know today; core actors in the supranational context often knew each other from debates in other international fora.32 With regard to third country nationals, the Council of Europe pursued some harmonisation in the form of the European Convention on Establishment of 1955 and other instruments, especially on medical and social services.33 However, these measures gathered few ratifications and were overtaken by EU enlargement. In contrast to human rights, the conventions apply reciprocally to nationals of state parties only, ie nationals of third countries cannot derive any rights from these multilateral instruments.

28 See Cindy Hahamovitch, ‘Creating Perfect Immigrants’ (2003) 44 Labor History 69. 29 See Migration for Employment Convention [No 66] (adopted 28 June 1939, never entered into force); the accompanying Recommendations Nos 61 and 66; and Chetail (n 4) 52–56. 30 See Ryszard Cholewinski, ‘International Labour Migration’ in Brian Opeskin and others (eds), Foundations of International Migration Law (CUP 2012) 283. 31 See Andreas Paulus, ‘Treaties of Friendship, Commerce and Navigation’ in The Max Planck Encyclopedia of Public International Law (OUP, March 2011) MN 9–13 http://www.mpepil.com (accessed 1 March 2023). 32 See Simone A Goedings, Labor Migration in an Integrating Europe (SDU Uitgevers 2005) 43–58, 74–84. 33 See ch 5.1.2.

20 Building an Area of Freedom, Security and Justice

1.1.5 Free movement within the single market Legal methodology was one of the reasons why EU integration in the single market overtook coordination in the framework of the Council of Europe. Reliance on international conventions with lengthy ratification procedures was supplanted by supranational law- making. The case law of the Court of Justice ensured coherent—and often dynamic— interpretation. The ‘new legal order’ had tangible advantages. However, none of this was known when the representatives of the six founding members came together to negotiate the Treaties establishing the European Coal and Steel Community (ECSC) of 1951 and the European Economic Community (ECC) of 1957. In particular, it was not self-evident that economic cooperation would embrace the free movement of workers, since trade agreements dealt mainly with goods at the time. Things turned out differently. Historic studies have shown that the determination of the Italian government explains why the final text recognised the ‘four freedoms’: the free movement of goods, services, capital, and persons.34 Once the principle of free movement of persons had been accepted, the negotiations turned to the conditions. It was governments further north which insisted on the equal treatment of workers to prevent downward pressure on domestic wage levels and working conditions as a result of migration. Agreement was facilitated by regional free movement among the Benelux countries and existing patterns of labour migration within Europe, which meant that the new obligations would not change much in practice.35 Moreover, important questions were relegated to implementing legislation.36 Throughout the 1960s and 1970s, the Council agreed on several regulations, including a wide definition of family member and the elimination of labour market tests, even though the wording of primary law did not prescribe this outcome.37 These rules were interpreted dynamically by the Court, thereby further advancing the status of migrant EU workers. Contemporary studies on Union citizenship habitually focus on Court of Justice of the European Union (CJEU), thereby sidelining the fact that the political institutions had prepared the ground for the case law by means of generous legislation.

1.1.6 Colonial migration: the downside of free movement States made a conscious choice to limit free movement to nationals. Doing so was in line with the international practice at a time when agreements were mostly reciprocal in nature, as shown in Chapter 5.1.1 on international law. True, the wording of the EU Treaties does not convey that limitation explicitly, since Article 45 TFEU refers to ‘workers’ in the abstract. 34 See today’s TFEU, art 26(2). 35 See Goedings (n 32) chs 1–2; and Simone Poli, ‘Migration in European Integration’ (2016) 22 J Eur Int Pol 279, 279–90. 36 There was a transitional period of 12 years, which was later shortened somewhat; see Treaty establishing the European Economic Community (adopted 25 March 1957, entered into force 1 January 1958) 298 UNTS 4300, arts 8, 38–41. 37 Goedings (n 32) chs 3–5 traces the negotiations on the initial Regulation No 15 (1961), the intermediate Regulation No 38/64/EEC, and the final Regulation (EEC) No 1612/68 on freedom of movement for workers [1968] OJ Spec Ed I-1968(II)/475; see also Síofra O’Leary and Sara Iglesias Sánchez, ‘Free Movement of Persons and Services’ in Paul Craig and Gráinne de Búrca (eds), The Evolution of EU Law (3rd edn, OUP 2021) 506, 512–16.

From International to EU Migration Law 21 However, criticism of ‘revisionist interpretation’38 by the EU institutions is not confirmed by archival research. During negotiations, there was agreement that only nationals should be covered. An explicit nationality clause was discarded in reaction to Franco-Italian disputes over the status of workers from Algeria and German concerns about compatriots from communist East Germany.39 Resolution of the dispute was entrusted to a working group on overseas countries and territories, which evaded a definite statement on Algerians by means of an open-ended formulation.40 A later proposal by the European Economic and Social Committee to include refugees and stateless persons via implementing legislation was not taken up by the Council for lack of competence.41 The single market was never supposed to promote universal free movement. The historic background to the Franco-Italian dispute was a short period of extensive emigration from the colonies to France, the UK, Belgium, and the Netherlands in the aftermath of the Second World War. These late colonial movements were not the result of a deliberate policy choice in favour of widespread immigration; instead, they came about as the by-product of attempts progressively to transform the colonial franchise on the basis of seemingly equal relations within the British Commonwealth, the Tripartite Kingdom of the Netherlands, and the Union française. These initiatives entailed comparatively liberal entry practices in the slipstream of the normative rebranding of colonial domination and exploitation.42 Few people would have predicted at the time that these initiatives would falter quickly. Relations with the African continent were a central question when the common market was set up in the 1950s and 1960s.43 Against the background of the early significance of colonialism, the free movement of persons in the single market can be presented in different lights. On the one hand, European integration was an attempt at overcoming the territorial nation state after two devastating world wars by means of extensive legal guarantees for cross-border movement, which were reconstructed, in the early 1990s, as a prototype of post-national citizenship and belonging.44 On the other hand, the exclusion of the (former) colonies reveals a darker side, which Etienne Balibar famously denounced as ‘European apartheid’.45 Restrictive tendencies have accompanied the integration process from the very beginning. Originally, however, migration was governed by domestic rules. The single market was not accompanied—unlike in the field of goods—by a common external migration policy mirroring the common commercial policy.46 For several decades, migration from third states remained a sovereign prerogative of the Member States. 38 Elspeth Guild and Steve Peers, ‘Out of the Ghetto?’ in Steve Peers and Nicola Rogers (eds), EU Immigration and Asylum Law (Martinus Nijhoff 2006) 81, 114. 39 The Italian government considered Algerians to be ‘competitors’ at a time when France had introduced restrictions for Italians to support entry from overseas. 40 See Goedings (n 32) 128–35. 41 See the intergovernmental Declaration 64/805/EEC of national representatives on refugees [1964] OJ 78/ 1225; in response to the opinion of the EESC [1962] OJ 62/64; neither document is available in English. 42 See Randall Hansen, ‘Globalization, Embedded Realism, and Immigrants to Europe’ (2000) 35 Comp Pol Stud 259; and Frederick Cooper, Citizenship between Empire and Nation (Princeton UP 2014). 43 See Peo Hansen and Stefan Jonsson, Eurafrica (Bloomsbury 2014). 44 See Yasemin Soysal, Limits of Citizenship (University of Chicago Press 1994) 145; and Dora Kostakopoulou, ‘European Union Citizenship: Writing the Future’ (2007) 13 ELJ 623, 628–33. 45 See Etienne Balibar, We, the People of Europe? (Princeton UP 2003); and generally Lucy Mayblin and Joe Turner, Migration Studies and Colonialism (Polity 2021). 46 In the terminology of international economic law, the free movement of persons established an (internal) free trade agreement, not an (external) customs union.

22 Building an Area of Freedom, Security and Justice Other bodies similarly did not pay much attention. Conventions in the context of the Council of Europe concerned reciprocal rights of nationals. In a remarkable episode of postcolonial legal history, the European Court of Human Rights rendered its first prominent judgment ever on migration in a case concerning nationals of the former colonies. Domestic rules erected strict hurdles for family reunification with spouses from abroad—a move judges found to comply with human rights, since the applicants could have ‘establish[ed] family life in their own or their husbands’ home countries’.47 The judgment introduced the well-known formula that states have ‘as a matter of well-established international law’ the right to control the entry of non-nationals as the starting point of the human rights inquiry. Later developments were more lenient, in particular with regard to protection against expulsion of former ‘guest workers’ and late colonial migrants.48

1.1.7 Theoretical ambiguity of Union citizenship The generic distinction between ‘Union citizens’ and ‘third country nationals’ has defined EU law and policy ever since 1992. The Treaty of Maastricht popularised the very notion of third country national when it introduced two separate legal regimes for the new category of ‘Union citizenship’49 and intergovernmental cooperation on ‘nationals of third countries’.50 The term ‘third country national’ supplanted this formulation in the years thereafter.51 It was used in the Tampere Conclusions of 1999 and by the Treaty of Lisbon, which established today’s primary law and fortified the ‘area of freedom, security, and justice’ as an overarching frame of reference. It will be used throughout this volume. Notwithstanding this straightforward legal distinction, the theoretical implications of both EU migration law and Union citizenship are contested. Among academic observers, the added value of the ‘citizenship’ label was and continues to be the subject of controversies. It had been chosen, at the time of the Maastricht Treaty, to signal a vague determination so to advance the integration process towards political union through electoral rights across borders and free movement of those who do not work.52 Yet, early commentators highlighted the ‘weakness’53 of the legal framework and castigated the new status as a misnomer bound to remain an ‘empty gesture’.54 The situation changed in the late 1990s when the Court started interpreting citizens’ rights dynamically,55 before embracing a more sceptical approach, as judges backtracked on earlier innovations and

47 Abdulaziz and others v United Kingdom App nos 9214/80, 9473/81 and 9474/81 (ECtHR, 28 May 1985) § 67; see also Marie-Bénédicte Dembour, When Humans Become Migrants (OUP 2015) ch 3; see also ch 15.7.2 in this volume. 48 See ch 15.2.2. 49 EC Treaty, art 8, as amended by the Treaty of Maastricht. 50 EU Treaty, art K.1, as amended by the Treaty of Maastricht. 51 See Kees Groenendijk, ‘Categorizing Human Beings in EU Migration Law’ in Saskia Bonjour and others (eds), The Others in Europe (Institut d’études européennes 2011) 21. 52 Today’s TFEU, arts 20(2)(a), (b), 21 were introduced by the Maastricht Treaty. 53 Síofra O’Leary, The Evolving Concept of Community Citizenship (Kluwer 1996) 304–307. 54 Joseph HH Weiler, ‘European Citizenship and Human Rights’ in Jan Winter and others (eds), Reforming the Treaty on European Union (TMC Asser 1996) 57, 68. 55 For an overview see Damian Chalmers, Gareth Davies, and Giorgio Monti, European Union Law: Text and Materials (3rd edn, CUP 2019) ch 11; and Dimitry Kochenov, ‘The Essence of EU Citizenship Emerging from the Last Ten Years of Academic Debate’ (2013) 62 ICLQ 97.

From International to EU Migration Law 23 limited the equal treatment of Union citizens who are not working.56 For our purposes, a rough sketch of the debate and the underlying motivation serves as an important point of reference for anyone trying to understand why legal rules on third country nationals differ from the category of Union citizenship. The introduction of Union citizenship built upon earlier conceptions of free movement within the single market as an ‘incipient form of European citizenship’.57 The ‘market citizen’58 was never supposed to be a factor of production only whose mobility sustained the common market. Social policy objectives had accompanied economic considerations from the beginning. Part-time workers, for instance, benefited from free movement even if they received social assistance.59 Furthermore, freedom to provide services was interpreted to embrace passive service reception, counter to the international practice.60 Free movement had always transcended a purely economic rationale; it served wider political purposes. Free movement was, in the words of the Schuman Declaration, ‘a first step in the federation of Europe’.61 That is not to say that there was a consensus on the end point. Behind abstract invocations of ‘citizenship’ and ‘political union’ lay diverse and potentially contradictory visions. The debate indicates, however, that many actors, especially within the EU institutions and among the elite, pursued the dream of further integration. This motivation gained momentum during the 1970s when heads of state or government promoted further integration through a focus on citizens’ rights. Calls for a ‘passport union’, known today as the Schengen area, and the free movement of everyone prepared the ground for the introduction of Union citizenship by the Maastricht Treaty.62 Union citizenship was a tangible expression of the well-known vocation of the EU Treaties ‘to lay the foundations of an ever closer union’.63 This teleological surplus of the Treaty regime holds two lessons for the analysis of migration law. First, the distinction between citizens’ rights and third country nationals is embedded in the design of the European project. We cannot expect the EU institutions to abandon it in the foreseeable future. Secondly, the experience of Union citizenship demonstrates that contextual factors influence legal developments. The Court’s famous dictum that Union citizenship ‘is destined to be a fundamental status’64 hints at the potential of dynamic interpretation, which judges have realised in numerous judgments. At the same time, the cautious approach of recent years reflects growing Euroscepticism and the limited practical impact of Union citizenship upon social practices and identificatory patterns.65

56 See Daniel Thym (ed), Questioning EU Citizenship (Bloomsbury/Hart Publishing 2017); and Agustín Menéndez and Espen DH Olsen, Challenging European Citizenship (Palgrave 2020) chs 4–5. 57 Commissioner Lionelli Levi-Sandri, ‘Free Movement of Workers in the European Community’ (1968) 11 Bulletin of the European Communities 5. 58 Hans-Peter Ipsen, Europäisches Gemeinschaftsrecht (Mohr Siebeck 1972) 250–54. 59 See Case 139/85 Kempf EU:C:1986:223. 60 See Andrew Evans, ‘European Citizenship’ (1982) 45 MLRev 497; and Case C-371/08 Ziebell EU:C:2011:809, paras 72–78. 61 ‘The Schuman Declaration’ (9 May 1950). 62 See Antje Wiener, Building Institutions (Westview 1998) chs 2–3; and Willem Maas, Creating European Citizens (Rowman & Littlefield 2007) chs 2–3. 63 Recital 1 of the Preamble of today’s TFEU, in line with the EEC Treaty 1957 (n 36). 64 Case C-184/99 Grzelczyk EU:C:2001:458, para 31; and Daniel Thym, ‘Frontiers of EU Citizenship’ in Dimitry Kochenov (ed), EU Citizenship and Federalism (CUP 2017) 705. 65 See Daniel Thym, ‘The Evolution of Citizens’ Rights in Light of the EU’s Constitutional Development’ in Thym, EU Citizenship (n 56) 111.

24 Building an Area of Freedom, Security and Justice For migration law, it remains to be seen whether the ‘area of freedom, security, and justice’ or fundamental rights can serve as a trajectory supporting dynamism.

1.2 Towards Distinct Policies for Third Country Nationals The EU is known for having secured peace and prosperity by integrating sovereign states into a supranational community. This did not imply, as we have seen, that the integration process aimed at universal free movement. There exists an ‘outside’ realm towards which the Union has to position itself. In doing so, political actors have always been torn between the post-national impulse to overcome borders and the competing vision of asserting common interests. Inspection of the development of EU migration law from the late 1970s onwards shows that such debates were a recurring feature and that policy priorities changed over time. Cooperation within the intergovernmental Schengen framework proved particularly productive (1.2.1), notwithstanding parallel coordination of justice and home affairs before and under the Treaty of Maastricht (1.2.2). The Treaty regime we know today emerged gradually, with the Treaties of Amsterdam and Lisbon as decisive occurrences (1.2.3 and 1.2.4). Political guidelines by the European Council were decisive for early developments, before they ran out of steam (1.2.5). Contemporary public debates see the juxtaposition of opposing narratives for migration policy (1.2.6).

1.2.1 ‘Schengen’: compensatory logic It was rather unlikely that an idyllic Luxembourgish town on the river Moselle would become a household name across Europe and beyond. And yet, that is precisely what happened to the municipality of Schengen for the sole reason of its geographic proximity to where the borders of France, Germany, and Luxembourg meet. At this location in 1985, representatives of the three above-mentioned states, together with Belgium and the Netherlands, signed an essentially political commitment gradually to abolish border checks.66 Participation of the Benelux countries was facilitated by the internal travel area they had established twenty-five years earlier. The first Convention of 1985 paved the way for the detailed prescriptions on the abolition of border controls and flanking measures in the Convention Implementing the Schengen Agreement of 1990.67 Border controls were discontinued within the ‘Schengen area’ on 26 March 1995, once the participating countries had concluded that the factual requirements for that step had been fulfilled. Other Member States followed the avant-garde before the Treaty of Amsterdam integrated the Schengen rulebook into the EU framework. The legal status of the Schengen acquis and its application to new Member States will feature in Chapter 12.3 on border controls.

66 The Agreement on the Gradual Abolition of Checks at their Common Borders (adopted and entered into force 14 June 1985) [2000] OJ L239/13 was an agreement of political intent to future action by means of a second treaty. 67 On the contents see David O’Keeffe, ‘The Schengen Convention’ (1991) 11 YB Eur L 185.

Towards Distinct Policies for Third Country Nationals 25 Like many other initiatives in the EU context, the motivation for change combined substantive policy considerations with the vague vision of ‘ever closer union’. Heads of state or government had discussed plans for a ‘People’s Europe’ and a ‘European Union’ in the 1970s, which included the project of a ‘passport union’ anticipating core tenets of the later Schengen cooperation.68 The move towards an internal market buttressed the desire for the abolition of border controls and customs checks.69 At a bilateral summit in May 1984, French President François Mitterrand and German Chancellor Helmut Kohl took up the idea and promised border-free travel.70 The Benelux countries joined the initiative, which was realised on the basis of public international law outside the EU framework. As often with projects of differentiated integration, relations between the core group and the non- participating countries were ambiguous. While Schengen is commonly remembered as a laboratory nowadays, the original debate embraced darker narratives. The exclusion of the founding member Italy was symbolically important; some governments were doubtful as to whether it would control its external borders effectively.71 Historical studies demonstrate two features of the early cooperation which inform our understanding of European migration law to date. First, support by national governments was critical for the Schengen project. Political scientists speak of ‘venue shopping’ when like-minded peers in national interior ministries promote their agenda via international fora.72 In that respect, ‘Schengen’ complemented the informal TREVI group, which had discussed internal security and selected asylum matters among the Member States from the 1970s onwards.73 Early justice and home affairs cooperation within the Schengen framework and other fora unfolded with the explicit support of national governments, not against their will. Secondly, there had always been disputes about the relative weight of national governments and EU institutions. The Commission had tried to get a metaphorical foot in the door of migration policy via supranational competences on social policy, only to be called off by the Member States. They successfully challenged modest consultation and cooperation requirements.74 Another strand of the legal debate concerned the question of whether the promise of the internal market as ‘an area without internal frontiers’ in today’s Article 26(2) TFEU brought about a legislative competence for border controls, or even a directly applicable obligation to do so.75 Discussions were about more than institutional prerogatives. Behind the reticence of national capitals to realise ‘Schengen’ within the framework of the Brussels-based institutions lay the concern that the freedom-enhancing rationale of the single market would trump security concerns.76 In the eyes of interior ministries, the 68 See Commission, ‘Communication: A Passport Union’ COM(75) 322 final; the terminology goes back to the historic model of the Nordic Passport Union of 1952/54. 69 See Commission, ‘White Paper: Completing the internal market’ COM(85) 310 final Nos 24–32. 70 See Angela Siebold, ZwischenGrenzen (Ferdinand Schöningh 2013) 40–59. 71 On the French debate see Simone Paoli, ‘France and the Origins of Schengen’ in Elena Calandri and others (eds), Peoples and Borders (Nomos 2017) 255, 258–65; and on Germany see Siebold (n 70) 301–12. 72 See ch 2.1.2. 73 TREVI was set up after the terrorist attacks by left-wing radicals, such as the Italian Brigate Rosse, and secessionists like the Spanish-Basque ETA or the Irish Republican Army; see Eva Oberloskamp, Codename TREVI (De Gruyter 2016). 74 Decision 85/381/EC setting up a prior communication and consultation procedure on migration policies [1985] OJ L217/25, adopted on the basis of today’s TFEU, art 153, was annulled by Joined Cases 281/85, 283, 284, 285 and 287/85 Germany and others v Commission EU:C:1987:351. 75 See Georgia Papagianni, Institutional and Policy Dynamics of EU Migration Law (Martinus Nijhoff 2006) 6–13. 76 See Ruben Zaiotti, Culture of Border Controls (University of Chicago Press 2011) chs 4–6.

26 Building an Area of Freedom, Security and Justice political emphasis should not be unhindered mobility but the realisation of effective security by other means than internal border controls. Contemporary debates about the need to ‘protect’ the external borders can be traced back to that period. Indeed, the Schengen rulebook follows a simple quid pro quo: the abolition of internal border controls was considered to require ‘flanking measures’ compensating national authorities for the loss of control options at the internal borders through inter-state cooperation on policing, criminal matters, and migration.77 Common standards were deemed indispensable for the control of the external borders, legislative harmonisation on questions such as asylum jurisdiction or tourist visas, and the introduction of novel instruments, in particular the Schengen Information System (SIS). Having said this, the compensatory logic was not about achieving internal freedom at the ‘price’ of external restrictions. Rather, national governments, not least among the original Schengen countries, pursued external controls as an end in itself to buttress domestic asylum reforms mentioned in Chapter 13.1. There were different views about the relative weight of openness and closure in the design of these measures but the consensus on their necessity was widely shared.78 Judges confirmed that view retroactively when they found the basic guarantee of ‘an area without internal frontiers’ not to be directly applicable without prior agreement on flanking measures.79 The gradual emergence of Schengen was a classic case of spillover, which occurs when supranational policies prepare the ground for political initiatives on related policy fields. The initial move towards EU migration law was the result of trickle-down effects of the political decision to abolish internal border controls in the broader context of the single market programme and the move towards political union. First agreed upon by interior ministers of the five original Schengen countries, the flanking measures soon transcended the compensatory logic and turned into ends in themselves.80 Schengen served as a laboratory for the design of new control instruments. We shall see in Chapters 9 and 11 that Europeanisation was critical in promoting the use of databases and pre-arrival measures, in particular visa requirements and carrier sanctions.

1.2.2 Treaty of Maastricht: fragile intergovernmental cooperation From the beginning, the Schengen area had been linked to the European Economic Community politically and legally.81 This connection gained momentum in the 1990s when the original idea to establish a ‘European Union’ with an enhanced degree of political integration lead to the Treaty of Maastricht. The Treaty entered into force at the end of 1993 after a turbulent ratification process. Unease with justice and home affairs in the UK and Denmark proved significant for later opt-outs, which will be discussed in Chapter 2.4 77 ibid ch 5; and Siebold (n 70) ch 3. 78 See Commission (n 68) 11–15, 15–19; Schengen Agreement 1985 (n 66) arts 7–9, 17–20; Commission (n 69) No 28; and Political declaration by the Governments of the Member States on the free movement of persons, attached to the Single European Act [1987] OJ L169/25. 79 See Case C-378/97 Wijsenbeek EU:C:1999:439, para 40, which goes back to an ‘activist’ Dutch MEP; see John Morijn, ‘Personal Conviction and Strategic Litigation in Wijsenbeek’ in Fernanda Nicola and Bill Davies (eds), EU Law Stories (CUP 2017) 178. 80 See Ferruccio Pastore, ‘Visas, Borders, Immigration’ in Neil Walker (ed), Europe’s Area of Freedom, Security and Justice (OUP 2004) 89, 94–98. 81 CISA, arts 134 and 140 recognised that Community law takes precedence in cases of conflict and that other Member States can join via an accession treaty.

Towards Distinct Policies for Third Country Nationals 27 on institutional practices. For our purposes, ‘Maastricht’ should be remembered as the moment when immigration and asylum policy generally fell within the ambit of the EU institutions. The legal construction of the Maastricht Treaty was complex and should only be studied by those with a generic interest in the institutional processes that led to today’s area of freedom, security, and justice. The newly founded EU was based on three ‘pillars’.82 The first pillar was the original EEC Treaty, which was rebranded as the Treaty establishing the European Community (EC Treaty); it contained the single market and related policy fields and was subject to the traditional characteristics of supranational law-making. The EC Treaty was complemented by the Treaty establishing the European Union (EU Treaty) with overarching provisions and two distinct sets of rules on foreign and security policy (second pillar) and justice and home affairs (third pillar). The second and third pillars had intergovernmental features. In justice and home affairs, the EU Treaty did not allow for the adoption of directives and regulations, only non-binding joint positions and international conventions subject to ratification by national parliaments.83 The direct outcome of intergovernmental decision-making was meagre, at least from a legal perspective. The Council unanimously agreed on a substantial number of non- binding resolutions on diverse matters such as unaccompanied minors, labour migration, admission of students, or the refugee definition,84 thus complementing prior informal cooperation on an ad hoc basis outside the EU framework.85 The Council also discussed international conventions to replace the Schengen framework, which were not pursued further when the Treaty of Amsterdam came into sight.86 The well-known Dublin Convention of 1990 on asylum jurisdiction, discussed in Chapter 13.3, proved the most tangible output. It was the only treaty on migratory matters that came into force after several years of a drawn- out ratification process. Even though the Maastricht Treaty’s third pillar produced little legally binding output, the informal arrangements provided a bedrock of common standards the EU institutions could build upon later.

1.2.3 Treaty of Amsterdam: gradual supranationalisation In 1997, the Treaty of Amsterdam made a decisive step towards the European migration law as we know it today. Heads of state or government decided to transfer the provisions on visas, border controls, asylum, and immigration (as well as private international law) from the intergovernmental ‘third pillar’ to the supranational ‘first pillar’.87 This allowed the EU institutions to adopt directives and regulations. Important instruments, which remain 82 The term was widely used in academic and political circles but did not feature in primary law. 83 See EU Treaty, as amended by the Treaty of Maastricht, art K.3(2); and Richard McMahon, ‘Maastricht’s Third Pillar’ (1995) 22 L Issues Eur Int 51. 84 See Elspeth Guild, Immigration Law in the European Community (Kluwer 2001) 255–73; Kay Hailbronner, Immigration and Asylum Law and Policy of the European Union (Kluwer 2000) 47–52; and Jörg Monar and Roger Morgan (eds), The Third Pillar of the European Union (Peter Lang 1995). 85 On the London Resolutions of 1992 and the ‘Ad hoc Group on Immigration’ see Eileen Denza, The Intergovernmental Pillars of the European Union (OUP 2002) ch 3. 86 See David O’Keeffe, ‘The New Draft External Frontiers Convention and the Draft Visa Regulation’ in Monar and Morgan, The Third Pillar (n 84) 135. 87 See EC Treaty, arts 61–69, as amended by the Treaty of Amsterdam; on the negotiations see Guild, Immigration Law (n 84) 295–340.

28 Building an Area of Freedom, Security and Justice in force until today, were adopted on the basis of the new Treaty framework, especially the Family Reunification Directive 2003/86/EC and the Long-Term Residents Directive 2003/ 109/EC. Having said this, the ‘communitaurisation’ of decision- making by the Treaty of Amsterdam remained incomplete. Its institutional provisions have been aptly described as a ‘ghetto’88 within the supranational legal order. Governments were not ready to abandon a tight control of decision-making on this sensitive matter.89 The European Parliament was consulted only, the Commission shared the agenda-setting power with the Member States, not all domestic courts could make preliminary references to the Court of Justice, and, crucially, the Council acted unanimously.90 National governments held a veto power and could block agreement unless the others accommodated their concerns. That is how optional clauses on integration requirements found their way into rules on family reunification and long-term residents. The Treaty of Nice, which entered into force in 2003, extended today’s ordinary legislative procedure to certain segments of migration law.91 One year later, the Council activated a bridging clause rendering more areas subject to qualified-majority voting, as well as the co-decision powers of the Parliament.92 This was done once the Member States had unanimously agreed on core pieces of legislation in order not to give the Parliament leverage to influence the original design. Full supranationalisation was brought about by the Treaty of Lisbon, which also abolished other remnants of intergovernmentalism, such as the limitation of preliminary references to national apex courts.93 Chapter 3 will demonstrate that the number of Court judgments grew substantially thereafter. From today’s perspective, we may conclude that the former ‘ghetto’ has been gentrified. When it came to the policy substance, the Treaty of Amsterdam left a two-fold mark. First, it introduced the notion of an overarching ‘area of freedom, security, and justice’, to be discussed below. Secondly, the Treaty of Amsterdam incorporated the body of rules that make up the Schengen ‘acquis’94 into the EU framework. That was achieved on the basis of country-specific opt-outs for the UK (before Brexit) and Ireland, along with rather complex rules on the association of Denmark, to be discussed in Chapter 2.4. The Schengen Protocol effectively turned the intergovernmental rules of the Schengen acquis into supranational legislation, which can be amended in the ordinary legislative procedure.95 Today’s Schengen Borders Code Regulation and the Visa Code Regulation are prominent successor instruments.

88 Steve Peers, EU Justice and Home Affairs Law (1st edn, Longman 2000) 2. 89 See Papagianni (n 75) 25–51; and Neil Walker, ‘In Search of the Area of Freedom, Security and Justice. A Constitutional Odyssey’ in Walker, Europe’s Area (n 80) 3, 16–20. 90 See EC Treaty, arts 67 and 68, as amended by the Treaty of Amsterdam; after a five-year-period, the Commission obtained a monopoly of initiative for most policy areas and qualified-majority voting in the Council became possible on selected matters; for details see Hailbronner (n 84) 92–103. 91 See EC Treaty, art 67(5), as amended by the Treaty of Nice (adopted 26 February 2001, entered into force 1 February 2003) [2006] OJ C321E/37; and Protocol [No 35] on Article 67 [2006] OJ C321E/317; for comments see Papagianni (n 75) 88–100. 92 See Decision 2004/927/EC [2004] OJ L396/45; and Pieter Jan Kuijper, ‘The Evolution of the Third Pillar from Maastricht to the European Constitution’ (2004) 41 CML Rev 609. 93 cf EC Treaty, art 68(1), as amended by the Treaty of Amsterdam. 94 The French noun ‘acquis’ (for ‘acquired’ rights and obligations) is widely used in EU jargon to describe a set of rules, in particular in the context of accession. 95 See ch 12.3.3; and Daniel Thym, ‘The Schengen Law’ (2002) 8 ELJ 218.

Towards Distinct Policies for Third Country Nationals 29

1.2.4 Treaty of Lisbon: autonomy of migration law Besides aligning decision-making with the orthodoxy of the supranational integration method, the Treaty of Lisbon confirmed the autonomy of migration law as a policy field in its own right. What had started off as a spillover of the abolition of internal border controls turned into a free-standing Treaty objective: the Union ‘shall offer its citizens an area of freedom, security, and justice’ comprising, amongst others, ‘appropriate measures with respect to external border controls, asylum, immigration’.96 The Treaty abandoned the earlier reference to ‘flanking measures’97 compensating Member States for the loss of control at the domestic borders. Article 3(2) TEU lists the area of freedom, security, and justice among the EU’s top priorities. Migration law towards third country nationals became a self-sufficient policy field. What is more, the Treaty of Lisbon consolidated the scope of Union competences in line with the proposal of the erstwhile Constitutional Treaty, which had been discarded after negative referenda in France and the Netherlands. The Convention, which had drafted the text of the Constitutional Treaty, was particularly active in the field of justice and home affairs; its recommendations were integrated in the Lisbon Treaty without major changes.98 It can be helpful, therefore, to consult the drafting documents to understand the meaning of formulations such as the ‘integrated management system for external borders’ in Article 77(2)(d) TFEU, which serves as the legal basis for Frontex. Preparatory documents can be found online,99 and academic literature supports their identification.100 The drafters of today’s Treaty text in the European Convention made a conscious choice to lay down broad competences, thereby allowing the institutions to develop EU migration policy gradually without the need to revert to politically cumbersome Treaty revision. The scope of the competences will be discussed in the sections on the constitutional foundations throughout the second part of this book.

1.2.5 European Council: political programming The project of an ‘area of freedom, security, and justice’ was conceived first by the Treaty of Amsterdam as a grand design mirroring the realisation of the customs union, the single market, and monetary union. All these projects were phased in over time on the basis of a political vision fulfilled through legislation, for which Treaty amendments introduced additional competences subject to streamlined decision-making.101 The area of freedom, 96 TEU, art 3(2), as amended by the Treaty of Lisbon (adopted 13 December 2007, entered into force 1 December 2009) [2008] OJ C115/13. 97 EC Treaty, art 61(a), as amended by the Treaty of Amsterdam. 98 TFEU, arts 77–80 essentially reiterate Treaty establishing a Constitution for Europe (adopted 24 October 2004, never entered into force) [2004] OJ C310/1, arts III-265–268; and the Convention’s Draft (adopted 18 June 2002) [2003] OJ C169/1. 99 See http://european-convention.europa.eu (accessed 1 March 2023). 100 See Clemens Ladenburger and Stephane Verwilghen, ‘Policies Relating to the Area of Freedom, Security and Justice’ in Giuliano Amato and others (eds), Genèse et destinée de la Constitution européenne (Bruylant 2007) 743; Henri Labayle, ‘L’espace de liberté, sécurité et justice dans la Constitution pour l’Europe’ [2005] Revue trimestrielle de droit européen 437; and Daniel Thym, ‘The Area of Freedom, Security and Justice in the Treaty establishing a Constitution for Europe’ WHI Paper 12/2004 http://ssrn.com/abstract=4117783 (accessed 1 March 2023). 101 See Jörg Monar, ‘The Area of Freedom, Security and Justice’ in Armin von Bogdandy and Jürgen Bast (eds), Principles of European Constitutional Law (2nd edn, Hart Publishing 2009) 551, 554–55.

30 Building an Area of Freedom, Security and Justice security, and justice followed this model with the important difference that the Treaty of Amsterdam laid down an abstract constitutional skeleton for which no distinct political vision existed. Policy objectives and countervailing interests were to be recalibrated step by step. Article 68 TFEU recognises the need for political guidance when it invites the European Council to ‘define the strategic guidelines for legislative and operational planning’. That provision was introduced by the Treaty of Lisbon after years of experience with agenda-setting at the highest political level. At the beginning, the European Council was influential. A promising start was the meeting in Tampere in October 1999, which kick-started supranational law-making after the entry into force of the Treaty of Amsterdam.102 Later five-year programmes were adopted at the meetings of the European Council in The Hague (2004),103 Stockholm (2009),104 and Ypres (2014).105 These programmes established a policy framework and helped overcome resistance when the European Council ‘urged’ hesitant interior ministers to ‘speed up’ legislation.106 In addition, other actors presented policy guidelines, such as the intergovernmental ‘Pact on Immigration and Asylum’ of 2008,107 the Commission’s ‘European Agenda on Migration’ of 2015,108 or the ‘New Pact on Migration and Asylum’ of 2020.109 These additional documents were influential, even though they did not necessarily reflect the views of all Member States. Throughout the years, the focus of attention has shifted. Such change of direction is no disadvantage, as it is the purpose of strategic guidelines to respond to changing circumstances. While the Tampere Conclusions were full of youthful enthusiasm echoing the optimistic outlook on globalisation after the end of the cold war, the Programme adopted in The Hague was dominated by the fight against terrorism following the September 2001 attacks. A few years later, the Stockholm Programme made an effort to find a middle ground between control imperatives and human rights concerns.110 Nevertheless, the Tampere Conclusions had similarly embraced a mixed approach already, ranging from the ‘control of external borders to stop illegal immigration’ to the ‘fair treatment’ of those residing legally.111 Debates on the Stockholm Programme of 2009 proved particularly heated, with skirmishes between the Commission and the Council about the weight of the different institutions and the road to be taken. While the Commission favoured further harmonisation on the basis of the broad set of competences under the Treaty of Lisbon, such as the vision of an ‘Immigration Code’, mentioned in Chapter 6.2.2 on the doctrinal foundations, the Council insisted on a cautious approach and objected to proposals which had emphasised

102 See European Council, ‘Presidency Conclusions of the Meeting in Tampere’ (15/16 October 1999); on the basis of the initial ‘Vienna Action Plan’ [1999] OJ C19/1. 103 ‘The Hague Programme’ [2005] OJ C53/1. 104 ‘The Stockholm Programme’ [2010] OJ C115/1. 105 European Council, ‘Conclusions of the Meeting in Ypres’ (EUCO 79/14, 27 June 2014) Nos 1–13. 106 See Presidency, ‘Seville European Council, 21/22 June 2002’ (Council doc 13463/02, 24 October 2002) No 37; and Papagianni (n 75) 215–20. 107 See ‘European Pact on Immigration and Asylum’ (Council doc 13440/08, 24 September 2008), which had been spearheaded by France, Germany, and the United Kingdom. 108 Commission, ‘Communication: A European Agenda on Migration’ COM(2015) 240 final. 109 See Commission, ‘Communication on a New Pact on Migration and Asylum’ COM(2020) 609 final. 110 See Cian C Murphy and Diego Acosta Arcarazo, ‘Rethinking Europe’s Freedom, Security and Justice’ in Diego Acosta Arcarazo and Cian C Murphy (eds), EU Security and Justice Law (Hart Publishing 2014) 1, 4–9. 111 European Council (n 102) Nos 3, 18, 21–22.

Towards Distinct Policies for Third Country Nationals 31 the rights of migrants. In symbolically important moves, it ditched the idea of an immigration code and replaced the call for ‘fairness’, in the Commission’s draft, with an insistence on ‘well-managed’ migration.112 The desire to evade renewed interinstitutional confrontation may have played a role in the decision to limit the Ypres Guidelines of 2014 to a few paragraphs with little substantive guidance.113 Moreover, the very idea of five-year plans may have lost its initial appeal, once EU institutions had succeeded in filling the original tabula rasa with important pieces of legislation. The sheer complexity of secondary legislation leaves little space for overarching guidance. It need not be the worst outcome, therefore, if the European Council discontinued the earlier practice of abstract programming.114 The five-year cycle was not taken up in 2019, unless one counts the ‘New Strategic Agenda 2019–24’ which dealt with migration on merely half a page.115 Controversial topics may always be taken up by the European Council, independent of five-year programmes. From a legal perspective, the guidance by the European Council remains political in the sense that the EU institutions are free to deviate during the legislative procedure. An example is the Long-Term Residents Directive, discussed in Chapter 15.6 on integration, which departed from the Tampere Conclusions by laying down strict conditions for intra- EU mobility, instead of replicating the free movement of Union citizens. The weight of the programmes stemmed from the underlying political consensus, not doctrinal attributes. Earlier programmes expired after five years, although they may serve as a contextual element elucidating the telos and the drafting history of the instruments adopted at the time. The loss of significance demonstrates that the area of freedom, security, and justice has reached a state of maturity. Change remains possible, but the centre of attention moves from abstract guidelines to the ‘nitty-gritty’.

1.2.6 Public discourse between ‘fortress Europe’ and ‘common values’ Debates about migration are not confined to the supranational institutions. The general public engages with supranational decision-making, which increasingly rises above the proverbial ‘closed doors’ of diplomatic negotiations. A turning point was the asylum policy crisis of 2015/16. An upsurge of public attention triggered a reconfiguration of the conceptual foundations in a sort of ‘constitutional moment’.116 The outcome of such fundamental realignment is not predetermined. Public discourse can push the debate in diverse directions. Corresponding public debates will not usually follow a quasi-academic style of reasoning: symbolic issues often develop a life of their own and short slogans with an

112 See Dora Kostakopoulou, Diego Acosta Arcarazo, and Tine Munk, ‘EU Migration Law’ in Acosta Arcarazo and Murphy, Justice Law (n 110) 129, 133. 113 See Philippe De Bruycker, ‘The Missed Opportunity of the “Ypres Guidelines” of the European Council’ EUI Migration Policy Centre Blog (29 July 2014). 114 The wording of TFEU, art 68 prescribes neither timeframes nor the adoption of a single document, ie it could possibly be complied with by means of regular interventions whenever the need for strategic guidance arises. 115 See European Council, ‘A New Strategic Agenda 2019–2024’ (Annex, EUCO 9/19, 20 June 2019) 7; apparently, draft guidelines were discussed later but no agreement was reached; see ‘Outcome of the 3756th Council meeting’ (Council doc 6582/20, 13 March 2020) 3. 116 See generally Bruce Ackerman, We, the People, vol 2 (Harvard UP 1998); and, on the EU, Daniel Thym, ‘The “Refugee Crisis” as a Challenge of Legal Design and Institutional Legitimacy’ (2016) 53 CML Rev 1545, 1566–69.

32 Building an Area of Freedom, Security and Justice emotional appeal embody the position of the speaker.117 In debates about migration, several of such symbolic markers stand out. A standard feature is a commitment to ‘a Europe that protects’.118 Similar language is used by those supporting ‘European sovereignty’,119 in an attempt not to leave the concept of sovereignty to those questioning the European project.120 Neither of these catchphrases necessarily comprises a clear-cut vision of migration policy, also considering that the object and purpose of protection and sovereignty will often be defined in an open manner, including values. Even so, the language of protection and sovereignty expresses the desire to improve state control over migratory movements. An emphasis on protection and sovereignty is often combined with semantic framing stressing border controls, ‘illegal’ migration, and security risks. Such a focus on control need not be equated with restrictions. It can be a sovereign choice to advance migration for humanitarian, economic, or political reasons, or as a French interior minister put it before he became President: immigration choisie et non subi.121 Chapter 18.1.1 will demonstrate that this is how the notion of ‘migration management’ was understood after the millennium change. Recent years have witnessed the emergence of a radical alternative. The desire for ‘management’ is increasingly being confronted with calls for ‘zero immigration’ by populist politicians and governments, in particular with respect to asylum. Hungary’s Prime Minister Victor Orbán spearheaded fence building and border closures, criticising the EU institutions: ‘they would like to manage migration, and not to stop the migrants’.122 Chapter 4.3.3 will show how such discursive framing may influence the policy outcome from an interdisciplinary perspective. A revealing example was Commission President von der Leyen thanking the Greek government for being the ‘ασπίδα’ (shield) of Europe when it closed its external borders in March 2020, after President Erdoğan had encouraged foreigners to leave Turkey in what initially appeared to be a potential mass exodus.123 Poland and Lithuania were praised by the European Council in October 2021 for rebutting a ‘hybrid attack’124 by means of an incitement of migratory movements by the Belarusian dictator. In both cases, martial language accompanied the denial of access to the asylum procedure. Critics of restrictive migration policies often use the antipodal slogan ‘fortress Europe’ to express discontent. While the term is widely understood to refer to migration nowadays, it emerged in the 1980s to describe the element of external economic closure when the single market programme harmonised internal standards, thus rendering it more difficult for 117 See Talcott Parsons, ‘The Theory of Symbolism in Relation to Action’ in Talcott Parsons and others (eds), Working Papers in the Theory of Action (The Free Press 1953) 31–62. 118 Former Commission President Jean-Claude Juncker, ‘Towards a Better Europe’ (State of the Union Address, 14 September 2016). 119 French President Emmanuel Macron, ‘Initiative pour l’Europe’ (Discours à la Sorbonne, 26 September 2017). 120 See Ségolène Barbou des Places, ‘Taking the Language of “European Sovereignty” Seriously’ (2020) 5 European Papers 287; and Simone Penasa and Graziella Romeo, ‘Sovereignty-based Arguments and the European Asylum System’ (2020) 22 EJML 11, 12–22. 121 See ‘Pour Nicolas Sarkozy, ‘L’immigration choisie est un rempart contre le racisme’ Le Monde (27 April 2006). 122 Eszter Zalan, ‘Visegrád Countries Immediately Push Back on New Migration Pact’ EUobserver (25 September 2020). 123 See Ursula von der Leyen, ‘Remarks at the Joint Press Conference with Kyriakos Mitsotakis, Prime Minister of Greece, Andrej Plenković, Prime Minister of Croatia, President Sassoli and President Michel’ (Statement 20/ 380, 3 March 2020). 124 European Council, ‘Conclusions’ (ECUO 17/21, 22 October 2021) No 19; see also ch 10.7.

Distinguishing Union Citizenship from Third Country Nationals 33 foreign companies to sell their produce.125 The slogan built upon Nazi vocabulary for the fortification of the French coast against an Allied invasion.126 Those criticising the prospect of ‘fortress Europe’ often emphasise an alternative vision of European integration that promotes common values instead of collective interests. Reference to ‘European values’ is widespread in policy debates, not only about migration. Chapter 4.3.4 will highlight that an emphasis on values and human rights may have radiating effects on the policy substance and court rulings. In doing so, human rights can serve as a counterpoint to the public references to ‘illegal’ migration and security threats. Note that the binary character of many debates need not prevent an accommodation of the countervailing positions, both theoretically and practically. Such room for alignment is particularly relevant for the EU institutions, which have traditionally sought to balance diverse positions in a mixture of interest-based realism and value-oriented idealism, as traced in Chapter 2.1 on the institutional practice. A telling example of how the search for compromise can go astray was the designation of the portfolio of Commission Vice-President Schinas for the 2020–24 period, which includes the coordination of internal and external migration policies. The original proposal spoke of ‘Protecting our European way of life’127 and was criticised as shorthand for restrictions, despite the small-print of the job description referring to values underlying the way of life. In the end, the phrasing was changed to the positive sounding motto of ‘promoting’128 the European way of life—a classic supranational compromise formula.

1.3 Distinguishing Union Citizenship from Third Country Nationals Anyone studying migration law at a time of fierce criticism of the EU’s asylum policy will find it difficult to grasp the enthusiasm with which observers had approached Europeanisation previously. There was a widespread assumption that the supranational institutions would counter securitarian biases among interior ministries when intergovernmental decision- making subsided.129 This prediction was correct to a certain extent, as the legislation described throughout this book will demonstrate. Nevertheless, many had hoped for further approximation to Union citizenship. The constitutional design of the area of freedom, security, and justice is one explanation why that did not happen. Instead of Treaty-based guarantees to free movement (1.3.1), the legislature determines the degree of openness or closure (1.3.2). Scenarios of overlap with Union citizenship were visible in early debates (1.3.3), before the adoption of migration-specific instruments gradually established an alternative policy rationale for migration law (1.3.4).

125 See C Michael Aho, ‘Fortress Europe’ (1994) 29 Columbia J World Bus 32. 126 See Cornelia Schmitz- Berning, ‘Festung Europa’ in Cornelia Schmitz- Berning (ed), Vokabular des Nationalsozialismus (De Gruyter 2007) 232. 127 Ursula von der Leyen, ‘A Union that strives for more. My agenda for Europe’ (July 2019) section 4. 128 Ursula von der Leyen, ‘Mission Letter: Margaritis Schinas’ (1 December 2020). 129 By way of example see Elspeth Guild, The Legal Elements of European Identity (Kluwer 2004) chs 7–12; and Anja Wiesbrock, ‘Granting Citizenship-Related Rights to Third-Country Nationals’ (2012) 14 EJML 63.

34 Building an Area of Freedom, Security and Justice

1.3.1 Primary law: open-ended Treaty objectives The introduction of ‘Union citizenship’ and distinct rules for ‘third country nationals’ in the Treaty of Maastricht were parallel phenomena, which built on the exclusion of foreigners, including colonial subjects, from free movement. The harmonisation of migration law was the functional equivalent of the external tariff for the importation of goods, which followed a different rationale from the vision of unhindered circulation within the single market. It did not come as a surprise, therefore, that the intergovernmental conferences and the drafters of the Constitutional Treaty, which later became the Treaty of Lisbon, opted for flexibility. Articles 77–79 TFEU depart from the internal mobility regime when laying down diverse— and potentially contradictory— objectives. The abolition of internal borders within the single market gives way to the ‘enhanced measures to combat illegal immigration’,130 which must be in ‘compliance with the principle of non-refoulement’.131 Generally, ‘the efficient management of migration flows’132 should be accompanied by ‘fair[ness] towards third-country nationals’.133 The vague notion of an ‘area of freedom, security, and justice’ does not shed light on how to operationalise this conglomerate of objectives. ‘Freedom’, ‘security’, and ‘justice’ are essentially contested theoretical concepts, which can be realised in diverse ways.134 Most political actors can subscribe to them in the abstract, although they may disagree quite fundamentally about the design of specific instruments. From a constitutional perspective, the open texture coincides with limited practical impact. Objectives in Articles 67 and 77–80 TFEU are legally binding on the EU institutions and can be points of reference in the policy debate. However, they will not usually translate into judiciable yardsticks for secondary legislation135—mirroring the constitutional status of the values and objectives enshrined in Articles 2, 3, and 21 TEU. Academics may criticise the predominance of securitarian approaches, but such criticism remains external to the law as long as the policy instruments do not fall foul of judiciable constitutional standards, such as human rights. The legislature benefits from a principled discretion on how to implement and balance policy objectives. For that reason, the abstract designation of ‘justice and home affairs’ can be considered more appropriate than the term ‘area of freedom, security, and justice’, which can be understood to hint at an underlying normative vision. Indeed, two Directorates General of the Commission and the European Parliament’s well-known Committee on ‘civil liberties, justice, and home affairs’ (LIBE) continue referring to ‘justice and home affairs’ in line with earlier Treaty language.136 The notion of an ‘area of freedom, security, and justice’ can only be understood against the backdrop of the compensatory logic of the original Schengen cooperation. The triad freedom—security—justice presents us with a rough designation of the ‘flanking measures’

130 TFEU, art 79(1). 131 ibid art 78(1). 132 ibid art 79(1). 133 ibid art 67(2); similarly, ibid art 79(1). 134 See generally Jeremy Waldron, Law and Disagreement (Clarendon Press 1999). 135 See further Anna Kocharov, Republican Europe (Hart Publishing 2017) 84– 90; and Jürgen Bast, Aufenthaltsrecht und Migrationssteuerung (Mohr Siebeck 2011) 141–44. 136 At present, there are two DGs on ‘migration and home’ (HOME) and ‘justice and consumers’ (JUST), which indirectly take up the title for EU Treaty, arts K.1–9, as amended by the Treaty of Maastricht.

Distinguishing Union Citizenship from Third Country Nationals 35 on internal mobility and migration (freedom), judicial cooperation (justice),137 and criminal matters (security). These themes defined the Convention Implementing the Schengen Agreement and the intergovernmental ‘third pillar’ of the Treaty of Maastricht. Similarly, Chapter 7.1.1 will explain that the term ‘area’ does not have a fixed meaning in the EU context and loses the territorial connotations it can have in English in other language versions, such as the French espace. There is, in short, no forward-looking policy vision behind the ‘area of freedom, security, and justice’. The loss of relevance of strategic planning through the European Council’s five-year programmes reaffirms that an overarching concept did not develop over time. Reference to the ‘fair treatment of third country nationals’138 may be read to invoke basic notions of social justice, but it appears in a different light upon closer inspection. ‘Fairness’ is not only, like ‘justice’, an essentially contested concept whose operationalisation requires policy choices on what it means. Moreover, it marks a symbolic distinction from the ‘equal treatment’ that Union citizens enjoy with nationals.139 The expression was first used by the Tampere Conclusions as the standard formula for foreigners residing legally.140 Historically, the term ‘fair treatment’ resonates with customary rules on the treatment of foreigners.141 Invocation of ‘fair treatment’ indicates that the EU Treaties aim at a level of protection transcending the minimum requirements of fundamental rights, while staying short of ‘equal treatment’ with nationals in the tradition of Union citizenship. Courts will not usually deduce judiciable standards from the abstract notion of fairness.142 The legislature is primarily responsible for deciding what it considers fair, while judges will focus on the interpretation of secondary legislation and human rights.

1.3.2 Legislative leeway within the confines of human rights Third country nationals cannot invoke constitutional guarantees to cross-border movement in the same way as Union citizens. Judges even found migration law and Union citizenship to be ‘entirely different’.143 Nevertheless, our analysis should beware false dichotomies. Acknowledging differences between justice and home affairs and the single market does not imply unfettered state discretion. There are alternatives to the binary juxtaposition of free movement and traditional notions of alienage.144 Articles 77–80 TFEU entrust the legislature with operationalising the abstract policy objectives. Moreover, third country nationals 137 Note that the English and French versions employ the normative term ‘justice’, while other languages use the technical word ‘law’ (German Recht; Dutch recht). 138 TFEU , art 79(1); similarly, TFEU , art 67(2); French traitement équitable; German angemessene Behandlung. 139 See Sara Iglesias Sánchez, ‘Constitutional Identity and Integration’ (2017) 18 GLJ 1797, 1803; and Andreas Funke, ‘Primärrechtliche Grundlagen’ in Ferdinand Wollenschläger (ed), Enzyklopädie Europarecht, Band X (2nd edn, Nomos 2021) § 16 MN 13, 58. 140 See European Council (n 102) No 18; and, argumentum e contrario, higher standards for long-term residents in No 21. 141 Other language versions, mentioned in n 138, demonstrate a direct linguistic similarity with the ‘Hull Doctrine’ requiring ‘adequate’ compensation in case of expropriation; see ch 5.1.1. 142 Contra Ester Herlin-Karnell, The Constitutional Structure of Europe’s Area of ‘Freedom, Security and Justice’ and the Right to Justification (Hart Publishing 2019) chs 2–4; and Anna Magdalena Kosińska and Barbara Mikołajczyk, ‘Does the Right to Migration Security Already Exist?’ (2019) 21 EJML 83, 106–109. 143 Case C-718/19 Ordre des barreaux francophones et germanophone and others EU:C:2021:505, para 53; see also Case C-930/19 État belge EU:C:2021:657, paras 71–76; and Iglesias Sánchez (n 139) 1798–807. 144 See Daniel Thym, ‘Citizens and Foreigners in EU Law’ (2016) 22 ELJ 296.

36 Building an Area of Freedom, Security and Justice can rely on alternative constitutional safeguards in the Charter of Fundamental Rights, which generally presents itself as an avant-garde catalogue.145 Human rights may be less generous than citizens’ rights but their protective credentials overcome classic notions of state sovereignty nonetheless. Chapter 5 will demonstrate that, in the field of migration, the Charter reaffirms existing guarantees under the European Convention on Human Rights (ECHR), while adding new guarantees, such as the rights of the child and the right to good administration. The degree of protection varies depending on the subject matter, but a common thread stands out: human rights do not usually—unlike Union citizenship—embody a right to entry, with the notable exception of refugees who may rely on Articles 4 and 18 of the Charter of Fundamental Rights of the European Union (CFR) to challenge non-admission at the border if they face a real risk of inhuman treatment.146 By contrast, the right to private and family life under Article 7 CFR limits state discretion when it comes to expulsion, but it ‘cannot be interpreted as denying States a certain margin of appreciation when they examine applications for family reunification’.147 Secondary legislation may establish a higher level of protection but such an outcome is not mandatory under human rights law. With regard to economic migration, Article 15 CFR reiterates the distinction between third country nationals and Union citizens. The provision starts with a reminder of citizens’ rights to seek employment anywhere,148 before reaffirming the absence of a generic right of third county nationals to enter Union territory: ‘Nationals of third countries who are authorised to work in the territories of the Member States are entitled to working conditions equivalent to those of citizens of the Union.’149 This sounds progressive, but the substance is meagre given that equal working conditions are largely uncontroversial in today’s Europe, as discussed in Chapter 15.4 on integration. Within the limits of human rights, the choice whether external borders shall be ‘open’ or ‘closed’ for labour migration rests with the political process. In a similar manner, the Charter differentiates between the intra-EU mobility of Union citizens and third country nationals. Think, by way of example, of an Albanian residing in Greece who wants to work in Slovakia. While Union citizens are entitled to free movement in such a scenario, Article 45(2) CFR maintains that similar rights ‘may be granted’ to third country nationals in accordance with the Treaties. The scope of intra-EU mobility is determined in the legislative process.150 The legislature may opt for a generous mobility regime but is not constitutionally obliged to do so. Distinctions between ‘citizens’ rights and the ‘human’ rights define many constitutions around the globe—with the rights to vote and unconditional entry and stay being prime examples of the insider privilege.151 EU primary law replicates this model.

145 CFR, recital 4 emphasises that the Charter responds to ‘changes in society, social progress and scientific and technological developments’. 146 See chs 12.2.4 and 13.2.3. 147 Case C-540/03 Parliament v Council EU:C:2006:429, para 59; as well as chs 14.2.3 and 16.2.3. 148 CFR , arts 52(2), (7) and the official Explanations [2007] OJ C303/17, 23 confirm that CFR , art 15(2) is to be interpreted in line with TFEU , arts 45, 49, 56. 149 CFR , art 15(3) (emphasis added). 150 As a ‘principle’, CFR , art 45(2) can be relied upon only indirectly, in line with art 52(5), once legislation has specified the conditions of free movement. 151 See Jo Shaw, The People in Question (Bristol UP 2020).

Distinguishing Union Citizenship from Third Country Nationals 37 A classic critique of the EU Treaties concerns what scholars of constitutional law describe as ‘overconstitutionalisation’, which occurs when primary law pre-empts policy choices, for instance in the single market and monetary policy.152 Judicial deference to political choices in economic and monetary issues may have mitigated this criticism in recent years.153 For our purposes, however, the argument highlights that the absence of ‘overconstitutionalisation’ on migratory matters can be described as a virtue from a theoretical perspective. The EU Treaties leave room for political choices.154 That change of direction in comparison to the Treaty design for the single market and monetary union may signal the maturity of the European project. Constitutional prescriptions give way to legislative leeway within the confines of human rights.

1.3.3 Overlap with Union citizenship Political debates, court judgments, and academic research are often dominated by path- dependency. Entrenched earlier patterns influence new developments until, possibly, a new paradigm is being established.155 Against this background, political decisions and the case law on migration can be described as a process of discovery, which gradually recognised the specificity of the Treaty regime for third country nationals. While the legislature was comparatively quick to adapt in line with the comments hereinafter, the Court started interpreting migration law in the 2010s. In doing so, judges were influenced by several dozen judgments dealing with grey areas of overlap between Union citizenship and the status of third country nationals which had been delivered previously. Four such areas of legal and conceptual intersection stand out. First, family members of Union citizens who have exercised the right to free movement benefit from generous legislation and Court judgments. Family members holding the passport of a third state are covered by these guarantees in the form of ‘derived rights’. A similar spillover occurs whenever European companies post workers from a third state to deliver a service in another Member State.156 A series of cases throughout the early 2000s increased the level of protection for family members of Union citizens. This process culminated in the Ruiz Zambrano judgment, before judges changed course and emphasised the limits of the earlier case law, as described in Chapter 14.3.1 on legal migration. Several of these follow- up judgments mentioned the different outlook of the EU migration law instruments,157 thereby recognising that third country nationals cannot be addressed from the angle of Union citizenship alone. Judges ‘discovered’ the specificities of the area of freedom, security, and justice in these rulings.

152 See Fritz Scharpf, ‘The European Social Model’ (2002) 40 JCMS 645; and Dieter Grimm, ‘The Democratic Costs of Constitutionalisation’ (2015) 21 ELJ 460. 153 See Nicole Scicluna, European Union Constitutionalism in Crisis (Routledge 2015) ch 5; and Loïc Azoulai, ‘The Court of Justice and the Social Market Economy’ (2008) 45 CML Rev 1335. 154 See also Kocharov (n 135) ch 3. 155 See generally James Mahoney, ‘Path Dependence in Historical Sociology’ (2000) 29 Theory and Society 507; for the CJEU see Susanne K Schmidt, ‘Who Cares about Nationality?’ (2012) 19 JEPP 8. 156 See ch 14.5.1. 157 See Case C-256/11 Dereci and others EU:C:2011:734, paras 71–72; Joined Cases C-356/11 and C-357/11 O and S EU:C:2012:776, paras 61–81; Case C-40/11 Iida EU:C:2012:2405, paras 78–81; and Case C-82/16 K.A. and others EU:C:2018:308, paras 44–46, 98–107.

38 Building an Area of Freedom, Security and Justice Secondly, the freedom-enhancing effect of enlargement was significant. Ever more states joined the European Union, which compromised around 500 million inhabitants on the eve of Brexit—approximately 130 million more than at the time of the Maastricht Treaty. Sixteen states including Austria, Finland, Poland, and Latvia, as well as Romania and Croatia joined the Union between 1995 and 2014, thus gradually extending free movement rights to an ever larger number of countries. Hence, people who would have had to apply for a visa only a generation ago can look for jobs across Europe nowadays, without the need for prior authorisation. To date, the prospect of free movement is an important motivation for states in the Western Balkans and Georgia to seek membership. However, the snowball effect of free movement via enlargement did not embody a general commitment to universal mobility, also considering that accession implied the export of control instruments to the new external borders.158 From a constitutional perspective, the attraction of membership reaffirms that it makes a difference whether you are an ‘insider’ or an ‘outsider’, precisely because the rules on third country nationals are different. Thirdly, the Court considerably enhanced the level of protection for former Turkish ‘guest workers’ and their families in a series of more than seventy judgments on the basis of an association agreement, which today’s EU had signed with Turkey in 1963. Judges assumed that these rules should be interpreted ‘so far as is possible’159 in light of the EU’s internal free movement regime. It seemed as if a ‘domino effect’ would extend the privileges of Union citizens to different categories of third country nationals.160 However, that proved to be a premature conclusion. Judicial dynamics were reversed when the Court distinguished the privileges of Union citizens from the status of Turkish nationals, while maintaining the protective credentials of the earlier case law. Chapter 17 will show that relations with neighbours moved away from the single market paradigm, instead embracing the specificities of the area of freedom, security, and justice. Fourthly, there remains room for subtle forms of overlap, approximation, and distinction between citizens’ rights and migration law. A case in point is the legislative ‘public policy’ exception for the denial or withdrawal of residence permits. The Court assumes that corresponding guarantees for third country nationals and Union citizens are to be interpreted similarly to start with, even though interpretation can result in differentiation, as we shall see in Chapter 10.3. Moreover, the relationship between case law concerning Union citizens and third country nationals is not a one-way street with the former having radiating effects on the latter. Similarly, the reverse scenario remains possible when judges emphasise the limits of free movement and equal treatment under the Free Movement Directive 2004/38/ EC, thereby effectively approximating Union citizenship to migration law.161 It was and is not a foregone conclusion that rules for third country nationals would follow the historic model of free movement.

158 See ch 11.3.2; and Rosemary Byrne, Gregor Noll, and Jens Vedsted-Hansen, New Asylum Countries? (Kluwer 2002). 159 Case C-303/08 Bozkurt EU:C:2010:800, para 20; and ch 17.4. 160 See Kees Groenendijk, ‘Recent Developments in EU Law on Migration’ (2014) 16 EJML 313, 321–24; and Wiesbrock (n 129) 85–89. 161 See Daniel Thym, ‘Supranational Courts in Europe’ (2021) 47 JEMS 4534, 4539–40.

Distinguishing Union Citizenship from Third Country Nationals 39

1.3.4 Tipping points in the legislative process While the Court is responsible for the ‘final’ interpretation of the Treaties, EU institutions have the ‘first’ word. For the political debate, the early 2000s were a foundational period. Negotiations were defined by occasionally heated debates, and the instruments adopted at the time presented a distinct outlook. Comparatively generous Commission proposals were altered or even rejected by the Council on several occasions. Unanimity meant that some governments exercised considerable influence. There were at least three symbolically important ‘tipping points’162 when Member States left their mark on the basic outline of EU migration law. First, conditions for family reunification were hardened during the legislative process. Member States, especially those with a sizeable community of former ‘guest workers’ and late colonial migrants, insisted on stricter provisions and optional clauses to protect domestic rules. Chapter 14.1.2 on legal migration will illustrate that Europeanisation served as a catalyst for informal horizontal policy transfer, thus promoting the introduction of restrictions in more and more Member States, including pre-departure language tests. These developments reversed the equality-based integration concept that had defined Union citizenship. Parallel negotiations of the Family Reunification Directive 2003/86/EC and the Free Movement Directive 2004/38/EC reinforced the impression that migration law was decoupled from the single market. Chapter 15.4 will explain that the Court accepted the legality of integration requirements in two landmark judgments, albeit subject to a hardship clause to the benefit of individuals. Secondly, the final version of the Long-Term Residents Directive 2003/109/EC considerably watered-down the promise of the Tampere Conclusions to ‘approximate’ the status of long-term residents to that of Union citizens. Chapter 14.4.5 on legal migration will mention that the rules on intra-European mobility for third country nationals require prior authorisation by the second Member State which can insist, amongst other things, on labour market tests and integration measures. Moreover, the legislature introduced caveats limiting equal treatment with nationals, instead of replicating the broad scope of Article 18 TFEU. To be sure, the Long-Term Residents Directive enhanced residence security and equal treatment. For our purposes, however, the departure from the original promise of alignment with Union citizenship stands out. Thirdly, the Commission proposed ambitious legislation promoting legal channels for economic migration in the early 2000s. The Council flatly rejected this initiative, as explained in Chapter 14.1, thereby laying the ground for lengthy consultation processes which finally resulted in the sectoral approach focusing on selected categories. The Treaty of Lisbon confirmed the EU’s competence for the harmonisation of labour migration, subject to a caveat in Article 79(5) TFEU that serves as a symbolic marker of national prerogatives. Entry and stay for economic purposes reaffirms the distinct outlook of EU migration law, precisely because it deviates from the single market model.

162 On the game theoretical concept of ‘tipping points’ see Thomas Schelling, The Strategy of Conflict (Harvard UP 1980).

40 Building an Area of Freedom, Security and Justice

1.4 Summary European migration law did not appear out of the blue but developed on the basis of political choices and pre-existing cooperation patterns. Inter-state collaboration emerged on the European continent during the early twentieth century, for refugees in the framework of the League of Nations and for migrant workers on the basis of bilateral agreements and within the ILO. While not all of these early initiatives proved successful, they provided a conceptual template for more robust cooperation after the Second World War. To highlight instances of international cooperation before the advent of European integration illustrates that governments recognised the reality of migratory movements in Europe and beyond. Indeed, the formation of the sovereign nation during the late nineteenth and early twentieth centuries was a period of heightened and highly stratified mobility. Mass emigration to the Americas and internal movements on the old continent coincided with fierce colonial domination of the Global South from where few people reached Europe. It was not self-evident that the creation of a common market would embrace the free movement of people, since international trade law has traditionally been limited to goods and services. The political determination of the Italian government was a decisive factor as to why the Treaty of Rome recognised the free movement of workers, which the EU institutions interpreted generously when adopting implementing legislation. Historic studies show that free movement was consciously limited to nationals of the Member States, although relations to the (former) colonies played a prominent role in the early years; the institutions established a generous internal mobility regime from which non-Europeans were deliberately excluded. That episode had widely been forgotten when the Treaty of Maastricht introduced ‘Union citizenship’ as a signpost for the political aspirations of ever closer union. It was during that period that the EU institutions started dealing with the entry and stay of third country nationals. Migration law entered the picture on the basis of a simple quid pro quo: the abolition of internal border controls within the Schengen area was considered to require ‘flanking measures’ compensating national authorities for the loss of control options at the internal borders. The evolution of the Treaty regime was complex and should primarily be studied by those interested in understanding how the rules we know today came into being. Originally, there were two parallel tracks: cooperation within the Schengen area among five Member States and the intergovernmental ‘third pillar’ with rudimentary coordination powers of the EU institutions. Both pathways were merged when the Treaty of Amsterdam integrated the Schengen acquis into the EU framework and established a supranational competence for law-making. Ever since, the institutions have been adopting regular directives and regulations on the entry and stay of third country nationals. The transformation of the Treaty regime came full circle with the Treaty of Lisbon, which eliminated the remnants of intergovernmental decision-making and established a broad set of supranational competences in today’s Articles 77–80 TFEU. The Lisbon Treaty also strengthened the autonomy of migration law as a policy filed in its own right, as an integral part of the area of freedom, security, and justice. Unlike in the case of the single market or monetary union, the EU Treaties refrained from enshrining a specific policy design within primary law. The choice for how ‘open’ or ‘closed’ the external borders shall be for third country nationals is determined in the legislative process, within the confines of human rights.

Summary 41 In the early years, the European Council guided decision-making in the form of five-year programmes adopted in Tampere, The Hague, and Stockholm. These programmes reflected distinct political priorities in light of changing circumstances, such as the fight against terrorism. They have lost momentum in recent years, partly as a result of profound disagreement on the way forward. These different positions are reflected in the public discourse by symbolic markers embodying the position of the speaker: commitment to ‘a Europe that protects’ contrasts with criticism of ‘fortress Europe’. EU institutions have to hammer out compromises on how to reconcile the diverse and potentially contradictory policy objectives in Articles 77–80 TFEU. The first decade of decision-making saw several important tipping points where the Council left its mark on the design of migration law by introducing rules that differed markedly from the freedom-enhancing rationale of Union citizenship and the single market paradigm. We shall discuss elsewhere the extent to which human rights and the Court may serve as a counterpoint.

2

Institutional Prerogatives and Decision-making Legal rules do not materialise out of thin air. Legislation is the result of political compromises and occasional struggles on the road to be taken. That is why an institutional framework is more than a formalistic skeleton: it assigns privileges to shape the outcome to some actors, while limiting the influence of others. Corresponding mechanisms are particularly relevant for the EU which, as a supranational polity, has always employed institutional design for unity building. The preamble of the former Treaty establishing the European Coal and Steel Community declared the objective ‘to lay the bases of institutions capable of giving direction to the future common destiny’.1 Institutions can be essential for the gradual approximation of national interests and the formation of a collective impulse. At the same time, the institutional architecture is subject to change, contestation, and challenges. As mentioned previously, adaptation of the Treaty regime to the orthodoxy of the supranational integration method occurred gradually; the transformation of hitherto intergovernmental decision-making came full circle with the Treaty of Lisbon. Against this background, this chapter will describe institutional prerogatives and the entrenched organisational culture, thus presenting a contextually and theoretically embedded analysis of everyday operations in the field of migration. Increasing criticism of the policy output, notably on asylum, and protracted political debates about the way forward render such contextual inspection pertinent. Our analysis begins with factors that help to explain, from the perspective of political science, both the success of the European venture and contemporary challenges at a time when the migration law instruments directly affect salient issues (2.1). A review of the official tasks and the established informal workings of the Commission, the European Council, the Council, and the European Parliament will focus on their position on migratory matters (2.2). Awareness of everyday practices is central for the appreciation of interinstitutional decision- making on the migration law instruments discussed throughout this book (2.3). A distinct feature of justice and home affairs is the non-participation of Denmark and Ireland (2.4).

2.1 Driving Forces behind Europeanisation European integration has fascinated observers from the beginning. It seemed unlikely that sovereign states would voluntarily ‘surrender’ their sovereignty to supranational institutions. Nevertheless, today’s EU proved surprisingly robust and expanded considerably over the decades. What started off as a primarily economic undertaking has gradually incorporated ever more substantive competences. Migration law is a telling example of the 1 Treaty establishing the European Coal and Steel Community (adopted 18 April 1951, expired 1 July 2002) 1754 UNTS 594, preamble recital 5. European Migration Law. Daniel Thym, Oxford University Press. © Daniel Thym 2023. DOI: 10.1093/oso/9780192894274.003.0003

Driving Forces behind Europeanisation 43 astounding success in embracing policy areas that are usually considered a manifestation of state sovereignty. Legal experts occasionally take the binding force of EU legislation for granted and expect— rightly so from a legal perspective—that political actors and administrative authorities should respect it. The precarious state of asylum policy reminds us that legal obligations are not always applied effectively and that difficulty can arise in reforming dysfunctional rules. Key findings within the political science literature help legal experts understand underlying dynamics. Different explanatory factors exist as to how EU migration law came about in the first place (2.1.1) and why the policy output fluctuates between protective and restrictive elements (2.1.2). Enhanced politicisation complicates the search for compromise (2.1.3) and explains both the criticism of a democratic deficit and the danger of institutional blockage (2.1.4).

2.1.1 The choice for Schengen as an example For legal analyses, the roles of the ‘good’ and the ‘bad’ guy are often straightforward. Supranational institutions pursue the common good and defend the rule of law, while Member States insist on national interests and try to circumvent their obligations. Political science presents us with a nuanced picture, fed by an in-built scepticism against the normative self-sufficiency of the law. Why would the states consent to extend the EU’s powers in successive Treaty amendments and the rights of migrants via legislation? Why do most domestic authorities and courts apply the latter despite the supranational institutions being ill-equipped to enforce it? Basic arguments of the political science literature can be elucidated by the origins of Schengen, which was traced in Chapter 1.2. At the heart of most arguments lies the somewhat artificial juxtaposition of two theoretical schools on how to explain EU integration in line with basic assumptions of international relations theory. On the one hand, intergovernmental (or realist) positions emphasise the national interest.2 France and Germany considered it to be in their mutual interest to prevent a lapse into economic protectionism and to promote further liberalisation by means of the single market programme; in the slipstream of these initiatives, internal border controls were abolished in the Schengen area.3 States ‘selfishly’ realised the venture in an intergovernmental setting. Over the years, interior ministers understood that Schengen allowed them to promote their agenda, thus spearheading the spread of control instruments. Once again, Europe appeared as the ‘rescue of the nation state’,4 by allowing governments to pursue objectives they could not effectively realise alone. Involvement of the supranational institutions was a solution to a prisoners’ dilemma, with supranational supervision ascertaining that others do not freeride by disrespecting common rules.5 Gradual supranationalisation of justice and home affairs can be rationalised on that basis.6 2 See generally Mark A Pollack, ‘Realist, Intergovernmentalist, and Institutionalist Approaches’ in Erik Jones and others (eds), The Oxford Handbook of the European Union (OUP 2012) 3. 3 See Andrew Moravcsik, The Choice for Europe (Ithaka Press 1998) 253–60; and Ruben Zaiotti, Culture of Border Controls (University of Chicago Press 2011) 6–10. 4 Alan S Milward, The European Rescue of the Nation State (2nd edn, Routledge 2000). 5 Moreover, smaller countries have traditionally insisted on strong supranational institutions to counterbalance potential dominance by bigger Member States. 6 See Stephan Stetter, ‘Regulating Migration’ (2000) 7 JEPP 80.

44 Institutional Prerogatives and Decision-making On the other hand, various strands of neo-functionalist, historical or sociological institutionalism, and constructivist analyses question the predominance of a seemingly unitary national interest.7 Instead, they emphasise spillover effects of earlier decisions and the role of domestic and private actors, other than governments, in influencing the policy outcome. According to neo-functionalists, the single market and the move towards political union provoked the abolition of internal border controls, thereby requiring states to cooperate on external borders, visas, and asylum. Even though interior ministries had originally been critical of ‘Schengen’, regular exchange among themselves promoted the emergence of a collegiate spirit and the transformation of control practices, which will be discussed in Chapter 12.1.3. The normative ideal of integration supported the gradual accession of ever more countries to the Schengen area. In the analysis of policy developments, both perspectives can be combined. Moreover, the relative weight of interests, normative ideas, and the institutional framework can change over time and according to the subject matter.8 For our purposes, these contrasting views on the formation of the Schengen area emphasise that our approach to migration law should be wary of false dichotomies. EU integration is about normative ideals and national interests at the same time. While the period after the cold war was defined by economic liberalisation, an optimistic outlook on globalisation, and an upsurge of idealism, state interests appear to have gained ground recently, also reflecting the changing geopolitical environment. Nevertheless, neither position explains the policy output single-handedly; explanatory factors will always overlap.

2.1.2 Continuity of mixed results in migration policy The public discourse tends to highlight symbolic markers that embody the position of the speaker: criticism of ‘fortress Europe’ stands for generous entry rules, whereas emphasis on a ‘Europe that protects’, or ‘European sovereignty’, accentuates migration management.9 Interaction of these outlooks has characterised EU migration policy from the beginning, although it is a new phenomenon that the debates make front-page news.10 We have seen in Chapter 1.2.1 that the intergovernmental set-up of the Schengen area was motivated, in part at least, by the desire to exclude the supranational institutions, which, at the time, focused on cross-border mobility rather than migration control. Dominance of interior ministers was a defining feature well into the early 2000s, on the basis of the Treaty of Maastricht’s intergovernmental ‘third pillar’ and, later, incomplete supranationalisation under the Treaties of Amsterdam and Nice. These early years have been described as a period of ‘securitisation’.11 Interior ministries used intergovernmental collaboration with like-minded peers to promote their agendas. Doing so can be rationalised as ‘venue shopping’, with governments side-stepping national

7 See Antje Wiener and Thomas Diez (eds), European Integration Theory (3rd edn, OUP 2019). 8 See Florian Trauner and Ariadna Ripoll Servent, ‘The Analytical Framework’ in Florian Trauner and Ariadna Ripoll Servent (eds), Policy Change in the Area of Freedom, Security and Justice (Routledge 2015) 11. 9 See ch 1.2.6. 10 The initial decision to abolish internal border controls was hardly noticed by the media; see Angela Siebold, ZwischenGrenzen (Ferdinand Schöningh 2013) 40–44. 11 See ch 4.3.4.

Driving Forces behind Europeanisation 45 parliaments and the public discourse, which, at the time, hardly took notice of the supranational developments.12 Such ‘two-level games’ have always existed, not only in the field of migration: national governments employ European fora to tame national politics and to support unpopular reforms.13 In a similar fashion, interior ministers ‘escaped’ to Europe to promote migration control. The historic background of the restrictive early years was what might be termed Europe’s first ‘asylum policy crisis’ when the fall of the Iron Curtain coincided with an increase in asylum applications. Neo- functionalist and constructivist accounts generally assumed that the supranationalisation of decision- making would strengthen liberal tendencies. The Commission, the European Parliament, and the Court of Justice were expected to champion individual rights, mirroring internal free movement. While the distinct role of these institutions will be discussed separately, contemporary literature recognises that the original expectation was only met to a limited extent for several reasons. To start with, European migration policy had never been defined by a clear-cut ‘race to the bottom’. The main destination countries mostly succeeded in ‘uploading’ domestic standards to the EU level. Doing so avoided changes to national legislation and prevented other countries from ‘competing’ by means of lesser levels of protection.14 The trend towards migrants’ rights gained momentum when the powers of the supranational institutions expanded. They are shielded, to a degree, from domestic scrutiny and often promote an idealistic vision of cross-border movements.15 Comments throughout this book will highlight the protective credentials of many pieces of legislation. This move towards moderately liberal policies was supported by the depoliticised environment of the early law-making, which implied greater influence of non-governmental organisations (NGOs).16 Nevertheless, comments that follow will demonstrate how the Commission and the Parliament struggled to shape migration policies. An important factor was the late move towards comprehensive supranationalisation under the Treaty of Lisbon. Member States had defined core legislation before the European Parliament gained institutional leverage. The intergovernmental origins of migration law had created a path-dependency follow-up legislation found difficult to deviate from.17 On the whole, the effects of Europeanisation cannot be expressed in a binary manner of either securitisation or individual rights. The output varies considerably over time and between policy areas.18 In the field of legal migration, national governments reined in the Commission’s proposals to adopt broad and comparatively generous directives on economic migration, family reunification, and long-term residents, which, nevertheless, establish

12 See Virgine Guiraudon, ‘European Integration and Migration Policy’ (2000) 38 JCMS 251; and Sandra Lavenex, The Europeanisation of Refugee Politics (Ashgate 2002). 13 European integration, including the European Convention on Human Rights, fostered democracy and (capitalist) economic reform in the face of the Communist ‘threat’; see Tony Judt, Postwar. A History of Europe since 1945 (Penguin 2005) pt I. 14 See Natascha Zaun, EU Asylum Policies (Palgrave 2017). 15 See Gallya Lahav, Immigration and Politics in the New Europe (CUP 2004) ch 5; and Eiko Thielemann and Natascha Zaun, ‘Escaping Populism: Safeguarding Minority Rights’ (2018) 56 JCMS 906. 16 See ch 2.3.6; and Andrew Geddes, The Politics of Migration and Immigration in Europe (1st edn, Sage 2003) 144. 17 See Florian Trauner and Ariadna Ripoll Servent, ‘The Communitarization of the Area of Freedom, Security and Justice’ (2016) 54 JCMS 1417. 18 See Saskia Bonjour, Ariadna Ripoll Servent, and Eiko Thielemann, ‘Beyond Venue Shopping and Liberal Constraint’ (2017) 25 JEPP 409.

46 Institutional Prerogatives and Decision-making important safeguards for individuals. They will be described in Chapters 14 and 15. In the field of asylum, the Parliament prioritised internal procedural safeguards, including for vulnerable groups, while being more lenient when it came to border controls. EU asylum policy can be described as a combination of liberal internal rules and external closure, as explained in Chapter 13.1 on asylum. Recent emphasis on the external dimension, described in Chapter 18, can be conceptualised as a remake of ‘venue-shopping’. Cooperation with third states provides an avenue to bypass internal constraints.

2.1.3 From ‘permissive consensus’ to ‘constraining dissensus’ European integration depends—much more than the nation-state—on the law and judges to ensure the effective realisation of common policies, since political mechanisms and administrative capabilities for conflict resolution and rapid response are less developed. Joseph Weiler famously described the formula for success of the supranational integration method as a dichotomy between enhanced legal supranationalism and limited political integration.19 EU integration prospered not despite but because of consensual decision- making, which allowed governments to keep a grip on the policy output: the move towards qualified majority voting did not fundamentally reverse the situation as a result of the prevailing consensus culture in line with subsequent comments. A standard explanation for the success of the EU has been described as ‘permissive consensus’: citizens generally agreed with the integration process but did not care much about the policy output; the EU dealt with critical structural questions and left the salient ‘bread and butter’ topics to the Member States.20 Low participation rates in European elections, combined with the visibility of domestic matters during the election campaign, embody a largely depoliticised integration process. Decisions were taken behind the closed door of the diplomatic negotiation room, where inter-state bargaining took place and a technocratic style of argumentation prevailed. Until this day, the strength of the EU institutions is the search for compromise on the basis of a permanent ‘grand coalition’ of political movements and national interests.21 For many years, EU migration policy was defined by a relatively low degree of public attention, which supported moderately liberal policy outcomes, notwithstanding sceptical public opinion.22 It is a defining feature of European politics that the ‘permissive consensus’ has given way to a ‘constraining dissensus’.23 Important areas of decision-making have become what political scientists refer to as salient issues which define the party-political contest and feature prominently in the public debate.24 This trend had been in the making since the Treaty of Maastricht, when it first became apparent that the EU institutions dealt with questions involving value judgments or having redistributive effects. The euro crisis of the early 2010s and the peak of asylum applications during 2015/16 reinforced the politicisation of EU 19 See Joseph HH Weiler, ‘The Transformation of Europe’ (1990/91) 100 Yale LJ 2403, 2410–31. 20 See Stuart Scheingold and Leon N Lindberg, Europe’s Would-Be Polity (Prentice-Hall 1970). 21 See Giandomenico Majone, Dilemmas of European Integration (OUP 2005); and Marco Dani, ‘Rehabilitating Social Conflicts in European Public Law’ (2012) 18 ELJ 621, 628–32. 22 See ch 4.2.3. 23 Liesbet Hooghe and Gary Marks, ‘A Postfunctionalist Theory of European Integration’ (2009) 39 Brit J Pol Sci 1. 24 cf Andrew Moravcsik, ‘In Defence of the “Democratic Deficit”’ (2002) 40 JCMS 603.

Driving Forces behind Europeanisation 47 affairs generally and migration specifically.25 Citizens are concerned with what they perceive, rightly or wrongly, as the impact of migration on collective identities, the labour market, or public services; moreover, migration often serves as a projection sphere for wider unease at globalisation.26 Anyone following the discourse on EU migration policy will realise an inbuilt asymmetry between numbers and public attention. In quantitative terms, most migration occurs legally, with third country nationals entering as tourists, workers, family members, or otherwise. Nevertheless, public attention has focused on asylum following 2015/16. The interdisciplinary Chapter 4.2.3 and 4.3.3 will explain that the use of semantic framing such as ‘smuggling’ or quasi-natural metaphors like ‘wave’ or ‘flow’ reinforce the impression that states have lost control. Interdisciplinary research acknowledges that such governance deficits, real or perceived, force political actors to concentrate on countering the impression that migration policies are out of control.

2.1.4 The danger of institutional blockage Increased politicisation is particularly worrying for the EU. Politicisation does not only, like in the nation state, affect the policy outcome in line with empirical findings that politicisation will usually result in more restrictive migration policies.27 Political salience of migratory movements has repercussions for the ability of the EU institutions to agree on the way forward. Politicisation weakens the weight of the supranational institutions and can harden governmental positions, thus complicating policy reform. Asylum is a case in point. For many years, classic destination countries in Western and Northern Europe had dominated decision-making, since they cared more about the draft legislation and had the technical experience on how to organise asylum systems.28 States in Southern Europe initially took a lesser interest due to lower numbers, widespread informality, and a lack of experience as destination countries.29 ‘New’ Member States in Central and Eastern Europe were obliged to transpose supranational rules whose contents they had not influenced effectively before or after enlargement.30 Events in recent years have changed this overall setting profoundly. Member States which may have followed decision-making rather passively in the past started showing a genuine interest—as the activities of the Visegrád countries illustrate. The second phase of asylum policy harmonisation during the early 2010s saw a confrontation of Northern and Southern governments on the contents of the Dublin III Regulation, which, as a result of the disputes, continued existing rules path-dependently instead of addressing the structural weaknesses of the Common European Asylum System.31 Reform proposals 25 See Giandomenico Majone, Rethinking the Union of Europe Post-Crisis (CUP 2014); and Edgar Grande, Tobias Schwarzbözl, and Matthias Fatke, ‘Politicizing Immigration in Western Europe’ (2019) 26 JEPP 1444. 26 To recognise the elements of perception and projection does not unmake the political effects. 27 See Marc Howard, The Politics of Citizenship in Europe (CUP 2009) ch 3; and Rhonda Givens and Adam Luedtke, ‘The Politics of European Union Immigration Policy’ (2004) 32 Policy Stud J 145. 28 See Zaun (n 14). 29 See Andrew Geddes and Peter Scholten, The Politics of Migration and Immigration in Europe (2nd edn, Sage 2016) chs 8, 10. 30 See Rosemary Byrne, Gregor Noll, and Jens Vedsted-Hansen, ‘Understanding the Crisis of Refugee Law’ (2020) 33 Leiden J Intl L 871, 881–88. 31 See ch 13.1.3 and 13.3.2.

48 Institutional Prerogatives and Decision-making of 2016 and 2020 have been blocked despite desperate efforts to find a compromise. EU institutions appear unable to broker a compromise between juxtaposing interests, which are hardened as a result of public scrutiny. Note that the underlying danger of impasse differs between policy fields. While asylum is the prime example of politicisation rendering legislative reform difficult if not impossible, other rules will rarely feature in the public debate. Databases, visa requirements, or cooperation with third states may be highly relevant for expert analyses but they do not usually become salient in the democratic contest. With regard to these subject matters, EU institutions can continue what they have traditionally done with great success: search for compromise packages among different actors, interests, and values on the basis of a technocratic style of argumentation. The situation can be different for migrant integration, family reunification, and economic migration, which interrelate with redistributive social policies or debates about collective identities, including the role of Islam. It is no coincidence that the level of ambition of the supranational legislation on economic migration and integration was cut back considerably by the Member States. Rumour has it that the Commission has refrained from proposing a revision of the Family Reunification Directive, since it does not want to provide national governments with an opportunity to insist on stricter conditions during the negotiations.32 Directive 2003/86/EC has not been amended ever since its adoption almost thirty years ago. Denial of legislative change can be an instrument to forestall toxic debates about migration, even though it is problematic from the perspective of democratic theory if statutory rules are effectively set in stone.33 A defining feature was the appearance of populist parties, which employ migration as a strategic tool to mobilise and extend their support base. The campaign leading up to the Brexit referendum, hostility towards relocation of asylum applicants by the Visegrád countries, disputes about search and rescue, and mobilisation against family reunification with spouses from Muslim countries are prime examples of how populist and right-wing parties at the domestic level have an indirect impact on supranational decision-making. Such policy entrepreneurs frame the policy debate, often through wrong or exaggerated claims.34 Divisive topics dominate the public discourse and leave the public opinion with the impression that governments have lost control and that migration has primarily negative repercussions for European societies. The resulting scenario can be dangerous for the European Union, whose legitimacy has traditionally rested on the ‘output’ of successfully policies, not on the ‘input’ of strong democratic credentials.35 The track record in problem-solving capacities is not replicated in all areas of migration policy nowadays. Visibility of the failure to reform asylum legislation, in particular, threatens to become a stigma. In a worst-case scenario, popular contestation could result in a vicious circle of further politicisation preventing the necessary

32 Various governments, not least from the Netherlands, have called upon the Commission to table a proposal for legislative reform. 33 That limitation is rarely described as a deficit in the field of migration, unlike in monetary union; see Catherine Colliot-Thélène, ‘What Europe Does to Citizenship’ in Damian Chalmers and others (eds), The End of the Eurocrats’ Dream (CUP 2016) 127. 34 See Vladislava Stoyanova and Stijn Smet (eds), Migrants’ Rights, Populism and Legal Resilience in Europe (CUP 2022). 35 See generally Fritz Scharpf, Governing in Europe (OUP 1999).

Influence on the Policy Output 49 reform. Institutional blockage might reinforce, in turn, unease at output deficits and render the policy debate even more toxic.36

2.2 Influence on the Policy Output Public debates occasionally use ‘Brussels’ as a synonym for EU politics. That is appropriate insofar as the institutions are essential for the European project. They provide the organisational infrastructure for the design of common rules and supported gradual expansion to ever more subject areas. Awareness of institutional idiosyncrasies and everyday practices is essential for anyone studying the evolution of European migration law. The outlook of supranational actors is defined by entrenched organisational patterns and a distinct ‘institutional culture’.37 Our comments begin with the unique position of the Commission in terms of agenda-setting and supervision (2.2.1), which occasionally compete with the strategic vision of heads of state or government in the European Council (2.2.2). Meetings of national ministers and their staff within the Council and preparatory bodies are the epicentre of political decision-making (2.2.3), notwithstanding the essential role of the European Parliament as a counterweight (2.2.4).

2.2.1 Commission The Commission is commonly described as the ‘motor of integration’, together with the Court of Justice. Political scientists emphasise that its ‘agenda-setting’ function gives the Commission considerable influence over the conceptual design and the legislative detail; the Council and the European Parliament will not usually reverse the structure of the Commission’s proposals fundamentally. The legal foundation of the agenda-setting power is the monopoly of initiative for new legislation.38 The Commission has traditionally used this monopoly skilfully by presenting itself as the source of technical expertise and as a repository of knowledge.

2.2.1.1 Civil servants between expertise and politics Human resources are the backbone of the influence. The Commission comprises the college of Commissioners and more than 30,000 civil servants working for various Directorates General (DGs). Although the college habitually encompasses one person per Member State, Commissioners are fully independent in the exercise of their function and do not represent ‘their’ country.39 Independence and a commitment to ‘promote the general interest of the Union’40 are firmly embedded in the institutional ethos. Recent years have seen an internal reorganisation: the President has gained influence, and several Vice Presidents 36 See Daniel Thym, ‘The “Refugee Crisis” as a Challenge of Legal Design and Institutional Legitimacy’ (2016) 53 CML Rev 1545, 1571–72. 37 On theoretical models see Peter A Hall and Rosemary CR Taylor, ‘Political Science and the Three New Institutionalisms’ (1995) 44 Political Studies 936. 38 See TEU, art 17(2); in justice and home affairs, it gained this monopoly five years after the entry into force of the Treaty of Amsterdam in most subject areas. 39 ibid art 17(3). 40 ibid art 17(1).

50 Institutional Prerogatives and Decision-making coordinate the work of the portfolios on related themes. One Commissioner oversees DG HOME, which serves as an embryonic interior ministry. She will not, however, direct migration policy single-handedly. Decisions are taken by the college as a whole, if necessary by majority. The influence of individual Commissioners is based on political skills, the power of persuasion, and an edge in information. Commissioners depend on the input of civil servants who are generally well-trained, speak different languages, and are defined by a supranational ethos and a broadly cosmopolitan worldview.41 They harbour the knowledge base and the technical expertise, thus exercising considerable influence on the contents of legislative proposals. Even so, the thematic scope of migration law exerts pressure on the supranational bureaucracy, which constantly has to design new proposals on diverse subject matters, while keeping an oversight of existing policies and supporting effective implementation.42 Protracted compliance deficits, notably in the field of asylum, indicate that the Commission does not always live up to this challenge effectively. Indeed, supervision of the correct implementation is a critical function, which is often referred to as the ‘guardian of the Treaties’. The sharpest instruments are infringement proceedings, which have been used discreetly on migratory matters—notwithstanding prominent proceedings against Hungary outlined in Chapter 3.3.2. As an additional instrument, agencies and financial support can foster compliance. While the Commission plays a prominent role in allocating the budgetary funds, it cannot usually decide autonomously how to spend money.43 Agencies like Frontex and the Asylum Office, discussed in Chapter 8, are similarly not direct subsidiaries; they are controlled by an executive board in which Member States decide jointly with the Commission.

2.2.1.2 Technocratic posture on migration law On the whole, one may hardly describe the Commission as a ‘government’, since it combines political executive powers with administrative functions and shares political leadership with the European Council.44 The institutional ethos is to be an honest broker and an incarnation of the common interest; it often hid political ambitions behind the surface of technical expertise and left strategic guidance to the heads of state or government.45 In recent years, the Commission has struggled to define its role between the tradition as technocratic expert body and the self-declared objective to be ‘very political’.46 The ensuing tension between political ambition and technocratic authority is particularly pronounced in migration law. Increasing politicisation complicates, as we have seen, the accommodation of countervailing interests on the basis of technical expertise alone, as illustrated by the failure to reform asylum legislation. Inspection of previous initiatives demonstrates how the Commission usually responds to politicisation. It may present ambitious proposals at first but adapts when realising that the Council is unwilling to go along. For instance, the Commission changed course with 41 See Hussein Kassim and others, The European Commission of the Twenty-First Century (OUP 2013) chs 2–4. 42 See generally Brigid Laffan, ‘From Policy Entrepreneur to Policy Manager’ (1997) 4 JEPP 422. 43 See ch 7.4. 44 See Deirdre Curtin, Executive Power of the European Union (OUP 2009) chs 4–5. 45 See Philipp Dann, ‘The Political Institutions’ in Armin von Bogdandy and Jürgen Bast (eds), Principles of European Constitutional Law (2nd edn, Hart Publishing 2009) 233, 257–61. 46 Former President Jean-Claude Juncker, ‘Time for Honesty, Unity and Solidarity’ (State of the Union Address, 9 September 2015).

Influence on the Policy Output 51 technocratic pragmatism when Member States flatly rejected a high level of harmonisation for economic migration and refused to approximate the status of long-term residents to that of Union citizens.47 It also stood ready to repeal the original draft legislation and to table new proposals on divisive topics such as family reunification, taking up governmental preferences.48 More recently, it respected red lines when presenting the latest initiative to reform the Dublin system, which abstained from the mandatory relocation of asylum seekers.49 Some may find such realpolitik confusing, with the Commission trimming the sails to the wind. Nevertheless, elasticity suits the traditional function as a ‘motor of integration’. Its priority is to advance the European project.

2.2.2 European Council ‘Summits’ of heads of state or government render the often Byzantine world of the Brussels- based institutions somewhat more accessible. The European Council takes crucial political decisions, mitigates crises, tries—often successfully—to bridge seemingly irreconcilable differences, and responds to unforeseen developments. Doing so corresponds to its function to ‘provide the Union with the necessary impetus’ and to ‘define the general political directions and priorities’.50 Heads of state or government have done so repeatedly ever since their first meeting at The Hague in 1969. Milestones of the integration process were associated with summits in towns across the continent, until Brussels became the regular meeting place. It is convened at least twice a year, although the interval can be much shorter when political events so require. Public visibility of the ‘summits’ corresponds to the political character of the European Council, which only became a formal institution with the Treaty of Lisbon. It does not take formal decisions with the exception of constitutional matters, such as the appointment of top personnel or Treaty change. Such formal decisions require a quasi-unanimous consensus in the sense of no-one disagreeing with the outcome.51 If the European Council deals with legislative files, it does so informally. Involvement of heads of state or government does not prevent the Council from adopting draft legislation by qualified majority in the ordinary legislative procedure. In relations with the other institutions, the European Council remains primus inter pares. Influence stems from the political weight of an intergovernmental consensus, not legal authority. While the other institutions will usually follow the strategic guidance of heads of state or government, they eagerly defend prerogatives in the design of secondary legislation. Latent interinstitutional tensions and occasional turf wars define the Brussels outlook. They can be described positively as an element of horizontal checks and balances.52 Episodes of 47 See ch 1.3.4. 48 See ch 14.1.2; and Georgia Papagianni, Institutional and Policy Dynamics of EU Migration Law (Martinus Nijhoff 2006) 241–45. 49 See ch 13.1.5; and Daniel Thym, ‘Never-Ending Story?’ in Daniel Thym (ed), Reforming the Common European Asylum System (Nomos 2021) 11, 13–15. 50 TEU, art 15(1). 51 See ibid art 15(4) in line with the international practice of formal consensus, notably within the World Trade Organization, which differs from legal unanimity insofar as unanimity commands a positive vote in favour, not the absence of active objection. 52 See Wolfgang Wessels, ‘The European Council’ in Takis Tridimas and Robert Schütze (eds), The Oxford Principles of European Union Law, vol I (OUP 2018) 490.

52 Institutional Prerogatives and Decision-making heightened tensions between the Commission and the European Council, for instance on the level of ambition of the Stockholm Programme,53 usually give way to pragmatic cooperation. To do so is in their mutual interest; one cannot succeed without the other. The Commission will often cede, as we have seen, publicly visible leadership to heads of state or government, while trying to shape the policy substance by using the strength inherent in the technocratic knowledge base. Raising a matter to the level of the European Council can help overcome political deadlock, since it involves a change of perspective. Narrow technical considerations give way to strategic reflection about the direction of the European project, and cross-sectoral compromises may emerge, based on reciprocal give-and-take on diverse subject matters. At the same time, involvement of heads of state or government runs the risk of deepening tensions through clashes at the highest level. It remains a question of diplomatic skills when to engage the institution.

2.2.3 Council Notwithstanding the official description of the Council as the collective of national ministers, the institution has many faces that effectively turn it into a quasi-permanent structure, convening on an almost daily basis in the Justus Lipsius building in Brussels. Ministers will not usually discuss the legislative detail. They leave this job to the numerous preparatory bodies, in which civil servants discuss in depth how to formulate specific articles. In justice and home affairs, there are more than a dozen working parties, doing the essence of the legislative discussions.54

2.2.3.1 The ‘backroom’: working parties of national civil servants Participants in the working parties usually fly to Brussels from national capitals and tend to have a high level of expertise. Anyone participating in the debate learns that technical knowledge plays a critical role and that the standing (and influence) of individual actors mainly depends on the quality of the statements.55 That is not to say that civil servants decide autonomously; they receive instructions from national ministries. Meetings take place behind closed doors, even though some documents can be accessed via the public register of the Council, especially after the adoption of the instrument in question56 (the latest documents are often leaked and can be found online57). Strategic matters are discussed in the ‘Strategic Committee for Immigration, Frontiers and Asylum’ (SCIFA), which supports coherence among the legislative portfolios and can help overcome disputes as a body composed of higher-ranking officials.58 53 See ch 1.2.6. 54 For an updated list see www.consilium.europa.eu/en/council-eu/preparatory-bodies/?filters=2023 (accessed 1 March 2023). 55 The author was a stagiaire with the German Permanent Representation for six months during 2004, in the framework of the Judicial Service Traineeship (Referendariat). 56 See www.consilium.europa.eu/en/documents-publications/public-register (accessed 1 March 2023); I recommend to search via the document number. 57 See www.statewatch.org (accessed 1 March 2023). 58 See ‘Strategic coordination of the work of the Council preparatory bodies in the area of migration’ (Council doc 12516/15, 1 October 2015).

Influence on the Policy Output 53 Working parties report to the Committee of Permanent Representatives (COREPER), which prepares the gatherings of the Justice and Home Affairs Council and serves as a diplomatic clearing house; political disputes are often resolved at COREPER level.59 Some pieces of legislation may never be discussed by ministers in the Council, as they are adopted as ‘A points’ without debate.60 It has become standard practice for the Justice and Home Affairs Council to meet over two days to allow interior ministers to focus on migration and security on one day, while ministers for justice will discuss criminal matters and judicial cooperation the other day.61 Some business has to be coordinated with other Council formations, especially on external relations. Meetings of the ministers and preparatory bodies are coordinated by the Council Presidency, which rotates every six months. By tradition, it serves as an honest broker and exercises little political leadership.62 The Presidency is supported by the Council Secretariat, which, like the Commission’s Directorates General, often plays a critical role behind the scenes.

2.2.3.2 Consensus culture in the shadow of majority voting Transition towards qualified- majority voting featured prominently in several Treaty amendments and finally became ubiquitous in migration law under the Treaty of Lisbon.63 A qualified majority requires a double majority comprising at least 55 per cent of the Member States (i.e. 15 out of 27) which jointly represent at least 65 per cent of the population of the Union.64 The widespread emphasis on majority voting among external observers does not imply, however, that conflictual voting is standard practice. Deliberations in the Council and preparatory bodies are defined, to the contrary, by an entrenched consensus culture. Discussions do not come to a standstill only because the double majority requirement would be met if the Presidency called a formal vote. Negotiations among the Member States habitually strive to take everyone on board.65 Studies of sociological institutionalism inform us that an esprit de corps is essential for internal cohesion and facilitates the resolution of conflicts; new members are socialised with the consensus culture, thus preserving it over time.66 Nevertheless, prevalence of compromise-building does not mean that the constitutional switch towards more qualified majority was irrelevant.67 Empirical studies demonstrate that the behaviour of national representatives changes when they cannot simply block decisions by means of a veto. The ‘shadow of the vote’ renders negotiations more flexible; the Council’s institutional ethos necessitates that one does not insist on minor questions.68 Similarly, there will typically be 59 See Papagianni (n 48) 221–34. 60 See Annex to Council Decision 2009/937/EU adopting the Council’s Rules of Procedure [2009] OJ L325/35, arts 3(6), 19, with later amendments. 61 De facto, there are two sub-formations on justice and home affairs. 62 See Christof Roos, ‘The Council and European Council in EU Justice and Home Affairs Politics’ in Ariadna Ripoll Servent and Florian Trauner (eds), The Routledge Handbook of Justice and Home Affairs Research (Routledge 2018) 421, 427–28. 63 See ch 1.2.4. 64 A blocking majority must include at least four Member States, thus precluding, by way of example, Germany, France, and the Netherlands, which jointly represent 37.4 per cent of the population, from exercising a veto; see TEU, art 16(4). 65 See Fiona Hayes-Renshaw, ‘The Council of Ministers’ in Dermot Hodson and John Peterson (eds), Institutions of the European Union (4th edn, OUP 2017) ch 4. 66 See Jeffrey Lewis, ‘The Janus Face of Brussels’ (2005) 59 IO 937. 67 Do not confuse the generic use of the term ‘consensus’ as an expression for a generally accepted outcome with the formal consensus requirement for decisions of the European Council, discussed previously. 68 See Jonathan Golub, ‘In the Shadow of the Vote?’ (1999) 53 IO 733.

54 Institutional Prerogatives and Decision-making strong pressure to reach a compromise, also on sensitive topics. By way of example, the former Dublin II Regulation (EC) No 343/2003 was adopted because Greece and Italy eventually gave in to a compromise proposal.69 A similar agreement has not emerged on the successor instrument so far, despite years of debate. It is in the nature of negotiations that participants search for allies to increase leverage. Such coalitions are rarely formalised, although special links, common convictions, and shared problems may result in loose formations.70 The Benelux countries, Nordic states, and, more recently, the Visegrád group of Poland, the Czech Republic, Slovakia, and Hungary often presented similar positions, notably on asylum and border controls.71 The influence of Franco-German cooperation has traditionally been based less on common interests than on the ability to find a compromise most Member States could support precisely because both countries often had different preferences to start with.72 Spain, Italy, Greece, Cyprus, and Malta have coordinated their position in light of shared geographic concerns; they even created the label ‘MED5’, mirroring the abbreviation ‘V4’ for the Visegrád countries. Austria, the Netherlands, and Germany were the driving force behind stricter rules on legal migration during the early 2000s. A recurring argument in the debate about controversial legislative proposals says that the Council should abandon the search for compromise and overcome resistance by means of majority vote, as it famously did on 22 September 2015 when ministers agreed on the relocation of asylum seekers from Italy and Greece against the express will of some governments.73 The Court rejected an action of annulment, thus confirming that the prevalent consensus culture is not legally binding.74 Chapter 3.3.2 on infringement proceedings will demonstrate that ‘rogue’ Member States refusing to comply with legislation adopted by qualified majority can be brought to Court, even though EU institutions are ill-equipped, besides the threat of penalty payments under Article 260(2) TFEU, to compel governments to comply with a judgment. That may be the deeper logic behind the consensus culture: promoting compliance with legal obligations prevents the escalation of divisive constitutional conflicts, which might damage the European project.

2.2.4 European Parliament Parliamentarians habitually emphasise that they belong to the sole EU institution with direct democratic legitimacy, since they are elected directly by Union citizens. Over the years, the former ‘assembly’ turned into a veritable legislative powerhouse. In justice and home affairs, it became an equal co-legislature with the Council under the Treaty of Lisbon.75 Parliament’s strength is the careful inspection of Commission proposals. Members of the 69 See ch 13.3.2. 70 See Papagianni (n 48) 204–208. 71 See www.visegradgroup.eu (accessed 1 March 2023). 72 This arguably explains, ex negativo, why a Franco-German proposal will not usually resolve protracted disputes about migration policy nowadays given the sheer diversity of preferences among ever more Member States. 73 See Relocation Decision (EU) 2015/1601, which was adopted against the votes of the Czech Republic, Hungary, Romania, and Slovakia, with Finland abstaining. 74 See ch 13.3.2; and Joined Cases C-643/15 and C-647/15 Slovakia and Hungary v Council EU:C:2017:631, paras 143–50. 75 See ch 1.2.4.

Influence on the Policy Output 55 European Parliament (MEPs) regularly table long list of amendments before entering into interinstitutional negotiations with the Council. Doing so follows the tradition of working parliaments, which, like the US Congress, serve as an institutional counterpart to the government they do not necessarily support.

2.2.4.1 Committees as the ‘engines’ of parliamentary impact The organisational backbone of the Strasbourg-based assembly is the committee structure, which allows MEPs to develop a high level of expertise.76 Draft legislation is discussed in the Committee on Civil Liberties, Justice and Home Affairs (LIBE Committee), whose composition reflects the relative weight of the transnational political groups in the plenary.77 Whereas the election of MEPs takes place at the domestic level, their supranational status is primarily defined by the transnational political groups that dominate the internal organisation. Traditionally, the Christian Democrats (EPP), the Social Democrats (S&D), and the Liberals (now: Renew Europe) have been the three biggest groups, although the elections of 2014 and 2019 saw an increase in the vote for smaller parties and, hence, political fragmentation of the parliamentary work. A particularly powerful figure is the ‘rapporteur’ which the LIBE Committee appoints for each legislative file; she presents the matter to the committee and drafts the legislative report, which forms the basis of interinstitutional negotiations.78 While she cannot force its opinion on other MEPs who ultimately have to approve each proposal for an amendment, the rapporteur exercises sizeable influence. Rapporteurs benefit from superior knowledge, the speakers’ platform, the consensus-building function within the committee, and the privileged role in interinstitutional negotiations. Their weight was further increased, as we shall see, by the move towards informal ‘trilogue’ meetings as the focal point of interinstitutional discussions. 2.2.4.2 Shifting positions on migration law Unlike most national parliaments, the European Parliament does not support the ‘government’. Rather, it forms an opinion of its own and defends it towards the Council and the Commission. On the basis of previous activities, many observers expected that MEPs would insist on more liberal legislation that advances the rights of migrants when MEPs gained co-decision powers in the late 2000s. While this expectation was met in some respects, prominent legislative files, such as the Return Directive, indicated a change of perspective. Increased powers seem to have triggered a shift to the ‘mainstream’, for which several explanations exist. First, earlier legislation, which had been adopted unanimously by the Council, created a path-dependency that MEPs were unable to reverse fundamentally.79 Secondly, co-decision entailed an institutional self-interest in presenting opinions which had a realistic chance to form the basis of an interinstitutional compromise. The prospect of bargaining drove the Parliament towards the centre, in the same way as the Council may have refrained from

76 The main seat is in Strasbourg where it holds at least a monthly period of plenary sessions, while most committee work takes place in Brussels. 77 See Rules of Procedure of the European Parliament, 9th Parliamentary Term, r 209 (February 2020). 78 ibid r 51(2). 79 See Trauner and Ripoll Servent (n 17) 1417–32.

56 Institutional Prerogatives and Decision-making insistence on stringent migration control in anticipation of the Parliament’s position.80 Empirical studies confirmed this hypothesis, especially for members of the social-democrat and the liberal groups.81 Thirdly, elections matter. Composition has changed over the years; centre-left parties lost ground, while right-wing and populist parties gained seats. Since 2014, a stable majority has required a ‘grand coalition’ involving the EPP, which puts a greater emphasis on migration control than the centre left.82 Finally, politicisation and the upsurge of populism contributed to a change of direction, in line with previous comments. MEPs may be less directly involved in public disputes at national and European levels than national parliamentarians, but they are certainly not detached from the views and preferences of the electorate they represent.

2.3 Interinstitutional Practices The Court famously stated that the EU Treaties strive for an ‘institutional balance’, which resembles classic conceptions of the separation of powers, albeit with a distinct European flavour of compromise-prone semantics hinting at an equilibrium.83 The search for compromise is entrenched in the institutional culture in line with previous comments, and it similarly defines interinstitutional relations. An overarching constitutional principle of institutional balance does not supersede, however, more specific procedural provisions in the EU Treaties. This section presents the formal procedures and informal practices of interinstitutional relations. Readers should distinguish the ordinary legislative procedure from Treaty change, which has little relevance to migration law (2.3.1). When it comes to secondary legislation, informal ‘trilogues’ dominate interinstitutional negotiations (2.3.2). Implementing and delegated acts have usually been confined to highly technical matters, unlike in other areas of EU activity (2.3.3), whereas ‘soft law’ has gained some prominence in response to recent policy crises (2.3.4). Notwithstanding far-reaching internal powers, the European Parliament is confronted with legal and practical challenges when supervising international cooperation (2.3.5). Our inspection will conclude with the role of private actors, which have traditionally used ‘Brussels’ as a forum to exercise considerable influence, prior to the politicisation of decision-making (2.3.6).

2.3.1 Treaty change: limited relevance A basic terminological cleavage concerns the distinction between the ‘primary law’ in the EU Treaties and ‘secondary legislation’ adopted by the institutions, on the basis of the Treaties, in the form of directives or regulations. While Treaty amendments follow public 80 See Ariadna Ripoll Servent, ‘The European Parliament in Justice and Home Affairs’ in Ripoll Servent and Trauner, Routledge Handbook (n 62) 385, 388–91. 81 On the late 2000s see Esther Lopatin, ‘The Changing Position of the European Parliament on Irregular Migration and Asylum under Co-decision’ (2013) 51 JCMS 740. 82 See Ripoll Servent (n 80) 389–90. 83 First used by Case 9/56 Meroni EU:C:1958:7: ‘balance of power’.

Interinstitutional Practices 57 international law, the adoption of secondary legislation mirrors domestic law-making. Treaty amendments require unanimity under Article 48 TEU and must be ratified by the Member States, which usually foresee national parliamentary consent (or a referendum). Several constitutional courts have established limits to the transfer of competences. By contrast, secondary legislation is adopted by the European Parliament and the Council, and the Court of Justice serves as the main judicial authority, whereas national parliaments have a marginal role to play. When it comes to migration, most initiatives can be realised without recourse to politically cumbersome Treaty revision. Chapter 1.2.4 explained that the Convention drafting the erstwhile Constitutional Treaty laid down broad competences, which found their way into today’s Articles 77–80 TFEU. The scope of these powers will be discussed in the section on the constitutional foundations at the beginning of the chapters on sectoral policies. We shall see that there are few limitations, since the constitutional caveats in Article 72 and Article 79(4) and (5) TFEU have little impact in practice. Only some initiatives on migratory matters would require Treaty change: in particular the centralisation of administrative decision-making, presented in Chapter 8.2.1 on agencies, and harmonisation of nationality law, mentioned in Chapter 15.7 on integration.

2.3.2 Secondary legislation: prevalence of informal ‘trilogues’ Articles 77–79 TFEU declare that directives and regulations shall be adopted in the ‘ordinary legislative procedure’, which is often referred to under its old name ‘co-decision’. There is no need to grasp the procedural intricacies with two readings in each institution, followed by a ‘conciliation’ phase in accordance with Article 294 TFEU; it is rarely followed through in practice. Compromise-building would be hindered if the European Parliament and the Council were officially obliged to reject the positions of each other by means of formal votes in first and second reading, before finally convening in a conciliation committee to find a middle ground. Moreover, the official course of action would erect an additional hurdle in terms of parliamentary majority requirements.84 Exceptionally, formal adoption of first reading positions may be tempting before elections to ensure continuity of the parliamentary work.85 In regular circumstances, however, the interinstitutional practice concentrates on the ‘trilogue’ format. Trilogues are informal gatherings of representatives of the Commission, the Council, and the European Parliament to reach common accord. They can take place at different stages of the first or second reading, possibly as early as after the LIBE Committee’s vote on the draft report.86 Recourse to informal negotiations can be effective, but it is problematic from a constitutional perspective. Doing so wilfully creates a grey zone where crucial political decisions are taken in the proverbial back room. Trilogues leave the European Parliament’s 84 Rejection or amendment of the Council position by the Parliament at second reading requires ‘a majority of its component members’ under TFEU, art 294(7)(b), (c), ie a positive vote in favour of at least 353 of the 705 MEPs, while the (simple) majority at first or third reading under TFEU, art 294(3), (13) is met whenever more ‘yes’ than ‘no’ votes are cast, ie abstention does not hinder the formation of a majority. 85 On (dis)continuity see Rules of Procedure (n 77) r 240; examples were negotiations on the second generation of asylum legislation. 86 See Rules of Procedure (n 77) r 71; and Joint declaration on practical arrangements for the codecision procedure [2006] OJ C145/5, Nos 11–23.

58 Institutional Prerogatives and Decision-making Rapporteur and the Council Presidency with much influence. Of course, the Parliament and the Council will have to consent to the compromise, but the informality of trilogue meetings gives the agents practical leeway the principals in the Council and the Parliament cannot control effectively.87 Arguably, the veil of obscurity typifying the backroom is one of the reasons explaining the success of the format. There are so many actors trying to influence the legislative process that an informal inner circle can be a value in itself. Negotiations on the recast of the Reception Conditions Directive 2013/33/EU were a rare case of the Council disagreeing with a compromise reached in a trilogue meeting.88

2.3.3 Curtailment of delegated and implementing acts ‘Delegated’ or ‘implementing’ acts play a minor role in EU migration law, also because the European Parliament habitually insists that politically relevant questions are being determined in the ordinary legislative procedure. MEPs successfully challenged an attempt by the Council to sideline co-decision for the subsequent adoption of common minimum lists of safe third countries and countries of origin under the former Asylum Procedures Directive.89 It also convinced the Court to annul a first body of rules on sea border surveillance, which had been agreed upon in a ‘comitology’ procedure by the Commission, in collaboration with a committee of national representatives.90 Annulment led to the adoption of today’s Sea Borders Regulation (EU) No 656/2014 in the ordinary legislative procedure two years later. Delegated and implementing acts play a lesser role in migration than in other policy areas where comitology procedures, involving expert bodies in decision-making, are widely used (and often criticised for lack of transparency).91 ‘Implementing’ acts are defined by the participation of committees of national or private experts, whereas ‘delegated’ acts are adopted by the Commission—or the Council, exceptionally92—without the involvement of other institutions.93 They employ the adjective ‘delegated’ or ‘implementing’ in the official title. The difference to ‘regular’ legislation lies in the decision-making procedure, whereas legal effects in domestic legal orders are the same as for any other supranational legislation. By way of example, implementing acts concern the specification of details for visa stickers, and delegation is used for changes to visa fees or the exceptional reintroduction of visa requirements for countries refusing reciprocal visa-free travel for all Member States.94 Judges rejected a legal challenge by the Commission, which had sought not to be given that politically treacherous task on how to respond to the initial refusal of the US to lift 87 Speaking to government representatives, I was told on serval occasions that deals struck at trilogue meetings can be difficult to oversee; note that the political groups have more influence through ‘shadow rapporteurs’, who participate in trilogue meetings. 88 See ‘State of play and guidance for further work’ (Council doc 5458/19, 21 January 2019). 89 See Case C-133/06 Parliament v Council EU:C:2008:257, with regard to Directive 2005/85/EC, arts 29, 36(3). 90 Case C-355/10 Parliament v Council EU:C:2012:516, paras 68–85 found the rules to modify essential elements of the original Regulation (EC) No 562/2006. 91 See Curtin (n 44) ch 5. 92 Case C-257/01 Commission v Council EU:C:2005:25 confirmed that the Council could reserve details on the visa procedure to a Council decision—the practice was later discontinued. 93 See TFEU, arts 290, 291. 94 See Visa Code Regulation (EC) No 810/2009, as amended by Regulation (EU) 2019/1155, arts 16(9), 27(1), 29(1a).

Interinstitutional Practices 59 visa requirements for some countries in Central and Eastern Europe.95 Rare examples of politically sensitive implementing acts concern sanctions, in the form of disadvantages in the visa procedure, against third states not cooperating on return, which will feature in Chapter 18.3.2 on international cooperation, or the specification of risks underlying the algorithm for the automated assessment of prior travel authorisation under the Article 33 ETIAS Regulation (EU) 2018/1240, which will be presented in Chapter 9 on databases.

2.3.4 Ancillary role of ‘soft law’ ‘Soft law’ is a fascinating subject of study precisely because it is not legally binding. Advantages lie in the flexibility, spontaneity, and additional leverage it gives the Commission, as the main author. Disadvantages include non-transparency, legal uncertainty, and circumvention of decision-making procedures. Non-binding instruments are widely used by the Commission in competition law, economic policy, or state aid. They were celebrated as an alternative form of governance in the early 2000s, although this enthusiasm has given way to nuanced assessments that highlight inbuilt risks and drawbacks.96 Moreover, judges have elaborated on indirect legal effects of soft law, or lack thereof.97 Such informal instruments have traditionally played a marginal role in migration law, before gaining some traction in cooperation with third states, during the Covid-19 pandemic, and in response to the policy crisis of 2015/16. At an intermediate level, we may distinguish three forms of soft law. First are formal non-binding recommendations, adopted by the Commission or by the Council, in areas of Union competence in accordance with Article 292 TFEU. Second are informal cooperation frameworks with third states, discussed in Chapter 18.2.2 on international cooperation. Thirdly, Commission communications may contain the legal viewpoint of the supranational executive on how to interpret secondary legislation. Such political communications carry less weight than formal recommendations but may influence national implementation nonetheless. Soft law instruments fulfil different functions. One the one hand, non-binding instruments may complement existing policies as an alternative to legally binding measures. Such substitution effect is particularly widespread in cooperation with third states, notably in the form of informal return arrangements or non-binding mobility partnerships. One step further, soft law can prepare the ground for the adoption of legislation later on. Chapter 13.10 will demonstrate that Commission Recommendations on resettlement served as tentative steps towards common action in the form of the Proposal for a Resettlement Framework Regulation. On the other hand, soft law instruments seek to influence the interpretation of legislation. Measures may include profane issues of everyday application, including a summary of the case law, notably in the form of the ‘Schengen Handbook’ and the ‘Return Handbook’, 95 See ch 12.2.3; and Case C-88/14 Commission v Parliament and Council EU:C:2015:499. 96 See Mark Dawson, ‘Three Waves of New Governance in the European Union’ (2011) 37 EL Rev 208; and Kenneth Armstrong, ‘New Governance and the European Union’ in Gráinne de Búrca and others (eds), Critical Perspectives on Global Governance (Hart Publishing 2013) 249. 97 See Federico Casolari, ‘The Unbearable “Lightness” of Soft Law’ in Francesca Ippolito and others (eds), Bilateral Relations in the Mediterranean (Edward Elgar Publishing 2020) 215, 216–17.

60 Institutional Prerogatives and Decision-making which contain useful information for national officials.98 Other initiatives go beyond codification and seek proactively to shape how national authorities, and possibly courts, handle existing rules. Such efforts to mould the interpretation of legislation informally are not legally binding but can be practically relevant, nonetheless; they were mentioned by the Court on at least one occasion, as a supplementary argument to support a specific conclusion.99 Prominent examples include a dynamic reading of the conditions for reintroducing internal border controls, discussed in Chapter 12.4.2, and of the ‘risk of absconding’ under the Return Directive, mentioned in Chapter 16.7.2. A notorious example of soft law transforming secondary legislation was the unprecedented external travel ban for third country nationals during the Covid-19 pandemic, which was emulated two years later when the Baltic States and Finland, in particular, effectively closed their borders to Russians during the armed attack against Ukraine.100 The travel ban during the pandemic illustrated the advantages and drawbacks of soft law exemplarily.101 The main benefit was spontaneity and flexibility. A first short Commission Communication was written in haste and could be adapted easily by means of follow-up documents in the weeks thereafter.102 The travel ban was formalised somewhat when the Council adopted a recommendation, updated regularly, to define countries covered by the entry ban.103 None of these measures was legally binding, but they seem to have been reasonably successful in coordinating national practices applying the Schengen Borders Code Regulation (EU) 2016/399. Nevertheless, the downside was obvious. The most drastic travel restrictions in the history of the Schengen area were decided informally on shaky legal grounds without parliamentary involvement. Important questions, such as the status of non-married partners of Union citizens, were left unaddressed. Travel bans by means of soft law are no model for accountable rule-making. It should be welcome, therefore, that the Commission proposed to formally introduce legally binding implementing acts on travel bans.104

2.3.5 International cooperation: side-effects of informalisation Parliamentary prerogatives with regard to international agreements on migratory matters are comparatively strong on paper but much weaker in practice. Diplomatic negotiations and the spread of informal arrangements diminish the Parliament’s authority. To start with, procedures for the conclusion of legally binding agreements in Article 218 TFEU deviate from the ordinary legislative procedure. Negotiations will usually be conducted by the Commission, cooperating closely with the Council, which may adopt negotiating 98 See Commission Recommendation (EU) 2017/2338 establishing a common ‘Return Handbook’ [2017] OJ L339/83; and Commission Recommendation establishing a common ‘Practical Handbook for Border Guards’ (C(2019) 7131, 8 October 2019), which both replaced earlier versions. 99 See Case C-635/17 E EU:C:2019:192, paras 63–69. 100 See Commission, ‘Commission: 1. Updating guidelines on general visa issuance in relation to Russian applicants; and 2. Guidelines on controls of Russian citizens at the external borders’ (C(2022) 7111, 30 September 2022). 101 See further Daniel Thym and Jonas Bornemann, ‘Schengen and Free Movement Law during the First Phase of the Covid-19 Pandemic’ (2020) 5 European Papers 1143, 1155–61. 102 See Commission, ‘Communication: COVID-19. Temporary restriction on non-essential travel to the EU’ COM(2020) 115 final, and follow-up documents. 103 See Council Recommendation (EU) 2020/912 on the temporary restriction on non-essential travel into the EU [2020] OJ C208I/1, and follow-up documents. 104 See Proposal for a Regulation amending the Schengen Borders Code, COM(2021) 891 final, art 21a.

Interinstitutional Practices 61 guidelines.105 By contrast, the European Parliament shall merely be ‘immediately and fully informed’.106 MEPs do not have a direct say on the EU’s position, let alone a seat at the negotiating table. They can, at most, exercise influence during the negotiations with the threat of non-ratification—a nuclear option that cannot be used extensively.107 To be sure, the end result requires ratification,108 but this leaves Parliament with a binary choice either to approve or reject an agreement which has been negotiated by others. The spread of informal cooperation frameworks and non-binding working arrangements with third states, discussed in Chapter 18.2.2 on the external dimension, further undermines the Parliament’s role. While informal collaboration has always existed at the international level and can serve legitimate functions, the extent of contemporary informal practices on migration is problematic from an institutional perspective. Judges confirmed that informal cooperation must emanate from supranational competences and that the Commission may represent the Union, subject to a political mandate by the Council defining essential aspects of the EU’s position.109 By contrast, MEPs are sidelined, although proposals for dynamic interpretation have been put forward by scholars.110 Moreover, Parliament may possibly create leverage by means of issue linkage, by politically connecting informal cooperation to legislative files subject to co-decision.111 For the time being, however, external migration governance remains by and large the domain of the executive. It allows political actors to circumvent institutional constraints that would exist internally.

2.3.6 Private actors: a democratic virtue for migration law? Union law has traditionally revolved around technical matters. This gave private actors an important role, for instance via expert committees designing implementing acts. The Commission’s self-perception as an expert body drawing authority from a sizeable knowledge base further enhanced symbiotic relations with private actors providing valuable input. As an expression of ‘better law-making’, the Commission committed to perform stakeholder consultations and to present impact assessments.112 Corresponding material can be found in ‘Staff Working Documents’ (SWDs) accompanying the official explanatory memorandum of the Commission’s legislative proposals.113 Private actors have traditionally

105 See TFEU, art 218(2)–(4). 106 ibid art 218(10). 107 See Christina Eckes, ‘External Relations Law’ in Diego Acosta Arcarazo and Cian C Murphy (eds), EU Security and Justice Law (Hart Publishing 2014) 186, 190–93. 108 See TFEU, art 218(6)(a)(v), read in conjunction with arts 77–79. 109 Case C-660/13 Council v Commission EU:C:2016:616, paras 36ff had recourse to TEU, arts 16–17, not the provision on Treaty negotiations in TFEU, art 218. 110 See Paula García Andrade, ‘The Duty of Cooperation in the External Dimension of the EU Migration Policy’ in Sergio Carrera and others (eds), EU External Migration Policies in an Era of Global Mobilities (Brill 2019) 299, 317. 111 See Tineke Strik, ‘The European Parliament’ in Paul Minderhoud and others (eds), Caught in Between Borders (Wolf 2019) 279, 285–89. 112 See Interinstitutional Agreement on better law-making [2016] OJ L123/1, Nos 12–19, saying that impact assessments are not mandatory. 113 Semi-official SWDs often supplement official proposals in ‘COM’ documents, adopted by the college of Commissioners; both can be retrieved via https://eur-lex.europa.eu/collection/eu-law/pre-acts.html (accessed 1 March 2023).

62 Institutional Prerogatives and Decision-making held much influence in the depoliticised environment of supranational law-making, even though it cannot be measured exactly. For our purposes, the diversity of private actors stands out. While experts in migration law will intuitively think of NGOs, multinational companies and industry associations are at least as active in Brussels. Stakeholder participation and informal contacts often amount to classic ‘lobbying’ by actors with substantial resources and leverage. By contrast, NGOs— like any other civil society organisation114—struggle with lesser funding and were originally newcomers to the supranational arena; professionalisation was a precondition for visibility and influence.115 A well-known success story is the European Council on Refugees and Exiles (ECRE), which brings together more than 100 organisations from across the continent. However, NGOs are not the only players. Trade unions and employer organisations influence debates on labour migration,116 private companies participate in discussions about border surveillance, and profit-oriented consulting companies compete for large research and evaluation studies.117 Involvement of private actors in decision-making can be conceptualised in different ways. One strand of the debate emphasises stakeholder participation as a valuable source of information that may improve the quality of legislation and facilitate compliance. These arguments gained momentum in the early 2000s, thus informing the introduction of today’s Article 11 TEU on stakeholder consultation and the Commission White Paper on European Governance.118 The underlying theoretical model of ‘deliberative democracy’ often goes hand in hand with an interest in new forms of decision-making, such as the ‘open method of cooperation’ (OMC) or private law-making, which play virtually no role in the field of migration. Having said this, stakeholder consultation can be of theoretical relevance, as third country nationals have no right to vote and depend on others, therefore, to represent their interests in the legislative process.119 Another strand of the debate emphasises the uncertain conceptual foundations of stakeholder participation between classic notions of information gathering and novel visions of deliberative democratic legitimacy.120 Such less enthusiastic views of private actors seem to have gained ground in recent years.121 Social scientists emphasise the ambivalence of their internal accountability and political legitimacy, as well as the danger of sidelining other forms of political agency and protest.122 While NGOs are rightly proud of their altruistic stance, profit-oriented consulting companies are in a different position—let alone private companies competing for contracts on technological border surveillance. Critical scholars

114 The meaning of ‘NGO’ and ‘civil society organisations’ does not follow clear terminology and differs between countries; this book generally employs the term NGO. 115 See Emily Gray and Paul Statham, ‘Becoming European?’ (2005) 43 JCMS 877. 116 See Georg Menz, The Political Economy of Managed Migration (OUP 2008). 117 See Oleg Korneev and Olga Kluczewska, ‘The Globalised Third Sector in the Migration Policy Field’ in Anna Triandafyllidou (ed), Handbook of Migration and Globalisation (Edward Elgar Publishing 2018) 54. 118 See Commission, ‘European Governance. A White Paper’ COM(2001) 428 final. 119 See generally Alexander Somek, ‘The Darling Dogma of Bourgeois Europeanists’ (2014) 20 ELJ 688. 120 See Christoph Möllers, ‘European Governance’ (2006) 43 CML Rev 313. 121 See Mark Dawson, ‘Better Regulation and the Future of EU Regulatory Law and Politics’ (2016) 53 CML Rev 1209. 122 See Meike Rodekamp, Their Members’ Voice (Springer 2014); and David Kennedy, The Dark Sides of Virtue (Princeton UP 2004) ch 1.

Differentiated Integration: Opt-outs after Brexit 63 emphasise that the Commission, in particular, participates in the production of the knowledge that forms the basis of its activities.123 Empirical studies have found that private actors were quite successful in the early years of supranational justice and home affairs cooperation, thereby countering the effects of ‘venue shopping’ of interior ministries, discussed previously, which had employed transnational cooperation to bypass domestic constraints. Private actors were particularly influential when their input was taken up by political actors within the institutions, which channelled the views of NGOs and others into the interinstitutional deliberations.124 By contrast, the increasing politicisation of decision-making on migratory matters seems to have limited their impact. Non-state actors are more influential when technical arguments dominate the agenda.

2.4 Differentiated Integration: Opt-outs after Brexit An institutional idiosyncrasy of justice and home affairs are country-specific opt-outs, which are often described as an expression of ‘multiple speeds’, ‘variable geometry’, integration ‘à la carte’, or ‘coalitions of the willing’.125 Schengen remains, together with monetary union, the most prominent example of differentiated integration. The opt-outs discussed below do not embody a generic model of differentiated integration. They resulted from political compromises, enshrined at Treaty level, to secure unanimous consent to Treaty revision.126 In areas not covered by the opt-outs, all Member States participate in legislative initiatives, which can be adopted by qualified majority against their will. The possibility of ‘enhanced cooperation’ exists on paper but is rarely used in practice.127 To date, the option of enhanced cooperation has not been activated for a single new initiative on the basis of Articles 77–80 TFEU. As a result of Brexit, the opt-outs lost significance, although they remain intact for Ireland and Denmark. The position of the UK is governed by international agreements nowadays, to be mentioned in Chapter 17.3 on association agreements. Unfortunately, the opt-outs do not follow a uniform rationale and are defined by heightened complexity. We need to distinguish arrangements for Denmark (2.4.1) and Ireland (2.4.2). Moreover, we are faced with two sets of rules for each country: first, measures building upon the Schengen acquis, whose scope will be elaborated upon in Chapter 12.3.3 on border controls; secondly, justice and home affairs beyond the thematic scope of Schengen. Altogether, we need to distinguish four distinct opt-out scenarios for Ireland and Denmark on the one hand and for measures (not) belonging to the Schengen framework on the other hand.128 In practical terms, 123 See Christina Boswell, The Political Uses of Expert Knowledge. Immigration Policy and Social Research (CUP 2009) ch 8. 124 See Christian Kaunert, Sarah Léonard, and Ulrike Hoffmann, ‘Venue-Shopping and the Role of Non- Governmental Organisations in the Development of the European Union Asylum Policy’ (2013) 1 Comp Migration Stud 179; and Georg Menz, ‘Stopping, Shaping and Moulding Europe’ (2011) 49 JCMS 437. 125 See Daniel Thym, ‘Supranational Differentiation and Enhanced Cooperation’ in Tridimas and Schütze, Oxford Principles (n 52) 848–55, which forms the basis of the comments in this section. 126 In accordance with TEU, art 51, protocols attached to the Treaties, such as those on the opt-outs, form an integral part of primary law. 127 See ibid art 20; TFEU, arts 326–34. 128 Allegedly, two working groups designed the rules and forgot to align their substance; see Pieter Jan Kuijper, ‘Some Legal Problems Associated with the Communitarization of Policy on Visas, Asylum and Immigration’ (2000) 37 CML Rev 345, 352.

64 Institutional Prerogatives and Decision-making recitals indicate whether Ireland and Denmark are bound and whether the legislation belongs to the Schengen acquis.

2.4.1 Denmark: the long shadow of the referenda Denmark did not object to the abolition of border controls and had subscribed to the Convention Implementing the Schengen Agreement prior to the Treaty of Amsterdam. Nonetheless, the government asked for an opt-out, since supranationalisation called into question compromise formulae the Danish government had relied upon in the campaign for a ‘yes’ vote in the second referendum on the Treaty of Maastricht.129 At the same time, Denmark did not want to leave Schengen and negotiated a ‘political opt-in’ in combination with a ‘legal opt-out’.130 Denmark remained a Schengen member without subscribing to supranationalisation. Unlike Ireland, Denmark cannot opt in to decision-making case by case, even though it could switch to the more flexible Irish position. Doing so has been made politically conditional on another referendum. The government called—and lost— such a referendum in late 2015.131 In line with the original compromise, Denmark participates in new measures that are building upon the Schengen acquis on the basis of ‘an obligation under international law’.132 This peculiar construction entails that the Court has no jurisdiction and that general principles, such as direct effect and supremacy, do not apply. Moreover, Denmark retains the theoretical option not to implement new legislation. In return, partner countries will ‘consider appropriate measures to be taken’, thus possibly triggering the exclusion from Schengen.133 In practice, collaboration has unfolded without major hiccups so far.134 Reasons for the trouble-free performance include the participation of Danish officials, albeit without a vote, in Council deliberations, close cooperation with neighbouring countries, and the tradition of border-free travel among the Nordic countries, which equally participate in Schengen.135 The Danish opt-out presents itself as an historic relict perpetuating the compromise for the ratification of the Treaty of Maastricht. Rejection of the supranational integration method means that Denmark cannot, on the basis of the present opt-out, participate in measures transcending the Schengen framework. By way of example, Danish rules on family reunification can be stricter than under the Family Reunification Directive 2003/86/EC, and the country does not contribute to solidarity measures, such as the 2015 Relocation Decisions, which were considered not

129 See David Howarth, ‘The Compromise on Denmark and the Treaty on European Union’ (1994) 31 CML Rev 765. 130 Monica den Boer, ‘Justice and Home Affairs Cooperation in the Treaty on European Union’ (1997) 4 MJECL 310, 311. 131 See Protocol [No 22] on the position of Denmark [2008] OJ C115/299, arts 7 and 8, which replaced an earlier protocol [1997] OJ C340/299; and https://en.wikipedia.org/wiki/2015_Danish_European_Union_opt-out_ref erendum (accessed 1 March 2023). 132 ibid art 4(1). 133 ibid art 4(2); and Daniel Thym, Ungleichzeitigkeit und Europäisches Verfassungsrecht (Nomos 2004) 110–14 http://www.ungleichzeitigkeit.de (accessed 1 March 2023). 134 In 2011, the Danish government announced plans to reintroduce limited permanent border controls at the insistence of the populist peoples’ party; however, the project was abandoned when the general elections in 2012 brought a different government into power. 135 See Thym (n 125) 876–78.

Differentiated Integration: Opt-outs after Brexit 65 to build on the Schengen acquis. Nevertheless, Denmark was associated with the Dublin acquis in the form of an international agreement.136 Indirect participating via international agreements is legally problematic in light of a judgment, concerning Gibraltar, that low- threshold intergovernmental cooperation should not evade formal participation under the opt-out protocols.137 Politically, EU institutions were willing to associate Denmark in light of the political peculiarities underlying the opt-out. Note that the association agreement is confined to the Dublin III Regulation (EU) No 604/2013 and the Eurodac Regulation (EU) No 603/2013, as well as any successor instruments in accordance with the procedure set out in Article 3. By contrast, Denmark is not bound by other asylum instruments, such as the Asylum Procedures Directive 2013/32/ EU or the Qualification Directive 2011/95/EU. This entails that it can proceed, subject to human rights law, with controversial projects for the external processing of asylum applications in third states, which would be illegal under EU legislation.138 References to the ‘Member States’ in these instruments do not cover Denmark either. As a result, Germany could not reject an asylum application as inadmissible in light of a previous decision in Denmark, irrespective of the level of protection there.139

2.4.2 Ireland: relic of British reticence The Irish position can only be understood against the background of British Euroscepticism, which had always objected—unlike Denmark—to the political project of border-free travel. Consecutive British governments had maintained that the geographical position of an island nation, the traditional absence of domestic identification requirements (ID cards), and the symbolism of Schengen cooperation as an instrument of ‘ever closer union’ argued against British participation.140 Nonetheless, the Labour government under Tony Blair was willing to consent to the integration of the Schengen acquis into the EU framework under the condition that Britain retained a flexible opt-out, including the option voluntarily to sign up to individual projects. Ireland was obliged to follow its neighbour to maintain the Common Travel Area for passport-free travel in the British Isles, including between the Republic and Northern Ireland.141 Ireland may have originally ‘declare[d]‌its firm intention to exercise [the opt-in option] to the maximum extent’,142 but it did not always live up to this promise. Upon closer inspection, the British/Irish opt-out is the most prolific expression of the à la carte logic of a principled freedom. First, both countries retain the right to ‘notify . . . that they wish to take part’ in the adoption of a proposal.143 This option of ex ante participation 136 See the Council Decision 2006/188/EC on the conclusion of the Agreement extending to Denmark the provisions of the former Dublin II Regulation (EC) No 343/2003 [2006] OJ L66/37. 137 See Case C-44/14 Spain v Parliament and Council EU:C:2015:554, paras 30–42. 138 Seech 13.4.7. 139 See Case C-497/21 Bundesrepublik Deutschland EU:C:2022:721, paras 40–51. 140 See Antje Wiener, ‘Forging Flexibility. The British “No” to Schengen’ (1999) 1 EJML 441, 456–59; and House of Lords, ‘Schengen and the United Kingdom’s Border Controls’ (Select Committee, 7th Report, Session 1998/99). 141 See Elaine Fahey, ‘Swimming in a Sea of Law’ (2010) 47 CML Rev 673, 679–82. 142 Declaration (No 56) by Ireland on Article 3 of the Protocol on the position of the United Kingdom and Ireland [2008] OJ C306/268. 143 Protocol [No 19] on the Schengen acquis integrated into the framework of the European Union [2008] OJ C115/290, art 5(1); and its predecessor [1997] OJ C340/93; and Protocol [No 21] on the Position of the United

66 Institutional Prerogatives and Decision-making was used extensively for core asylum legislation and for measures against illegal entry and stay, whereas both countries refused to join legal migration initiatives. In addition, Ireland ‘may at any time request to take part in some or all of the provisions’ of legislative instruments ex post.144 On this basis, both countries subscribed to important segments of the original Schengen cooperation, in particular the Schengen Information System.145 The combined effect of ex ante and ex post participation was constructive, albeit with a focus on control-oriented instruments. Ireland’s freedom of choice is not absolute. The Schengen Protocol limits participation to initiatives which are ‘capable of autonomous application’.146 This meant, in the eyes of the Council, that British requests for membership in Frontex and participation in the Passports and Travel Documents Regulation (EC) No 2252/2004 had to be rejected, since both initiatives were intractably linked to the abolition of internal border controls, to which the UK had refused to sign up.147 The Court of Justice confirmed this standpoint in three judgments demonstrating a certain willingness to ensure that the opt-outs do not result in ‘cherry picking’.148 During the intergovernmental conference drafting the Lisbon Treaty, the British government demanded—and obtained—further flexibility.149 London insisted on the right to opt out of amendments of instruments in whose adoption it had decided to participate.150 Again, Ireland followed its neighbour. By a simple declaration, it may withdraw from ongoing legislative procedures, although it remains bound by the instrument under revision— an option Ireland used for the Asylum Procedures Directive 2013/32/EU,151 while Britain opted out of the Asylum Qualification Directive 2011/95/EU. From the perspective of legal certainty, it is regrettable that Ireland will continue to be bound by previous rules, which are repealed with regard to all other Member States.152 Judges may be asked to interpret provisions applying to Ireland only.153 If the Irish exclude themselves from an amendment, the Council can vote against continued participation in related instruments if such ‘rump’ legislation cannot be applied effectively.154 The example of the dispute over Frontex shows that the Court may side with the Council in cases of conflict. Kingdom and Ireland in respect of the area of freedom, security, and justice [2008] OJ C115/295, art 3(1), which builds upon a previous version [1997] OJ C340/295. 144 ibid art 4. 145 Decision 2002/192 concerning Ireland’s request to take part in some of the provisions of the Schengen acquis [2002] OJ L64/20; together with Council Implementing Decision (EU) 2020/1745 [2020] OJ L393/3; Decision 2009/350/EC [2009] OJ L108/53; as well as Decision 2000/365 concerning the request of the United Kingdom to take part in some of the provisions of the Schengen acquis [2000] OJ L131/43. 146 AG Verica Trstenjak, Opinion in Case C-77/05 United Kingdom v Council EU:C:2007:419, point 107. 147 Legally, both disputes concerned the delimitation of the Schengen Protocol (n 143) arts 4 and 5; as well as rules in Protocol No 21 (n 143), which are more flexible. 148 See ch 13.3.3; Case C-77/05 United Kingdom v Council EU:C:2007:803; Case C-137/05 United Kingdom v Council EU:C:2007:805; Case C-482/08 United Kingdom v Council EU:C:2010:631; and Maria Fletcher, ‘Schengen, the European Court of Justice and Flexibility under the Lisbon Treaty’ (2009) 5 EuConst 71. 149 See European Council, ‘Presidency Conclusions’ (Council doc 11177/1/07, 26 June 2007) No 19(l); and Clemens Ladenburger, ‘Police and Criminal Law in the Treaty of Lisbon’ (2008) 4 EuConst 20, 28. 150 See Schengen Protocol (n 143) art 5(2)–(5); and Protocol No 21 (n 143) art 4a. 151 Contrast Asylum Procedures Directive 2013/32/EU, recital 58 with the former Asylum Procedures Directive 2005/85/EC, recital 33. 152 The repeal of the earlier measure through new legislation does not extend to the Ireland, since it is not bound by the amendment; see also Steve Peers, EU Justice and Home Affairs Law, vol 1 (4th edn, OUP 2016) 29–33. 153 By way of example see Case C-616/19 Minister for Justice and Equality EU:C:2020:1010. 154 See Schengen Protocol (n 143) art 5(3); and Protocol No 21 (n 143) art 4a(2).

Summary 67 The UK stopped being bound by rules it had opted into when the transitional period under the Withdrawal Agreement ended on 1 January 2021. Brexit had a paradoxical effect in the field of migration: even though the desire to ‘take back control’ played a prominent role in the referendum campaign, withdrawal may complicate British migration policies. The UK will no longer participate in the Schengen Information System and the Dublin system. Instead, London may have to pay a political ‘price’ to convince Brussels to join existing initiatives on the basis of an international agreement.155 Negotiations on the Trade and Cooperation Agreement confirmed the reversal of previous policy dynamics when the EU rejected the British demand for the conclusion of a readmission agreement that could have served as a partial replacement for the Dublin system.156 Future relations will be discussed in the chapter on association agreements.

2.5 Summary Institutional provisions have a direct impact on the policy outcome by determining the weight of different actors. Originally, intergovernmental cooperation resulted in ‘venue shopping’, with interior ministries promoting a securitarian agenda on justice and home affairs. Today, the European Parliament serves as an equal co-legislator, besides the Council acting by qualified majority. In this overall context, political scientists examine the relative weight of national interests, normative convictions, and other factors in the construction of European migration law. Inspection of the driving forces helped overcome binary descriptions of reticent Member States insisting on policy restrictions versus supranational institutions supporting the rights of migrants. Justice and home affairs were supranationalised with the support of national governments, and the Council adopted important pieces of legislation unanimously. By contrast, the institutional practice of the Commission, in particular, displays practical elasticity. A critical variable for supranational politics is the degree of politicisation, which has increased in recent years, partly triggered by the resurgence of populism. Politicisation complicates supranational decision-making that has traditionally relied on a technocratic style of reasoning and compromise-building behind closed doors. The ‘permissive consensus’ gradually gave way to a ‘constraining dissensus’ when migration became a salient issue in public discourses at the national and supranational levels. In such an environment, policy outcomes tend to become more restrictive, and the danger of institutional blockage is real (as the example of asylum illustrates). A crucial contextual factor is the ‘institutional culture’, which determines how the relevant actors manage their decision-making prerogatives. EU institutions are complex bodies whose working practices are informed by entrenched patterns of behaviour. This chapter shed light on the ‘black box’ that the everyday functioning of the institutions is to many observers. We aimed to show, in other words, the ‘living constitution’ in the making of migration law. The Commission thrives on the basis of a unique knowledge base in an environment of technocratic decision-making. Its proposals are highly influential, yet it often hides 155 See Daniel Thym, ‘The Irony of Brexit for Immigration Control’ EULawAnalysis (20 October 2017). 156 See Ben Hulme, ‘Analysis of the United Kingdom’s Proposal for a UK-EU Readmission Agreement’ EULawAnalysis (20 August 2020).

68 Institutional Prerogatives and Decision-making behind the political initiatives of heads of state or government meeting within the European Council. The Commission has repeatedly changed course with technocratic pragmatism when it realised that Member States were unwilling to support an initiative. Constant meetings of working groups and other preparatory bodies turn the Council into a quasi- permanent institution, where national officials discuss the proverbial comma of new proposals. Internal deliberations are defined by a deep-rooted culture consensus which tries to take everyone on board. Votes rarely happen, although the option of a majority vote means that Member States are bound to be flexible. As a working assembly, the European Parliament exercises considerable influence via the high level of expertise within the influential LIBE Committee. In contrast to parliamentary democracies, where debates are often defined by conflicts between the government and the opposition, supranational decision-making is characterised by permanent consensus-building and compromise-seeking. This is one of the factors explaining why the majority view appears to have turned towards the ‘mainstream’ to ensure greater impact on the policy outcome, when it finally obtained full co-decision powers under the Treaty of Lisbon. In practice, most legislation is finalised in the secretive ‘trilogue’ format nowadays: representatives of the Parliament, the Council, and the Commission meet behind closed doors to hammer out the final deal. The weight of the European Parliament in the ordinary legislative procedure is an important reason why delegated and implementing acts did not gain much relevance in migration law. By contrast, MEPs hold less sway to influence international agreements concluded with third states, which was weakened further by the spread of informal arrangements. Soft law plays an ancillary—although increasing—role in migration law. It serves as a flexible tool to guide the interpretation of existing rules or to advance new initiatives as a substitute to legally binding measures. Soft law raises concerns over accountability and legal certainty when used on sensitive issues. Private actors have traditionally been powerful in Brussels; they include NGOs, established organisations like trade unions or churches, and profit-seeking consultancy groups or companies applying for valuable contracts. Increasing politicisation arguably undermines the influence non-governmental actors, which depend on institutional actors to take up their views. An inspection of law-making would be incomplete without the country-specific opt-outs of Denmark and Ireland (as well as the UK, before Brexit). They stem from political compromises enshrined at Treaty level and do not establish a general blueprint for ‘multiple speeds’ or ‘coalitions of the willing’. The operation of the opt-outs is defined by an excessive complexity, since we have to distinguish four distinct opt-out scenarios for Ireland and Denmark on the one hand and for measures (not) belonging to the Schengen framework on the other hand. While Denmark participates in Schengen, it cannot join other initiatives. Ireland, by contrast, did not subscribe to the abolition of border controls but participates in various other projects.

3

Court of Justice: Achievements and Limitations The European Union has been described as a ‘lawyers’ paradise’1 to highlight the vital role the law has traditionally played in the construction of the European polity. Judgments of the Court of Justice are an essential point of reference for legal practice and academic analyses, and they are a central theme of this volume. Reference to the ‘limitations’ in the title of this chapter does not question the pivotal role of the Court in Luxembourg. Rather, we intend to draw attention to contextual factors that define the success in terms of guiding the uniform interpretation and application of EU migration law. A statistical survey of all judgments interpreting the legislation discussed in this volume will demonstrate, together with other considerations, the relative weight of different procedural avenues, pertinent actors, and subject matters. In doing so, our comments will combine the procedural basics with a critical inspection of the judicial practice. Our analysis begins with the influence of the court architecture, such as the option of specialised tribunals to oversee Frontex and the Asylum Office (3.1). Notwithstanding the track record of dynamic interpretation on constitutional matters, a bird’s eye view on the judgments on migration will demonstrate an upsurge in statutory interpretation focusing on the outcome of the legislative process (3.2). Institutional practices are critical for anyone trying to understand access to justice. While infringement proceedings and validity disputes play a secondary role in migration law, preliminary references by domestic courts are, statistically and conceptually, the hallmark of the Court’s profile (3.3). Protracted compliance deficits in asylum and border controls indicate that the control function should not depend on judges alone. This chapter will conclude with alternative accountability fora that may support respect for supranational rules (3.4).

3.1 Influence of the Court Architecture English language literature fluctuates between the classic abbreviation ‘ECJ’ and the comparatively new ‘CJEU’, which responds to the change in the official designation from ‘Court of Justice’ to ‘Court of Justice of the European Union’ under the Treaty of Lisbon. Crucially, this volume does not ignore the change of name when it occasionally uses the traditional label of ‘ECJ’. Doing so emphasises internal differentiation, since the holistic label ‘CJEU’ embraces three separate bodies: the ‘Court of Justice’, the ‘General Court’, and, possibly in future, ‘specialised courts’.2 Judges at the Court of Justice are the apex of the judicial architecture, and they decide the overwhelming majority of cases on migration, so far at least. The EU’s top court should not be confused with the European Court of Human Rights

1 2

Antoine Vauchez, Brokering Europe (CUP 2015) 1. See TEU, art 19; and TFEU, arts 251–57.

European Migration Law. Daniel Thym, Oxford University Press. © Daniel Thym 2023. DOI: 10.1093/oso/9780192894274.003.0004

70 Court of Justice: Achievements and Limitations (ECtHR) in Strasbourg, whose impact on EU migration law will be traced in Chapter 5.3 on human rights.

3.1.1 Collective and multi-lingual adjudication In contrast to the US Supreme Court, names of judges sitting on the bench of the CJEU are not usually mentioned outside expert circles (and even the latter rarely discuss their role). This difference reflects the lesser politicisation of the judicial function in Europe, as well as the internal workings. Judges do not proclaim individual opinions. Instead, the Court agrees on a single text. The final judgment does not even indicate, in contrast to the ECtHR, whether it embodies a consensus or a majority view; separate opinions do not exist. While the reporting judge (rapporteur) has practical weight, she cannot succeed without the support of others. Participants confirm that the inner workings support consensus-building, mirroring the judicial tradition on the European continent.3 The need to agree on a single text may explain the opacity of the judicial reasoning in some cases, when judges concur on the outcome but disagree on the reasons. In contrast to the International Court of Justice, there are no ‘national’ judges in cases relating to a specific Member State. Assignment to a chamber is not influenced by the nationality of judges, which are fully independent in the exercise of their function. In contrast to individual judges, Advocates General tend be widely known, in expert circles at least. Their non-binding opinions are important points of reference for legal analyses, precisely because they articulate an argument that may be more coherent than the final judgment. While the Court often follows the Advocate General (AG), the final verdict may differ, especially in legally difficult or politically sensitive cases. Even if the end result is the same, lacunas in the Court’s reasoning cannot automatically be filled with the position of the AG, unless judges say so explicitly (as they regularly do). The last British AG Eleanor Sharpston gained a certain celebrity status owing to her outspoken opinions that often proposed advancing the rights of migrants. Judgments are—like secondary legislation—translated into all the twenty-four official languages, even though the official version of the file is defined by the language of the case.4 In practice, the French version has an aura of quasi-authenticity, since the Court still holds the internal délibéré in the former lingua franca of the EU institutions. The ‘lawyer-linguists’ will usually produce a high-quality translation into English, but it can be useful, for those mastering French, to compare language versions whenever a statement remains ambiguous. The same applies to those habitually reading judgments in other languages. Consultation of the French or English text can help overcome uncertainties about how to interpret ambiguous passages.

3 See Sophie Turenne, ‘Institutional Constraints and Collegiality at the Court of Justice of the European Union’ (2017) 24 MJECL 565, 576–80. 4 See Rules of Procedure of the Court of Justice [2012] OJ L265/1, arts 36, 37(3), with later amendments.

Influence of the Court Architecture 71 25 20 15 10 5

Grand Chamber

2

1

20 2

0

20 2

9

20 2

8

20 1

7

20 1

6

20 1

5

20 1

4

20 1

3

20 1

2

20 1

1

20 1

0

20 1

9

20 1

20 0

20 0

8

0

Chamber (5 judges)

Chamber (3 judges)

Figure 3.1 Designation to different formations within the CJEU* *The statistics include judgments listed in the digest of CJEU case law, the Newsletters on European Migration

Issues (NEMIS) and Asylum Issues (NEAIS) of the Centre for Migration Law at Radboud University Nijmegen, and some additional rulings; it excludes cases that were removed from the register and those dealing with Union citizenship, as well as association agreements described in ch 17; I am grateful to Jonas Bornemann, Kilian Umbach, and student assistants for support in categorisation.

3.1.2 Horizontal outlook beyond migration law Experts in migration law should apprehend that the judges’ outlook transcends migration law. There are no specialised chambers for particular themes. There is constant variation in the assignment of subject matters, although a chamber may occasionally deliver several judgments on related themes: for instance, a series of family reunification cases or a sequence of Dublin rulings.5 Moreover, jurisdiction fluctuates between the Grand Chamber of fifteen judges and chambers of five or three judges. There are, for example, five different chambers of five judges each, which, taken together, deliver most judgments on migration. Figure 3.1 shows which formations deal with migration law. Bigger formations within the Court are generally chosen for (politically) important or (legally) difficult cases,6 and the formation may change exceptionally during the procedure.7 An instructive example is infringement proceedings: the high percentage of judgments by (small) chambers of three judges in the late 2000s reflected, as we shall see, the prevalence of uncontentious late transposition cases, whereas more recent and controversial proceedings against Hungary were decided by the Grand Chamber. Generally speaking, our statistical survey shows that smaller formations have gained relevance over the years, indicating that the case law has grown to maturity, although an increase in the political sensitivity of migration may result in more Grand Chamber judgments, such as between 2016 and 2018 and in 2021/22. Preliminary references on recurring questions, such as the status of Palestinian

5 Jurisdiction is assigned by the President under the Rules of Procedure (n 4) art 15; see further Daniel Thym, ‘A Bird’s Eye View on ECJ Judgments on Immigration, Asylum and Border Control Cases’ (2019) 21 EJML 166, 168. 6 See Rules of Procedure (n 4) art 60(1). 7 As it did in Case C-181/16 Gnandi EU:C:2018:465.

72 Court of Justice: Achievements and Limitations refugees, discussed in Chapter 13.5.6, will be given to smaller formations, on the basis of lead judgments by bigger formations. Having said this, the sheer complexity of the legal material entails that new questions are bound to arise frequently, thus complementing ‘settled case law’ with innovative rulings. EU migration law will remain exciting. Lack of specialisation supports the coherence of the supranational legal order but has the side effect of hindering the emergence of high levels of expert knowledge. The constant shift of jurisdiction helps to explain, as a contextual factor, why EU migration law was not aligned further with international refugee law in the Court’s practice, as demonstrated in Chapter 5.6.2 on the Refugee Convention. Judges may have to decide cases on diverse matters such as asylum, competition, tax law, and consumer protection in parallel. As a result, they will approach migration law as an integral part of the supranational legal order. Anyone trying to understand the judicial output should be aware of feedback loops between the migration law instruments and the overarching doctrinal foundations of Union law, which will be addressed in Chapter 6.

3.1.3 Frontex, Asylum Agency, and the pitfalls of specialised tribunals EU Treaties explicitly allow for the creation of ‘specialised courts’ in accordance with Article 257 TFEU. This option was introduced to allow for an increase in judicial capacity by means of organisational differentiation but has not been activated so far. The former Civil Service Tribunal ceased to exist in 2016, following a dispute over the reform of the General Court; its expanded size rendered the specialised tribunal unnecessary.8 Nevertheless, specialised tribunals might possibly become relevant for EU migration law, if the powers of Frontex or the Asylum Agency were strengthened substantially. Chapter 8.5 will explain that, at present, the powers of the agencies are essentially limited to support domestic authorities. If Frontex or the Asylum Agency were given autonomous decision-making authority, legal remedies before a supranational court would have to be foreseen. The sheer number of individual cases involved would effectively require the creation of a specialised tribunal— with the option of appeal, on points of law, to the General Court and, exceptionally, the Court of Justice.9 Doing so would pose formidable logistical challenges in terms of hiring new staff and organising effective working structures, including that of the definition of the geographical seat(s) of the specialised tribunals. Asylum applicants, or their legal representatives, could hardly be expected to travel to Luxembourg to appeal against a decision by the Asylum Agency delivered on a Greek island. While practical considerations would argue in favour decentralised tribunals, possibly in different locations across the Union, constitutional considerations support a uniform seat in Luxembourg.10 Moreover, specialised tribunals are disliked by judges at the Court who worry about coherence.11 These problems reflect, with a view to the judicial function, the difficulties in holding the migration agencies to account for wrongdoing, as discussed in Chapter 8.5. 8 Smaller Member States had insisted on one judge per country instead of expertise-driven nominations, thus inflating the number of judges on the bench. 9 See TFEU, arts 257(3), 256(2)(2). 10 Protocol [No 6] on the Location of the Seats of the Institutions [2008] OJ C115/265 can be interpreted to require the physical presence of the entire CJEU in Luxembourg. 11 See ‘Draft Amendments to the Statute of the Court of Justice’ (Council doc 8787/11, 7 April 2011) 7–9.

Influence of the Court Architecture 73 In light of organisational challenges, EU institutions might consider an intermediate solution. Regulatory agencies in other domains often have an independent internal appeals structure for review by a body of technical experts.12 Procedures may follow quasi-judicial patterns, but the appeals bodies retain an administrative character nonetheless. Such internal review adds value in areas of technical regulation, such as chemicals, where technical disputes may be resolved internally, before a legal remedy is lodged before the General Court (or, potentially, a specialised tribunal); such judicial remedies may be limited to points of law.13 In the areas of asylum and border controls, such double administrative and judicial appeals structure would be impracticable: the number of cases involved is too high and long procedures should be avoided. Some might propose that external judicial oversight should be replaced by a fully independent first instance of legal remedies within the agencies. Indeed, Article 13(1) Return Directive 2008/115/EC was deliberately drafted in abstract language: remedies shall be decided by a ‘judicial or administrative authority or a competent body composed of members who are impartial and who enjoy safeguards of independence’, in line with ECtHR case law that accepts the option of internal administrative oversight.14 In the EU context, however, Article 47 CFR offers a higher level of protection, in light of which the Return Directive must be interpreted. States have to guarantee judicial independence—a requirement administrative review does not usually meet.15 This argument can be projected upon Article 14(3) Schengen Borders Code Regulation (EU) 2016/399, whereas Article 46 Asylum Procedures Directive 2013/32/ EU leaves no room for administrative review. A fully fledged centralisation of administrative decision-making under the auspices of Frontex or the Asylum Agency would require supranational judicial appeal bodies, most likely via decentralised specialised tribunals. For the time being, this option remains elusive.

3.1.4 Non-governmental organisations and the limits of third party intervention Court judgments are not academic articles written in some study room. Rather, they present the endpoint of highly formalised proceedings, which usually comprise a written part and an oral hearing.16 Reading judgments, one realises that the Court refers to the arguments put forward during the proceedings, even though observers can learn about these views indirectly only, via explicit references in the judgment; the ‘dossier’ with the submissions will only become public knowledge 30 years after the judgment. Ignorance of submissions on the part of external observers should not, however, be equated with irrelevance. Judges

12 By way of example see Regulation (EC) No 1907/2006 concerning the registration, evaluation, authorisation and restriction of chemicals (REACH) [2006] OJ L396/1, arts 89–94, with later amendments; and Regulation (EU) 2017/1001 on the European Union trade mark [2017] OJ L154/1, arts 66–72. 13 See Regulation (EC) No 1907/2006 (n 12) arts 94, 72. 14 Article 13 ECHR does not, unlike Article 6(1) ECHR on civil rights and criminal charges, require an independent tribunal; see Kudła v Poland App no 30210/96 (ECtHR [GC], 26 October 2000) § 157. 15 See Joined Cases C-924/19 PPU and C-925/19 PPU Országos Idegenrendészeti Főigazgatóság Dél-alföldi Regionális Igazgatóság EU:C:2020:367, paras 128–35. 16 For a comprehensive discussion, see Koen Lenaerts, Ignace Maselis, and Kathleen Gutman, EU Procedural Law (OUP 2014).

74 Court of Justice: Achievements and Limitations frequently respond—explicitly or implicitly—to the arguments put forward.17 They have a significant impact on the judicial output, thus turning the right to intervene into a crucial procedural bottleneck. In the case of preliminary references, intervention is limited to the parties of the main proceedings, Member States, and most EU institutions.18 While the former appears self- evident, the latter mirrors the central role of national governments and the supranational institutions for the European polity. The Commission, in particular, is omnipresent. Representatives of its Legal Service, who have an imposing knowledge of Union law, systematically participate in all cases, thus exercising considerable influence. Member States intervene less systematically, mainly in response to references from ‘their’ courts and on politically sensitive subjects.19 More than a dozen governments intervened in high-profile cases such as X and X on humanitarian visas or N.S. on human rights in the Dublin procedure. Intervention gives governments a privileged avenue to influence the output. By contrast, third parties cannot usually intervene, with the exception of Norway, Iceland, and Switzerland in cases relating to the Schengen acquis.20 Neither non-governmental organization (NGOs) nor the UNHCR have an official say, and the rules of procedure do not permit amicus curiae submissions below the threshold of formal intervention, which are widely used in the US and before the ECtHR; a potential exception may be actions for annulment or damages initiated by private parties against Frontex, where attempts by academics to intervene have failed on purely procedural grounds.21 At best, private actors may get an indirect foot in the door of the courtroom via domestic proceedings, or they must make their views known otherwise.22 NGOs and third parties are, so to speak, the lone voice in the wilderness. Visibility of their position in the public debates contrasts with the relative exclusivity of the judicial procedure in Luxembourg.

3.2 Constitutional Authority of the Supranational Judiciary Judgments of the Court were essential for the success of the integration project. In the words of an early commentator: ‘Tucked away in the fairyland Duchy of Luxembourg and blessed, until recently, with benign neglect by the powers that be and the mass media, the Court of Justice . . . has fashioned a constitutional framework for a federal-type structure in Europe.’23 This pedigree of ‘judicial activism’ raised the question in how far the Court would take an active stance on migration law. Assessing the judicial output, we should recognise that the Court serves as a constitutional court and administrative tribunal simultaneously. 17 For a rich empirical analysis, see Marie De Somer, Precedents and Judicial Politics in EU Immigration Law (Routledge 2018) ch 8. 18 See Rules of Procedure (n 4) art 96(1). 19 See Jonas Bornemann, ‘The Role of Member State Governments in Migration Litigation before the ECJ’ (2020) 22 EJML 541. 20 See Protocol [No 3] on the Statute of the Court of Justice [2008] OJ C115/210, art 23(4). 21 See Sergio Carrera and Bilyana Petkova, ‘The Potential of Civil Society and Human Rights Organizations through Third-Party Interventions before the European Courts’ in Mark Dawson and others (eds), Judicial Activism at the European Court of Justice (Edward Elgar Publishing 2013) 233, 241–44; and Case T-600/21 WS and others v Frontex EU:T:2022:474 and EU:T:2022:464, involving Violeta Moreno-Lax and Mariana Gkliati. 22 By way of example, see the UNHCR statement on a pending case, taken up by AG Evgeni Tanchev, Case C- 507/19 Bundesrepublik Deutschland EU:C:2020:768, points 41–45. 23 Eric Stein, ‘Lawyers, Judges, and the Making of a Transnational Constitution’ (1981) 75 Am J Intl L 1.

Constitutional Authority of the Supranational Judiciary 75 While the reputation as a ‘motor of integration’ rests on the constitutional judgments, including human rights (3.2.1), many rulings on migration are defined by an administrative mindset (3.2.1). The Court accentuates classic questions of statutory interpretation and makes an effort to follow the position of the legislature (3.2.3). Experts in migration law must stand ready to chart the often Byzantine waters of doctrinal hermeneutics.

3.2.1 Reputation of ‘constitutional imagination’ Ever since their first term at university, experts in EU law have been told that the Court is a prominent player; the EU as we know it today is, partly at least, a creation of judges in Luxembourg.24 Judges famously emancipated the new legal order from public international law by establishing direct effect and supremacy and have forcefully defended the autonomy of Union law ever since. The Court established a set of unwritten general principles, including fundamental rights, to buttress the doctrinal self-sufficiency of the supranational legal order. Dynamism did not halt at constitutional matters. Judges similarly embraced a dynamic interpretation of Union citizenship and the fundamental freedoms, thereby limiting statal control imperatives. This left the Court with a reputation as a champion of the individual. Nonetheless, critics highlight an inbuilt ambiguity of the case law on individual rights, which followed a functional rationale supporting the effet utile (effective application) of Union law. Direct effect allowed citizens to enforce Union law via domestic courts, and unwritten fundamental rights foreclosed criticism of domestic constitutional courts.25 Judges strengthened individuals, since doing so supported the European project. Tellingly, the Court refrained from enhancing their status when the outcome might have hindered the effet utile. By way of example, Chapter 13.3.6 will explain that judges hesitated to suspend Dublin transfers in light of human rights. At the same time, the case law developed a life of its own and empowered individuals to promote their agenda via Union law. Chapter 5.4 will demonstrate that the Court increasingly refers to the Charter to justify dynamic rulings. In addition, Chapter 6.5 will present examples of how the concept of individual rights, enshrined in directives and regulations, can support third country nationals in the absence of far-fetched constitutional guarantees. Notwithstanding this track record, awareness of the ambiguity between support for the effet utile and individual’s rights is crucial for those studying migration law. It was and is no foregone conclusion that judges would interpret the migration law instruments dynamically. As mentioned in Chapter 1.3, the constitutional foundations of the area of freedom, security, and justice differ from the freedom-enhancing paradigm of the single market and Union citizenship. They lay down diverse and potentially contradictory objectives for law-making that leave the legislature much leeway.

24 See the classic account of Joseph HH Weiler, ‘The Transformation of Europe’ (1990/91) 100 Yale LJ 2403. 25 See Bruno de Witte, ‘Direct Effect, Primacy and the Nature of the Legal Order’ in Paul Craig and Gráinne de Búrca (eds), The Evolution of EU Law (3rd edn, OUP 2021) 187.

76 Court of Justice: Achievements and Limitations 30 25 18

15

25 25

23

22

20

19

17 13

10

9

8 5

5

3

7

8

9

3

0

20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 20 16 20 17 20 18 20 19 20 20 20 21 20 22

0

28

26

Figure 3.2 Number of judgments on EU migration law per year* *See Figure 3.1 for the methodology.

3.2.2 Migration law: ‘administrative mindset’ As times change, so does the case law. Experts in Union citizenship often deplore that the dynamism early judgments gave way to an uncertain outlook, which similarly defines some of the judgments on Union citizenship and free movement.26 Within this overall context, the Court had to define its approach towards migration law when it started dealing with the matter in earnest in the early 2010s, as illustrated in Figure 3.2. At the time, the transposition period for the first wave of legislation had expired, and all domestic courts had obtained the right to make preliminary references. Early judgments may be rationalised as a process of discovery. They gradually embraced the specificities of secondary legislation on justice and home affairs. Doing so distinguished EU migration law from situations of overlap with the single market and Union citizenship, including association agreements.27 The sheer number of judgments meant that the case law on migration law gained momentum, gradually stepping out from under the shadow of other areas of Union law. A bird’s view on the judicial output will rarely be able to present clear-cut conclusions as to whether judges advance state interests or tend to side with migrants. Nevertheless, several trends at an intermediate level of abstraction are discernible. Overall assessments of the judicial output changed other the years, mirroring the increasing self-sufficiency of migration law. Early judgments exhibited classic features of the case law on Union citizenship and the single market: frequent reliance on general principles, predominance of teleological interpretation, and emphasis on individual rights.28 From the mid-2010s onwards, judges increasingly deferred to the position of the legislature and recognised a certain leeway of the Member States, notwithstanding instances of

26 See ch 1.1.7. 27 See ch 1.3.3. 28 See Kees Groenendijk, ‘Recent Developments in EU Law on Migration’ (2014) 16 EJML 313, 329–34; and Diego Acosta Arcarazo and Andrew Geddes, ‘The Development, Application and Implications of an EU Rule of Law in the Area of Migration Policy’ (2013) 51 JCMS 179.

Constitutional Authority of the Supranational Judiciary 77 dynamic innovation.29 A noticeable trend was careful and often scrupulous inspection of secondary legislation, which I have described as an ‘administrative outlook’ elsewhere.30 As starting point, the Court frequently emphasises that, in relation to a specific provision, ‘it is necessary to consider its wording, its origin, its context and the objectives pursued by the legislation’31. Multiple rulings demonstrate the vigour of such analyses, which can stretch over several dozen paragraphs. The newfound prominence of statutory interpretation coincided with a serious of controversial rulings on legal aspects of the asylum policy crisis of 2015/16, which have been poignantly described as an expression of ‘judicial passivism’32. Judges reaffirmed the structural integrity of the Dublin III Regulation (EU) No 604/2013 in the Jafari judgment, evaded a decision on humanitarian visas in the X and X case, and rejected an action for annulment against the EU-Turkey Statement.33 The outcome of all three judgments can be defended from a doctrinal perspective, but the judicial reasoning left observers with the impression that judges had evaded important constitutional questions. The ‘administrative mindset’ was accompanied by a lack of ‘constitutional imagination’34. Politically sensitive judgments on asylum in the aftermath of 2015/16 need not be representative. Chapter 6.5.3 will present evidence that several judgments in the early 2020s have explored the potential of the Charter of Fundamental Rights, which had often been ignored in previous rulings. Nevertheless, the overall picture will, also in future, exhibit shades of grey, no straightforward choice of black or white. The Court may continue to tread carefully on matters that are salient in the public discourse. Moreover, statutory interpretation in light of the wording and the general scheme can be expected to define the majority of judgments. It becomes visible once we shed light into less prominent corners of the multi-faceted case law discussed throughout this volume. Nevertheless, some areas may be defined by dynamism.

3.2.3 Interaction with the legislature Political scientists often describe courts as semi-autonomous actors trying to maximise their power. If that assumption was correct, the analysis would concentrate on ‘court curbing mechanism’ through which other actors mitigate the autonomy of judges.35 However, such focus on court autonomy tends to underestimate procedural and substantive constraints. That is not to say that extra-legal factors are irrelevant, as highlighted by the critical legal studies movement, which will be mentioned in Chapter 4.4. However, external factors do coalesce with classic arguments of legal interpretation, which define the official reasoning.36 29 See Jean-Yves Carlier and Luc Lebœuf, ‘Droit européen des migrations’ [2015] Journal de droit européen 111; and Cian C Murphy and Diego Acosta Arcarazo, ‘Rethinking Europe’s Freedom, Security, and Justice’ in Diego Acosta Arcarazo and Cian C Murphy (eds), EU Security and Justice Law (Hart Publishing 2014) 1, 8–15. 30 See Daniel Thym, ‘Between “Administrative Mindset” and “Constitutional Imagination”’ (2019) 44 EL Rev 138, 148–51. 31 Case C-647/16 Hassan EU:C:2018:368, para 40. 32 Iris Goldner Lang, ‘Towards “Judicial Passivism” in EU Migration and Asylum Law?’ EU Immigration and Asylum Law (24 January 2018). 33 See chs 13.2.4, 11.5, and 3.3.1.1. 34 Thym (n 30) 153–55. 35 See R Daniel Kelemen, ‘The Political Foundations of Judicial Independence in the European Union’ (2012) 19 JEPP 43. 36 See Gunnar Beck, The Legal Reasoning of the Court of Justice of the EU (Hart Publishing 2012).

78 Court of Justice: Achievements and Limitations Judges belong to an epistemic community taking seriously the tradition of doctrinal hermeneutics that are held in high regard among judges, professionals, and academics in the civil law jurisdictions on the European continent. In doing so, the Court simultaneously serves as a constitutional court and as an administrative tribunal. Whereas ‘constitutional’ adjudication authorises judges to correct the outcome of the legislative process in light of higher-ranking norms, such as human rights, ‘statutory’ (or ‘legislative’) interpretation is bound to follow the position of the legislature as a matter of principle. There is ample evidence that judges in Luxembourg generally aim at securing their ‘external’ legitimacy towards the legislature,37 despite the political institutions being in a weak position to overturn Court rulings through legislative change or Treaty amendment because of the procedural rigidity of decision-making.38 Quantitative studies by political scientists have shown that the Court tends to follow the preferences expressed by the other institutions ex ante in the form of legislation or during court proceedings.39 The pertinence of an ‘administrative mindset’ in the case law on migration law fits into this overall picture. From a theoretical perspective, emphasis on statutory interpretation arguably reflects the politicisation of migration law, discussed in Chapter 2.1.4. Judges might feel that it would overstretch their institutional legitimacy to be perceived as ‘activist’ on salient issues. Constitutional theorists generally emphasise that thick constitutional values can hardly be realised by judges alone. Innovative jurisprudence must be embedded in political processes and social practices if it does not want to remain a ‘hollow hope’.40 Accordingly, jurisprudential dynamism in the field of migration may require a broader imagination, which legal academics and judges cannot bring about single-handedly.41 In the absence of a widely shared pan-European vision, judges ‘tread along’ and orientate their reasoning at the legislative output. Having said this, an ‘administrative mindset’ harbours more potential than seemingly technical arguments about the wording, the general scheme, the telos, and the drafting history might suggest. Chapter 6.4 on the doctrinal foundations will discuss numerous examples of the inbuilt open-endedness of many migration law instruments, which often comprise compromise formulae. An ‘administrative mindset’ should not be confused with stasis. Statutory interpretation will often support the position of migrants, and if the legislation remains unclear, judges will vindicate the views of individuals in some cases, while siding with states on other occasions. Instead of a binary conclusion of the Court being predominantly ‘migrant-friendly’ or ‘statist’, countless judgments bear testimony of mixed outcomes: the proverbial shades of grey.

37 See the Court’s President Koen Lenaerts, ‘The Court’s Outer and Inner Selves’ in Maurice Adams and others (eds), Judging Europe’s Judges (Hart Publishing 2013) 13, 17–28; and Vassilis Hatzopoulos, ‘Actively Talking to Each Other’ in Dawson and others (n 21) 102. 38 See Susanne K Schmidt, The European Court of Justice and the Policy Process (OUP 2018). 39 See Olof Larsson and Daniel Naurin, ‘Judicial Independence and Political Uncertainty’ (2016) 70 IO 1; and Clifford Carrubba and Matthew Gabel, International Courts and the Performance of International Agreements (CUP 2014). 40 Gerald N Rosenberg, The Hollow Hope (2nd edn, University of Chicago Press 2008). 41 See Loïc Azoulai, ‘“Integration Through Law” and Us’ (2016) 14 ICON 449, 455–60; and, generally, Seyla Benhabib, ‘Claiming Rights across Borders’ (2009) 103 Am Pol Sci Rev 691.

Access to and Output in the Area of Migration 79

3.3 Access to and Output in the Area of Migration Article 47 CFR gives ‘everyone’ whose rights under Union law have been violated the ‘right to an effective remedy before a tribunal’. While the details of that provision will be discussed in Chapter 7.2 on the administrative dimension, comments hereafter concentrate on access to the Court of Justice. Narrow interpretation of rules for standing mean that only the EU institutions and the Member States will usually be able to access the Court directly (3.3.1). Notwithstanding the visibility of infringement proceedings in the expert debate, inspection of the Commission’s everyday practices demonstrates limited practical relevance (3.3.2). Instead, statistics reaffirm the essential function of preliminary references by domestic courts (3.3.3), although thematic asymmetries in the judicial output indicate that the ‘demand’ for supranational guidance differs markedly between the Member States and regarding subject matters (3.3.4).

3.3.1 Validity disputes Article 263 TFEU does not allow individuals to seize the supranational judiciary to declare national laws to be incompatible with Union law. To do so remains the prerogative of domestic courts, which have to consult judges in Luxembourg in cases of doubt. It is less obvious, however, whether individuals can challenge the validity of supranational legislation. The wording of Article 263(4) TFEU could possibly be interpreted broadly. Nevertheless, the CJEU has defended a narrow reading ever since the seminal Plaumann judgment of 1963. Judges have reaffirmed that position in the new millennium, while emphasising that Member States are under an obligation to establish effective legal remedies domestically.42

3.3.1.1 Direct actions by individuals on migratory matters In practice, the narrow definition of standing means that individuals will rarely be in a position to access the Court directly to challenge compliance of secondary legislation with human rights or other guarantees. Chapter 8.5.3 on agencies will discuss to what extent there is limited room for direct actions when Frontex or the Asylum Agency engage in operational activities. A rare case in which the General Court found a direct action to be admissible—but unfounded—concerned access to documents on border guard ships deployed in an ongoing Frontex operation.43 Controversially, the Court found another action for annulment against the EU-Turkey Statement to be inadmissible, which it qualified as an informal (political) commitment of the Member States meeting within the European Council.44 Even if judges had attributed—correctly, in my view—the Statement to the EU institutions, they would have had to answer follow-up questions. A first admissibility threshold would have been easier to cross, since judges generally interpret jurisdiction to ‘review 42 See Case 25/62 Plaumann EU:C:1963:17; and Case C-583/11 P Inuit Tapiriit Kanatami and others v European Parliament and Council EU:C:2013:625, para 104. 43 See Case T-31/18 Izuzquiza and Semsrott v Frontex EU:T:2019:815. 44 See Joined Cases C-208–210/17 P NF and others v European Council EU:C:2018:705, para 16; confirming Case T-192/16 NF v European Council EU:T:2017:128, paras 46–75; in line with the parallel decisions in Case T- 193/16 and Case T-257/16.

80 Court of Justice: Achievements and Limitations the legality of acts’ under Article 263 TFEU broadly to include any measure intended to produce legal effects45—a condition the EU-Turkey Statement might possibly have fulfilled.46 Claimants would have had more difficulties demonstrating that they are individually concerned in accordance with the Plaumann formula. Be it as it may, the inadmissibility of a direct legal challenge entailed that Greek courts held the responsibility for ensuring compliance with Union law. We shall see that this indirect route of decentralised adjudication proved ineffective in practice.

3.3.1.2 Privileged access by the EU institutions Member States, the European Parliament, the Council, and the Commission can access the Court directly, mirroring privileged standing rights of the political institutions before domestic constitutional courts. Many of the early judgments on migration were such interinstitutional proceedings: several disputes concerned institutional prerogatives with regard to the adoption of implementing or delegated acts, and the limits of the British-Irish opt-out;47 Hungary and Slovakia challenged—unsuccessfully—the mandatory relocation of asylum seekers;48 and the action for annulment, which the European Parliament had brought against the Family Reunification Directive, was critical for the doctrinal foundations of EU migration law.49 On the whole, however, such spectacular proceedings have remained singular events. Member States and EU institutions will not usually challenge the validity of acts adopted with their consent. Moreover, the judgment on family reunification is representative insofar as judges will rarely find secondary legislation to be in outright violation of human rights and, hence, null and void. Rather, they will insist on consistent interpretation, thus weaving human rights into the meaning of secondary legislation. Interpretation in light of human rights can be of immense practical importance, but it is an unattractive outcome for the plaintiff, since it loses the case officially. That is why actions for annulment are a practical rarity before the Court.

3.3.2 Infringement proceedings In public international law, conflicts are usually settled outside courts by means of dispute settlement or seemingly archaic forms of countermeasures. Even if an international tribunal holds jurisdiction, it will usually be seized by states directly. Against this background, we understand a critical innovation of the EU legal order. Instead of states taking legal actions against each other, the Commission gained the reputation as ‘guardian of the Treaties’ through the repeated and strategic use of infringement proceedings against reticent Member States.

45 See Case C-27/04 Commission v Council EU:C:2004:436, para 44. 46 See Mauro Gatti and Andrea Ott, ‘The EU- Turkey Statement’ in Sergio Carrera and others (eds), Constitutionalising the External Dimension of EU Migration Policies in Times of Crisis (Edward Elgar Publishing 2019) 175. 47 See ch 2.3.3 and 2.4.2. 48 See ch 13.3.2. 49 See ch 6.5.2.

Access to and Output in the Area of Migration 81 Table 3.1 Procedure underlying Court judgments on migration law up until 2022* Preliminary References

231 (85.7%)

Infringement Proceedings

22 (8%)

Actions for Annulment

14 (5%)

Opinion (Art. 218 TFEU)

1 (0.3%)

*See Figure 3.1 for the methodology.

3.3.2.1 Importance of the pre-litigation stage Infringement proceedings have generally lost relevance. Between 1990 and 2020, the number of Member States more than doubled and Union law embraced ever more subject areas, including migration. Nevertheless, the overall number of infringement proceedings remained stable: it stood at 903 in 2020.50 Closer inspection unveils remarkable details. Almost two-thirds of the new files concerned ‘late transposition’ of new directives. Such cases undoubtedly exercise useful pressure but are, by and large, a formality; moreover, less than 5 per cent of the infringement proceedings are referred to the Court.51 The reasons are diverse: Member States comply belatedly, the Commission’s legal assessment might change, or it decides not to pursue the matter for political or other reasons. Settled case law holds that the Commission has largely unfettered discretion whether to initiate proceedings and whether to refer a file to the Court. Having said this, the pre-litigation stage is much more than a formality. It provides the Commission with a platform to apply pressure. If Member States follow suit, the Commission achieves the objective in a quick and pragmatic manner. Court judgments may be visible to the external observer, but they consume time, can lead to political tensions, and cost scarce resources.52 Complaints by private actors, including NGOs, used to be an effective way to draw the Commission’s attention to potential breaches of Union law. In the late 1980s, almost 20 per cent of all complaints lead to formal infringement proceedings, while the ratio is much lower nowadays. In 2020, roughly 5 per cent of all complaints were pursued further, and migration was not among the priority areas where the Commission took action.53 Complaints by individuals and NGOs with the Commission are not particularly successful in practice anymore. 3.3.2.2 Changing dynamics in migration law A statistical survey, summarized in Table 3.1, confirms that the Commission does not take the lead in enforcing migration law (the data excludes cases resolved in the pre-litigation phase). In the early years, there was quite a number of judgments, but they mainly concerned late transposition, often for instruments of minor political importance, such as transit for purposes of removal by air. Designation to chambers of three judges confirms 50 Contrast the Commission, ‘General Statistical Overview’ SWD(2021) 212 final; with the ‘Eighth Annual Report to the European Parliament on Commission Monitoring of the Application of Community Law’ COM(91) 321 final. 51 ibid 21–24. 52 See Jos Hoevenaars, A People’s Court (Eleven 2018) 238–42. 53 See ‘Eighth Annual Report’ (n 50) 146; and ‘General Statistical Overview’ (n 50) 17–18.

82 Court of Justice: Achievements and Limitations that these early infringement proceedings had little relevance in practice, despite being visible in our statistical survey. More than two-thirds of all (!) infringement proceedings that have resulted in a judgment until the end of 2021 concerned late transposition.54 For many years, the Commission evaded bringing Member States to Court on legally difficult or politically sensitive topics. In the early years, only one infringement procedure resulted in a judgment that left a mark on the evolution of EU migration law with regard to fee levels.55 This hesitation left observers with a sense of disappointment,56 even if we figure in that the significance of infringement proceedings has decreased generally. The period of passive supervision came to an end during the policy crisis of 2015/16 when it became apparent that administrative practices on diverse subject areas such as reception conditions and the speed or quality of asylum procedures were deficient, if not dysfunctional. Infringement proceedings for failure to effectively apply EU legislation on the ground were a novelty; they provided the Commission with an opportunity to repel criticism of inertia.57 Hungary became a recurring ‘client’ of the infringement procedure when it openly refused to comply with the Council Decisions on the relocation of asylum seekers, and when it introduced asylum laws in flagrant violation of EU legislation. Several of these files ended up before the Court.58 These activities are important, yet the practical drawbacks are evident as well. Chapter 13.3.2 on asylum will demonstrate that infringement proceedings for non-compliance with the Relocation Decisions on quota-based mandatory relocation did not have practical effects, as the two-year period for relocation had elapsed when the Court delivered the judgment. Another ruling on Hungarian transit zones was delivered no less than five years after the Commission’s letter of formal notice and two years after referral to the Court.59 It came shortly after a judgment on the same matter, in response to a preliminary reference by a Hungarian court. Notwithstanding these practical limitations, the example of Hungary illustrates why infringement proceedings are a critical instrument the Commission can—and should— use strategically whenever a government wilfully defies Union law.60 Intentional non- compliance by self-declared ‘rebels’61 go beyond ‘regular’ breaches, which, unfortunately, are widespread across Europe in the domains of asylum and border controls. It is an act of symbolic disobedience that undermines the normative integrity of EU migration law.62 To counter such flagrant violations is an end in itself. Having said this, strategic proceedings against ‘rogue’ governments have risks. The burden of proof for administrative practices on 54 See the judgments in Case C-448/04; Case C-454/04; Case C-455/04; Case C-462/04; Case C-476/04; Case C- 4/07; Case C-5/07; Case C-3/07; Case C-34/07; Case C-57/07; Case C-58/07; Case C-59/07; Case C-256/08; Case C-293/08; Case C-272/08; Case C-322/08; and Case C-139/13. 55 See ch 10.1.3. 56 See Philippe De Bruycker and Henri Labayle, ‘Chronique de jurisprudence consacrée à l’Espace de liberté, de sécurité et de justice’ (2014) 49 Cahiers de droit européen 719, 720. 57 See Commission, ‘More responsibility in managing the refugee crisis’ (Press Release IP/15/5699, 23 September 2015); and the follow-up action (Press Release IP/15/6276, 10 December 2015; IP/17/5002, 7 December 2017). 58 At the end of 2022, five such cases had been referred to the Court (not all had been decided); see Case C-718/ 17; Case C-808/18; Case C-761/19; Case C-821/19; Case C-823/21; and https://ec.europa.eu/atwork/applying-eu- law/infringements-proceedings/infringement_decisions/?lang_code=en (accessed 1 March 2023). 59 See Case C-808/18 Commission v Hungary EU:C:2020:1029, paras 60–67. 60 See also No 3 Commission, ‘EU law: Better results through better application’ [2017] OJ C18/10. 61 AG Eleanor Sharpston, Joined Cases C-715/17, C-718/17 and C-719/17 Commission v Poland and others EU:C:2019:917, point 141. 62 See Evangelia (Lilian) Tsourdi, ‘Relocation Blues’ (2021) 58 CML Rev 1819, 1839–42.

Access to and Output in the Area of Migration 83 the ground rests with the Commission,63 thus complicating, by way of example, measures against pushback allegations. Moreover, judgments need not be respected, thus raising the conflict to a new level. When Hungary refused to change domestic legislation on transit zones, the Commission went to Court again and asked for financial penalties under Article 260(2) TFEU.64 We can only hope that the Hungarian government will respect the outcome.

3.3.3 Preliminary references Judges in Luxembourg are indispensable for the uniform interpretation of EU migration law. Nevertheless, the Court cannot handle, for reasons of capacity, more than the well- known tip of the iceberg. Domestic courts are essential actors in the European variant of federalism, whose multi-level system will be portrayed in Chapter 7.1 on the administrative dimension. Statistical data confirms the pivotal function of domestic courts. Two-thirds of all new cases before the Court are preliminary references.65 We have seen on the previous pages that the ratio is even higher in migration law: more than 86 per cent of the judgments delivered so far have responded to preliminary references, which are the ‘keystone’66 of the court architecture. Involvement of domestic courts has advantages, which help explain why the CJEU interprets rules on the direct access of individuals narrowly, while insisting on low admissibility thresholds for preliminary references. Judges in Luxembourg appreciate the division of labour with domestic courts, which serve as gatekeepers. They do the essential groundwork in terms of fact-finding, have an intimate knowledge of national laws, and, crucially, sit in courthouses across Europe. The Court concentrates on the interpretation of Union law, while leaving its application to the main proceedings to the referring courts.67 Preliminary references are the procedural equivalent of tentacles of sea creatures that permit the Court to reach out to judges in the Spanish Extremadura or the Romanian Transylvania. That does not mean, however, that all courts engage in active dialogue.

3.3.3.1 Stark discrepancies between the Member States Preliminary references on migration started reaching the Court in greater numbers after the Treaty of Lisbon had discontinued an earlier limitation to the highest courts.68 Readers who are not familiar with the Court practice might be surprised that so many references come from a few countries, as illustrated in Figure 3.3. In the field of migration, two jurisdictions (Germany and the Netherlands) account for 40 per cent of all references. If we add Belgium, Italy, and France, five countries account for more than two-thirds of all references. By contrast, the courts of more than half of the Member States (15) have sent three references or less. That contrast is significant even if it mirrors the overall picture of wide discrepancies; German and Italian courts are generally the most willing collaborators.69 63 See Case C-441/02 Commission v Germany EU:C:2006:253, paras 48–49. 64 Commission, ‘Migration: The Commission refers Hungary to the Court’ (Press Release IP/21/5801, 12 November 2021). 65 See CJEU, Annual Report 2021. Judicial Activity (March 2022) 243. 66 Opinion 2/94 Accession to the ECHR EU:C:1996:140, para 176. 67 See Case C-235/95 Dumon and Froment EU:C:1998:365, para 25. 68 See ch 1.2.4; and Steve Peers, ‘Mission Accomplished?’ (2011) 48 CML Rev 661, 681–85; on inadmissibility earlier see Case C-51/03 Georgescu EU:C:2004:200, paras 28–33. 69 See Annual Report 2020 (n 65) 264–68.

84 Court of Justice: Achievements and Limitations 47

50

47

45 40 35 30

24

25 15 10 5 0

19 14

13 7

12

9 3 2

0 1

7 3

1 0

6 1 2 2 0

2 0 0

4

6 0

AT BE BG CZ CY DE DK EE ES FI FR GR HR HU IE IT LT LU LV MT NL PL PT RO SE SI SK UK

20

Figure 3.3 Preliminary reference on EU migration law per country during 2006–22* *Abbreviations follow the internet top level domain; on the methodology see Figure 3.1.

This incongruent practice implies that certain perspectives and themes do not reach the Court. While the substantial number of references from the founding members reflects both the general orientation of the judiciary towards Union law and the practical importance of migratory issues, the relative insignificance of other countries invites comments. The low numbers for Denmark ( 0), Ireland (7), and the UK (6, before Brexit) can be explained by their opt-outs. In addition, the lesser visibility of the domestic courts from Central and Eastern Europe—with the exception of Bulgaria (7) and Hungary (12)—can be rationalised by the comparatively low number of third country nationals living there.70 Nonetheless, one might have expected larger countries like Spain (9), Poland (2), or Romania ( 0) to play more than an ancillary role. Most striking, however, is the marginal role of Greek courts, which referred only one case on the Return Directive.71 Not a single Greek court had consulted judges in Luxembourg on how to interpret the asylum acquis in the aftermath of 2015/16 until 2022.72 In a remarkable move, the Greek Council of State had declared return of Syrians to Turkey on the basis of the safe third country provision to be compatible with Union law without asking the Court of Justice, until it finally referred a related question on whether Turkey still qualified as safe in light of an enduring refusal to take anyone back in 2023.73 The original decision not to refer was incompatible with the well-known obligation, on the part of the highest courts, to make a reference under Article 267(3) TFEU in cases of doubt. The Greek example may be an extreme case, but it is illustrative of a more general divergence between the theory of mandatory referral and widespread factual flexibility. It is quite simply inconceivable that

70 Most references from these countries concerned asylum. 71 See Case C-184/16 Petrea EU:C:2017:684. 72 On the practice of Greek courts see Angeliki Papapanagiotou-Leza and Stergios Kofinis, ‘The Case of Greece’ in Madalina Moraru and others (eds), Law and Judicial Dialogue on the Return of Irregular Migrants from the European Union (Hart Publishing 2020) 281. 73 See the original 13:12 vote against referral by the (Greek) Council of State (Συμβούλιο της Επικρατείας), App nos 2347/2017 and 2348/2017 (judgments of 22 September 2017); and ibid, App no 177/2023 (decision of 3 March 2023).

Access to and Output in the Area of Migration 85 the highest courts dealing with migration law are not confronted with more questions to which the answers are not evident.

3.3.3.2 Carrots and sticks for loyal cooperation Theoretically, the Commission could have initiated infringement proceedings against Greece when the highest administrative court refused to refer an important case, but this option should not detract from the necessity of voluntary cooperation. Empirical studies show that divergences between the Member States can be explained (not: justified) by differences in the organisation and self-image of the judiciary, knowledge of Union law, pragmatic considerations like length of proceedings or workload, personal circumstances such as interest in European affairs, and the role of parties.74 Political scientists highlight, furthermore, that the CJEU activated domestic courts by pleading to their self-interest. Primacy of Union law enhanced their power in countries which have traditionally known weaker court supervision. Telling examples include a Polish court seeking support from the Court against the exclusion of visa refusal from its jurisdiction and Hungarian references extending the scope of judicial oversight.75 Lower courts can be particularly fruitful partners, since they may use cooperation with the Court of Justice to circumvent domestic judicial hierarchies.76 An avalanche of preliminary references from Spanish courts reached Luxembourg after the Supreme Tribunal had interpreted domestic rules in contradiction with the Long-Term Residents Directive.77 Impulses from lower courts remain essential for the dynamic evolution of EU migration law. Moreover, the option of decentralised referrals is a motivation for apex courts to take the obligation to make a reference seriously. After all, referring courts exercise influence by framing the question in a specific manner.78 Publication of the order for reference of domestic courts, from 2020 onwards, will make available an important document external observers had previously been unaware of. Non-referral need not signal trouble. As a rule of thumb, preliminary references will usually be an indication of loyalty and compliance, but the reverse statement need not be correct. Silence can mean many things: for the one, national legislation and judicial practice may comply with EU migration law; for the other, domestic courts may listen carefully, in passive obedience, to what the Court says in response to referrals from elsewhere.79 In many cases, however, silence will indicate potential problems, also considering that domestic courts may misread the meaning of secondary legislation whose ‘legal concepts do not necessarily have the same meaning [than in the legal orders of the] Member States’.80

74 See Jasper Krommendijk, ‘The Preliminary Reference Dance between the CJEU and Dutch Courts in the Field of Migration’ (2018) 10 Eur J Leg Stud 101, 122–41. 75 See Case C-403/16 El Hassani EU:C:2017:960, paras 39–41; and Case C-585/16 Alheto EU:C:2018:584, paras 107–13. 76 See generally Karen J Alter, ‘The European Court’s Political Power’ (1996) 19 Western Eur Pol 458, 466–71. 77 See Case C-448/19 Subdelegación del Gobierno en Guadalajara EU:C:2020:467; parallel references in cases Case C-531/19; Case C-533/19; Case C-534/19; Case C-549/19; and Case C-567/19 were repealed after the judgment; further Diego Acosta Arcarazo, ‘Court of Justice of the European Union (Ninth Chamber), WT v Subdelegación del Gobierno en Guadalajara’ (2020) 22 EJML 457. 78 See also Daniel Thym, ‘Friendly Takeover, or: the Power of the “First Word”’ (2020) 16 EuConst 187, 204–206. 79 See Jasper Krommendijk, National Courts and Preliminary References to the Court of Justice (Edward Elgar Publishing 2021) ch 7; and Galina Cornelisse and Madalina Moraru, ‘Introduction’ in Moraru and others (n 72) 17. 80 See Case 283/81 CILFIT EU:C:1982:335, para 19.

86 Court of Justice: Achievements and Limitations

3.3.3.3 Between micromanagement and vagueness Notwithstanding the positive description of the preliminary reference procedure as a ‘dialogue’81, national judges occasionally lament the absence of a genuine conversation and the ‘ivory tower’ logic of the supranational judiciary. The sense of frustration will be particularly strong if domestic courts are left with the impression that the Court merely provides abstract answers, or did not adequately grasp the legal and practical specificity of the preliminary reference.82 Such frustration may even be a reason, for domestic courts, not to make another reference. This does not suggest, however, that the Court should always give detailed instructions.83 To respond to preliminary references means to walk a tightrope between micromanagement and vagueness. The appropriate balance is difficult to keep and ultimately depends on the circumstances. Generally, the CJEU should defer to domestic courts whenever the answer to a question requires an intimate knowledge of either the facts or domestic law. Having said this, the application of that standard can differ markedly, as the case law on asylum decisions illustrates. While some judgments engage in detailed prescriptions, such as a ‘strong presumption’84 that refusal to perform military service in Syria during the second half of the 2010s supports refugee status, other rulings retreat to abstract formulations. A telling example is the Elgafaji formula on a gliding scale about indiscriminate violence in civil wars, which will be discussed in Chapter 13.6.2 on subsidiary protection. The Court provided guidance at such a high level of abstraction that the common standard will rarely result in synchronised assessments of the situation in specific countries. Divergences between recognition rates for different countries of origin can be traced back, amongst others, to the open-endedness of the supranational case law. Experts in migration law should recognise that the intuitive call for further refinement need not be the optimum. Limited resources require judges in Luxembourg to tread carefully, instead of answering every question in detail. A resilient working relationship with domestic courts is essential for the effectiveness of EU migration law. The Court of Justice should strive for thorough instructions whenever a case concerns politically or legally important questions. It should, in particular, invest time and energy in coherent rulings on horizontal questions that define the evolution of EU migration law, while leaving domestic courts to micromanage situations which primarily relate to one specific country or require an intimate knowledge of the facts.85 A classic example of political salience are controversial pre-departure language tests as a precondition for family reunification, which will be analysed in Chapter 15.4 on integration and where the Court gave detailed instructions. Meanwhile, a fine example of conceptual importance is found in the meaning of ‘public policy’ justifying the rejection or withdrawal of residence permits, discussed in Chapter 10.3 on overarching themes of secondary legislation.

81 Accession to the ECHR (n 66) para 176. 82 See Krommendijk (n 79) chs 5–6. 83 See Daniel Sarmiento, ‘Half a Case at a Time’ in Catherine van de Heyning and Maartje de Visser (eds), Constitutional Conversations in Europe (Intersentia 2012) 13, 26–40. 84 Case C-238/19 Bundesamt für Migration und Flüchtlinge EU:C:2020:945, para 57. 85 See also Alexander Kornezov, ‘The New Format of the Acte Clair Doctrine and Its Consequences’ (2016) 53 CML Rev 1317.

Access to and Output in the Area of Migration 87 10

46

27

Institutions

7

Border controls Visas Asylum

61

Legal Migration Return

117

Asylum cases dealt with:

117 (44%)

Dublin/Relocation Procedures Directive Qualification Directive and temporary protection

38 33

Reception Conditions Directive

6

Legal migration cases concerned:

40

61 (23%)

Family reunification Long-term residents

25 22

Economic migration, incl. students

10

Other Areas

4

Figure 3.4 Subject areas of Court judgments on migration law up until 2022

3.3.4 ‘Demand’ side Legal experts tend to overestimate the influence of courts. While some subject matters rarely end up in court, other domains feature prominently in the judicial output. Statistics from Germany illustrate that administrative courts had to deal with veritable ‘waves’ of litigation on asylum, depending on the number of applications, while legal migration for economic and other purposes ends up before domestic courts much less frequently.86 There are similar discrepancies in the number of CJEU judgments on different themes.

3.3.4.1 Thematic asymmetries in migration law Our statistical survey of the case law confirms that certain issues are more frequent than others. Figure 3.4 shows that more than four out of ten judgments concerned asylum, reflecting the controversial nature of many proceedings and the high number of asylum applications in Member States whose domestic courts make many references. Within Germany, the ratio of negative asylum decisions being challenged stands just below 90 per cent,

86 Approximately 50 per cent of all decisions by administrative courts concerned asylum during 2020, compared to 7 per cent for other aspects of migration law (including return) and 1.7 per cent for environmental protection (to mention another area); see Destatis, Justizgeschäftsstatistik 2020 (Fachserie 10 Reihe 2.4, August 2021) 12.

88 Court of Justice: Achievements and Limitations Table 3.2 Subject areas of Court judgments on migration law up until 2022* Asylum cases dealt with: Dublin/Relocation Procedures Directive Qualification Directive and temporary protection Reception Conditions Directive Legal migration cases concerned: Family reunification Long-term residents Economic migration, incl. students Other Areas

117 (44%) 38 33 40 6 61 (23%) 25 22 10 4

*For

the methodology see Figure 3.1; again, I am grateful to Jonas Bornemann, Kilian Umbach, and student assistants for support in the categorisation.

considerably higher than for other segments of migration law.87 Explanations involve the often existential significance of the outcome for individuals, the legal and factual complexity of many proceedings, as well as benefits like the suspension of return during the court procedure. The high proportion of CJEU rulings on the Dublin system, asylum procedures, and refugee status determination, as opposed to reception conditions, summarized in Table 3.2, may be rationalised by the density of supranational prescriptions. By contrast, those living in a third state are much less likely to seize courts: the number of judgments on visas and border controls is low, despite the high number of third country nationals concerned and the density of legislative harmonisation. Allegations of pushback practices at the external borders received much attention, and the consulates of all Schengen countries reject more than one million visa applications per year. Nevertheless, few references reach the Court on these matters, reflecting the absence of litigation at the domestic level.88 Reasons include logistical, financial, and practical difficulties in accessing domestic courts from abroad, a low chance of success, and the absence of immediate benefits during the court procedure. The example of the Return Directive illustrates the relational character of preliminary references. A first judgment on the criminalisation of illegal entry or stay, which will be presented in Chapter 16.3.3, set in motion a ‘snowball effect’,89 with several follow-up references testing the limits of the first ruling. Similar effects can be observed for take back requests under the Dublin system, the scope of legal remedies against asylum decisions, and the use of specialised detention facilities.90 A first judgment may kick-start a process of discovery. References from different countries indicate that Court judgments are read across the continent. We can expect similar developments in the future. Step by step, new subject matters gain prominence.

87 See (German) Federal Government (Bundesregierung), ‘Ergänzende Informationen zur Asylstatistik’ (Bundestag doc 20/2309, 17 June 2022) 50. 88 During 2019 (before the pandemic), the German complaint ratio for refusal of short-term visas stood at 0.4 per cent; and was only slightly higher with 3.7 per cent for long-term visas; see (German) Federal Government (Bundesregierung), ‘Visaerteilungen im Jahr 2018 und im ersten Halbjahr 2019’ (Bundestag doc 18/14701, 4 November 2019) 6, 35. 89 Galina Cornelisse, ‘The Scope of the Return Directive’ in Moraru and others (n 72) 41, 58. 90 See chs 13.3, 7.2.3, and 13.7.4.

Access to and Output in the Area of Migration 89

3.3.4.2 Strategic litigation to the benefit of migrants The Court’s reputation of dynamic rulings rested, as we have seen, on the doctrine of direct effect and the preliminary reference procedure providing legal entrepreneurs with an avenue to challenge state practices. It was attractive for individuals to involve the supranational judiciary. Judges in Luxembourg often approached domestic laws from a different perspective than the highest national courts. Case law on Turkish nationals, mentioned in Chapter 17.4 on association agreements, is an excellent example of how committed individuals could disrupt deep-rooted domestic patterns under recourse to a seemingly obscure body of rules. A loose group of lawyers and academics from the Netherlands, Germany, the United Kingdom, and Austria succeeded in bringing dozens of cases to the CJEU, thereby effectively decoupling the legal status of Turkish nationals under the association acquis from ‘regular’ migration law towards third country nationals.91 More recently, an active practitioner from Copenhagen built upon this tradition by orchestrating several Danish references on the Turkish association acquis.92 Sociolegal studies on the making of preliminary references within the area of freedom, security, and justice shed light on underlying dynamics. European rules establish a structure of opportunity not anyone will use. To do so requires agents in the form of lawyers, NGOs, or support networks to identify prospective lead cases and to convince domestic courts to refer the matter. A thorough understanding of Union law, the practice of domestic judges, political determination beyond the individual case, as well as logistical, academic, and financial support can be relevant factors that define the potential of strategic litigation.93 If successful, individual actors exercise considerable sway over the evolution of EU migration law. At the same time, however, success is no foregone conclusion. Qualitative studies on strategic litigation before the CJEU have indicated mixed results, especially with regard the long-term transformation of the case law.94 To bring a case to Luxembourg, will rarely be enough. Ideally, it needs to be well-chosen in terms of factual background and legal circumstance.95 Moreover, states ‘fight back’ by intervening strategically in important cases, thereby exploiting their privileged procedural status. Increasing politicisation arguably renders dynamic interpretation less likely at a time when an administrative outlook defines, as we have seen, the supranational judicial output. Experience from the US highlights that transformative litigation should be embedded in broader political strategies to succeed in the medium run.96 91 Proactive lawyers and advisers include: Dirk Schaap and Kees Groenendijk from the Netherlands; Rolf Gutmann, Hagen Lichtenberg, and Rainer Hofmann from Germany (the German government was frequently advised by Kay Hailbronner); Nicola Rogers from the UK; and Wilfried Ludwig Weh from Austria; see Kees Groenendijk, ‘The Court of Justice and the Development of EEC-Turkey Association Law’ in Daniel Thym and Margarite Zoeteweij-Turhan (eds), Rights of Third-Country Nationals under EU Association Agreements (Brill Nijhoff 2015) 39, 57–60; and Kees Groenendijk, Merle Rondhuis, and Tineke Strik, ‘Klagen bij de Europese Commissie’ (2015) 6 Asiel & Migrantenrecht 24. 92 Thomas Ryhl is mentioned as the representative in many cases; for a list of all judgments see Part 4 NEMIS Newsletter https://cmr.jur.ru.nl/nemis (accessed 1 March 2023). 93 See Hoevenaars (n 52) chs 5–7; and Virginia Passalacqua, ‘Legal Mobilization via Preliminary Reference’ (2021) 58 CML Rev 751, 764–75. 94 See Moritz Baumgärtel, Demanding Rights (CUP 2019); and also De Somer (n 17) ch 8. 95 Baumgärtel (n 94) ch 8. 96 See Austin Sarat and Stuart Scheingold, ‘The Dynamics of Cause Lawyering’ in Austin Sarat and Stuart Scheingold (eds), The Worlds Cause Lawyers Make (Stanford UP 2005) 1; and also Daniel Thym, ‘The End of Human Rights Dynamism?’ (2020) 32 IJRL 569, 588–95.

90 Court of Justice: Achievements and Limitations

3.4 Alternative Accountability Mechanisms Flagrant deficits in core areas of asylum law and border controls stark discrepancies in the number of judgments on different themes accentuate the need for alternative accountability mechanisms besides judicial oversight: think of the inadequate reception conditions in the ‘hotspots’, the failure of the takeback procedure under the Dublin system, or pushbacks allegations against border guards. Moreover, cooperation with third states can be difficult to get a hold of from a legal perspective. Court proceedings are essential, but they should be complemented—not replaced—by alternative instruments to ensure that the law is being applied in practice (3.4.1). EU migration law benefits from the involvement of the Ombudsperson (3.4.2), financial and managerial accountability via the Court of Auditors (3.4.3), political oversight by parliaments (3.4.4), and access to documents as a catalyst for public scrutiny (3.4.5).

3.4.1 Accountability for wrongdoing beyond courts Most legal accounts intuitively focus on courts as the primary, if not sole, forum for holding public authorities to account. That focus on the law and courts often conceals the added value of other accountability fora and control standards. Such alternative mechanisms attracted much attention after the change of the millennium, as an integral part of the new modes of governance.97 In the field of migration, accountability beyond courts is about complementary avenues for ensuring compliance, not about replacing legally binding rules and court oversight by soft law instruments, as it occasionally did in other domains. Accountability need not be confined to respect for legal rules. It is generally conceived of as an umbrella concept embracing non-judicial control fora and non-legal control standards. At an abstract level, we can distinguish diverse forms of political (towards parliaments, the media, etc.), financial (vis-a-vis auditors or anti-fraud bodies), administrative (ombudspersons, controllers, complaint bodies), legal (via courts), and social (towards civil society and the general public) channels of accountability.98 These different mechanisms and standards are not mutually exclusive and may reinforce each other. Thus, the Ombudsman prepared the ground for the introduction and reform of the internal complaints mechanism for Frontex. Inquiries by international expert bodies, under human rights law, concerning state practices at the external borders can unearth useful information that can be used in litigation before courts later-on. Having said this, the specific outlook of the accountability research is not confined to preparing the ground for legal remedies before courts. Non-judicial mechanisms can be an instrument in their own right to draw attention to existing shortcomings and to activate diverse actors and resources to overcome them. The added value of political, financial, administrative, and social accountability becomes evident when misconduct does not amount to a violation of judiciable legal standards. Cooperation with third states is a prime example

97 See ch 2.3.4, 2.3.6. 98 See Carol Harlow, Accountability in the European Union (OUP 2002) ch 1; and Mark Bovens, Deirdre Curtin, and Paul ’t Hart, ‘Studying the Real World of EU Accountability’ in Mark Bovens and others (eds), The Real World of EU Accountability (OUP 2010) 31, 41–56.

Alternative Accountability Mechanisms 91 of a subject matter where recourse to extra-legal standards in non-judicial fora can be particularly useful. Comments that follow concentrate on formal mechanisms that oblige an actor to respond to inquiries about potential misbehaviour. Such institutionalised mechanisms will often result in a decision, which will not usually be legally binding. Accountability in this sense concentrates on ex post supervision, not diverse forms of ex ante influence, such as training, appointment procedures, or reporting requirements, which will be mentioned in Chapter 8.5.2 with regard to the governance of the migration agencies. Our inspection of administrative, financial, political, and social accountability will present one example each. Focus on the supranational level is not meant to ignore the practical relevance of additional domestic mechanisms.

3.4.2 Ombudspersons as vehicles of administrative control Informal complaints with (semi-)independent inspectors or watchdogs have existed for centuries, but the institution of the ‘European Ombudsman’ is very much a proliferation of the Scandinavian model. It was formalised under the Treaty of Maastricht and has been emulated domestically, not least in Central and Eastern Europe. The added value of the European Ombudsman stems from easy accessibility via a complaints mechanism open to all Union citizens and third country nationals residing within Union territory.99 There are no restrictive standing rules, which complicate direct access to the Court of Justice.100 Inquiries result in a ‘decision’, which may not be legally binding but can exercise political and administrative pressure. A fine example are own-initiative reports on the initial absence and later dysfunctionality of the internal complaints mechanisms for Frontex. They played a critical role in establishing and improving a mechanism, which effectively replicated the Ombudsman’s administrative accountability function within the Warsaw-based agency.101 In addition, proposals were put forward how Frontex could improve the planning and conduct of operations.102 Notwithstanding broad jurisdiction for ‘maladministration’, which can cover misconduct below the threshold of illegality, many reports on migration concentrate on legal standards. An important instrument are own-initiative inquiries, which allow the Ombudsman to draw attention to structural weaknesses court judgments often struggle to address systematically in the absence of claimants. Most activities in relation to Frontex emanate from such own-initiative inquiries. Informality can be another advantage, allowing the Ombudsman to get active before another institution takes a final decision. This may prevent tensions from escalating, also considering that the Ombudsman can intervene at short notice, whereas 99 See TFEU, art 228(1). 100 See Decision 94/262/ECSC, EC, Euratom governing the performance of the Ombudsman’s duties [1994] OJ L113/15, arts 2–4, with later amendments. 101 See Ombudsman, ‘Implementation by FRONTEX of its fundamental rights obligations’ (Decision OI/5/ 2012/BEH-MHZ, 12 November 2013); ‘Respect for fundamental rights in joint return operations (JRO)’ (Decision OI/9/2014/MHZ, 4 May 2015); and ‘Frontex complaints mechanism for alleged breaches of fundamental rights’ (Decision OI/5/2020/MHZ, 15 May 2021). 102 See Ombudsman, ‘How Frontex ensures accountability in relation to its enhanced responsibilities’ (Decision OI/4/2021/MHZ, 17 January 2022).

92 Court of Justice: Achievements and Limitations court proceedings often take years. The Ombudsman can make—unlike courts—proposals how existing legislation or administrative practices may be changed, as it did for Frontex. Jurisdiction concerns the EU institutions and bodies, not domestic authorities. That is an important limitation, notably in the areas of border controls and asylum. More than half of the complaints made to the Ombudsman fall outside its jurisdiction; they are usually forwarded to national ombudspersons or similar bodies.103 A related limitation concerns the complaints mechanism within Frontex. The behaviour of national border guards is beyond its reach, although the overwhelming number of complaints concerned their performance.104 Additional ex ante and ex post mechanisms supporting human rights compliance by Frontex staff and national personal include training, involvement of fundamental rights officers, a code of conduct, and reporting obligations.105 Ongoing debates about a future Screening Regulation are considering whether to oblige Member States to set up an independent monitoring mechanism whose jurisdiction ratione materiae remains to be hammered out during the negotiations.

3.4.3 Financial and managerial accountability Money and administrative effectiveness are often overlooked in legal analyses, although much money is regularly dispersed on the basis of the funds mentioned in Chapter 7.4 on the administrative dimension. Even more money is being spent on cooperation with third states; it serves as an essential sweetener to enlist neighbouring countries into collaborative migration management, discussed in Chapter 18.3. Control of budgetary spending and administrative effectiveness can be powerful tools to oversee policies, which are difficult to grasp via classic accountability mechanisms. A crucial institution to do so is the European Court of Auditors (ECA), which serves as the Union’s public audit body.106 Oversight covers technical financial matters of regular and reliable reporting to prevent fraud as well as the sound and cost-effective administration of EU funds beyond the prevention of financial irregularities. Similar functions are performed by the European Anti-Fraud Office (OLAF), which serves as an internal watchdog of the Commission to prevent corruption and mismanagement.107 Both institutions are active in the field of migration. Concerns about mismanagement in terms of achieving policy objectives motivated special reports by the Court of Auditors on (dis)functionalities of the hotspot approach in Greece and Italy, which received almost 2bn EUR in funding during the second half of the 2010s.108 During 2021/22, ECA and

103 See Ombudsman, Annual Report 2020 (May 2020) 30. 104 See Frontex Regulation (EU) 2019/1896, art 111; and David Fernández Rojo, EU Migration Agencies (Edward Elgar Publishing 2020) 191–201. 105 Frontex Regulation (EU) 2019/1896 (n 104) arts 62, 80–81, 108–10, Annex V; and Thomas Groß, ‘Defizite des Grundrechtsschutzes bei FRONTEX-Einsätzen’ [2020] Zeitschrift für Ausländerrecht 51. 106 See TFEU, arts 285–87; and Harlow (n 98) ch 5. 107 OLAF stands for Office européen de lutte antifraude; its powers are defined by Commission Decision 1999/ 352/EC, ECSC, Euratom establishing the European Anti-Fraud Office [1999] OJ L136/20, with later amendments. 108 See ECA, ‘Asylum, relocation and return of migrants’ (Special Report No 24/2019, 13 November 2019); and ‘EU response to the refugee crisis’ (Special Report No 06/2017, 25 April 2017).

Alternative Accountability Mechanisms 93 OLAF censured Frontex, thus triggering the resignation of the former executive director.109 Additional examples will be mentioned in several chapters throughout the book. Audits of spending on external cooperation are relevant, since there are few other and effective accountability fora. NGOs and academics have tried to activate the Court of Auditors to engage in closer scrutiny of the funds channelled into Northern Africa, especially Libya,110 building on an earlier, damning report.111 Similar criticism has been brought forward against the EU Emergency Trust Fund for Africa.112 It should be noted, however, that there is no guarantee of success. The Court of Auditors is not obliged, as a public interest body, to take up complaints by individuals. Moreover, reports are, like the decisions of the Ombudsman, not legally binding; their effectiveness depends on the willingness of other actors to take on board the recommendations, if need be as a result of public pressure (‘naming and shaming’).

3.4.4 Political oversight by parliaments While the European Parliament is known best for co-decision powers in law-making, it exercises an important control function over the Commission and other executive bodies. Leaving aside mechanisms of ex ante control in the appointment of personnel, instruments of ex post supervision include written and oral questions, committees of inquiry, or radical measures such as a motion of censure against the college of Commissioners.113 In 2021, the LIBE Committee set up an extraordinary ‘Frontex scrutiny group’ on allegations of involvement, or tacit acquaintance, of the border agency with pushback practices by Greek border guards.114 The latest version of the Frontex Regulation (EU) 2019/1896 extended parliamentary supervision through reporting obligations. MEPs publicly discuss their plans with the candidates for the position of executive director, can request them to appear in parliament, and serve as one of the budgetary authorities for setting up and discharging the budget.115 Similar powers exist vis-a-vis the Asylum Agency. Moreover, the European Parliament commissions reports written by its research staff or by external academics. They can be useful tools for public scrutiny and policy debates. A telling example was a report on the impact assessment of the Return Directive, which the Commission had failed to put forward.116 All parliamentary reports are publicly available, thus facilitating their reception by expert circles and the general public.

109 See ECA, ‘Frontex’s support to external border management’ (Special Report No 08/2021, 7 June 2021); and OLAF, ‘Final Report Olaf.03(2021)21088’ (2021) https://fragdenstaat.de/dokumente/233972-olaf-final-report- on-frontex (accessed 1 March 2023). 110 See the (informal) complaint to ECA by the Global Legal Action Network (GLAN), ‘EU Financial Complicity in Libyan Migrant Abuses’ (April 2020). 111 See ECA, ‘Frontex’s Support to External Border Management’ (Special Report No 9/2016, 17 March 2016). 112 See Thomas Spijkerboer and Elies Steyger, ‘European External Migration Funds and Public Procurement Law’ (2019) 4 European Papers 493. 113 See TFEU, arts 226, 230(2), 234. 114 See LIBE Committee, ‘Report on the Fact-finding Investigation on Frontex Concerning Alleged Fundamental Rights Violations’ (Rapporteur: Tineke Strik, EP Working doc, 14 July 2021). 115 See Frontex Regulation (EU) 2019/1896, arts 6, 65, 102, 106(2), 107(2), 116. 116 See ch 16.5.

94 Court of Justice: Achievements and Limitations

3.4.5 Access to documents as a source of public debates Information is a precondition for meaningful public debate. The younger generation may have difficulties imagining how the world of EU decision-making looked like up until the 1990s, before the arrival of the digital age and rules on access to documents (for instance under the secretive intergovernmental Schengen cooperation117). Nowadays, ‘transparency’ is a constitutional obligation of the EU institutions.118 They have established public registers, which can be accessed via the internet. Such a public register is particularly relevant for the Council, whose internal workings had traditionally been defined by the habits of diplomatic confidentiality.119 Rules on access to documents can be a useful source of information for those engaging in research on subject matters for which documents are not publicly available. That does not mean, however, that applications for access to documents will always succeed. Confidentiality of the legislative procedure, the effectiveness of operational activities, and the sensitivity of diplomatic relations can be legitimate reasons to reject access.120 As mentioned previously, journalists failed, before the General Court, to oblige Frontex to make information about domestic ships employed in an ongoing border control operation available. Moreover, publication of many official documents means that delicate matters may be discussed in informal ‘non papers’, which have become standard practice in discussions about asylum reform. Unofficially, the website of the non-profit organisation Statewatch serves an invaluable source of documents on justice and home affairs that cannot be accessed via regular channels.121

3.5 Summary Judgments are the hallmark of many legal analyses. Their interpretation and critique are crucial for the conceptual foundations, evolutionary trends, and practical effects of EU migration law. At the same time, widespread focus on the Court of Justice coincides with a lesser interest in the inner workings, including inherent limitations and weaknesses. This chapter set out to lift the veil of judicial decision-making by introducing readers to the internal structure, procedural rules, and practical relevance of the CJEU in migration law. The organisational structure and thematic outlook invite judges to approach EU migration law as an integral part of the supranational legal order; the theoretical option of a specialised tribunal remains elusive for the time being. EU institutions and Member States have privileged access to the court room, while NGOs will only be heard officially when they are associated with a party to the proceedings. Closer inspection of the judicial output demonstrates that the Court serves as a constitutional court and as an administrative tribunal in parallel. Judgments on the constitutional foundations of the European project earned it a reputation of an ‘activist’ court, for 117 See Daniel Thym, ‘The Schengen Law’ (2002) 8 ELJ 218, 221–26. 118 See TFEU, art 15. 119 See ch 2.2.3.1. 120 See Regulation (EC) No 1049/2001 regarding public access to documents [2001] OJ L145/43; and Deirdre Curtin, Executive Power of the European Union (OUP 2009) ch 8. 121 See https://www.statewatch.org (accessed 1 March 2023).

Summary 95 instance with regard to Union citizenship. By contrast, the bulk of the rulings on migration law is defined by a thorough examination of the wording, the general scheme, or telos of secondary legislation. There is less dynamism, although the open-endedness of many provisions makes surprise outcomes possible. Moreover, the Court appeared to attach greater significance to the Charter in the early 2020s, which allows judges to correct the outcome of the legislative process. The preliminary reference procedure is the jewel in the crown of the European court architecture. An empirical inspection of the judicial output in the field of migration shows that 85 per cent of all rulings responded to questions by domestic courts. Narrow rules on standing are the flip side of the emphasis on preliminary references, thus limiting direct access by individuals. Infringement proceedings play a marginal role in the Commission’s practice, notwithstanding symbolically important action against ‘rebels’ who wilfully disrespect EU migration law. References by domestic courts are, also for individuals, the most effective way to obtain a judgment. It is problematic, therefore, that our empirical examination exhibits marked discrepancies in the practice of domestic courts. References from some countries, including Greece, rarely reach Luxembourg, while domestic courts from other Member States are recurring players. Similarly, there are manifest differences in the number of judgments on specific themes. While asylum jurisdiction and procedures, return, or family reunification are often dealt with, legislation on economic migration, visas, or reception conditions is less visible or virtually absent from the docket. That is one reason why alternative accountability mechanisms play an important role, especially in areas which judges have difficulties to get hold of. Cooperation with third states and external border controls are prime examples of the added value of non-judicial accountability instruments via the Ombudsman, parliamentary oversight, financial audits, or public discourse. It does not diminish the pivotal role of judges in Luxembourg to recognise that there are limitations to what supranational adjudication can achieve.

4

Interdisciplinary Perspectives and Methodology European migration law profoundly affects the lives of numerous individuals. Its analysis inevitably raises delicate questions of normative significance which preoccupy citizens, academics, and politicians. Moreover, state policies do not always succeed. Generous rules on the admission of highly qualified migrants will not necessarily translate into abundant blue card applications, and the EU institutions are trying hard to construct an asylum policy that works reasonably well in practice. Inspection of the legal rules alone does not allow us to comprehend underlying dynamics. To do so requires a basic knowledge of interdisciplinary perspectives, which come in diverse forms. The thematic breath of the interdisciplinary debate entails that there is not one sister discipline but multiple perspectives employing different methodologies. Our presentation will concentrate on those aspects which are particularly relevant. Some theories and research findings will be taken up elsewhere. Different conceptions of migrant integration, for instance, will be presented in Chapter 15.1.2 on settlement, and the interdisciplinary outlook on the institutions and the Court was discussed in Chapters 2 and 3. Most interdisciplinary discussions about migration do not usually focus on the EU. They often discuss state policies independent of the relative weight of domestic decisions and supranational legislation. Debates about how to improve supranational or domestic legislation benefit from an awareness of the driving forces behind migratory movements. Numerous ‘push’ and ‘pull’ factors at the micro-, meso-, or macro-level of analysis underlie the choice of destination and the decision whether to leave (4.1). Social scientists have coined the notion of ‘control gap’ to designate failure, on the part of Western democracies, to achieve policy objectives, notably in the field of asylum. Corresponding debates concern the role of diverse actors, public opinion, and individual agency. Their inspection enlightens the exploration of contextual factors influencing policy-making (4.2). Critical scholars will appreciate post-structural theory in the French tradition, which brings to the light ideological predispositions of the law as an instrument of government and critically examines the use of language (4.3). While the critique will often be external to the law, interdisciplinary research can occasionally support the reconstruction of the legal material from within. Doing so remains a formidable challenge. Academics can employ different methodologies ranging from contextual and interdisciplinary analyses to different facets of the continental tradition of doctrinal hermeneutics, which defines the supranational case law (4.4).

4.1 Drivers of Migratory Movements Public discourse, political debates, and academic analyses often refer to ‘migration’ as a seemingly uniform phenomenon. That is beneficial to identify overarching themes, such European Migration Law. Daniel Thym, Oxford University Press. © Daniel Thym 2023. DOI: 10.1093/oso/9780192894274.003.0005

Drivers of Migratory Movements 97 as the topics elaborated upon in the first part of this volume. At the same time, grouping together varied circumstances under the abstract heading of ‘migration’ can conceal differences and specificities. An Indian IT specialist applying for a blue card to work in Dublin is confronted with different legal rules and practical challenges than a Moldovan construction worker with a temporary residence permit for Bulgaria or the grandchildren of Turkish ‘guest workers’ who had moved to the Netherlands decades ago. EU migration law concerns diverse situations which are defined by distinct legal rules and factual dynamics. Even ostensibly uniform groups may have different backgrounds, as the example of asylum applicants arriving on the Italian island of Lampedusa illustrates: Tunisians typically escape economic hardship, Eritreans have often spent thousands of euros to flee from a repressive dictatorship, and Yemenis may have worked in the Libyan oil industry before deciding to leave the country during the civil war. Against this background, we cannot expect the social sciences to provide us with a simple formula determining the choice of destination and the decision whether to leave in the first place. There is not one theory but overlapping theories operating against a backdrop of distinct theoretical propositions and on the basis of divergent methodologies. Abstract references to ‘push’ and ‘pull’ factors are a shorthand for a continuum of numerous explanatory factors influencing the size and direction of migratory movements (4.1.1). Interdisciplinary studies have warned against the ‘sedentary bias’ which conceptualises migration as a simple sequence of departure, journey, and settlement. Migration—as human life in general— habitually goes through multiple episodes and can change over time (4.1.2). Family and ethnic networks, historically embedded corridors, and infrastructure are important factors shaping the volume and direction of movements, or the absence thereof (4.1.3). Of particular relevance for the legal analysis is the overlap of ‘voluntary’ and ‘forced’ migration, which contrasts with an ostensibly clear-cut distinction in the policy discourse and the legal material at the European and international levels (4.1.4).

4.1.1 Numerous ‘push’ and ‘pull’ factors Less than 4 per cent of the global population are estimated to be international migrants in the sense of habitually residing beyond the borders of the country of birth, thus excluding the second and third generation. The percentage has been surprisingly stable over the decades, growing from 2.3 per cent to 3.5 per cent between 1970 and 2020.1 Invocation of ‘push’ and ‘pull’ factors in countries of origin, destination, and transit is useful insofar as it supports the identification of multiple drivers of migration. They do not, however, translate into a quasi-mathematical formula that allows for precise predictions or the exact identification of the variables determining previous movements. Migration is almost always highly context-specific, and the social sciences apply diverse theories and methodologies. While some concentrate on economic considerations, others highlight social, political, cultural, or historical factors; quantitative surveys coexist with qualitative methods, such as interviews or ethnographic participatory observation.2 1 See World Migration Report 2022 (IOM 2021) ch 2; during the same period, absolute numbers have increased from 84 to 272 million as a result of population growth. 2 See Hein de Haas, Stephen Castles, and Mark J Miller, The Age of Migration (6th edn, Palgrave 2020) ch 3.

98 Interdisciplinary Perspectives and Methodology Economic theories were particularly influential in the early debate, which concentrated on rational cost-benefit calculations. Income differentials are an important explanation, although they will rarely be the sole or decisive factor.3 Remember that most people decide not to migrate despite massive global inequalities. We shall see in Chapter 18.4.1 on the external dimension that the simple idea of eradicating ‘root causes’ by means of economic development cannot be expected to work in a straightforward manner. It is not usually the very poor who migrate, since doing so requires financial resources and organisational skills. Human capital models of economic theory account for such contextual factors; they conceive of migration as an investment individuals expect to pay off in the long-run, for either themselves or the families supporting them in the hope for financial remittances. Of course, plenty of other features have to be considered as well.4 Chapter 14.4.1 on economic migration will highlight several factors studies have found to be significant, such as language, the political and social climate, tax levels, or families and friends. Even sunshine can be relevant, notably for the rich and highly skilled, as demonstrated by pensioners residing in Portugal or IT specialists working in California. One element social scientists often do not pay much attention to is the law. That ignorance can be a useful challenge to the intuitive assumption, on the part of legal experts, that the law is a decisive factor. For highly skilled economic migrants, in particular, legal pathways are a necessary but insufficient variable; generous entry rules alone will not necessarily attract the best and the brightest. In the field of asylum, however, the reverse conclusion holds true, to a certain extent at least. While some are deterred by visa requirements, carrier sanctions, and border controls, others find irregular channels. It would be unrealistic to expect states to stop migration, even though the impact of policies cannot usually be measured precisely. The counterfactual question of how many people would decide to go the Europe if visa requirements and entry restrictions were abandoned cannot be answered in a methodologically satisfactory manner. One element obscuring the impact of legal rules is the comparatively low level of information on the part of many migrants. Empirical studies show that forced migration is often defined by diffuse knowledge and spontaneity; information obtained via social media or word-of-mouth marketing can be significant, irrespective of whether it is adequate.5 Moreover, state practices are complemented by other push and pull factors, whose relative weight fluctuates over space and time.6 Several of them will be discussed on the pages that follow. A prominent argument, in both academic writing and the policy debate, is the ‘welfare magnet’ hypothesis, according to which a high level of social benefits will attract fewer qualified migrants, while discouraging the qualified who prefer to pay lower taxes.7 It has been discussed controversially, with multiple studies finding some supportive indicators and many counterarguments.8 The ‘welfare magnet’ idea is generally considered to be simplistic 3 cf George Jesus Borjas, ‘The Economics of Immigration’ (1994) 32 J Econ Lit 1667. 4 See Douglas S Massey and others, ‘Theories of International Migration’ (1993) 19 Popul Dev Rev 431. 5 See Heaven Crawley and Jessica Hagen-Zanker, ‘Deciding Where to Go: Policies, People and Perceptions Shaping Destination Preferences’ (2019) 57 Int Migr 20; and Joris Schapendonk, ‘Turbulent Trajectories’ (2012) 2 Societies 27. 6 See Eric Neumayer, ‘Asylum Destination Choice’ (2004) 5 European Union Politics 155; and Antonia Scholz, Warum Deutschland? (Forschungsbericht 19, BAMF 2013). 7 See George Jesus Borjas, ‘Immigration and Welfare Magnets’ (1999) 17 J Lab Econ 607. 8 See Corrado Giulietti and Jackline Wahba, ‘Welfare Migration’ (IZA Discussion Papers No 645, Institute of Labor Economics, March 2012).

Drivers of Migratory Movements 99 insofar as the level of benefits in home, transit, and destination countries may be one push or pull factor among many but not necessarily the decisive one. Migrants and refugees will rarely leave West Africa or the Middle East with the firm intention of cashing in on welfare payments; their primary aim is for economic success and personal security, with the welfare state as only one consideration amongst others.9 In addition, welfare policies may influence the choice of destination within Europe, even though asylum seekers will not always distinguish clearly between social benefits and public service quality generally.10 Politicians often support welfare cuts, since it is one of the few ‘pull’ factors they control, unlike, for instance, economic growth. Moreover, they feed a narrative of ‘abuse’ and ‘bogus’ applications that can prepare, as we shall see, the ground for restrictions. For our purposes, the debate about the ‘welfare magnet’ reiterates that push and pull factors are not a simple formula. We should acknowledge that the situation in Europe is complemented by individual and structural elements elsewhere: civil wars, political oppression, and lack of economic prospect or money. Such ‘hard’ factors are complemented by ‘soft’ elements. Among the 97 per cent of the world population that does not move, many prefer living with family and friends in a society where they feel at home despite massive problems: cultural identity, risk-aversion, age, and fear are additional factors.11 In short, there are multiple driving forces.

4.1.2 Migration as a process over time Political debates and academic analyses have been defined by what critics call ‘methodological nationalism’ for decades. Crucially, this critique is not about nationalism in the sense of favouring the interests of one state. It concerns, rather, the theoretical model of territorial statehood that underlies many academic analyses and policy debates as an axiomatic assumption, while cross-border movements and transnational social ties are treated as an exception that proves the rule. Moreover, migration is perceived of as a straightforward sequence of departure, journey, and settlement.12 Such ‘sedentary bias’13 tends to ignore elements of persisting transnationalism, both in terms of recurring physical movements and contacts with colleagues, friends, and families abroad. Moreover, migration is often multidirectional, with no predetermined end result. Decisions about geographical preferences and permanent settlement may change over a life course, and circular movements between states can be a choice some prefer to stasis.

9 See Jeremy Ferwerda and Justin Gest, ‘Pull Factors and Migration Preferences’ (2021) 55 IM Rev 431; and Ole Agersnap, Amalie Sofie Jensen, Henrik Kleven, ‘The Welfare Magnet Hypothesis’ (Working Paper No 26454, NBER, November 2019). 10 See Petra W de Jong and Helga AG de Valk, ‘Intra-European Migration Decisions and Welfare Systems’ (2020) 46 JEMS 1773, 1781–86. 11 See generally William Samuelson and Richard Zeckhauser, ‘Status Quo Bias in Decision Making’ (1988) 1 J Risk & Uncertainty 7; and for migration Kerilyn Schewel, ‘Understanding Immobility. Moving Beyond the Mobility Bias in Migration Studies’ (2020) 54 IM Rev 328. 12 See the classic critique by Andreas Wimmer and Nina Glick Schiller, ‘Methodological Nationalism, the Social Sciences, and the Study of Migration’ (2003) 37 IM Rev 576; see also Nina Glick Schiller and Noel B Salazar, ‘Regimes of Mobility Across the Globe’ (2013) 39 JEMS 183. 13 Oliver Bakewell, ‘Keeping Them in Their Place’ (IMI Working Paper 8, Oxford International Migration Institute 2007); and Daniel Thym, ‘Migrationsfolgenrecht’ (2017) 76 Veröffentlichungen der Vereinigung Deutscher Staatsrechtslehrer 169, 169–73.

100 Interdisciplinary Perspectives and Methodology The EU’s single market illustrates the significance of temporary migration. While some follow resolute plans, others relocate temporarily as students, workers, or family members— with intentions about the length of stay often being vague early on.14 Germany, for example, benefits from the circular migration of Romanian seasonal workers and Polish women who care for elderly family members in private households; several hundred thousand Union citizens enter and leave the country each year.15 Similarly, the majority of ‘guest workers’ returned home, most of them voluntarily after having earned money and acquired new skills. Others stayed, also because the ban on further recruitments closed the door for circular migration; moreover, the stagnating economy in Turkey and the military coup of 1980 rendered return unattractive.16 Historians highlight that even the archetypical period of ‘emigration’ from Europe to the United States during the late nineteenth and early twentieth centuries saw significant numbers of returns.17 Social scientists have countered the entrenched ‘methodological nationalism’ by proclaiming ‘transnationalism’ and ‘mobility’ new paradigms.18 Such analyses fit well into broader globalisation processes and will be of intuitive appeal to many experts in migration law. Yet, such paradigmatic revolution runs the risk of replicating the pitfalls of earlier debates by not questioning the conceptual underpinnings of both mobility and immobility.19 There can be a ‘mobility bias’ in the same sense as others are trapped in a ‘sedentary bias’, also considering the high percentage of people who do not migrate. Surveys in West Africa, for instance, have repeatedly found a staggeringly widespread aspiration to move, which contrasts with the comparatively small number of actual mobility.20 Mobility and immobility are both realities, and we should not generally prioritise the one over the other theoretically or in the research design. The example of asylum applicants challenges one-sided assumptions of migration as a simple sequence of departure, journey, and settlement. The term ‘transit’ suggests that movements are following a predetermined plan. That need not be the case. More than three million Syrians were staying in Turkey at the time of writing, who did not board vessels bound for Greece for the most part. Even those doing so will not necessarily have planned this trajectory when leaving Syria.21 Chapter 18.4.3 will mention that Morocco, Turkey, and Libya (before the civil war) are not just ‘transit’ countries but destinations in their own right; not every foreigner residing there wants to go to Europe, although preferences may, of course, change over time. The same applies to secondary movements within the European Union, which can be the result of experiences in countries of first arrival.22 Not only individual intentions may change. Some migrants are prevented from pursuing their journey by 14 See Renee Luthra, Lucinda Platt, and Justyna Salamońska, ‘Types of Migration’ (2018) 52 IM Rev 368. 15 In 2019 (before the Covid-19 pandemic), almost 600,000 entries contrasted with 430,000 exits; see Johannes Graf, Freizügigkeitsmonitoring. Jahresbericht 2019 (BAMF 2020). 16 See Cord Pagenstecher, ‘Die “Illusion” der Rückkehr’ (1996) 47 Soziale Welt 149. 17 See Wimmer and Glick Schiller (n 12) 586–92. 18 See Thomas Faist, Margit Fauser, and Eveline Reisenauer, Transnational Migration (Polity 2013); and Mimi Sheller and John Urry, ‘The New Mobilities Paradigm’ (2006) 38 Environ Plan A 207. 19 See Adrian Favell, ‘Rebooting Migration Theory’ in Caroline B Brettell and James F Hollifield (eds), Migration Theory (2nd edn, Routledge 2008) 259, 270–71. 20 See Joint Research Centre, Many More to Come? Migration from and within Africa (Commission 2018) 17–18. 21 See Franck Düvell, ‘The “Great Migration” of Summer 2015’ (2019) 45 JEMS 2227; and generally Justin Schon, ‘Focus on the Forest, not the Trees’ (2015) 28 J Refugee Stud 437. 22 See Markus Wagner, Jimy Perumadan, and Paul Baumgartner, ‘Secondary Movements’ (CEASEVAL Research on the CEAS No 34, 2019); and (Dutch) Advisory Committee on Migration Affairs, Secondary Movements of Asylum Seekers in the EU (ACVZ 2019).

Drivers of Migratory Movements 101 border controls or encampment, for instance in hotspots on the Greek islands.23 Limbo and entrapment are an integral part of the migratory experience in the same way as movement and the modification of original ideas.

4.1.3 Intermediate level: networks, regimes, and infrastructures Social scientists often distinguish the micro-, meso-, and macro-level of analysis. While the micro-perspective concentrates on personal aspirations and capabilities, macro-theories highlight abstract features beyond the control of individuals, such as global inequalities, legal rules, or the structure of the labour market.24 Both micro-and macro-elements struggle to explain the profound asymmetry of many movements: specific countries of origin (and cities or regions within them) often see considerably higher rates of emigration than neighbouring states with similar micro-and macro-characteristics. Migration processes tend to become partly self-perpetuating, with ‘pioneers’ venturing to new destinations that may be embedded over time resulting in what is often called—somewhat pejoratively from today’s perspective—‘chain migration’.25 Little Italy and Chinatown in Manhattan bear testimony to earlier episodes of feedback loops between pioneers and followers, as do neighbourhoods in Paris, Lisbon, or Brussels with a high percentage of residents whose families came from the former colonies some time ago. Three meso features help to explain the entrenched asymmetries of migratory movements. First, networks have been identified as an essential component. Pioneers serve as anchors and role models for family and kin, providing information on the journey and supporting newcomers after arrival.26 Such networks are relevant for economic migration and refugees alike. Afghans or Ukrainians entering the EU in the 2020s are more likely to aim to go to places, by means of irregular movement if need be, where compatriots, friends, and relatives are already staying. In the same way, family reunification in accordance with the Family Reunification Directive 2003/86/EC can be perceived as a specific expression of legally sanctioned ‘chain migration’. Nevertheless, we not to conceive of networks as an ostensibly unstoppable phenomenon. They interact with other push and pull factors, which may facilitate or undermine self-perpetuating effects.27 Drivers of migration are almost always highly complex and inter-dependent. Secondly, movements are often sustained by economic, cultural, and political structures beyond the direct control of individuals or states. Historical and structural theories emphasise the significance of ‘migration regimes’ which have developed over time.28 Chapter 1.1 mentioned two migration regimes: labour migration from Italy to the industrial heartland of the coal and steel industry throughout the twentieth century and colonial immigration towards France, Belgium, or the Netherlands. Another example is regional migration 23 See Tobias G Eule and others, Migrants before the Law (Palgrave 2019) ch 5. 24 See de Haas and others (n 2) 43. 25 John S MacDonald and Leatrice D MacDonald, ‘Chain Migration, Ethnic Neighborhood Formation and Social Networks’ (1964) 42 Milbank Mem Fund Q 82. 26 See Faist, Fauser, and Reisenauer (n 18) ch 3; and Paul Collier, Exodus: How Migration Is Changing Our World (OUP 2013) ch 2. 27 See Hein de Haas, ‘The Internal Dynamics of Migration Processes’ (2010) 36 JEMS 1587. 28 See John Salt, ‘A Comparative Overview of International Trends and Types, 1950–80’ (1989) 23 IM Rev 431; and Jan Lucassen, Migrant Labour in Europe, 1600–1900 (Croom Helm 1987).

102 Interdisciplinary Perspectives and Methodology patterns in western, southern, and eastern Africa across artificially drawn colonial borders, which account for the majority of cross-border movements south of the Sahara. Again, migration regimes interact with other factors supporting or hampering mobility. Late colonial migration to Western Europe came to an end as a result of policy change, and labour migration to the coal and steel industry did not survive the decline of these sectors. Historically, the Mediterranean has been an arena for commercial, human, and cultural exchange29—in contrast to today’s perception as an external border. Thirdly, mobility depends on infrastructure. Movements across longer distances are facilitated by cheaper transportation, such as the construction of railways and the introduction of steamships across the Atlantic in the late nineteenth century, or the arrival of cheap air travel and mobile phone communication more recently. At the same time, states use infrastructure to thwart movement. Telling examples are the equipment of Frontex and databases, to be traced in Chapters 8 and 9. In response to these restrictions, migrants often rely on ‘smugglers’ to cross borders, as our comments on migrant agency will illustrate. It is a general feature of infrastructure that it often seeps into the background and receives little consideration as a result, even though interdisciplinary analyses have turned their attention to them recently.30

4.1.4 Continuum of ‘voluntary’ and ‘forced’ migration The distinction between ‘voluntary’ and ‘forced’ migration is firmly embedded in the legal analysis. It underlies the Treaty regime with two distinct legal bases for asylum (Article 78 TFEU) and legal migration, including return (Article 79 TFEU). The distinction between ‘refugees’ and ‘migrants’ similarly defines the international debate: the Refugee Convention and UNHCR deal with forced migration, while the International Organization for Migration (IOM) and the International Labour Organization (ILO) cover voluntary movements for economic and other purposes. Chapter 1.1.3 explained that this binary distinction was not a foregone conclusion but developed during the first half of the last century. Today, however, it is so deeply ingrained in the legal material that legal experts tend to take it for granted, also considering that ethical reasons support dealing with those fleeing severe hardship in a different way. Social scientists emphasise that the complex interaction of multiple push and pull factors means than motives can rarely be disentangled neatly. ‘Forced’ and ‘voluntary’ migration are ideal-typical antipodes of a continuum of overlapping personal characteristics and macro-features. Poverty and authoritarian government may qualify the ‘voluntary’ character of labour migration in the same way as the aspiration of a better life can be one factor why refugees leave a country, although others do not do so.31 The picture becomes even more complex when we consider migratory movements over time. Afghans born in refugee settlements in Iran may never have been to their ‘home’ state, although hypothetical treatment there determines whether they will receive international protection or not; similarly, 29 See Braudel Fernand, La Méditerranée: L’espace et l’histoire (Arts et Métiers Graphiques 1977). 30 See Sheller and Urry (n 18) 210–13; and Biao Xiang and Johan Lindquist, ‘Migration Infrastructure’ (2014) 48 IM Rev S122. 31 See Jeremy Hein, ‘Refugees, Immigrants and the State’ (1993) 19 Ann Rev Sociol 43; and Anthony H Richmond, ‘Sociological Theories of International Migration’ (1988) 36 Curr Sociol 7.

Significance of State Measures 103 Syrians leaving Turkey after having resided there for five years will rarely do so out of immediate fear of persecution.32 Refugee status determination concentrates on selected aspects, thus giving preference to some elements over others. In this overall context, interdisciplinary studies tend to employ the term ‘mixed migration’ to designate the overlap between numerous push and pull factors. ‘Mixed migration’ in this sense concerns the absence of clear-cut differentiation of the drivers of individual decisions whether to leave and where to go.33 By contrast, UNHCR and the policy discourse habitually speak of ‘mixed migration’ when people with and without protection needs form two more or less distinct subgroups of cross-border movements.34 A paradigmatic example are Eritreans and Tunisians sitting side by side on a boat arriving in Lampedusa. The UNHCR worries that recognising the motivational overlap may undermine the willingness of states to protect refugees. Indeed, the policy discourse is full of examples of branding people with protection needs as ‘economic’ or ‘illegal’ migrants, for instance Syrians who move on to Greece after having resided in Turkey for several years. As described in the introduction, this contribution employs the term ‘migration/migrant’ as an overarching category comprising voluntary and forced migration. Underlying reasons are not directly concerned with the continuum of ‘forced’ and ‘voluntary’ movements from an interdisciplinary perspective, let alone the policy discourse on mixed migration. Rather, our choice of terminology presents an heuristic device to identify overarching themes of the legal analysis and feedback loops between migration law instruments. In some circumstances, however, interdisciplinary findings about the overlap of ‘voluntary’ and ‘forced’ migration will be of direct relevance for the legal analysis, as Chapter 13.5.3 on asylum will examine for the potential extension of the notion of persecution to economic hardship and ‘climate refugees’.

4.2 Significance of State Measures EU institutions aim ‘at ensuring . . . the efficient management of migration flows’.35 That objective remains counterfactual insofar as legislation will never coincide fully with real- life events. Interdisciplinary debates have addressed the relative weight of state measures from different perspectives. Normative theories of moral and political philosophy develop abstract arguments about how to justify borders (4.2.1). Political scientists engage in empirical studies about the ‘control gap’, thus trying to unearth reasons why governments fail to deliver on self-declared control imperatives, including the role of courts (4.2.2). Various disciplines examine the importance of public opinion (4.2.3). A defining feature of recent decades has been the ‘denationalisation’ of migration policy, through the involvement of diverse actors at local, supranational, global, and transnational levels (4.2.4). Ethnographers emphasise that migrants are not just the object of legal rules but individual agents navigating their way with creativity (4.2.5).

32 See Heaven Crawley and Dimitris Skleparis, ‘Refugees, Migrants, Neither, Both’ (2018) 44 JEMS 48. 33 See Nicholas Van Hear, ‘Managing Mobility for Human Development. The Growing Salience of Mixed Migration’ (Human Development Research Paper 2009/20, April 2009). 34 See UNHCR, ‘Refugee Protection and Mixed Migration. A 10-Point Plan of Action’ (December 2006). 35 TFEU, art 79(1).

104 Interdisciplinary Perspectives and Methodology

4.2.1 Normative background: the open borders debate Experts in migration law often have a good knowledge of human rights and tend to subscribe to universalist visions of social justice and political philosophy. That is perfectly legitimate and corresponds to a strand in the theoretical debate which highlights the equal moral worth of human beings and their joint possession of the surface of the world in the tradition of Immanuel Kant. From this perspective, state sovereignty and migration control appear suspicious precisely because they deviate from the starting point of normative universalism—a position that gained ground with the advance of globalisation in the 1990s. Theories of global justice addressed an important lacuna, since sovereign statehood had often been treated as an axiomatic absolute previously. A prominent example is the liberal philosopher John Rawls whose ‘veil of ignorance’ concerned particularistic communities to start with, not the world as a whole.36 Political theorists had questioned the legitimacy of migration control before the global justice debate. Positions put forward roughly reflected the contrast between ‘liberalism’ and ‘communitarianism’, which had defined Western political theory until the end of the cold war. Joseph Carens famously advanced a (liberal) case for open borders,37 while David Miller defended a (communitarian) vision of statehood.38 Studies thereafter have taken up these arguments from diverse perspectives and developed sophisticated arguments.39 Legal experts should not expect normative theory to deliver straightforward guidance on how to deal with specific questions, since many arguments remain abstract or build upon theoretical assumptions that are not universally accepted.40 Readers should also be aware of terminological false friends. An instructive example is Hanna Arendt’s much cited ‘right to have rights’ that is not primarily concerned with legal guarantees to be enforced by courts, as explained in Chapter 5.2.2 on human rights. Theoretical debates in recent years have concentrated on opening up the normative black box the concept of state sovereignty had traditionally been. They recognised that sovereignty may present itself as a proxy for complex arguments about the benefits of self- government and particularistic communities.41 Any analysis of these questions will inevitably be confronted with disputes about intricate questions to which no easy normative or empirical answers exists: the relationship between migration and redistributive policies in welfare states;42 the weight of culture or, more generally, societal self-perception as a

36 See John Rawls, The Law of Peoples (Harvard UP 2001); and Seyla Benhabib, The Rights of Others (CUP 2004) 74–94. 37 See Joseph H Carens, ‘Aliens and Citizens. The Case for Open Borders’ (1987) 49 Rev Politics 251; and Joseph H Carens, The Ethics of Immigration (OUP 2013) ch 1. 38 See David Miller, Strangers in Our Midst (Harvard UP 2016), in line with previous publications; and also Michael Walzer, Spheres of Justice (Basic Books 1983). 39 See Sarah Fine and Lea Ypi (eds), Migration in Political Theory (OUP 2016); and Christopher Heath Wellman and Phillip Cole, Debating the Ethics of Immigration (OUP 2011). 40 See Thomas Spijkerboer, ‘A Distributive Approach to Migration Law’ in Roland Pierik and Wouter Werner (eds), Cosmopolitanism in Context (CUP 2010) 249. 41 See Alexander Somek, The Cosmopolitan Constitution (OUP 2014); and Margaret Moore, ‘Cosmopolitanism and Political Communities’ (2006) 32 Soc Theory & Prac 627; see also Daniel Thym, ‘Citizens and Foreigners in EU Law’ (2016) 22 ELJ 296, 311–15. 42 See David Abraham, ‘Immigrant Integration and Social Solidarity in a Time of Crisis’ (2014) 1 Crit Hist Stud 215; and Markus ML Crepaz, ‘Rumors that Diversity is the Death of the Welfare State are Greatly Exaggerated’ in Gary P Freeman and Nikola Mirilovic (eds), Handbook on Migration and Social Policy (Edward Elgar Publishing 2016).

Significance of State Measures 105 justification for closure;43 or the distinction between access to the territory and treatment thereafter.44 Some of these arguments will surface elsewhere in this book, for instance with regard to the welfare state and the notion of integration in Chapters 14.1.6 and 15.1.2 on legal migration and settlement.

4.2.2 Explaining the ‘control gap’: the liberal paradox While legal experts intuitively consider state policies to be decisive factors, economic analyses, in particular, are defined by widespread ignorance of the regulatory environment. We are confronted with a somewhat perplexing contrast between scholarship describing states as omni-powerful actors and the alternative emphasis on control failure and individual agency.45 Political scientists have described the distinct experience of the classic destination countries in the Western hemisphere after the Second World War under the heading of ‘control gap’ to designate the widespread perception that states failed to deliver on policy objectives.46 The ‘control gap’ debate was fed by the experience of ‘guest worker’ schemes resulting in permanent settlement, the ineffectiveness of return, the self-reinforcing effects of family reunification, and the pertinence of irregular entry and stay. James Hollifield coined the term ‘liberal paradox’ to resolve the puzzle: the liberal normative self-image prohibited Western democracies from implementing the policies that would be necessary to realise the official policy objectives.47 EU integration played a prominent role in early accounts, which gained ground during the early 1990s when the move towards political Union, the prospect of enlargement, and dynamic court judgments left observers with the impression that the supranational institutions were about to revolutionise migration law. Free movement of Union citizens and the case law of the ECtHR were often cited as precursors of novel forms of post-or transnational membership.48 Others highlighted that the liberal paradox was primarily an internal phenomenon, with domestic legislation, constitutional norms, and national courts curtailing governmental preferences.49 Comprehensive and longitudinal perspectives hardly support any of these black-and-white accounts. We have seen in Chapter 1.2 that Treaty amendments established a distinct legal framework for third country nationals differing from the freedom-enhancing rationale of Union citizenship. At the same time, Chapter 2 demonstrated that EU institutions respond to diverse impulses. The liberal paradox presents itself as a combination of diverse factors whose relative weight may differ over space and time and depending on the subject matter: supranational 43 See Liav Orgad, The Cultural Defense of Nations (OUP 2015); and Samuel Scheffler, ‘Immigration and the Significance of Culture’ (2007) 35 Philos Pub Aff 93. 44 See Carens, Ethics (n 37); and Miller (n 38). 45 See generally James F Hollifield and Tom K Wong, ‘The Politics of International Migration’ in Brettell and Hollifield (n 19) 227, 235–45. 46 See Wayne Cornelius, Philip Martin, and James F Hollifield (eds), Controlling Immigration (1st edn, Stanford UP 1994). 47 See James F Hollifield, Immigrants, Markets and States (Harvard UP 1992) ch 1. 48 See Yasemin Soysal, Limits of Citizenship (University of Chicago Press 1994) ch 8; and Saskia Sassen, Losing Control? (Columbia UP 1996) 88–99. 49 See Christian Joppke, ‘Why Liberal States Accept Unwanted Immigration’ (1998) 50 World Polit 26; and Virginie Guiraudon, ‘The Marshallian Triptych Reordered’ in Michael Bommes and Andrew Geddes (eds), Immigration and Welfare (Routledge 2000) 72.

106 Interdisciplinary Perspectives and Methodology developments may stimulate reform in some Member States, while replicating domestic policies elsewhere; judges will promote the rights of migrants in some cases, while siding with government in others; legislation can advance migrants rights in the same way as courts.50 Chapter 3.2 showed that the CJEU makes an effort to balance countervailing interests, and Chapter 5.3 will find similar tendencies with regard to the ECtHR. Judges play a critical role in upholding the rights of migrants, yet their function is not intrinsically linked to the elimination of control imperatives. Interdisciplinary studies have highlighted other factors that help to explain the ‘control gap’. They will be discussed below with regard to the structure of domestic politics, migrant agency, and international cooperation. Moreover, some authors emphasise that the policy outcome can be explained, in part at least, as the functional continuation of sectoral policies. The inclusionary logic of the welfare state commanded that foreigners received social rights even though doing so might complicate migration control; the neoliberal rationale of the market economy supported the employment of those staying irregularly; and constitutionalism promoted human rights.51 In other words, the practical effects of migration policy need not necessarily be the result of migration-specific choices. The highly stratified character of modern societies can create feedback loops on the basis of policy preferences which do not concern migration primarily.

4.2.3 Public opinion and intergroup threat perceptions Interdisciplinary studies have highlighted a mismatch between sceptical public opinions and the moderately liberal policy output of Western democracies before the millennium change, for instance with regard to ‘guest workers’, colonial migration, and family reunification.52 Empirical surveys demonstrate that the views of the population has been surprisingly stable over the decades.53 That discrepancy constitutes a puzzle many social scientists have tried to rationalise. For European migration law, the depoliticised character of decision- making provides a first explanation, since the supranational institutions had traditionally been shielded from intense public scrutiny, as we have seen in Chapter 2.1.3 on the institutional practice. They advanced policies that might not have been adopted domestically. Experts in political economy emphasise additional domestic reasons. While the benefits of migration (such as cheap labour for employers and reunited families for migrant communities) are concentrated and can be reaped in the short term, potential costs and side-effects (such as competition on the labour market, increased welfare spending, or— controversially—cultural diversity) are diffuse and often occur in the long run. Politicians have traditionally been more responsive to better organised advocates of migration than to

50 See Daniel Thym, ‘Supranational Courts in Europe: A Moderately Communitarian Turn in the Case Law on Immigration and Citizenship’ (2021) 47 JEMS 4534; and Randall Hansen, ‘Globalization, Embedded Realism, and Path Dependency’ (2000) 35 Comp Pol Stud 259. 51 See Gary P Freeman, ‘Immigrant Incorporation in Western Democracies’ (2004) 38 IM Rev 945; and Christina Boswell, ‘Theorizing Migration Policy’ (2007) 41 IM Rev 75, 87–95. 52 See Jeannette Money, Fences and Neighbours (Cornell UP 1999). 53 See Anthony F Heath and Lindsay Richards, ‘Contested Boundaries’ (2020) 46 JEMS 489; and Alan E Kessler and Gary P Freeman, ‘Public Opinion in the EU on Immigration from Outside the Community’ (2005) 43 JCMS 825.

Significance of State Measures 107 the sceptical general public.54 Moreover, debates do not fit easily into the traditional left- right spectrum. Employer organisations and centre right parties will oppose cultural diversity but may support labour mobility, whereas trade unions and the centre left will support non-discrimination, while being sceptical of labour market liberalisation.55 In Europe, cultural aspects were and are influential. Experts in social psychology use ‘intergroup threat theory’ to analyse developments, while recognising that the definition of the relevant groups is socially constructed and, hence, subject to change.56 ‘Migrants’ or ‘Muslims’ are prime examples of out-groups in the contemporary European discourse, from which the general public tends to distinguish itself. Crucially, social psychology examines how perceptions of threat can increase—or decrease—tensions. Perceptions of threat can be either realistic (economic well-being, physical security) or symbolic (identity, value systems) and concern the individual level or society as a whole.57 Considerable research has applied these abstract findings to the European context to understand how and for which reasons perceptions of different migrant groups differ over space and time. For our purposes, ‘intergroup threat theory’ can serve as a bridge connecting the debate in the social sciences to critical studies on the use language, mentioned below, and our previous conclusion about the restrictive effects of politicisation, presented in Chapter 2.1. Terrorist attacks or violent crime committed by a tiny minority of migrants can feed an impression of latent threat—as do references to ‘waves’ or images of individuals ‘storming’ border fences. Populist parties exploit these images, thus nurturing a sense of ‘loss of control’. Moderate politicians find it difficult to counter such discursive tactics, let alone set the agenda with a counter-narrative. A potential, although double-edged, antidote can be visible forms of state control. Strengthening Frontex, reducing social benefits, or buttressing return will not necessarily deliver the policy output politicians promise, but it can send ‘control signals’.58 Politicians employ the element of symbolic closure to guide the public discourse into calmer waters.

4.2.4 Multiple actors of migration governance As a matter of principle, EU migration law follows classic administrative paradigms of hierarchical command and, if need be, coercive law enforcement. Softer forms of governance exist, for instance for migrant integration, but even there mandatory rules define the policy debate. Think of pre-departure language tests. Private actors can be relevant, such as companies sponsoring the stay of researchers or external service providers supporting visa processing, but such examples of ‘public-private partnership’ take place in the shadow of 54 See Gary P Freeman, ‘Modes of Immigration Politics in the Liberal Democratic State’ (1995) 29 IM Rev 861, 882–86; and Collier (n 26) chs 3–4. 55 See James Hampshire, The Politics of Immigration (Polity Press 2013) chs 2–3. 56 See Eva GT Green and Christian Staerklé, ‘Migration and Multiculturalism’ in Leonie Huddy and others (eds), The Oxford Handbook of Political Psychology (2nd edn, OUP 2013) 852. 57 See Walter G Stephan, Oscar Ybarra, and Kimberly Rios, ‘Intergroup Threat Theory’ in Todd D Nelson (ed), Handbook of Prejudice, Stereotyping, and Discrimination (2nd edn, Psychology Press 2016) 255. 58 See James F Hollifield, Philip Martin, and Pia M Orrenius, ‘The Dilemmas of Immigration Control’ in James F Hollifield and others (eds), Controlling Immigration (3rd edn, Stanford UP 2014) 3, 27; and Chris F Wright, ‘How Do States Implement Liberal Immigration Policies?’ (2014) 27 Governance 397.

108 Interdisciplinary Perspectives and Methodology compulsory legislation. With regard to the instruments discussed in this volume, migration law essentially presents itself as a conventional area of administrative top-down command and enforcement.59 Other disciplines support more ambivalent conclusions about how the contents of the law is being applied in practice. From such standpoints, EU migration law can be described as an expression of ‘governance’ comprising street-level agency, private parties, assorted transnational networks, and international bodies. ‘Governance’ became a catchword to describe such interdisciplinary analyses that transcend classic forms of top-down command and enforcement.60 Research in this tradition can be used to complement the ‘top- down’ legal analysis with the ‘bottom-up’ perspective of other disciplines. Doing so does not change the contents of the law, but it unearths valuable insights into how the law in the books translates into law in action. A classic approach of sociolegal research are case studies on the work of ‘street-level bureaucrats’.61 Telling examples are decision-making in asylum cases and the lived experience of visa practices in consulates abroad.62 State agents may interpret the legal framework creatively either to develop practical solutions for the benefit of migrants or to circumvent obligations. Sociolegal studies also depict the impact of individuals on legislative processes and court proceedings. We have seen in Chapter 2.3.6 on the institutional practice that diverse private actors, including NGOs, seek to influence the legislative debate or compete for lucrative contracts on border security. Chapter 3.3.4.2 demonstrated the potential of strategic litigation. Multiple actors are active at the international plane: UNHCR advises states how to apply the Refugee Convention; private vessels perform search and rescue in the Mediterranean; public and private bodies implement development projects; Moroccan border guards receive equipment and financial support from the EU budget; the firm Western Union has the largest market share for the transfer of financial remittances; and airlines are wary of the threat of carrier sanctions. These examples of transnational and international migration governance will be discussed in the relevant sectoral chapters throughout the second part of this book. While initial commentators had perceived globalisation to undermine migration control,63 there is an alternative reading. International cooperation can be presented as a response to bridge the ‘control gap’. States understand that many push and pull factors are beyond their direct reach, and they try to increase leverage by means of international cooperation. Doing so also allows them to diminish the ‘liberal paradox’, since international cooperation will not always be subject to human rights law and public scrutiny. Chapter 2.1 illustrated that the early phase of EU migration law can be described as ‘venue shopping’, with interior ministries dominating the policy agenda. Chapter 18 will explain that recent years have witnessed the externalisation of migration control to third states. States can be effective in trying to regain control.

59 See Daniel Thym, Migrationsverwaltungsrecht (Mohr Siebeck 2010) 31–48. 60 See David Levi-Faur, The Oxford Handbook of Governance (OUP 2012). 61 See Michael Lipsky, Street-Level Bureaucrats (Russell Sage 2010). 62 See Eule and others (n 23) chs 3–4; and Federica Infantino, ‘How Does Policy Change at the Street Level?’ (2021) 47 JEMS 1028. 63 See Saskia Sassen, A Sociology of Globalisation (WW Norton 2007) ch 5; and Stephen Castles, ‘The Factors that Make and Unmake Migration Policies’ (2004) 38 IM Rev 852, 861–65.

Significance of State Measures 109

4.2.5 Individual migrant agency Ethnographic research was instrumental in contrasting the claim to law enforcement with the practical experience of migrants. Third country nationals are actors with a certain degree of agency navigating their ways forwards.64 Emphasis on individual agency in ethnographic research resonates with interdisciplinary ‘mobility studies’, which criticise the sedentary stasis of methodological nationalism.65 It also overlaps with theoretical debates about political contestation as the basis for politics and rights, amongst others by Hanna Arendt.66 Finally, migrant agency contextualises the description of migration law as a purportedly omni-powerful instrument of repressive government in the Foucauldian tradition on the pages that follow. ‘Border studies’ have proven particularly productive, probably because the complexity and occasional obscurity of the factual situation on the ground entails better opportunities for circumventing state action.67 Hotspots, the ‘jungle’ in the French city of Calais from where migrants try to reach the UK, and irregular movements across the Western Balkans reveal significant degrees of migrant agency, thus calling into question the dystopia of ‘fortress Europe’.68 The evolution of control policies, traced in Chapters 11 and 12, can even be described as a sort of ‘cat-and-mouse game’:69 individuals evade restrictive instruments with innovative tactics, and states respond with another set of control measures.70 Irregular migrants residing within the territory similarly use different strategies to optimise their situation by means of informal work, marriage, or regularisation.71 Such accounts are stimulating to read precisely because they challenge conventional descriptions of state sovereignty in light of the lived experience of migrants. At the same time, one should be careful not to overstate, or even celebrate, ostensibly heroic episodes of migrant agency. Their action is habitually driven by the fear of persecution or economic hardship, embedded in highly unequal power relations, and accompanied by personal sacrifice and suffering.72 What is more, agency will not always succeed. Encampment, loss of life, and harsh poverty define the experience at the external borders in the same way as individual agency. A case in point is the ambivalent role of ‘smuggling’ and other businesses irregular migrants rely upon for diverse purposes ranging from informal employment to financial transfers. Smuggling will be perceived a useful ‘service’ by some, while others will experience exploitation, corruption, or violence.73 The EU response 64 See Dawn Chatty, ‘Anthropology and Forced Migration’ in Elena Fiddian-Qasmiyeh and others (eds), The Oxford Handbook of Refugee & Forced Migration Studies (OUP 2014) 74; and Irene Bloemraad, ‘Theorising the Power of Citizenship as Claims-Making’ (2018) 44 JEMS 4. 65 See n 18 and accompanying text. 66 See ch 5.2.2. 67 See generally Sandro Mezzadra and Brett Neilson, Border as Method, or, the Multiplication of Labor (Duke UP 2013). 68 See Vassilis Tsianos and Serhat Karakayalı, ‘Transnational Migration and the Emergence of the European Border Regime’ (2010) 13 Eur J Soc Theory 373; and Leonie Ansems de Vries and Elspeth Guild, ‘Seeking Refuge in Europe’ (2019) 45 JEMS 2156. 69 Thomas Gammeltoft- Hansen and James C Hathaway, ‘Non- Refoulement in a World of Cooperative Deterrence’ (2015) 53 Columbia J Transnatl L 235, 246. 70 See Dimitris Papadopoulos, Niamh Stephenson, and Vassilis Tsianos, Escape Routes (Pluto Press 2008). 71 See Masja van Meeteren, Godfried Engbersen Weitaus, and Marion van San, ‘Striving for a Better Position’ (2009) 43 IM Rev 881. 72 See Vicki Squire, ‘Unauthorised Migration Beyond Structure/Agency?’ (2017) 37 Politics 254. 73 Contrast the nuanced outlook by Nina Perkowski and Vicki Squire, ‘The Anti-Policy of European Anti- Smuggling as a Site of Contestation in the Mediterranean Migration “Crisis”’ (2019) 45 JEMS 2167; to the positive narrative of Ruben Andersson, Illegality, Inc (UC Press 2014).

110 Interdisciplinary Perspectives and Methodology to the facilitation of irregular entry and trafficking in human beings will be discussed in Chapter 16.3 on return.

4.3 Law as an Instrument of Government Legal experts intuitively acknowledge that the law is not neutral but reflects political, social, economic, and normative preferences. Critical studies in the tradition of post-structural theory and social constructivism highlight this ideological predisposition (4.3.1). The distinction of diverse migration statuses demonstrates how public authorities employ labels as an instrument of government to distinguish between (un)desirable foreigners (4.3.2). The use of language partakes in how we see the world and can influence the policy outcome, as the example of the ‘crisis’ metaphor illustrates (4.3.3). The notion of ‘securitisation’ brings together critical perspectives, as does the counter-narrative of a rights-based approach (4.3.4).

4.3.1 Starting point: Foucault and constructivism Michel Foucault is the figurehead of interdisciplinary analyses that put the accent on legal rules and administrative practices being instruments of social control. Foucault popularised concepts like ‘biopolitics’ or the ‘panopticon’, which designate classic forms of disciplinary power, as well as ‘governmentality’ for subtle form of internalised self-control.74 One step further, the situation at the external borders is often described in categories of ‘bare life’,75 exposing migrants to a state of emergency where their basic physical existence is questioned.76 Detention and deportation are paradigmatic expressions of sovereign power, as are databases as an epitome of modern technology. Such focus on the law as an instrument of government contrasts with the emphasis on individual migrant agency in contemporary border studies. Post-structural theory in the French tradition stressed that the categories in which we describe the world are socially situated and constantly reconstructed. The critique of ‘methodological nationalism’ and the ‘sedentary bias’, mentioned previously, are classic examples of how abstract findings about the relativity of our analytical categories can inform the analyses of specific questions of supranational migration policy. Constructivism can be combined, as we shall see, with critical legal studies, and it may even influence the doctrinal reconstruction of the legal material, in particular when interpreting open-ended human rights norms.

74 See https://en.wikipedia.org/wiki/Michel_Foucault#Philosophical_work (accessed 1 March 2023). 75 76

Giorgio Agamben, Homo Sacer: Sovereign Power and Bare Life (Meridian 1998). See Itamar Mann, Humanity at Sea (CUP 2016) 6–10.

Law as an Instrument of Government 111

4.3.2 ‘Labels’ as a means of government Article 79(2)(a) TFEU mentions family reunification as one ‘purpose’ for authorising entry and stay, and we will discuss multiple other purposes throughout this book. These migration statuses are so deeply ingrained into supranational and domestic legislation that legal experts tend to take them for granted, although they are far from self-explanatory from an interdisciplinary perspective.77 Indeed, the very distinction between ‘nationals’ and ‘foreigners’ (or ‘Union citizens’ and ‘third country nationals’) is the result of fundamental choices that led to the international legal order as we know it today.78 The distinction of ever more migrant statuses in recent decades need not be interpreted as an extension of state control or even xenophobia.79 Alternatively, it can be described as the increasing legalisation of status acquisition, with earlier notions of largely unfettered administrative discretion giving way to detailed rules comprising legal guarantees for migrants.80 Europeanisation was an integral part of this trend. A prominent critique of the conditionality of the different migration statuses, notably in North America, concerns the continuation of earlier forms of openly racist admission criteria. Seemingly neutral yardsticks, such as wage levels, education, or language skills, are considered to embody institutionally embedded racism.81 Others, by contrast, emphasise that the proliferation of stratified admission criteria coincided with the rise of neoliberal thinking. Having to prove one’s worth influenced the reform of social policies and migration laws, thus promoting an idea of ‘earned’ membership for foreigners and nationals alike.82 Even the rise of non-discrimination policies can be described as a result of the novel emphasis on individual merit.83 Points-based admission systems for the highly skilled are the embodiment of neoliberal migration policies, which will be mentioned in Chapter 14.4.1 on legal migration. The element of control inherent in the use of migration categories encompasses the basic distinction between ‘forced’ and ‘voluntary’ migration. Interdisciplinary analyses do not only highlight, in line with previous comments, that the driving forces behind migratory movements contradict the seemingly clear-cut juxtaposition of (forced) refuge and (voluntary) migration. They also demonstrate that these categories are socially constructed and have emerged over decades, if not centuries.84 Legal experts are bound to work with these labels regularly for the simple reason that they define the legal material, but doing so is not a

77 See Loïc Azoulai, ‘Le droit européen de l’immigration, une analyse existentielle’ [2018] Revue trimestrielle de droit européen 519. 78 See ch 1.1; Frank Schulze-Engler, ‘Irritating Europe’ in Graham Huggan (ed), The Oxford Handbook of Postcolonial Studies (OUP 2013) 669; and Nicholas De Genova, ‘Introduction’ in Nicholas De Genova (ed), The Borders of ‘Europe’ (Duke UP 2017) 1. 79 cf Diego Acosta Arcarazo and Jacopo Martire, ‘Trapped in the Lobby: Europe’s Revolving Doors and the Other as Xenos’ (2014) 39 EL Rev 362. 80 On the status quo ante see Tomas Hammar, ‘Comparative Analysis’ in Tomas Hammar (ed), European Immigration Policy (CUP 1985) 239, 249–62. 81 See ch 1.1.6; David S FitzGerald and others, ‘Can You Become One of Us?’ (2018) 44 JEMS 27; and Kevin R Johnson, ‘Race Matters’ (2000) 2 U Ill L Rev 525. 82 See Bridget Anderson, Us and Them? (OUP 2013); and Antje Ellermann, ‘Human-Capital Citizenship and the Changing Logic of Immigrant Admissions’ (2020) 46 JEMS 2515. 83 See Alexander Somek, Engineering Equality (OUP 2011); and Christian Joppke, Is Multiculturalism Dead? (Polity 2016) ch 5. 84 See Christiane Harzig and Dirk Hoerder, What is Migration History? (Polity 2009) 66–72; and Anderson (n 82) ch 2.

112 Interdisciplinary Perspectives and Methodology neutral exercise from an interdisciplinary perspective.85 Migration law, like most other systems of categorisation, creates hierarchical systems of rights, which states use to reinforce sovereign control. Critical studies have highlighted that even benevolent categories, like the spread of non- refoulement, combine tangible benefits for migrants with the ideological stabilisation of the sovereignty-based world order.86 Article 1A(2) of the Refugee Convention reinforces the conceptual nexus between citizens and states, when it protects anyone who fears being persecuted and is ‘unable . . . to avail himself of the protection of [his home] country’. One step further, critical studies have criticised refugee law for extensively relying on legal rules depicting individuals as seemingly passive objects in need of humanitarian support.87 Asylum legislation, with sophisticated procedures and multiple classifications ranging from safe third countries to membership of a particular social group, can be presented as a means of governmental control transforming highly political choices into depoliticised and technical administrative issues.88

4.3.3 Language between ‘abuse’, ‘crisis’, and ‘normality’ Language is not simply a means of communication using words with a pre-defined meaning. Our speech acts partake in the constant and performative (re)construction of the concepts with which we describe the world. These abstract findings of analytical philosophy prepared the ground for modern discourse analysis, which, in its post-structural appearance, critically examines how the use of language evolves over time and has psychological framing effects upon social and political choices.89 Feminism spearheaded the cautious use of language, which has received considerable attention in the context of migration recently. A classic example is the term ‘stranger’, which can have positive or negative connotations in line with the Latin distinction between ‘guest’ (hospes) and ‘enemy’ (hostis), thus being open to subtle variances and gradual shifts.90 How we speak about migration influences our thinking and can have an impact on the policy outcome. In the British debate, the term ‘bogus’ asylum seeker, or the generic notion of ‘abuse’, have traditionally been employed by those supporting restrictive measures. Similar semantics exist in other countries; a popular argument concerns putative ‘welfare migrants’ who supposedly enter a country to obtain social benefits. These examples have in common that they seek to delegitimise the motivation of individual migrants. Doing so will rarely have direct legal consequences, since the general principle of ‘abuse’, discussed in Chapter 10.8 on general features of secondary legislation, has a more limited meaning in Union law. 85 See Marta Bivand Erdal and Ceri Oeppen, ‘Forced to Leave? The Discursive and Analytical Significance of Describing Migration as Forced and Voluntary’ (2018) 44 JEMS 981; and Oliver Bakewell, ‘Research beyond the Categories’ (2008) 21 J Refug Stud 432. 86 See Nevzat Soguk, States and Stranger (Minnesota UP 1999); and Emma Haddad, The Refugee in International Society (CUP 2008) ch 1. 87 See Simon Behrman, ‘Legal Subjectivity and the Refugee’ (2014) 26 IJRL 1. 88 See Roger Zetter, ‘Labelling Refugees’ (1991) 4 J Refug Stud 39; and Simon Behrman (ed), Law and Asylum: Space, Subject, Resistance (Routledge 2018). 89 See Michel Foucault, L’ordre du discours (Gallimard 1971); and Amos Tversky and Daniel Kahnemann, ‘The Framing of Decisions and the Psychology of Choice’ (1981) 30 Science 453. 90 See Jacques Derrida, De l’hospitalité (Calmann-Lévy 1997); and Bonnie Honig, ‘Proximity and Paradox’ in Hans Lindahl (ed), A Right to Inclusion and Exclusion? (Hart Publishing 2009) 209.

Law as an Instrument of Government 113 Nevertheless, framing effects may direct the political debate towards restrictions. The same applies to the notion of ‘illegal migrant’, which negates legal personhood, and the less crude designation of ‘illegal stay’ referred to in Chapter 16.1.1 on return. Those supporting generous migration policies will speak of ‘irregular’ migration, or evade the negative connotations of ‘illegality/irregularity’ altogether. Events during 2015/16 were often referred to as a ‘migration’ or ‘refugee crisis’ (the terminology differed between countries). This choice of language has been criticised for portraying migrants and refugees as the origin of the ‘crisis’. It has been suggested that we should speak rather of a crisis of refugee protection or human rights, in light of the immense loss of life in the Mediterranean,91 or a crisis of inter-state solidarity.92 This contribution either avoids the crisis terminology or refers to an ‘asylum policy crisis’ to draw attention to the failure of the supranational institutions at overcoming structural deficits discussed in Chapter 13.1. In any case, we confine the use of the crisis metaphor to events during 2015/16, instead of insinuating an ongoing state of emergency. However, that is precisely what some political actors suggest by constantly talking about irregular arrivals or anonymous ‘waves’ or ‘flows’ that leave the general public with the impression that migration is some kind of unstoppable natural phenomenon. Such images feed a sense of ‘loss of control’, which can have negative repercussions, as we have seen, on public opinion. To what extent different frames succeed in influencing public perceptions remains a question of empirical analyses whose findings may differ over time and space.93 Counter-narratives exist. During 2015, the media paid much attention to terrible shipwrecks and the death of the toddler Alan Kurdi, who died when his family tried to reach the Greek island of Kos.94 Others highlight that migration policy is not only about public interests but also about the ‘rights’ of refugees and the ‘values’ of the European Union, thus opposing the widespread focus on negative language about ‘illegality’, ‘abuse’, or ‘threats’. Comparatively common is the assertion that cross-border movements characterise the human experience and have shaped European history. In the words of Commission President von der Leyen: ‘Migration has always been a fact for Europe—and it will always be. Throughout centuries, it has defined our societies, enriched our cultures, and shaped many of our lives. And this will always be the case.’95 Nevertheless, such positive language seems to have fallen behind the crisis metaphor on the whole, although it remains an option to direct the public debate in a different direction.

91 See Cathryn Costello and Minos Mouzourakis, ‘The Common European Asylum System: Where Did It All Go Wrong?’ in Maria Fletcher and others (eds), The European Union as an Area of Freedom, Security and Justice (Routledge 2018) 263, 264; and Gemma Marolda Gloninger, ‘From Humanitarian Rescue to Border Security’ (2019) 21 EJML 459. 92 See Cecilia Rizcallah, ‘Facing the Refugee Challenge in Europe’ (2019) 21 EJML 238, 241–48. 93 See Rens Vliegenthart, ‘Framing Citizens in the Field of Immigration Politics’ in Marco Giugni and Maria Grasso (eds), Handbook of Citizenship and Migration (Edward Elgar Publishing 2021) 269–83. 94 See https://en.wikipedia.org/wiki/Death_of_Alan_Kurdi (accessed 1 March 2023). 95 Ursula von der Leyen, ‘Building the World We Want to Live in’ (State of the Union Address, 16 September 2020).

114 Interdisciplinary Perspectives and Methodology

4.3.4 Securitisation versus rights-based approach To associate migration with security has a long history. Think of the wartime internment of ‘enemy aliens’ or the expulsion of ‘unwanted’ foreigners for diverse reasons. More recent interdisciplinary studies have coined the term ‘securitisation’ for the migration-security nexus. Such analyses assume that it can have discursive framing effects to connect migration semantically to latent security threats.96 Debates about ‘securitisation’ gained ground after the terrorist attacks of 2001 and the ensuing spread of restrictive measures, which were buttressed by The Hague Programme in the EU.97 Having said this, the debate about securitisation is not intrinsically linked to classic threats such as terrorism. Similarly, it encompasses the impact on prosperity (as a result of alleged ‘welfare migration’), the loss of control over the external borders (by means of ‘illegal migration’), and the identity of European societies (in response to the ‘Muslim other’).98 Security concerns are closely related to the language of ‘European sovereignty’ and a ‘Europe that protects’, used at the supranational level in line with comments in Chapter 1.2.6. This wide conceptualisation means that the notion of ‘securitisation’ can be applied to diverse policy initiatives at the external borders or within the territory; there is no clear- cut definition of where it starts or ends. Similarly, securitisation rarely translates into legal-doctrinal categories, although the ‘public policy’ caveat, used in most legislative instruments, and the generic safeguards for the maintenance of internal and external security in Articles 72 TFEU can be relevant. They will be discussed in Chapter 10.3 and 10.7 on general features of secondary legislation. Interdisciplinary debates about securitisation primarily concern the proliferation of databases, Frontex, the fortification of border controls, detention, forced removal, carrier and employer sanctions, and the criminalisation of illegal entry and stay. These measures will be discussed in the relevant sectoral chapters of this volume. A related argument concerns the ‘criminalisation’ of migration law (or ‘crimmigration’99), which is not limited to criminal law sensu stricto but equally encompasses the entire arsenal of coercive law enforcement.100 The added value of the securitisation viewpoint is to connect disparate phenomena and to show how the securitarian discourse can coincide with the spread of restrictive measures. Like in the case of language, there are alternative narratives accentuating the economic, moral, or social benefits of migration. Particularly relevant for legal experts is what is often called a ‘rights-based approach’, which emphasises that migration law is about promoting human rights as much as it concerns security. To do so is more than a reminder of mandatory constitutional requirements judges are bound to respect. Sociolegal studies have shown that discursive framing highlighting the human rights of migrants can have radiating effects

96 See Barry Buzan, Ole Wæver, and Jaap de Wilde, Security. A New Framework for Analysis (Rienner 1998); and Christian Kaunert and Ikrom Yakubov, ‘Securitization’ in Ariadna Ripoll Servent and Florian Trauner (eds), The Routledge Handbook of Justice and Home Affairs Research (Routledge 2018) 30. 97 See ch 1.2.5; and Didier Bigo, ‘Criminalisation of “Migrants”’ in Barbara Bogusz and others (eds), Irregular Migration and Human Rights (Martinus Nijhoff 2004) 61. 98 See Ayse Ceyhan and Anastassia Tsoukala, ‘The Securitization of Migration in Western Societies’ (2002) 27 Alternatives 21; and also Zygmunt Bauman, Strangers at Our Door (Polity Press 2016) chs 2–3. 99 The term was coined in the US context; see Juliet Stumpf, ‘The Crimmigration Crisis’ (2006) 56 Am U L Rev 367. 100 See Valsamis Mitsilegas, The Criminalisation of Migration in Europe (Springer 2015); and Neža Kogovšek Šalamon (ed), Causes and Consequences of Migrant Criminalization (Springer 2020).

Legal Methodology between Doctrine and Critique 115 upon the political debate.101 Chapter 2 demonstrated that the human rights discourse was influential at the supranational level during the initial phase of legislative harmonisation in the new millennium, while security concerns left their mark on the earlier and more recent institutional practice.

4.4 Legal Methodology between Doctrine and Critique Questions of methodology are rarely discussed by legal practitioners or academics working on EU migration law. This silence need not be an expression of negligence. In contrast to other disciplines, there are unifying elements connecting the contributions of academics and practitioners: the legal material and court judgments. Doctrinal research in the continental tradition pervades university education on the European continent and the judicial practice of the Court of Justice, even though academic analyses are increasingly relying on alternative methods as well (4.4.1). At the same time, doctrinal research in the continental tradition can be more demanding than the pejorative designation as ‘black letter’ law suggests (4.4.2). Particularly challenging are approaches combining legal insights with the findings of other disciplines (4.4.3).

4.4.1 Academic discourse: from enthusiasm to scepticism Legal cultures are handed over from generation to generation through social practices and shared experiences, among which university education and court judgments play a prominent role. They support a broadly doctrinal outlook on the European continent, which comes in different forms. In this overall context, research on EU law has changed. It used to be defined by a pro-European impetus,102 which had defined much of the early scholarship on justice and home affairs. It was widely expected, at the time, that the EU would replicate the historic mission of overcoming inter-state borders by promoting the rights of third country nationals.103 This expectation became a reality to a limited extent only, as we have seen in Chapters 2 and 3 on the institutional and judicial practice. Recent years have witnessed a ‘critical turn’ of EU legal studies generally and migration law specifically in response to the profound policy crises that have engulfed monetary union and asylum policy.104 Many academics were and are frustrated by what they perceive to be a proliferation of restrictions. Disappointment is one reason that helps to explain the increasing turn towards interdisciplinary perspectives on legal developments. At the same time, academic research on EU migration law has undergone professionalisation. Political salience and practical weight entail that justice and home affairs have acquired a reputation as an essential area of expertise. This gain in reputation allowed migration law experts and research institutions that had often worked with little job security

101 See Leila Kawar, Contesting Immigration Policy in Court (CUP 2015); and Saskia Bonjour, ‘Speaking of Rights’ (2016) 38 Law & Pol 328. 102 See Francis Snyder, ‘New Directions in European Community Law’ (1987) 14 J L & Soc 167, 167–70. 103 See ch 1.3. 104 See Editorial Comments, ‘The Critical Turn in EU Legal Studies’ (2015) 52 CML Rev 881.

116 Interdisciplinary Perspectives and Methodology or institutional funding to gain some ground, although much precarity persists.105 In that respect, legal research lagged behind the humanities and the social sciences where the standing of migration studies had improved considerably from the 1990s onwards. This move to the centre goes along with a greater methodological sophistication, instead of practice-oriented output.106 Moreover, contextual research in the Anglo-Saxon tradition has generally advanced on the European continent in recent years, complementing the traditional focus on doctrinal hermeneutics.

4.4.2 Between ‘black letter’ and doctrinal constructivism Whereas judicial precedent takes centre stage in Anglo-Saxon countries, written legislation and the ideal of codification characterise the civil law tradition. There is much overlap in practice, but this does not unmake the discrepancy in outlook among legal practitioners and academics. Experts with a background in common law are often unaware that the concept of ‘legal order’ hints at an organic vision of coherence and perfection, as Chapter 6.1 on the doctrinal foundations will illustrate. Doctrinal research can be more than a crude exercise of legal positivism repeating the contents of secondary legislation or echoing the judicial reasoning. When discussing these matters, we employ the adjective ‘doctrinal’ as a broad description of the civil law tradition, although the term is used in the common law with a somewhat different meaning to designate a body of judicial precedent built over time. The alternative description ‘dogmatic’ may be frequent in civil law countries but can have negative connotations expressing discontent in English. Close cooperation with legal practitioners is one reason for the prevalence of doctrinal work on the continent. Journals and edited volumes will frequently publish contributions written by academics and practitioners, be it judges, lawyers, civil servants, or people working for as NGOs. Career paths overlap in many cases, especially in the early phases when many do not have a permanent position. Senior academics may assume advisory and other roles in legal practice, while some practitioners enjoy participating in academic debates.107 Research funding by private donors, national research bodies, and EU institutions promotes practice-oriented output.108 Writing studies for the European Parliament or participating in reports coordinated by private consultancy firms for the Commission can be a prestigious (and financially rewarding, unlike journal articles). For example, the transnational Odysseus Network for Legal Experts on Immigration and Asylum in Europe brings together more than forty members from academia and legal practice.109 Court judgments are a unifying element of practice and academia. They reiterate the significance of doctrinal arguments for the simple reason that the Court of Justice follows 105 See Stephen Castles, ‘Understanding Global Migration’ (2010) 36 JEMS 1565, 1572–73; and Michael Bommes, ‘Migration Research in Germany’ in Dietrich Thränhardt and Michael Bommes (eds), National Paradigms of Migration Research (Vandenhoeck & Ruprecht 2010) 127, 148–59. 106 See Rosemary Byrne and Thomas Gammeltoft-Hansen, ‘International Refugee Law between Scholarship and Practice’ (2020) 32 IJRL 181, 191–92; and Jo Shaw, ‘European Union Legal Studies in Crisis?’ (1996) 16 OJLS 231. 107 Byrne and Gammeltoft-Hansen (n 106) 182–88; and Harm Schepel and Rein Wesseling, ‘The Legal Community’ (1997) 3 ELJ 165. 108 See Christina Boswell, ‘The “Epistemic Turn” in Immigration Policy Analysis’ in Freeman and Mirilovic (n 42) 11, 14–16; and Christina Boswell, The Political Uses of Expert Knowledge (CUP 2009). 109 See https://odysseus-network.eu/members (accessed 1 March 2023).

Legal Methodology between Doctrine and Critique 117 the tradition of doctrinal hermeneutics. Many judges, or the référendaires working for them, read law journals. And articles written by academics may even be cited in the opinions of Advocates General, as well as in domestic judgments on EU migration law. On the European continent, the work of practitioners and academics forms, to some extent at least, an overarching epistemic community of legal experts working on similar themes and with a shared methodology. Having said this, doctrinal research comes in many forms, and the prestige of academia differs across Europe. Italy, Germany, and jurisdictions influenced by their traditions, including countries in Central and Eastern Europe, have a strong reputation in doctrinal research; judicial reasoning often develops doctrinal arguments at length and will frequently refer to academic contributions. By contrast, French judgments have traditionally been short statements of result, and the ‘doctrine’ is an analytical summary of core developments rather than an organic evolution of the law with original contributions by academics. Nordic countries and the Netherlands are defined by a tradition of legal formalism, assigning a largely passive role to courts and academics alike. Common law jurisdictions experience little overlap between academics and judges, and doctrinal work tends to focus on judgments; the principally uncritical summary of the facts and the judicial reasoning have little in common with the rich tradition of doctrinal constructivism in the German or Italian tradition. At the supranational level, these various outlooks coalesce into a unique blend.110 Brexit arguably entails that the continental tradition remains influential. This overall picture becomes even more complex if we include the domestic level. National parliaments transpose directives, and domestic authorities and courts decide individual cases, as elaborated upon in Chapter 7 on the administrative dimension. While secondary legislation and CJEU judgments serve as a common point of reference, the legal discourse at the domestic level (in the national language) is partly disconnected from the transnational debate (usually in English these days).111 That has a knock-on effect for legal practice. Supranational judges will normally have pursued a judicial, academic, or political career domestically before being appointed to the bench in Luxembourg; their outlook will often be multilingual and informed by domestic developments as a result.112 Similarly, preliminary reference are influenced by the national context from which they emerge more than by the specific outlook of the transnational debate in the English language. A fine example is the German Constitutional Court obliging a domestic court to refer a case to the CJEU about the effects of refugee protection, granted in Italy, on extradition procedures to the home state, since no less than six different authors, all writing in German, disagreed about the interpretation of Article 9 Qualification Directive 2011/95/EU.113 Anyone aspiring to have a holistic understanding of EU migration law should strive to take into account the linguistic, thematic, and methodological diversity of legal traditions. 110 See Bruno de Witte, ‘European Union Law: A Unified Academic Discipline?’ in Antoine Vauchez and Bruno de Witte (eds), Lawyering Europe (Hart Publishing 2013) 101; and Armin von Bogdandy, ‘National Legal Scholarship in the European Legal Area: A Manifesto’ (2012) 10 ICON 614. 111 See Daniel Thym, ‘The Limits of Transnational Scholarship on EU Law’ (Working Paper LAW 2016/14, EUI 2016) https://ssrn.com/abstract = 2785668 (accessed 1 March 2023). 112 See Hughes Bouthinon-Dumas and Antoine Masson, ‘Quelles sont les revues juridiques qui comptent à la Cour de justice de l’Union européenne?’ [2013] Revue trimestrielle de droit européen 781; and Antonin Cohen, ‘Ten Majestic Figures in Long Amaranth Robes’ in Vauchez and de Witte (n 110) 21–43. 113 See Federal Constitutional Court (Bundesverfassungsgericht), Case 2 BvR 2096/21 (decision of 30 March 2022) paras 46–52.

118 Interdisciplinary Perspectives and Methodology When it comes to doctrinal work, we may distinguish two ideal-typical approaches. On the one hand, the systematisation of the case law and the analysis of secondary legislation in light of the interpretative standards used by the Court (wording, telos, general scheme, drafting history). To do so can be influential in practice, also considering that the open- ended structure of many provisions often entrusts judges with fine-tuning their meaning. The sheer complexity of EU migration law entails that there is plenty of room for doctrinal interpretation. On the other hand, more sophisticated doctrinal research will explore the theoretical foundations and systemic coherence of the legal order and propose solutions at an intermediate level of abstraction. Doing so acknowledges that abstract legal norms cannot be captured fully by interpretative hermeneutics alone; they convey a set of normative values and express basic choices of societies, which can change over time.114 Such outlook can be described as contextually embedded doctrinal constructivism.115 It will usually concern constitutional matters and overarching themes.

4.4.3 Interdisciplinary and critical approaches Complexity is a defining feature of EU migration law. The number and length of legal instruments has multiplied over the years, thus confronting experts with a veritable ‘avalanche’ of secondary legislation, policy developments, practical changes on the ground, and court judgments. Such ‘data overload’ was identified as one reason why EU legal studies have traditionally pursued limited theoretical ambitions, focusing on legislation and court judgments instead.116 That may have changed. Widespread frustration with the policy output among academics and the general reorientation of legal research across Europe entail that academic research strives for interdisciplinary analyses, especially among younger academics writing doctoral theses. Comments throughout this chapter facilitate the identification of non-legal perspectives readers may find useful, and, yet, interdisciplinary embeddedness may come at a price. Time and energy invested into contextual inquiries cannot be used to decipher EU migration law at a time when the sheer complexity of the legal material increases the need for doctrinal categorisation. This may weaken the traditionally close connection between academia and practice. In a worst case scenario, academia might lose practical impact, while not being recognised by other disciplines as an equal sparring partner if it cannot explain the distinct outlook of legal research mirroring the methodological rigidity and theoretical finesse of the social sciences and the humanities. To prevent such an outcome, legal experts should ideally retain an awareness of doctrinal sophistication and develop methodologically sound arguments combining the legal outlook with non-legal perspectives. A number of comments may help readers to define their approach. First, methodological reflection is warranted, both for those performing doctrinal work and others. Interdisciplinary research is usually based on theoretical or methodological assumptions that require an advanced knowledge of other disciplines. Legal experts

114 See Robert R Cover, ‘Foreword: Nomos and Narrative’ (1983) 97 Harv L Rev 4; and Martti Koskenniemi, ‘Constitutionalism as Mindset’ (2007) 8 Theoretical Inq L 9, 23–36. 115 See Armin von Bogdandy, ‘Founding Principles of EU Law’ (2010) 16 ELJ 95. 116 See Neil Walker, ‘Legal Theory and the European Union’ (2005) 25 OJLS 581, 585.

Legal Methodology between Doctrine and Critique 119 should try to understand these propositions, including theoretical or methodological counterarguments supporting a different outcome.117 In doing so, they should be aware of false friends. Similar terminology or concepts may hide deeper discrepancies, in the same way as arguments about the United States or other countries need not be equally valid for the distinct situation of the European polity with its multiple levels of government.118 Of course, interdisciplinary work by legal experts cannot be expected to have the same level of theoretical depth and methodological sophistication as generic pieces from other disciplines. A certain degree of methodological flexibility is warranted, provided that legal experts are aware of the pitfalls of trans-disciplinary exploration. Secondly, it is the legacy of the critical legal studies movement to have highlighted the ideological predisposition of the law mentioned previously, thereby deconstructing its widespread self-description as a quasi-neutral set of rules. Feminism, post-colonial thought, and critical race theory follow these patterns. They are particularly relevant for anyone who is critical of policy developments. Critical approaches contextualise the legal material by highlighting theoretical assumptions, practical effects, or normative implications.119 Such criticism will often remain ‘external’ to the law and plead for policy change. At the same time, there may be room for ‘internal’ critique reintegrating the critical reflection into the interpretation of the law. EU migration law leaves room for such proposals at the level of statutory interpretation and constitutional adjudication, as a result of the open-endedness of many legislative instruments and abstract human rights norms.120 The contextualisation of the law need not be limited to critical approaches, as illustrated by the North-American tradition of ‘law and economics/literature/film/etc’. Thirdly, many legal experts will have unspoken normative and political preferences, even though doctrinal work has traditionally been successful in fencing in the practical impact of these opposing preconceptions.121 Depending on the issues at stake, the same author may call for literal interpretation based on the wording or prefer teleological dynamism when the wording conflicts with their normative convictions.122 Readers may wish to replace such implicit positions with an open and interdisciplinary reflection on underlying preferences. In doing so, the degree of normative thrust will vary among authors. An ideal-typical distinction may be made between legal experts perceiving their work as advocacy and those aspiring to be essentially analytical chroniclers of legal developments.123 Both approaches are equally legitimate and are faced with the same challenge to develop a sound methodological argument which should ideally combine robust doctrinal hermeneutics with an informed interdisciplinary outlook.

117 See Jost Halfmann, ‘World Society and Migrations’ in Michael Bommes and Ewa Morawska (eds), International Migration Research (Ashgate 2005) 129. 118 See Adrian Favell, ‘Migration Theory Rebooted?’ in Caroline B Brettell and James F Hollifield (eds), Migration Theory (3rd edn, Routledge 2015) 318, 321–24. 119 See Robert Cryer, Tamara Hervey, and Bal Sokhi-Bulley, Research Methodologies in EU and International Law (Hart Publishing 2011) 59–72; and the contributions to Siniša Rodin and Tamara Perišin (eds), The Transformation or Reconstitution of Europe (Bloomsbury/Hart Publishing 2018) 3. 120 See ch 6.4; and Thomas Spijkerboer, ‘Analysing European Case Law on Migration’ in Loïc Azoulai and Karin de Vries (eds), EU Migration Law (OUP 2014) 188. 121 See Martti Koskenniemi, From Apology to Utopia (2nd edn, CUP 2005) 513–614. 122 See Martijn van den Brink, ‘Justice, Legitimacy and the Authority of Legislation within the European Union’ (2019) 82 ML Rev 293, 295–300. 123 See also Byrne and Gammeltoft-Hansen (n 106) 189–93.

120 Interdisciplinary Perspectives and Methodology

4.5 Summary Cross-border movements are defined by multiple factors that do not translate into a quasi- mathematical formula allowing for the accurate identification of the decisive variables. Numerous push and pull factors interact, and their relative weight varies over space and time. Elements at the macro-level of analysis, such as wage levels or insecurity, interact with micro-considerations, for instance with regard to information levels, courage, or financial capacities. The law is one element amongst others influencing the decision whether to leave and where to go—not necessarily the most important one, as the example of the ‘welfare magnet’ hypothesis illustrates. Many drivers of migration are beyond the direct control of states. Moreover, preferences can change over time. Temporary stay may turn into settlement, and asylum applicants may decide leave a ‘transit’ country after having lived there for years. Family and ethnic networks play a crucial role, supporting what is often called— somewhat pejoratively—‘chain migration’. The complexity of drivers entails that the legal distinction between ‘voluntary’ and ‘forced’ migration presents itself as a continuum of explanatory factors from an interdisciplinary perspective. Whereas legal experts concentrate on the interpretation and application of the law, other disciplines explore factors which define the design and effectiveness of state policies. Normative theories develop abstract arguments about how to justify borders in a globalised world. Social scientists have coined the notion of ‘control gap’ to analyse the reasons why states often fail to achieve policy objectives. EU integration and court judgments play a prominent role in some explanations, while others focus on the structure of the public discourse and the political system. Politicisation is one factor which helps to explain why sceptical public opinion increasingly leaves a mark on policy developments. In this overall context, intergroup threat theory rationalises how perceptions of threat can influence public opinion. Such findings can be combined with contemporary discourse analysis emphasising that the use of language has framing effects on the way we think and, hence, the policy output. Classic examples of performative speech acts supporting restrictive tendencies are invocations of ‘abuse’ and ‘crisis’, which are widespread in asylum. Ethnographic research contrasts the ‘top-down’ perspective of the legal analysis with the ‘bottom-up’ experience of administrative authorities and migrants. Sociolegal studies unearth the degree of flexibility and creativity ‘street-level bureaucrats’ exercise when applying the law, for instance at the external border where the situation is often unclear. Similarly, migrants possess a certain degree of agency navigating their way with creativity and spontaneity, although we should be careful not to idealise their lived experience. Such accounts present a counterweight to descriptions of European migration law as a purportedly omni- powerful instrument of repressive government in the tradition of post-structural theory. To unmask the ideological predisposition of the law is the starting point of critical approaches that contend, for example, that seemingly neutral legislation on migration status conditionality may embody neoliberal thinking and institutionally embedded racism. ‘Securitisation’ has become a prominent umbrella concept to analyse the shift towards restriction in the domain of asylum in Europe and beyond. Recent years have witnessed the proliferation of interdisciplinary research among academics working on European Union law generally and supranational migration law specifically. Doing so complements the focus on doctrinal work in the continental tradition, which has traditionally constituted a unifying element for an overarching epistemic community of

Summary 121 legal academics and practitioners in many civil law jurisdictions. Frustration with policy outcome is one reason for the interdisciplinary turn, which can be methodologically challenging, especially if the author aims for the reintegration of non-legal perspectives into the legal argument. The latter embraces various forms of doctrinal work ranging from legislative interpretation and the systematisation of the case law to sophisticated arguments of contextually embedded doctrinal constructivism, especially on overarching themes and constitutional matters. All these approaches are equally legitimate, provided that authors reflect openly on what they are doing.

5

Human Rights and State Sovereignty ‘Migration’ is commonly understood to refer to cross-border movements—an international phenomenon by definition. Nevertheless, international law did not develop an extensive set of customary rules or contractual obligations on many topics discussed in this book. The diversity of factual backgrounds and conflicts of interests between home states and receiving countries has prevented the emergence of an overarching international migration law, thus leaving important aspects for states to deal with individually. Chapter 1 revealed that the history of European migration law did not see extensive internationalisation, until today’s European Union promoted internal free movement and began regulating the entry and stay of third country nationals. Yet, absence of a dense network does not imply that international law is irrelevant. There are important guarantees states must respect. The international legal context of European migration law includes both essential multilateral instruments, such as the Refugee Convention and international human rights law, as well as contractual commitments enshrined in bilateral treaties. This chapter will present core features of these instruments and explain how they interact with EU legislation. Indeed, the claim to autonomy inherent in the Court’s famous dictum that the EU Treaties establish ‘a new legal order’1 means that the legal effects of international obligations are not straightforward, in the same way as it can be difficult to determine the precise meaning of international rules. Interaction with supranational legislation brings to the fore various nuances, with repercussions for academic analyses and day-to-day decision-making. Our inquiry begins with the classic template of ‘alien’s law’, which informs the design and analysis of statutory obligations to date. Inter-state treaties on specific subject matters often protect individuals living abroad qua nationality (5.1). Meanwhile, human rights are the primary source for the protection of migrants nowadays; they serve as a conceptual counterweight to state sovereignty (5.2). Comments thereafter will concentrate on the parameters guiding the interaction of the main human rights instruments with EU migration law, while specific guarantees, such as the prohibition of refoulement, will feature in the chapters of Part II. We shall focus on the lead function of the European Convention on Human Rights (ECHR) (5.3), its impact on the Charter of Fundamental Rights of the European Union (CFR) (5.4), and the complementary role of international treaty bodies (5.5). The Refugee Convention serves as an essential point of reference for the Qualification Directive, even though the international practice has had, so far at least, limited practical impact on the judicial output of the Court of Justice (5.6).

1

Case 26/62 van Gend en Loos EU:C:1963:1.

European Migration Law. Daniel Thym, Oxford University Press. © Daniel Thym 2023. DOI: 10.1093/oso/9780192894274.003.0006

‘Aliens Law’ as Protection qua Nationality 123

5.1 ‘Aliens Law’ as Protection qua Nationality Conventional descriptions of state sovereignty had emphasised largely unfettered state control over migration: Apart from special treaties of commerce, friendship, and the like, no state can claim the right for its subjects to enter into, and reside on, the territory of a foreign State. The reception of aliens is a matter of discretion, and every state is by reason of its territorial supremacy competent to exclude aliens from the whole, or any part, of its territory.2

That is not to say, as we have seen in Chapter 1, that states necessarily pursued restrictive entry policies, rather that the choice whether to do so was not determined by international law. Few obligations dealt with side aspects of contemporary migration law, including minimum standards to be enforced horizontally via the rules of state responsibility (5.1.1). Conventions drafted in the framework of the Council of Europe build on this tradition and retain some relevance in specific circumstances (5.1.2), as do rules on the reduction of statelessness (5.1.3).

5.1.1 Historic minimum standards as inter-state obligations In a world defined by state sovereignty, individuals held few rights of their own. Classic international law, as it developed during the nineteenth century, protected individuals only indirectly via the coordination of conflicting sovereignty claims. The legal status of ‘aliens’—as non-nationals were widely called at the time—was construed as the intersection of the territorial sovereignty of the host country and the personal sovereignty of the home state. On that basis, political and legal conflicts were regulated along the horizontal axis of inter-state relations.3 Legal debates concentrated on ‘minimum standards’ for aliens, which did not lay down individual rights in the contemporary sense. Home states asserted their own rights when responding to the mistreatment of nationals by host countries, as opposed to individuals acting on their own behalf.4 Enforcement of minimum standards relied on diplomatic protection by home states in accordance with the rules on state responsibility, including the potential of compensation.5 International law distinguished two methods of how to regulate the status of foreigners, which inform migration law until today. States disagreed about whether to support ‘minimum standards’ establishing an autonomous level of protection defined internationally— or ‘national treatment’, which effectively prescribed parity of foreigners with nationals. A variant of ‘national treatment’ was the ‘most favoured nation’ principle, which deferred to domestic rules but guaranteed treatment that was no less favourable than that of nationals 2 Hersch Lauterpacht, Oppenheim’s International Law, vol I (8th edn, Longham 1955) 675–76; for nuances in the earlier debate see Vincent Chetail, ‘Sovereignty and Migration in the Doctrine of the Law of Nations’ (2016) 27 EJIL 901. 3 See Vincent Chetail, International Migration Law (OUP 2019) 59–61. 4 See Panevezys-Saldutiskis Railway Case (Estonia v Lithuania) (Permanent International Court of Justice, 28 February 1939) para 65. 5 See Kay Hailbronner and Jana Gogolin, ‘Aliens’ in The Max Planck Encyclopedia of Public International Law (OUP online edition, last updated September 2013) http://www.mpepil.com (accessed 1 March 2023) paras 25–28.

124 Human Rights and State Sovereignty of any other third country.6 Both the Refugee Convention and supranational legislation take up these formulae. Beneficiaries of international protection, for instance, have access to social welfare under the same conditions as nationals (national treatment); access to employment follows the rules on foreigners residing legally (most favoured nation); and identity papers or travel documents must be issued in accordance with international rules, irrespective of domestic law (minimum standard).7 While the methods of regulating the status of foreigners remain relevant, the contents of the historic minimum standards have little significance for migration law. Entry and stay were not generally dealt with—except for some provisions in bilateral treaties of friendship, commerce, and navigation, which were mentioned in Chapter 1.1.4 on the prehistory of Europeanisation. Minimum standards focused on economic guarantees for trade and investment; they were the forerunner of contemporary investment protection treaties rather than of modern human rights law. As a result, dynamics were reversed. Those supporting enhanced protection argued against national treatment, since parity with nationals often meant little or no compensation in case of expropriation. Against the opposition of Latin American countries, the United States insisted on ‘adequate’ compensation. Several language versions of Article 67(1) TFEU take up this formula when promising the ‘fair’ (French: équitable; German: angemessene) treatment of third country nationals. Moving beyond minimum standards, the idea of horizontally regulating the status of foreigners through the use of reciprocal guarantees for nationals of the state parties by means of bi-or plurilateral treaties remains pertinent until today. Free movement of Union citizens follows this paradigm in the same way as the EU’s association agreements with neighbouring states, which will be presented in Chapter 17. Their benefits are usually confined to the nationals of the state parties, not other foreigners. Secondary legislation on legal migration habitually embraces provisions authorising more favourable treatment for nationals of some countries under bilateral agreements.8 Moreover, the concept of personal sovereignty of home states over nationals residing abroad informs the obligation, enshrined in international custom, to readmit nationals.9 The non-binding Global Compact for Safe, Orderly and Regular Migration, adopted in 2018, reaffirmed ‘the obligation of States to readmit their own nationals’.10 Migration law as a domain of inter-state obligations retains some relevance in the age of human rights.

5.1.2 Limited impact of the conventions of the Council of Europe After the Second World War, the Council of Europe was founded as an umbrella organisation to promote ‘common action in economic, social, cultural, scientific, legal, and administrative matters’.11 To date, the Strasbourg-based organisation brings together forty-six state 6 See generally Hollin Dickerson, ‘Minimum Standards’ in Max Planck Encyclopedia (n 5) paras 6–10 (last updated October 2010). 7 See Refugee Convention, arts 21, 23, 27–28; and Qualification Directive 2011/95/EU, arts 24–25, 29, 32(1). 8 By way of example see Family Reunification Directive 2003/86/EC, art 3(4)(a); and Blue Card Directive (EU) 2021/1883, art 4(1). 9 See Kay Hailbronner, ‘Readmission Agreements and the Obligation on States under Public International Law’ (1997) 57 Heidelberg J Intl L 1, 11–15. 10 See UNGA, ‘Global Compact for Safe, Orderly and Regular Migration’ Objective 21 No 37 (Resolution No 73/ 195, 19 December 2018). 11 Statute of the Council of Europe (adopted 5 May 1949, entered into force 3 August 1949) ETS No 1, art 1(b).

‘Aliens Law’ as Protection qua Nationality 125 parties, including Azerbaijan and Turkey. It should not be confused with the EU institutions known as the ‘Council’ and the ‘European Council’. Contrasts between the Council of Europe and the European Union transcend questions of membership. Institutional design differs markedly: while the EU is commonly described as being ‘supranational’ with legislative functions and directly effective rules, the Council of Europe follows the paradigms of public international law. As a result, negotiations of international treaties often dragged on for years, ratification by national parliaments was delayed (and will rarely include all members), and application suffered from the absence of an international judicial body with generic jurisdiction to interpret common rules. Notwithstanding these limitations, several conventions on specific aspects of migration law were concluded in the framework of the Council of Europe before the EU started occupying the terrain.12 Although these conventions remain legally intact, their practical relevance was considerably diminished by successive rounds of EU enlargement. A defining feature was reciprocity, since the conventions covered nationals of state parties only, not third country nationals. As a result, comparatively progressive rules on protection against expulsion, access to social assistance, or the rights of migrant workers, including family reunification, lost ground when the beneficiaries became Union citizens (or benefited from association agreements). Today, the European Convention on Social and Medical Assistance of 1953,13 the European Convention on Establishment of 1955,14 and the European Convention on the Legal Status of Migrant Workers of 197715 have limited bearing primarily for Turkish nationals and—as a result of Brexit—British citizens living in another state party which ratified the convention in question.16 These multilateral conventions are complemented by a network of bilateral agreements on social security coordination and double taxation, which do not concern entry and stay but are practically relevant nonetheless. Analysis of these agreements generally requires an advanced knowledge of tax law or social security coordination. Application ratione personae is usually confined to nationals of the Contracting Parties. Moreover, these agreements are primarily concerned with the protection of acquired rights and the coordination of national social security systems in case of legal migration; they do not set out supranational prescriptions for benefit levels, as described in Chapter 15.3.4 on integration. Before the advent of EU harmonisation, the Council of Europe also engaged in political initiatives to coordinate asylum legislation. These initiatives did not succeed, although Chapter 13.3.1 will mention that they were one potential source of inspiration for the well-known ‘first entry rule’ in the Dublin III Regulation according to which Member States at the external borders shall typically assume responsibility of asylum seekers.

12 For an overview see Kees Groenendijk, ‘Long-Term Immigrants and the Council of Europe’ (1999) 1 EJML 275, 276–81; and Karin Oellers-Frahm, ‘The Contribution of the Council of Europe to the Legal Position of Aliens’ in Jochen A Frowein and Torsten Stein (eds), Die Rechtsstellung von Ausländern nach staatlichem Recht und Völkerrecht, vol 2 (Springer 1987) 1725. 13 Adopted 11 December 1953, entered into force 1 July 1954, ETS No 14; it applies to several EU Member States, as well as to Norway, Iceland, the UK, and Turkey. 14 Adopted 13 December 1955, entered into force 23 February 1965, ETS No 19; it is binding upon nine EU Member States, as well as Norway, Iceland, the UK, and Turkey. 15 Adopted 24 November 1977, entered into force 1 May 1983, ETS No 93; it has been ratified by eleven States, including six EU Member States and five third states. 16 While Turkey acceded to all three listed above, the UK ratified the first two; a full list of conventions, signatures, and ratifications can be found at https://www.coe.int/en/web/conventions/full-list (accessed 1 March 2023).

126 Human Rights and State Sovereignty

5.1.3 Reduction of statelessness as a legal obligation The classic template of legal protection qua nationality presupposed that home states stood ready to defend the rights of nationals. This was never straightforward. Effective diplomatic protection requires consular capacities, diplomatic clout, and political will. It was beyond reach for many foreigners and was often exercised to secure financial interests of economic actors. In spite of this, the regulatory function of nationality as a door opener for legal guarantees meant that states pursued the objective of closing loopholes by ensuring that ‘everyone has a right to a nationality’,17 while simultaneously avoiding double nationality.18 The Convention on the Reduction of Statelessness of 1961 is one of the most successful conventions concerning the rights of migrants; it has been ratified widely in Europe and beyond.19 More specifically, the Convention obliges states to establish domestic rules so that stateless children and adults may acquire nationality by birth or via naturalisation under certain conditions. This was given additional legal teeth by the European Court of Human Rights (ECtHR), which integrated core guarantees against statelessness into the case law on the right to respect for private life under Article 8 ECHR, for situations dealing with the breakup of Yugoslavia and the Soviet Union.20 These principles can possibly be relied upon in other circumstances as well, although the precise parameters would have to be sorted out. On the European continent, the legal framework for the reduction of statelessness is comparably strong, notwithstanding room for improvement.21 We shall see in Chapter 15.7 on integration that it will be governed by international law in the years to come. The wide scope of supranational competences excludes the harmonisation of nationality laws. That is not to say that Union law has nothing to say on statelessness. Article 67(2) TFEU defines the term ‘third country national’ to include ‘stateless people’; competences for law- making in Articles 77–80 TFEU embrace stateless people. EU institutions followed this guidance. By way of example, the introductory definitions in the Long-Term Residents Directive and the ICT Directive reaffirm that the rules therein apply to ‘any person who is not a citizen of the Union’.22 Stateless people may, therefore, acquire residence permits in accordance with that legislation. This subsidiary supranational safety net is irrelevant, however, whenever they qualify for naturalisation under the Convention on the Reduction of Statelessness or domestic laws. Recent years have witnessed a heightened interest in statelessness. That is laudable in many respects but runs the risk of overstating the significance of de jure statelessness. It seems that the increase in attention is informed, in part at least, by the symbolic visibility of the catchword ‘statelessness’. It is easier to urge states to counter statelessness than to engage in complex debates on, for instance, the procedural intricacies of the Return Directive. Yet, 17 Universal Declaration of Human Rights (UNGA Resolution 217 (III), 10 December 1948),, art 15(1), which is not legally binding. 18 See (European) Convention on the Reduction of Cases of Multiple Nationality and Military Obligations in Cases of Multiple Nationality (adopted 6 May 2002, entered into force 28 March 1968) ETS No 43, with later additional protocols. 19 Adopted 30 August 1961, entered into force 13 December 1975, 989 UNTS 175; it was ratified by all EU Member States except for Poland, Estonia, Slovenia, and France (they only signed it). 20 See Hoti v Croatia App no 63311/14 (ECtHR, 26 April 2018) §§ 119–24; and the previous ‘weaker’ case law on non-deportation, such as Sisojeva v Latvia App no 60654/00 (ECtHR [GC], 15 January 2007) §§ 90ff. 21 See Katia Bianchini, Protecting Stateless Persons (Brill/Nijhoff 2018). 22 Long-Term Residents Directive 2003/109/EC, art 2(a); and ICT Directive 2014/66/EU, art 3(a).

Human Rights qua Personhood 127 it is precisely these matters of de facto statelessness the interdisciplinary debate is primarily concerned with, by focusing on the legal status and the lived experience of those having a passport of a third state officially that, in reality, is not worth much.23 Legal experts should beware, therefore, of synchronising the legal rules on the reduction of de jure statelessness with the more general debate.

5.2 Human Rights qua Personhood Human rights often take centre stage in analyses of migration law. They do so for good reasons. Human rights feature prominently in political debates, and judges at the ECtHR have played an essential role in extending the rights of migrants.24 Furthermore, there is social reputation to be earned as an expert in human rights in the world of academia and beyond. Anyone intuitively understands their significance, while complex questions of legislative interpretation appear less attractive. Internationally, human rights were critical for the introduction of individual guarantees independent of nationality (5.2.1). Hanna Arendt coined the famous formula of the ‘right to have rights’, which legal academics occasionally misinterpret (5.2.2). It reminds us of the normative significance of human rights as a trajectory for the promotion of migrants’ interests (5.2.3).

5.2.1 Individual guarantees for ‘everyone’ A comparison to the nationality-based protection regime of former times reveals that international human rights law transcends the paradigm of sovereign statehood in at least three ways. First, human rights transcend the horizontal axis of inter-state relations. Their protective reach is not commonly limited to nationals of state parties, as in the case of most bilateral agreements. A Nigerian woman living in Italy can rely on human rights in the same way as Ukrainians working in Poland. The distinction between citizens’ rights and human rights is normally confined to specific guarantees, for instance in Articles 15, 39–40, and 45 CFR. Secondly, other guarantees can usually be invoked by ‘everyone’25 within the jurisdiction of a state. The jurisdictional yardstick requires scrutiny of the potential and limit of extraterritorial reach, which can be found in Chapter 12.2.2 on border controls. Very few provisions erect additional hurdles in terms of migration status. By way of example, social security benefits and advantages can be confined to those ‘residing and moving legally’,26 while freedom of movement may be restricted to anyone ‘lawfully within the territory’,27 thus probably excluding asylum applicants during the procedure.28 Such limitations ratione 23 See Katherine Tonkiss and Tendayi Bloom, ‘Theorising Noncitizenship’ (2015) 19 Citizenship Stud 837; and the other contributions to the special issue. 24 While experts in international law habitually refer to ‘human rights’, the term ‘fundamental rights’ is often preferred by scholars of constitutional law; this book uses both expressions interchangeably, mirroring the official titles of the ECHR and the CFR. 25 ECHR, art 1. 26 CFR, art 34(2). 27 Additional Protocol No 4 to the ECHR, art 2(1). 28 See ch 13.8.4 and ch 13.9.4.

128 Human Rights and State Sovereignty personae are the exception to the rule that human rights generally apply to ‘everyone’ irrespective of nationality. Thirdly, a central novelty was the international character of the instruments. They elevated matters that had previously been considered to be essentially in the domaine réservée of internal affairs to the international plane. Crucially, the conventions went beyond abstract commitments by establishing expert bodies or courts to supervise interpretation and application. They mark an important practical distance from the historic model of the minority treaties of the inter-war period, mentioned in Chapter 1.1.3, which proved ineffective in protecting ethnic and cultural minorities in Central and Eastern Europe against pervasive atrocities.

5.2.2 Hanna Arendt and the ‘right to have rights’ The historic context of the inter-war period is relevant for contemporary debates insofar as the plight of refugees and stateless persons informed the famous dictum by Hanna Arendt portraying citizenship as the ‘right to have rights (and) . . . to belong to some kind of organised community’—in contrast to abstract personhood based on international guarantees, which Arendt associated with rightlessness.29 She was critical, consequently, when the Universal Declaration of Human Rights promoted human rights without meaningful institutional infrastructure. Nevertheless, one should be careful not to misread Arendt’s statement as a vindication of nationality-based protection regimes. Her critique transcended the forum of protection. Arendt’s usage of the terms ‘right’ and ‘citizenship/nationality’ did not follow legal categories but was based on a theoretical understanding of ‘the political’ as being embedded in constant political and social struggles.30 It should not be equated, therefore, with a call for powerful courts. An important lesson of Arendt’s work is critical awareness of the context of international instruments. To be effective they depend on domestic courts and state authorities to take up their views. Capacities preclude international bodies from dealing with literally thousands of cases; they depend on international lead judgments to radiate and to shape domestic practices. Moreover, normative preference for the position of individuals, which human rights are generally perceived to embody, should not be confused with depoliticisation or conceptual primacy of judicial avenues over political and social emancipation.31 Notwithstanding the variety of theoretical positions, of which the views of Hanna Arendt represent one camp, constitutional theorists generally recognise that human rights raise profound questions of public authority, territorial ordering, and social solidarity, which do not become irrelevant in a post-Westphalian world.

29 See Hannah Arendt, The Origins of Totalitarianism (Harcourt, Brace & Company 1951) ch 9; for a contextualisation see Seyla Benhabib, The Rights of Others (CUP 2004) ch 2. 30 See Ayten Gündoğdu, Rightlessness in an Age of Rights (OUP 2015) chs 3–4; and also Dana Schmalz, Refugees, Democracy and the Law (Routledge 2020) 51–56, 71–82. 31 See the prominent critique by David Kennedy, The Dark Sides of Virtue (Princeton UP 2004) 13–24.

Human Rights qua Personhood 129

5.2.3 Normative counterweight to state sovereignty Human rights serve as a counterweight to statist claims to migration control, amongst others since their open-ended texture lends itself for dynamic interpretation. The ECHR famously conceives of the ECHR as ‘a living instrument, which . . . must be interpreted in the light of present-day conditions’.32 Human rights convey a set of normative values and express basic choices of societies that can change over time, thus serving as a trajectory for the promotion of social justice.33 Legal arguments inevitably overlap with normative considerations when courts define the protective scope of abstract human rights norms. This potential for dynamism renders human rights attractive for anyone trying to advance the legal status of migrants. At the same time, the potential for dynamic interpretation hints at a built-in limitation. Human rights serve as a counterweight to sovereign statehood but do not render it irrelevant. International courts habitually challenge established national practices from an external viewpoint, thus often advancing the rights of migrants. Such outcomes are not, however, based on the inevitable primacy of individual rights over public interests. The ECtHR commences most judgments with a generic recognition that ‘as a matter of well- established international law and subject to its treaty obligations, a State has the right to control the entry of non-nationals into its territory’.34 Academics have criticised the underlying ‘Strasbourg reversal’35 to ground judicial deliberations on statist control prerogatives as starting point. That criticism does not undo, however, the judicial practice. Judges considerably limited state sovereignty without invalidating it. Widespread inclination to focus on ground-breaking judgments, while ignoring more constrained rulings and unsuccessful applications, can support a one-sided impression of vigorous and continuous dynamism. Holistic accounts of the ECtHR case law tend to paint, as we shall see, a more nuanced overall picture. Chapter 3.2.2 came to a similar conclusion for the judicial output of the Court of Justice, which epitomises the proverbial shades of grey. The metaphor of balance that underlies the assertion of human rights as a counterweight to state sovereignty does not mean that a stable equilibrium will be reached. The balance is bound to remain precarious for doctrinal, theoretical, and contextual reasons. Doctrinally, the open-ended texture of human rights law complicates definite statements. Judges in Europe generally treat human rights not as a trump card but as the entry ticket for a complex balancing exercise.36 Doing so reiterates the basic tension between normative utopia and interest-based realism which, according to Martti Koskenniemi, constitutes the grammar of international law.37 International courts are obliged to orientate their reasoning at the object and purpose of the instrument that is human rights protection; at the same 32 Tyrer v United Kingdom App no 5856/72 (ECtHR, 25 April 1978) § 31. 33 See generally Robert R Cover, ‘The Supreme Court, 1982 Term –Foreword: Nomos and Narrative’ (1983) 97 Harv L Rev 4; and Martti Koskenniemi, ‘Constitutionalism as Mindset’ (2007) 8 Theor Inq Law 9, 23–36. 34 First used by Abdulaziz and others v United Kingdom App nos 9214/80, 9473/81 and 9474/81 (ECtHR, 28 May 1985) § 67. 35 See Marie-Bénédicte Dembour, When Humans Become Migrants (OUP 2015) chs 1, 4; similarly, the ‘statist assumption’ of Cathryn Costello, The Human Rights of Migrants and Refugees in European Law (OUP 2016) 10–11. 36 See generally Matthias Klatt and Moritz Meister, The Constitutional Structure of Proportionality (OUP 2012). 37 See Martti Koskenniemi, From Apology to Utopia (2nd edn, CUP 2005) chs 4, 8; and also Christian Tomuschat, Human Rights (3rd edn, OUP 2014) 1–11.

130 Human Rights and State Sovereignty time, they are also legally bound to take seriously the opinion of the Contracting Parties in accordance with the law of treaties.38 Theoretically, the abstract notion of sovereignty serves as a proxy for arguments about the value of particularistic self-government. Chapter 4.2 illustrated that political theorists disagree quite fundamentally about justifications of state borders, but it is precisely this variety of normative positions which provides the theoretical background for the unattainability of an equilibrium. Chapter 1.3.1 explained why the abstract notion of an ‘area of freedom, security, and justice’ in the EU Treaties does not provide straightforward answers either. Contextually, the politicisation of migration law, traced in Chapter 3.2.3, arguably weakens the relative weight of judges. Political disputes over contemporary ECtHR judgments on migratory matters exhibit political contestation that was unheard-of during the expansive phase of the case law a generation ago.

5.3 Lead Function of the European Convention Dynamic interpretation is a comparatively new phenomenon. Insofar as the ECtHR is concerned, it started in the early 1990s at a time of widespread optimism about globalisation. Many of the judgments delivered at the time retain critical relevance as a protective shield, notwithstanding a series of rulings refraining from further expansion (5.3.1). Experts in migration law should recognise institutional constraints preventing the Strasbourg-based Court from functioning as a transcontinental court of appeal. The sheer number of individual applications requires effective cooperation with domestic courts, which the practice of the Grand Chamber supports through general principles guiding the decentralised application of the Convention (5.3.2). Previous disputes between the ECtHR and the CJEU over the transfer of asylum seekers have subsided, with both courts pursuing a viable working relationship (5.3.3). While comments hereafter assess the judicial output holistically, specific aspects of the case law will be presented as an integral part of the constitutional foundations in the sectoral chapters in Part II.

5.3.1 Judicial dynamism in the field of migration Judges in Strasbourg started dealing with migration law in earnest in the 1990s, following more than thirty years of almost complete silence—with the exception of the Abdulaziz judgment that introduced the well-known formula of state control ‘as a matter of well- established international law’ in a case concerning family formation with applicants from the former colonies.39 Things changed quickly thereafter. Articles 3 and 8 ECHR turned into veritable hotbeds of dynamic interpretation during an epoch that was described as the ‘end of history’40 to illustrate the widespread expectation that globalisation would go 38 ND and NT v Spain App nos 8675/15 and 8697/15 (ECtHR [GC], 13 February 2020) § 182 reaffirmed that the ECHR must be interpreted in line with the Vienna Convention on the Law of Treaties, including, art 31(1), (3), (4) on the intention and subsequent practice of state parties as one—not necessarily the decisive—element influencing interpretation. 39 See ch 1.1.6; for earlier decisions of the former European Commission of Human Rights see Dembour (n 35) ch 4. 40 Francis Fukuyama, The End of History and the Last Man (Free Press 1992).

Lead Function of the European Convention 131 hand in hand with a predominance of liberal constitutionalism. An integral part of this normative and discursive shift was the universalisation of the human rights discourse, which had not originally been intended fundamentally to restructure cross-border situations.41 Optimism about globalisation, including new forms of governance and membership beyond the state,42 are among the contextual factors of the dynamism—in the same manner as increasing politicisation, doubts about globalisation, and the return of geopolitical rivalry inform contemporary debates about international law.43 Experts in migration law who begin studying the subject today will find it difficult to grasp the sense of surprise with which the innovative early judgments were received, provided that they were acknowledged at all, since the Convention did not play a major role at the time. A telling example was the warning of the German federal administrative court against the ‘conversion’44 of the prohibition of torture and inhuman or degrading treatment into a de facto right to asylum. Rulings which are taken for granted nowadays were met with astonishment three decades ago. These innovations turned ECtHR judgments into an essential point of reference for anyone dealing with migration law, be it as academics or practitioners. Examples of dynamism are the human rights-based prohibition of refoulement and the transformation of the prohibition of collective expulsion into a free-standing procedural guarantee, as shown in Chapter 12.2.4 on border controls and Chapter 13.2.2 on asylum. At the same time, instances of thwarted dynamism and judicial standstill equally define the case law. Chapters 10.4 on general features will explain that the Grand Chamber treads carefully when applying the prohibition of discrimination. Chapter 15.2.2 on settlement will illustrate that judges refrained from declaring an absolute ban on expelling second- generation migrants. Denial of humanitarian visas was prominently found not to amount to an exercise of state jurisdiction, since to conclude otherwise would ‘negat[e]‌’45 statist control imperatives. Against the backdrop of these diverse findings, observers paint nuanced overall pictures. While academics often lament built-in limitations,46 practitioners occasionally warn against a slippery slope of unbound interpretation.47 Moreover, state parties and several constitutional courts called upon judges to exercise constraint,48 reiterating our earlier finding about human rights as a normative and doctrinal counterweight to—not a replacement for—state sovereignty. By way of example, British Conservatives have promised to dilute the domestic role of the Convention for years,49 and the Swiss Peoples’ Party launched a 41 See Samuel Moyn, The Last Utopia (Harvard UP 2010). 42 See the optimistic outlook by Yasemin Soysal, Limits of Citizenship (University of Chicago Press 1994); and Rainer Bauböck, Transnational Citizenship (Edward Elgar Publishing 1994). 43 See ch 13.1.6. 44 (German) Federal Administrative Court (Bundesverwaltungsgericht), Case 9 C 38.96 (judgment of 15 April 1997); see also Kay Hailbronner, ‘Artikel 3 EMRK. Ein neues europäisches Konzept der Schutzgewährung?’ [1999] Die Öffentliche Verwaltung 617, 618. 45 MN and others v Belgium App no 3599/18 (ECtHR [GC], 5 May 2020) §§ 124–25; and further ch 12.2.2. 46 See Dembour (n 35); and Costello (n 35); for a more positive outlook see Sylvie Saroléa, Droits de l’homme et migrations (Bruylant 2006) 317–83. 47 See the former President of the Belgian Constitutional Court Marc Bossuyt, Strasbourg et les demandeurs d’asile (Bruylant 2010) 105–75. 48 See Angelika Nussberger, The European Court of Human Rights (OUP 2020) chs 4, 6; and Marten Breuer (ed), Principled Resistance to ECtHR Judgments (Springer 2019). 49 See Roger Masterman, ‘The United Kingdom’ in Patricia Popelier and others (eds), Criticism of the European Court of Human Rights (Intersentia 2016) 447.

132 Human Rights and State Sovereignty referendum to the same effect, even though the ‘self-determination initiative’ was defeated in 2018.50 In the same year, the Committee of Ministers of the Council of Europe called upon the Court not usually to substitute domestic balancing in complex subject matters, such as migration, by its own assessment.51 Absence of further innovation can be rationalised, in part at least, as a judicial response to such criticism. Pleas for caution stretch from domestic constitutional courts defending their prerogatives over governments across Europe wary of restrictions on migration control to outright populist challenges to the authority of the inter-and supranational judiciary. Irrespective of underlying motives and explanations, absence of further innovation in no way diminishes the protective credentials of the earlier case law. International standards provide a safety net against potential retrogression at times of heightened politicisation.

5.3.2 Focus on general principles and the Grand Chamber Those who are not familiar with EU law should not confuse the European Court of Human Rights (ECtHR) in Strasbourg with the Court of Justice of the European Union (CJEU) in Luxembourg. While the latter can apply supranational EU legislation to the twenty-seven Member States, the former has jurisdiction over the European Convention on Human Rights (ECHR), which is binding on the forty-six members of the Council of Europe (Russia left after the war of aggression against Ukraine during 2022). While judges in Luxembourg are usually consulted via preliminary references from domestic courts, as described in Chapter 3.3 on the judicial practice, the Court in Strasbourg can be seized directly by individuals.52 Individual applications are the hallmark of the judicial architecture, which can pose significant challenges nevertheless. Judges in Strasbourg have been overwhelmed by the sheer number of applications, stretching institutional resources and creating a backlog of pending cases.53 Workload is a concern experts in migration law should not underestimate. Judges simply do not have the capacity to be a final court of appeal in thousands of cases, literally speaking. The Court’s effectiveness depends on a constructive working relationship with domestic courts applying the Convention. Judges in Strasbourg will continue dealing with individual applications to ensure that justice is done and to respond to new developments; yet the tension between ‘individual’ justice for applicants and ‘constitutional’ justice focusing on general principles cannot be resolved easily.54 Judges have to balance countervailing claims to European supervision and the prerogatives of domestic institutions on a daily basis. Interim measures are not legally binding but they can influence developments on the ground nonetheless. They have gained visibility in the field of asylum in recent years. Additional Protocol No 16, which entered into force in 2022, officially integrated the principle of subsidiarity into the last recital of the preamble. Living the basic idea behind

50 See Stefan Schlegel, ‘Eine direkte Demokratie wird in der globalisierten Rechtsordnung erwachsen’ Verfassungsblog (3 December 2018). 51 See Committee of Ministers, ‘Copenhagen Declaration’ (13 April 2018) No 28(c) https://rm.coe.int/copenha gen-declaration/16807b915c (accessed 1 March 2023). 52 See ECHR, arts 34–35. 53 See Nussberger (n 48) 23–37. 54 See Steven Greer, The European Convention on Human Rights (CUP 2006) ch 3.

Lead Function of the European Convention 133 subsidiarity, the Court has made a deliberate effort to lay down predictable standards that can guide the decentralised application of the Convention.55 Judgments of the Grand Chamber are particularly relevant. They habitually decide politically sensitive or legally difficult questions. In doing so, the Grand Chamber replicates a practice spearheaded by the German constitutional court: it distinguishes between ‘general principles’ and the ‘application to the present case’, thereby allowing practitioners and academics to identify abstract standards that can be generalised.

5.3.3 Interaction with EU law Article 6(2) TEU allows for the EU’s accession to the ECHR, and a draft accession agreement had been negotiated before it was blocked by judges in Luxembourg. Human rights limitations to Dublin transfers were one of the issues of concern on the part of the CJEU.56 Political negotiations about a revision of the accession agreement were ongoing at the time of writing. Having said this, formal accession would primarily have procedural consequences. For instance, individuals could seize the ECtHR, after having exhausted local remedies before EU courts, concerning an alleged human rights violation that can be attributed to Frontex staff.57 Accession would not, by contrast, change the constitutional status of the Convention in the EU legal order, which is formally placed at a level below primary law, like any other international treaty. In practice, these hierarchical subtleties do not matter much as long as judges in Luxembourg and Strasbourg march to the beat of the same drum. Both courts seem determined to follow that line. Chapter 13.3.6 will explain that earlier tensions over limits of mutual trust have subsided. The Court of Justice accepted the basic premises of the human rights case law concerning Dublin transfers, while the ECtHR showed restraint in several follow-up rulings. Of course, there may always be discrepancies on specific questions. Such nuances are a necessary ingredient of any pluralistic setting of overlapping jurisdiction among judges in Strasbourg, Luxembourg, and at the highest national courts. What matters is that these differences and occasional wrangles unfold in an environment of dialog and trust.58 Preconditions for such constructive cooperation are good. Article 52(3) CFR aligns the meaning of the Charter with the Convention. From a doctrinal perspective, parallelism concerns the interpretation of human rights, not the contents of secondary legislation. Comments throughout this book will show that EU legislation goes beyond the minimum level of protection under human rights law in many respects. An abstract embodiment is the concept of individual statutory rights, discussed in Chapter 6.5 on the doctrinal foundations. The distinction of principle between human rights and legislation may become blurred when the Court in Strasbourg has recourse to the ‘consensus’ argument in deciding whether to interpret the Convention dynamically.59 The consensus argument might become problematic if judges turned to the 55 See Janneke Gerards, General Principles of the European Convention on Human Rights (CUP 2019) ch 2. 56 See Opinion 2/13 Accession to the ECHR EU:C:2014:2454, paras 192–95. 57 At present, the application would be inadmissible in line with Bosphorus Airways v Ireland App no 45036/98 (ECtHR [GC], 30 June 2005) §§ 149–58. 58 See generally Aida Torres Pérez, Conflicts of Rights in the European Union (OUP 2009); and Nico Krisch, Beyond Constitutionalism (OUP 2010) ch 4. 59 See MSS v Belgium and Greece App no 30696/09 (ECtHR [GC], 21 January 2011) §§ 250–51.

134 Human Rights and State Sovereignty contents of secondary legislation, or non-binding positions of political actors, to find a consensus on how to deal with a specific topic. Doing so would effectively transform legislative commitments into constitutional prescriptions. Judgments on reception conditions illustrate the underlying doctrinal uncertainties.60

5.4 Added Value of the Charter Under the Treaty of Lisbon, the Charter of Fundamental Rights became legally binding: rights and principles therein ‘have the same legal value as the Treaties’.61 Whereas individual judgments will be discussed throughout this volume, this section concentrates on constitutional questions, thus demonstrating the doctrinal specificities and the added value of the Charter. Particularly relevant are instances where the contents of the Charter transcends guarantees in the European Convention (5.4.1). A useful example are the rights of the child, which harbour the potential for surprise outcomes (5.4.2). A legally binding Charter did not transform the European Union into a human rights organisation sensu stricto. It must respect fundamental rights and can contribute to their effective protection, but the policy output may pursue other objectives as well (5.4.3). Absence of a generic human rights policy is supported by the limited scope of the Charter, which applies to national activities only when implementing Union law (5.4.4).

5.4.1 More generous protection An original intent behind the introduction of the Charter was to make existing obligations ‘visible’,62 building upon the unwritten general principles of Union law. The Convention drafting the Charter opted for a pragmatic solution to start with. Article 52(3) CFR states in relation to fundamental rights replicating the ECHR: ‘the meaning and scope of those rights shall be the same as those laid down by the said Convention’. Such parallel interpretation defines the judicial practice. Guarantees replicating the Convention include the prohibition of torture and inhuman or degrading treatment (Article 3 ECHR, Article 4 CFR),63 human rights-based non-refoulement (Article 3 ECHR, Article 19(2) CFR),64 respect for private and family life (Article 8 ECHR, Article 7 CFR),65 the prohibition of collective expulsions (Article 4 Additional Protocol No 4 to the ECHR, Article 19(1) CFR),66 and equal treatment (Article 14 ECHR, Article 21 CFR).67 Notwithstanding occasional disagreement, references to the case law of the Court in Strasbourg is standard practice in the reasoning of the Court

60 See ch 13.2.2; and the ECtHR President Síofra O’Leary, ‘L’ELSJ à la Cour européenne des droits de l’homme’ in Véronique Beaugrand and others (eds), Sa justice (Bruylant 2022) 575, 587–94. 61 TEU, art 6(1). 62 CFR, recital 4. 63 See Joined Cases C-411/10 and C-493/10 NS and others EU:C:2011:865, paras 86–88, 109–14. 64 Case C-239/14 Tall EU:C:2015:824, paras 53–54 confirmed that Article 19(2) CFR codifies ECtHR case law on Article 3 ECHR. 65 See Case C-400/10 PPU McB EU:C:2010:582, para 53. 66 Parallelism ensues from Article 52(7) CFR read in conjunction with the ‘Explanations relating to the Charter’ [2007] OJ C303/17, 24. 67 See Joined Cases C-199–201/12 X, Y and Z EU:C:2013:720, para 54.

Added Value of the Charter 135 of Justice in Luxembourg, despite the EU not having formally acceded, as we have seen, to the European Convention. At the same time, the Charter contains new guarantees that may develop a life of their own. Such dynamism will usually develop over time in a process of discovery. Procedural guarantees, for instance, have gained momentum in the case law on migration: the rights to good administration, to an effective remedy, and of the defence will feature in Chapter 7 on the administrative dimension. In other respects, the Charter is less dynamic. It reiterates the free movement of Union citizens, while simultaneously confirming that is does not cover third country nationals.68 With regard to some articles, it remains difficult to say how the Charter relates to the Convention. Chapter 13.2.3.2 will discuss ambiguities surrounding the right to asylum, and the question of extraterritorial effects will feature in Chapter 12.2.2 on border controls. At an intermediate level of abstraction, two distinctions stand out. While the Convention lists grounds for interferences in the wording of each provision, the Charter contains a horizontal rule, in Article 52(1) CFR, on when limitations can be justified. That is relevant, in particular, for detention where Article 5 CFR might establish stricter justification requirements, as argued in Chapter 16.2.2 on detention. Moreover, the Charter contains both individual rights and abstract ‘principles’. Principles cannot be relied upon autonomously, although legislation adopted to implement them may have to be interpreted in light of the Charter in accordance with Article 52(5) CFR. Principles are particularly relevant for social guarantees enshrined in Title IV on solidarity, although the Charter abstains from specifying which provisions exactly contain principles instead of individual rights.69 Social rights and principles have been mentioned occasionally in judgments on migration, albeit without transformative effects.

5.4.2 Rights of the child as an example Substantial dynamism may emanate from the rights of the child in Article 24 CFR, notably the obligation that ‘the child’s best interests must be a primary consideration’. Special attention to the child’s best interests may not add much substance doctrinally, since it replicates core guarantees under the United Nations Convention on the Rights of the Child of 1989,70 which the ECtHR integrated into the interpretation of Article 8 ECHR on a previous occasion.71 The Court in Strasbourg holds the best interests formula not to be a ‘trump card’72 obliging state parties to always do what is best for a child; countervailing interests can be considered. On the whole, judges take the rights of the child seriously but do not systematically prioritise their interests in the balancing exercise.73 A specific expression is the

68 On CFR, arts 15, 45 see ch 1.3.2. 69 See Niilo Jääskinen, ‘Fundamental Social Rights in the Charter’ in Steve Peers and others (eds), The EU Charter of Fundamental Rights. A Commentary (2nd edn, CH Beck/Hart Publishing/Nomos 2021) 1855. 70 Adopted 20 November 1989, entered into force 2 September 1990, 1577 UNTS 3. 71 See the dynamic interpretation in light of an international consensus by Nunez v Norway App no 55597/09 (ECtHR, 28 June 2011) § 84. 72 IAA and others v United Kingdom App no 25960/13 (ECtHR, 8 March 2016) § 46. 73 See Ciara Smyth, ‘The Best Interests of the Child in the Expulsion and First-Entry Jurisprudence of the European Court of Human Rights’ (2015) 17 EJML 70.

136 Human Rights and State Sovereignty administrative detention of minors, which must only be used as a ‘last resort’, as explained in Chapter 16.2.2 on return. Balancing between the child’s interests and countervailing policy objectives is ingrained in the ‘best interests’-formula as a result of careful drafting. Note that the wording refers to the best interests as ‘a’ (not the) ‘primary’ (not paramount) concern.74 The contrast stands out when you compare the careful wording of Article 3 Convention on the Rights of the Child to the stricter formulation in a previous, non-binding Declaration.75 Interpreting Article 24 CFR, the Court of Justice has followed this rationale to start with when judges in Luxembourg emphasised the need for balancing.76 On many occasions, Article 24 CFR was mentioned as one argument amongst others in the interpretative exercise, without explaining to what extent the outcome depended on human rights.77 Such vagueness mirrored a general tendency towards interpretative ambivalence, traced in Chapter 6.5.3 on the doctrinal foundations. However, several cases concerning family reunification with unaccompanied minors turning years old seemed to mark the potential beginning of judges drawing more specific conclusions from Article 24 CFR during the 2020s. They will be mentioned in Chapter 14.2.3 and 14.3.5 on legal migration. Judgments prohibiting the adoption of return or transfer decisions against unaccompanied minors, to be discussed in Chapter 16.5.5 on return and Chapter 13.3.3 on the Dublin system, appeared to signal an increasing momentum behind Article 24 CFR. Room for future innovation may be found in the domain of refugee status determination with regard to children.78 Introduction of a free-standing guarantee, as opposed to dynamic interpretation of Article 8 ECHR, arguably reinforced the prospect of unexpected effects. Having said this, such outcome will typically be context-specific, mirroring the factual and doctrinal specificities of the case at hand. In other scenarios, the balancing exercise inherent in the ‘best interests’-formula may tilt in a different direction. The legal effects of Article 24 CFR remain notoriously difficult to predict.

5.4.3 EU not a ‘human rights organisation’ It is beyond doubt that the EU must respect human rights and that many of its policies concern highly sensitive matters. In that respect, the supranational institutions have an important role to play in upholding and promoting human rights, in line with the basic commitment of Article 2 TEU that the Union ‘is founded on’ respect for human rights. Nevertheless, the constitutional duty to respect human rights does not turn the EU into a fully-fledged ‘human rights organisation’, as academic commentators had originally requested.79 Migration law demonstrates the difference. Articles 77–80 TFEU embrace 74 French: ‘doit être une considération primordiale’; see further Geraldine Van Bueren, The International Law on the Rights of the Child (Martinus Nijhoff 1998) 45–51. 75 See UNGA, ‘Declaration on the Rights of the Child’ (Resolution 1386 (XIV), 20 November 1959). 76 See eg Case C-635/17 E EU:C:2019:192, paras 56–67; and Case C-129/18 SM EU:C:2019:248, paras 62, 66–68. 77 See Case C-233/18 Haqbin EU:C:2019:956, paras 53–55; Case C-648/11 MA and others EU:C:2013:367, paras 57–60; Joined Cases C-356/11 and C-357/11 O and S EU:C:2012:776, paras 76–80; and Mark Klaassen and Peter Rodrigues, ‘The Best Interests of the Child in EU Family Reunification Law’ (2017) 19 EJML 191. 78 See ch 13.5.3; and Ciara Smyth, ‘The Human Rights Approach to “Persecution” and its Child Rights Discontents’ (2021) 33 IJRL 238, 265–75. 79 See Philip Alston and Joseph HH Weiler, ‘An “Ever Closer Union” in Need of a Human Rights Policy’ in Philip Alston and Joseph HH Weiler (eds), The EU and Human Rights (OUP 1999) 3.

Added Value of the Charter 137 diverse objectives for law-making and do not orientate supranational policies at the primary—let alone sole—objective of promoting human rights, mirroring the job description of some international organisations.80 That is not to say, crucially, that other policy considerations should take priority. Political processes do not, however, revolve around human rights as the alpha and omega of the EU’s mission statement. At a textual level, Article 78(1) TFEU fits into the overall picture when it calls on the institutions to develop common policies ‘ensuring compliance with the principle of non- refoulement’; secondary legislation ‘must be in accordance’ with the Refugee Convention. This prescription integrates, as we shall see, the substance of the Refugee Convention into the supranational legal order as a judicially enforceable benchmark. It does not, however, oblige EU institutions actively to promote a ‘spirit’ of the Convention beyond respect for legal commitments.81 To do so may be desirable in many circumstances, but Article 78(1) TFEU does not prevent the legislature from taking into account other objectives. EU asylum policy, in particular, is full of examples of how other policy objectives prevail over human rights concerns. Of course, actors within the institutions and other participants in the policy debate can petition for an active human rights policy, even in the absence of a constitutional prescription to go beyond the level of protection enshrined in the Charter. Reference to human rights and generic ‘values’ are standard practice in the public discourse on migration, also as a counterweight to the emphasis on a ‘European that protects’ or European ‘sovereignty’.82 The interdisciplinary Chapter 4.3 demonstrated that the use of language can have framing effects upon political preferences, thus presenting a semantic trajectory for a greater emphasis on human rights.

5.4.4 Limited scope of application of the Charter A specific expression of the EU’s functional outlook on human rights is the limited scope of the Charter, which applies to the Member States ‘only when they are implementing Union law’.83 The European Union is neither a human rights organisation nor a federal state with an overarching bill of rights covering any domestic activity. Settled case law reaffirms that Member States are bound by the Charter ‘within the scope of European Union law’.84 The Charter is, so to say, the ‘shadow’ of Union law; just as an object defines the contours of its shadow, the scope of Union law determines the field of application of the Charter.85 Despite that seemingly clear-cut starting point, the precise definition of the scope is subject to debate.

80 See Mark Dawson, The Governance of EU Fundamental Rights (CUP 2017); and Armin von Bogdandy, ‘The European Union as a Human Rights Organisation?’ (2000) 37 CML Rev 1307. 81 Similarly see Piotr Sadowski, ‘A Safe Harbour or a Sinking Ship?’ (2019) 11 EJLS 29, 56–57; for a different position see Violeta Moreno-Lax, ‘Life after Lisbon’ in Diego Acosta Arcarazo and Cian C Murphy (eds), EU Security and Justice Law (Hart Publishing 2014) 146, 148–49. 82 See ch 1.2.6. 83 TFEU, art 51(1). 84 Case C-617/10 Åkerberg Fransson EU:C:2013:280, para 19. 85 See Koen Lenaerts and José Antonio Gutiérrez-Fons, ‘The Place of the Charter in the EU Constitutional Ediﬁce’ in Peers and others (n 69) 1711, 1713.

138 Human Rights and State Sovereignty Judgments indicate that the Charter applies whenever state action responds to a supranational obligation. Thus, domestic legislation and administrative activities transposing or implementing a directive or regulation can be assessed in light of the Charter.86 By contrast, the Charter does not apply whenever secondary law does not regulate specific subject areas. One may activate national constitutions or the ECHR instead, but the Court of Justice does not hold jurisdiction.87 Any domestic application of the Charter requires, therefore, careful inspection of whether the factual circumstances and the domestic rules in question are covered by one of the EU migration law instruments ratione materiae, personae, temporis, and loci. Doing so is comparatively straightforward for national legislation. By way of example, the Charter applies to the living conditions of those with subsidiary protection as far as Articles 20–35 Qualification Directive 2011/95/EU regulate their status, while the same subject areas cannot be analysed in light of the Charter for those awarded complementary humanitarian protection under domestic laws not subject to EU harmonisation.88 Competences in Articles 77–80 TFEU do not bring about an application of the Charter before the adoption of secondary legislation; the option of future legislation is not enough.89 By contrast, the exercise of national implementing discretion can be judged in the light of the Charter whenever Member States have different options on how to achieve the objective prescribed in a directive or regulation.90 Such implementing discretion must be distinguished from optional clauses not to apply secondary legislation to specific scenarios; corresponding national provisions are not covered by the Charter.91 It may occasionally be difficult to determine whether a situation concerns implementing discretion or optional derogations, as the example of the discretionary clause in Article 17(1) Dublin III Regulation (EU) No 604/ 2013 illustrates.92 Procedural rules must respect the Charter within the scope of national procedural autonomy, including the right to an effective judicial remedy. Difficult to assess are operational activities. By way of example, the Charter applies only to those border control activities for which the Schengen Borders Code or related instruments lay down specific obligations, while other control measures cannot usually be scrutinised in light of the Charter.93 What this entails for border surveillance in-between the official crossing points remains largely uncharted territory at this juncture. By comparison, standards for the supervision of projects co-financed under the EU budget are straightforward: the Charter serves as a yardstick for the selection criteria when determining which projects receive funding, while the actual realisation of the projects is not usually covered.94

86 For typologies see Benedikt Pirker, ‘Mapping the Scope of Application of EU Fundamental Rights’ (2018) 3 European Papers 133; and Filippo Fontanelli, ‘Implementation of EU Law through Domestic Measures after Fransson’ (2014) 39 EL Rev 682, 689–97. 87 See Åkerberg Fransson (n 84) para 44. 88 See Case C-542/13 M’Bodj EU:C:2014:2452, paras 42–46; and Case C-312/12 Ajdini EU:C:2013:103, para 25. 89 See Joined Cases C-609/17 and C-610/17 TSN and AKT EU:C:2019:981, paras 46–52. 90 See Case C-540/03 Parliament v Council EU:C:2006:429, paras 22–23, 104–105. 91 See Case C-94/20 Land Oberösterreich EU:C:2021:477, paras 45–58. 92 See ch 13.3.4. 93 See Case C-23/12 Zakaria EU:C:2013:24, paras 39–42. 94 See Case C-401/11 Soukupová EU:C:2013:223, paras 28–31.

Complementary Role of International Bodies and Political Fora 139

5.5 Complementary Role of International Bodies and Political Fora Courts in Europe have traditionally been perceived as forerunners of human rights dynamism. Their visibility rested on institutional potency, was buttressed by innovative rulings, and reflected the prevalent hegemony of academics working on and in Western countries. Recent years have witnessed a shift of attention to developments elsewhere. For our purposes, treaty bodies at the European or international levels are particularly relevant (5.5.1). Their output has recently become a source of inspiration for critical scholars, although findings will not usually be legally binding (5.5.2). In addition to treaty bodies, international organisations and diplomatic initiatives are important (5.5.3), as exemplified by the Global Compacts on Refugees and for Migration, adopted in 2018 (5.5.4). Furthermore, foreign courts may have to be considered occasionally, when reviewing cooperation of third states with the EU.

5.5.1 Plethora of human rights treaty bodies On the European continent, the Court in Strasbourg is a primus inter pares but not the sole human rights body. The Council of Europe’s European Social Charter (ESC) complements the civil and political rights in the ECHR with ‘second generation’ guarantees about economic and social matters, which are subject to a monitoring mechanism by an international expert committee.95 Guarantees in the ESC must be activated by state parties on a case- by-case basis and are often narrowly construed so as not to bring about far-reaching assurances.96 Nevertheless, they can inform the status of economic migrants and other third country nationals with a residence permit. Guarantees in the ESC inspired the design of equal treatment provisions in secondary legislation. Chapter 16.2.3 will show that proposals have been put forward to overcome explicit limitations to people lawfully resident or working regularly. Asylum applicants and irregular migrants will often rely on human rights-based non- refoulement, as presented in Chapter 13.2.2 on asylum. In comparison to these entrenched guarantees, the added value of the United Nations Convention against Torture (CAT)97 and the European Convention for the Prevention of Torture (CPT)98 lies in the committees of independent experts. Both committees may express views in country reports, on the occasion of general comments, or in response to individual complaints in line with the procedural rules of the instrument in question. Additional procedural avenues exist under other multilateral human rights conventions, including the Human Rights Committee under the International Covenant on Civil and Political Rights (ICCPR). Other relevant treaty bodies 95 Revised European Social Charter (adopted 3 May 1996, entered into force 1 July 1999) ETS No 163, which has been ratified by many Member States, while the others are still bound by the original European Social Charter (adopted 18 October 1961, entered into force 26 February 1965) ETS No 35. 96 ibid art A(1) part III; and generally Karin Lukas, The Revised European Social Charter: An Article by Article Commentary (Edward Elgar Publishing 2021). 97 United Nations Convention against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment (adopted 10 December 1984, entered into force 26 June 1987) 1465 UNTS 85; and additional protocols. 98 Adopted 26 November 1987, entered into force 1 February 1989, ETS No 126, as amended by Protocol No 1 and No 2 (adopted 4 November 1993, entered into force 1 March 2002) ETS Nos 151 and 152.

140 Human Rights and State Sovereignty in the field of migration include the Committee on Economic, Social and Cultural Rights and the Committee on the Rights of the Child.

5.5.2 Potential source of dynamism in the field of migration An overall assessment of the output of treaty bodies is complicated by the predominance of abstract country reports and general comments. Not surprisingly, observers have found ambivalence. While international bodies follow—or stay behind—the ECtHR in many respects, they can serve as a source of innovation in other regards.99 There are indications that the restrictive turn of EU asylum policy may have resulted in a new vigour of international bodies. For instance, the Committee on the Rights of the Child (CRC) sanctioned the Spanish policy of ‘hot returns’, which the ECtHR found to be compatible with the prohibition of collective expulsions.100 Chapter 12.2.2 on border controls will mention that the Human Rights Committee (HRC) held Italy responsible for not having undertaken enough efforts to rescue migrants on a boat in distress, in what might possibly pave the way for a relaxation of the jurisdiction yardstick. It was not clear, at the time of writing, whether these instances remained isolated phenomena or heralded the emergence of an era where international treaty bodies would replace courts in Europe as forerunners of human rights dynamism. It was uncertain, moreover, whether courts in Europe would integrate such momentum into their reasoning, or whether we would observe a trend towards interpretative pluralism. Indeed, the position of international expert bodies is not usually legally binding. States made a conscious choice to speak of ‘committees’ delivering ‘views’ or ‘opinions’, instead of ‘courts’ and ‘judgments’. Their position is not legally binding inter partes, nor does it amount to authentic interpretation of human rights treaties.101 By way of example, the International Court of Justice (ICJ) did not follow the argument, put forward by a committee, that travel restrictions for nationals of specific states amount to direct or indirect discrimination on grounds of ethnic origin or race.102 Positions of international treaty bodies must be considered as one element amongst others when analysing international human rights law. To be effective, treaty bodies depend on domestic or supranational courts to translate their views into legally binding output. An inspection of the practice of the ECtHR and the CJEU indicates mixed results. Judges in Strasbourg regularly employ the consensus argument to integrate the contents of multilateral conventions into the interpretation of human rights. By contrast, views of international bodies feature less prominently in the reasoning, mirroring the practice of most domestic courts, which typically accommodate international findings selectively.103 Therefore, any newfound dynamism of treaty bodies in 99 For an overview see Başak Çalı, Cathryn Costello, and Stewart Cunningham, ‘Hard Protection through Soft Courts?’ (2020) 21 GLJ 355. 100 Contrast Committee on the Rights of the Child, DD v Spain (CRC/C/80/D/4/2016, 1 February 2019) Nos 14.4–14.6; to ND and NT v Spain (n 38). 101 See Case C-249/96 Grant EU:C:1998:63, para 46; Tomuschat (n 37) chs 6–8; and Georg Nolte, Fourth Report on Subsequent Agreements and Subsequent Practice in Relation to Treaty Interpretation (UN doc A/CN.4/694, 7 March 2016) 9–27. 102 See ch 10.4.4. 103 See Machiko Kanetake, ‘UN Human Rights Treaty Bodies Monitoring before Domestic Courts’ (2018) 67 ICLQ 201.

Complementary Role of International Bodies and Political Fora 141 the field of migration need not trickle down to legally binding judicial output at the national or European levels. In contrast to the ECtHR, the CJEU is known for an introverted style of reasoning. Whereas judges in Luxembourg regularly refer to the Court in Strasbourg nowadays, they hardly ever mention international instruments. One of the few examples to the contrary is the action of annulment the European Parliament had brought—unsuccessfully—against certain provisions of the Family Reunification Directive.104 At the time, international instruments compensated for the absence of a legally binding Charter before the Treaty of Lisbon. Today, the Court does not even discuss the international perspective if secondary legislation, such as the Return Directive 2008/115/EC, refers to international human rights law.105 We shall see a similar focus on European affairs in the case law on refugee status with regard to the position of UNHCR.

5.5.3 Special rapporteurs and diplomatic initiatives There is not always a strict separation between executive and judicial functions at the international level when it comes to human rights monitoring. Legal considerations may coalesce with policy advice, for instance in country reports of international treaty bodies. Other examples of practical overlap are special rapporteurs appointed within the UN system, for instance against torture or for the human rights of migrants, as well as the Commissioner for Human Rights within the Council of Europe. Their mission statements cover the drafting of reports, recommendations, or other output through a combination of legal considerations and policy advice. Position papers of UNHCR often follow this model. Such combination can be useful in many respects as long as observers do not confuse these positions with firm legal statements. Meanwhile, international diplomacy has moved away from the earlier practice of negotiating multilateral agreements. Lengthy ratification procedures, lack of consensus about the way forward, and reluctance on the part of states to submit to legally binding rules rendered the choice for ‘hard law’ unattractive.106 A telling example is the UN Migrant Workers Convention, which entered into force almost thirteen years after adoption in 1990. Even then, success was muted: not a single Member State of the EU ratified the Convention; they were unhappy with what they perceived to be one-sided obligations, including for workers who are illegally present.107 The fate of the UN Migrant Workers Convention reiterates our general conclusion, in Chapter 18.1 on the external dimension, about inter-state cooperation being complicated by power asymmetries and conflicts of interest between destination countries and home states.

104 See Parliament v Council (n 90) paras 37, 57. 105 See Tamás Molnár, The Interplay between the EU’s Return Acquis and International Law (Edward Elgar Publishing 2021) ch 3. 106 See Chetail (n 3) 280–99. 107 See ch 16.2.3; and Ryszard Cholewinski, Migrant Workers in International Human Rights Law (Clarendon Press 1997) 186–204.

142 Human Rights and State Sovereignty

5.5.4 Global Compacts for Migration and on Refugees Negotiations on the Global Compact for Safe, Orderly and Regular Migration and on the Global Compact on Refugees were emblematic of intergovernmental struggles. Both compacts were adopted in 2018 and are, as resolutions of the UN General Assembly, not legally binding.108 Whereas negotiations on the Global Compact on Refugees proved less troublesome, the contents evaded controversial topics, such as access to asylum, and failed to establish a robust framework for responsibility sharing.109 One step further, the Global Compact for Migration was caught by the political salience of migration policies in the European public discourse. Several Member States voted against it in the General Assembly, although the EU had been reasonably successful in influencing the basic contours and the small- print of the final outcome during negotiations.110 It deals with various aspects of migration law and touches upon important subject matters which are not subject to extensive international obligations so far. That broad thematic scope of the Global Compact for Migration was a novelty, transcending the traditional focus of the international debate on development issues, which will be traced in Chapter 18.4 on the external dimension. Adoption could possibly mark the beginning of a truly global discourse in the years to come. However, its lasting impact will ultimately depend on whether states can agree to put the abstract declarations of intent into practice by means of domestic laws or inter-state cooperation. Indeed, the Global Compact for Migration does not present itself as a blueprint for a legally binding universal convention. The text states explicitly that the objectives enshrined therein shall be realised at the regional level and in bilateral relations primarily.111 The UN stands ready, together with IOM and the UNHCR, to serve as a facilitator for regional and bilateral cooperation.112 International actors will not, however, develop solutions top-down for how states across the world are to respond to context-specific challenges. Two brief examples illustrate that the Global Compact for Migration is full of diplomatic compromise formulae. First, states ‘commit to ensure that nationals are duly received and readmitted’. This emphasis on the obligation to take back migrants staying irregularly is complemented by a promise of the receiving countries to ‘facilitate’ access to regularisation procedures, subject to ‘an individual assessment . . . on a case-by-case basis’.113 We do not learn when option A (return) should take priority over option B (regularisation), which, moreover, was formulated so as to leave receiving states much leeway. Even if the Compact was legally binding, which it is not, courts would find it difficult to deduce hard legal obligations from it. Secondly, climate change featured prominently in the negotiations, and the outcome may sound impressive. States pledge to ‘[c]‌ooperate to identify, develop, and strengthen solutions for migrants compelled to leave’ their home state owing to natural disasters, climate 108 See Global Compact for Migration (n 10); and UNGA, ‘Global Compact on Refugees’ (Resolution 73/151, 17 December 2018). 109 See James C Hathaway, ‘The Global Cop-Out on Refugees’ (2018) 30 IJRL 591; and Alexander Aleinikoff, ‘The Unfinished Work of the Global Compact on Refugees’ (2018) 30 IJRL 611. 110 See Juan Santos Vara and Laura Pascual Matellán, ‘The Global Compact on Migration’ in Elaine Fahey (ed), Framing Convergence with the Global Legal Order (Hart Publishing 2020) 163, 167–72. 111 See Global Compact for Migration (n 10) Nos 40–47. 112 See also Colleen Thouez, ‘Strengthening Migration Governance’ (2019) 45 JEMS 1242. 113 Global Compact for Migration (n 10) Objective 21 No 37, Objective 7 No 23(i).

Refugee Convention: Distant Lodestar 143 change, or environmental degradation.114 Such abstract commitment to cooperative solutions is a far cry from a clear-cut obligation to admit on an individual basis, let alone recognition of the legal category of ‘climate refugee’, discussed in Chapter 13.5.3.4 on asylum. Note that climate-induced movements featured primarily in the Global Compact for Migration, thus marking a symbolic distance from refugee protection in the parallel document.115

5.6 Refugee Convention: Distant Lodestar The Refugee Convention is commonly described as the ‘magna carta’ for refugee protection.116 That is appropriate insofar as the New York Protocol of 1966 turned the Convention into the founding document of refugee law by withdrawing earlier limitations in time and space. The Convention reminds the international community of the legal, ethical, and political importance of protecting refugees; it also serves as a unifying point of reference for legal rules and policy developments worldwide. At the same time, conceptual significance does not necessarily translate into tangible impact on everyday practices. Notwithstanding the mythical aura of perfection inherent in the designation as ‘magna carta’, the text remains largely silent on core issues of contemporary asylum policy, such as access to territory or status determination procedures. The Convention resembles a lodestar more than a constitutional document; it indicates the direction but does not usually give detailed guidance on how to answer specific questions. Article 78(1) TFEU states unequivocally that EU institutions must respect the Refugee Convention, meaning that non-compliance may be sanctioned by the Court of Justice (5.6.1). In spite of this, the Convention has a limited impact on everyday legal practices; it lacks the institutional support of international treaty bodies, and the position of UNHCR is not legally binding. Judges at the Court of Justice prefer classic arguments of statutory interpretation to active engagement with international practice (5.6.2). Human rights take centre stage in most contemporary debates about asylum law and policy in Europe nowadays (5.6.3).

5.6.1 Interaction with Union law Unlike the Member States, the EU has not formally acceded to the Refugee Convention. Primary law may encompass an exclusive external competence for most matters covered by it nowadays, but the Convention does not allow for the EU’s accession.117 Political initiatives to seek accession have been abandoned quietly.118 The option of a unilateral declaration, on the part of the EU, to commit itself formally on the international plane to adhere to

114 ibid Objective 5 No 21(i). 115 cf Global Compact on Refugees (n 108) Nos 8, 12, 63; and see Jane McAdam, ‘Swimming against the Tide’ (2011) 23 IJRL 2. 116 The formula goes back to the first UN Deputy High Commissioner James M Read, Magna Carta for Refugees (UN Department of Public Information 1951). 117 See Refugee Convention, art 39(2); on the scope of external treaty-making powers see ch 18.2.1. 118 cf The Stockholm Programme [2010] OJ C115/1, 32.

144 Human Rights and State Sovereignty the Refugee Convention has not been realised so far;119 neither has it assumed the responsibilities of the Member States as state parties by way of functional succession, following the earlier example of the GATT.120 As a result, Member States continue to appear separately in the conference of state parties and the related forum of the UNHCR Executive Committee. In doing so, they are legally bound to represent the EU’s viewpoint.121 The Council could even adopt formal decisions under Article 218(9) TFEU determining positions national governments would have to support.122 During negotiations on the Global Compacts, such coordination was mostly successful. There is little doubt that EU institutions are committed to respect the Refugee Convention as a matter of principle. Doctrinally, however, Article 78(1) TFEU sets out a free-standing obligation under Union law. Such internal commitment prevents a mismatch between the obligations of the Member States under public international law and supranational asylum legislation.123 Judges in Luxembourg are compelled, qua Article 78(1) TFEU, to respect the Refugee Convention when interpreting asylum legislation. In doing so, the Court serves as the ultimate arbiter of how to interpret the Convention within the EU legal order, and domestic courts are bound to follow its guidance. By contrast, domestic courts remain free to interpret the Refugee Convention and to determine its effects under national constitutional law whenever a subject matter is not harmonised at the EU level. A rare example of such residual authority of the Member States are sanctions for unlawful entry or presence under Article 31 Refugee Convention, which the Court found, controversially so, not to be covered by supranational legislation.124 Within the EU legal order, Article 78(1) TFEU elevates the Convention to the rank of constitutional law. Non-compliance constitutes an infringement of the Treaties requiring an interpretation of secondary legislation in conformity with the Convention, or possibly even annulment.125 This has been reaffirmed in welcome clarity by the Court repeatedly over the years.126 When interpreting the Refugee Convention judges have recourse to general principles on how to interpret international treaties.127 Incorporation into Union law does not alter its international legal characteristics, notwithstanding the potential of subtle changes if all language versions of the Qualification Directive are equally valid.128 Reference to the ‘Geneva Convention’ in the English wording of Article 78(1) TFEU mirrors the semantic tradition in many continental jurisdictions, while the international debate in the English language usually refers to the ‘Refugee Convention’, as does the author of this volume. It informs the contents and interpretation of the Qualification Directive 2011/95/EU, whose

119 For the parallel debate about international humanitarian law see Nicholas Tsagourias, ‘EU Peacekeeping Operations’ in Martin Trybus and Nigel White (eds), European Security Law (OUP 2007) 102, 115–18; the internal commitment under Article 78(1) TFEU must be distinguished from a commitment at the international level. 120 See Case C-481/13 Qurbani EU:C:2014:2101, para 23; and Hemme Battjes, European Asylum Law and International Law (Martinus Nijhoff 2006) 79–80. 121 See Case C-45/07 Commission v Greece EU:C:2009:81, paras 30–31 for the IMO. 122 See Case C-399/12 Germany v Council EU:C:2014:2258, paras 48–68 in relation to a different organisation. 123 See Battjes (n 120) 59–61, 167–68. 124 See ch 16.3.3. 125 See Joined Cases C-391/16, C-77/17 and C-78/17 M and others EU:C:2019:403, paras 75, 77. 126 It has been settled case law ever since Joined Cases C-175/08, C-176/08, C-178/08 and C-179/08 Abdulla EU:C:2010:105, paras 51–53. 127 See (German) Federal Constitutional Court (Bundesverfassungsgericht), Case 2 BvR 450/11 (decision of 8 December 2014) para 37; and generally Case C-63/09 Walz EU:C:2010:251, paras 22–23. 128 See ch 6.4.1.

Refugee Convention: Distant Lodestar 145 Recital 23 states explicitly that the Directive should ‘guide the competent national bodies of Member States in the application of the Geneva Convention’.

5.6.2 Supporting role in the judicial practice One of the structural weaknesses of the Refugee Convention is the absence of a judicial body, at the international plane, to interpret the Convention authoritatively in light of present- day conditions. Inter-state disputes before the ICJ in accordance with Article 39 have remained a theoretical option. Unlike contemporary human rights instruments, the Refugee Convention did not set up an independent expert body; accordingly, UNHCR has a crucial role to play in terms of ‘supervising the application’.129 It performs this executive function in a dynamic and often innovative manner; evident through interventions in domestic legislative debates, publication of guidelines and the Handbook on status determination, even the coordination of diplomatic initiatives as with Global Compact on Refugees.130 In addition, state parties represented in the UNHCR Executive Committee (ExCom) used to play an active part for many decades, before the body turned more passive as of recently. Valuable contributions on the part of the UNHCR cannot, however, compensate for the absence of a judicial body; its positions are not legally binding, also considering that they often combine questions of legal interpretation with political considerations of what an appropriate solution would be.131 In the absence of international judicial oversight, domestic courts and the Court of Justice interpret the Refugee Convention on the basis of the interpretative standards of the law of treaties. To do so requires tribute to the position of the UNHCR, consideration of state practice, and respect for the Convention’s telos to protect refugees effectively.132 Such decentralised adjudication of open-ended provisions inevitably results in differences of opinion among courts, as well as the academic debate, concerning the ‘correct’ meaning; different views will often co-exist. In light of this, one of the added values of Europeanisation stands out: legislative harmonisation in the Qualification Directive and its authoritative interpretation by the CJEU establishes a common understanding domestic authorities and courts are bound to follow. In doing so, judges in Luxembourg would ideally consider the views of international actors, in particular UNHCR, and other domestic courts across the world in an exercise of transnational judicial dialogue. In this respect, the supranational judicial output presents us with a nuanced picture. While the Court generally recognises, in line with Recital 22 Qualification Directive 2011/ 95/EU, that the opinions of UNHCR ‘are particularly relevant’,133 it occasionally presented a different outcome. Instances of divergence will be discussed in Chapter 13.5.4.6 on asylum, for instance with regard to the meaning of ‘membership of a particular social group’. Such parting is not problematic per se, given that the UNHCR cannot interpret the Convention 129 Refugee Convention, art 35(1); and also UNGA, ‘Statute of the Office of the United Nations High Commissioner for Refugees’ (Resolution 428 (V), 14 December 1950), No 8(a). 130 See generally Corinne Lewis, UNHCR and International Refugee Law (Routledge 2012). 131 See James C Hathaway, The Rights of Refugees under International Law (2nd edn, CUP 2021) 58–67; and Case 2 BvR 450/11 (n 127) para 45. 132 See generally Vienna Convention on the Law of Treaties, art 31. 133 Case C-720/17 Bilali EU:C:2019:448, para 57; see also Case C-369/17 Ahmed EU:C:2018:713, paras 56–57; and Case C-528/11 Halaf EU:C:2013:342, paras 44–47.

146 Human Rights and State Sovereignty authoritatively, but the judicial practice disappointed many observers, nonetheless, since judges did not reflect upon the discrepancies and underlying reasons openly.134 Of course, silence in the official reasoning need not be misunderstood as ignorance, especially if the Advocate General discussed the matter.135 We should also recognise that the Court does not usually cite legal material other than judgments of its own and of the ECtHR.136 Therefore, prevalent silence is not a sign of disrespect specifically for UNHCR, also considering its positions have occasionally been referred to.137 Nevertheless, one is bound to notice an introverted style of reasoning when reading the judgments. Interpretation of the Qualification Directive often concentrates on questions of statutory interpretation in light of the wording, general scheme, objectives, telos, or drafting history.138 A different outlook is presented by the European Court of Human Rights (ECtHR), which is generally receptive to international law. It regularly reproduces the positions of UNHCR and other international actors and takes them up in the reasoning, although it may depart from the views of the UNHCR when assessing the situation in a specific country or when deciding questions of principle. For example, judges disagreed with the UNHCR (without saying so explicitly) when they considered parts of Somalia to be an internal flight alternative or when they laid down comparatively generous standards for the detention of asylum seekers at the point of entry.139 By contrast, the Court relied extensively on the UNHCR when it found the Greek asylum system to suffer from systemic deficiencies in 2011 or when it considered Syria to be generally unsafe a few years later in light of human rights-based non-refoulement obligations.140 Many observers expected the Court of Justice to become an internationally recognised authority for the interpretation of the Refugee Convention.141 That is not to say that it would become a judicial hegemon, but the position of an international court speaking for twenty- six Member States (excluding Denmark, as a result of its opt-out) entails visibility, which is further enhanced by the translation of all judgments into the widely understood languages of English, French, Spanish, and Portuguese.142 Undoubtedly, EU legislation and Court judgments have a global reach anybody studying refugee law is bound to take into account. Europeanisation increased the visibility of views from continental Europe next to the position of courts from Commonwealth countries and the US, which had dominated

134 See Roland Bank, ‘The Potential and Limitations of the Court of Justice of the European Union in Shaping International Refugee Law’ (2015) 27 IJRL 213, 222–36; and Ségolène Barbou des Places, ‘Droit de l’asile et de l’immigration’ [2019] Revue trimestrielle de droit européen 175, 185–86. 135 By way of example see Case C-507/19 Bundesrepublik Deutschland EU:C:2021:3, paras 46–50; and AG Evgeni Tanchev, Opinion in ibid, EU:C:2020:768, points 41–45. 136 This would probably change if international organisations were given the right to intervene in the proceedings; see ch 3.1.4. 137 See Case C-91/20 Bundesrepublik Deutschland EU:C:2021:898, paras 56–57. 138 See ch 3.2.2; and Hemme Battjes, ‘Piecemeal Engineering’ in Vincent Chetail and others (eds), Reforming the Common European Asylum System (Brill 2016) 197, 214–28; and Lars Bay Larsen, ‘Quelques remarques sur les relations entre le droit international et le régime d’asile européen commun’ in Antonio Tizzano and others (eds), La Cour de justice de l’Union européenne sous la présidence de Vassilios Skouris (Bruylant 2015) 73, 78–80. 139 See Salah Sheekh v Netherlands App no 1948/04 (ECtHR, 11 January 2007) §§ 141, 100; and Saadi v United Kingdom App no 13229/03 (ECtHR [GC], 29 January 2008) §§ 65–74, 54–57. 140 ‘UNHCR’ was used no less than 78 times by MSS v Belgium and Greece (n 59); on Syria see LM and others v Russia App nos 40081/14, 40088/14 and 40127/14 (ECtHR, 15 October 2015) §§ 110–11. 141 See Guy S Goodwin-Gill and Hélène Lambert (eds), The Limits of Transnational Law (CUP 2010). 142 See Eleanor Drywood, ‘Who’s In and Who’s Out?’ (2014) 51 CML Rev 1115; and Hélène Lambert, ‘Transnational Law and Refugee Identity’ in Susan Kneebone and others (eds), Refugee Protection and the Role of Law (Routledge 2014) 203.

Refugee Convention: Distant Lodestar 147 the earlier transnational debate. Nevertheless, the ECJ’s introverted style of reasoning constrains the potential of transnational appeal.143 A crucial test will be the position of British courts in years to come. While they are no longer bound by the CJEU’s case law as a result of Brexit, they ‘may have regard’144 to it nevertheless. In addition, domestic courts of the Member States, which are often specialised on asylum matters, will continue to play an important role as points of reference for transnational judicial dialogue.

5.6.3 Practical priority of human rights The design of the Refugee Convention reflects the state of international law after the Second World War. At the time, the international community introduced conventions with broader normative aspirations, which were not yet supported by the institutional infrastructure of later treaty bodies. Human rights instruments transcend this limitation and are, consequently, at the centre of many contemporary debates. Position papers by UNHCR often refer to human rights in parallel to the Refugee Convention, although the latter mentions human rights only once, albeit prominently, in the first recital of the preamble. Recent years have seen a veritable turn of refugee law scholarship towards human rights as the core legal battleground. Their newfound prominence stems from practical significance, conceptual prominence, and professional appeal.145 We may distinguish between two expressions of this turn towards human rights. On the one hand, academics have argued forcefully that the Refugee Convention should be interpreted progressively in light of human rights.146 Article 9(1)(a) Qualification Directive 2011/95/EU supports this approach when defining persecution as a ‘severe violation of basic human rights’ and expressly covers persecution by non-state actors and for gender- specific reasons. Chapter 13.5 will explain that unanimous adoption of this definition was a remarkable exercise of progressive legislation. Court judgments have developed the legislative standard further, for instance with regard to homosexuality and religious freedom. On the other hand, the institutional weight of human rights exerts gravitational force. Judgments of the ECtHR and the opinions of international treaty bodies give answers to present-day problems, compared to which vague formulations in the Refugee Convention resemble an oracle more than precise guidance. Furthermore, human rights law may bridge gaps in the Convention, which remains silent on status determination procedures, legal remedies, or family reunification. Finally, human rights may extend the protective reach of ‘refugee law’ to additional push factors. Indiscriminate violence in situations of armed conflict is an example of further protection, transcending the notion of ‘persecution’ under the Refugee Convention. Chapter 13.6 will explain how subsidiary protection under the Qualification Directive and complementary protection schemes under national laws relate to human rights and the Refugee Convention.

143 See Bank (n 134) 237–41. 144 European Union (Withdrawal) Act 2018, s 6(1), (2), (5) on ‘retained legislation’, which reiterates former Union law in domestic British rules. 145 See Vincent Chetail, ‘Are Refugee Rights Human Rights?’ in Ruth Rubio-Marín (ed), Human Rights and Immigration (OUP 2014) 19. 146 See Hathaway (n 131) 128–74; and also Romit Bhandari, Human Rights and the Revision of Refugee Law (Routledge 2020).

148 Human Rights and State Sovereignty

5.7 Summary European migration law was not written on a clean slate. It developed against the backdrop of international rules which continue to be an essential point of reference. Nevertheless, interaction is not straightforward, especially in the practice of the Court of Justice, and experts with a background in international law will be disappointed by its claim to autonomy. EU migration law can have a meaning of its own, distinct from international law. Inspection of the judicial practice revealed an introverted style of reasoning. Whereas the Court affords constitutional status to the Refugee Convention in the EU’s legal order, the judicial reasoning focuses on the interpretation of secondary legislation, not international practice. Positions of UNHCR are not legally binding, but judges might wish openly to reflect reasons for not following them nevertheless. International treaty bodies are rarely mentioned, in contrast to the case law of the ECtHR. The approach taken by international law towards cross-border movements changed fundamentally after the Second World War. Awareness of this transformation helps to recognise specificities of the international legal context. Traditionally, foreigners—or ‘aliens’, as they used to be called—were protected qua nationality. Minimum standards were construed as inter-state obligations, to be enforced by means of diplomatic protection. Conventions negotiated in the framework of the Council of Europe emulated this model by laying down guarantees for nationals of state parties, as do contemporary association agreements of the EU with third states. Their impact is usually neatly circumscribed by treaty rules, whose contents can differ markedly depending on the negotiations. In an environment that protected foreigners qua nationality, the reduction of statelessness was a vital objective. EU law cannot regulate this question directly, but the international legal framework is comparatively strong and functions reasonably well. Thus, de jure statelessness is not a major area of practical concern in contemporary Europe. Human rights law has turned into an essential normative compass precisely because it rises above the inter-state paradigm. Human rights instruments usually apply to ‘everyone’ within the jurisdiction of a state, not only nationals of state parties; the ‘right to have rights’ was detached from nationality, with individuals being protected qua personhood. The open-ended texture of human rights means that they serve as a conceptual and doctrinal counterweight to state sovereignty, feeding the interests of migrants into decision-making. They will continue to play this role as a result of a built-in potential for dynamic interpretation. International human rights law would not have turned into such a beacon of hope for progressive scholarship without the institutional infrastructure of courts and treaty bodies pronouncing themselves on how to interpret human rights in specific scenarios and in light of present-day conditions. The European Court of Human Rights (ECtHR) is the centre of gravity of human rights law, on the European continent at least. Its judgments were critical in extending the rights of migrants over past decades—notwithstanding a genuine effort, on the part of judges, to balance the rights of migrants with statal control imperatives. Deference was reinforced in a series of more recent rulings, in which judges in Strasbourg refrained from extending further the protective reach of the Convention. Such statism receives much attention, but it does not undo the doctrinal innovations of the earlier case law. Applying these standards, institutional constraints prevent the Court from micro-managing state practice. Effective application of the Convention depends on loyal cooperation with domestic courts, which

Summary 149 the Grand Chamber supports through general principles that may guide decentralised application. Differences of opinion with the Court of Justice over the human rights limitations to the Dublin system have given way to a pragmatic working relationship between the two influential courts in Strasbourg and Luxembourg. Judgments interpreting the Charter of Fundamental Rights show that the CJEU generally replicates the findings of the ECtHR, as required by Article 52(3) CFR. One step further, some provisions in the Charter establish a higher level of protection, for instance on administrative procedures and legal remedies. The rights of the child enshrined in Article 24 exemplify the potential for surprise outcomes. Several judgments relied on the article to correct the outcome of the legislative process to the benefit of children and their parents. The visibility of the legally binding Charter does not, however, turn the EU into a fully- fledged human rights organisation. Primary law obliges the institutions to respect human rights, while simultaneously promoting other Treaty objectives. Member States are bound by the Charter only when implementing Union law, thus raising intricate questions about the demarcation of the Charter’s scope. In contrast to the European Convention, international human rights and the practice of treaty bodies play a marginal role in the practice of the Court of Justice. That is, in part at least, the result of limited legal sway, since the opinions of treaty bodies are not legally binding. Even so, experts of international law will find it frustrating that judges in Luxembourg often ignore international practice. This decoupling may be further reinforced if international treaty bodies continue along the path of dynamic interpretation at a time when judges in Europe seem to tread carefully on migratory matters. Legal considerations and policy advice coalesce in the job description of special rapporteurs, appointed within the UN system, and other international bodies with an executive vocation. A prime example of diplomatic initiatives are the Global Compacts on Refugees and for Safe and Orderly Migration, which are meant to guide, as non-binding soft law instruments, policy developments in years to come. Success is far from guaranteed, partly as a result of compromise formulae reiterating, rather than resolving, conflicts of interests. The Refugee Convention informs asylum practices as the ‘magna carta’ for refugee protection. Article 78(1) TFEU assigns constitutional status to the Refugee Convention within the EU legal order. Nevertheless, direct invocation has limited practical effects in the absence of a (quasi-)judicial body with the authority to interpret the Refugee Convention authoritatively at the international plane. Decentralised adjudication by domestic courts necessarily results in different views, which the Court of Justice canalises into a uniform interpretation on the European continent. Aside from judgments, the actual wording of the Qualification Directive was critical for establishing a uniform understanding of the refugee concept. Human rights take centre stage in most contemporary debates about asylum law and policy, complementing and superseding the Refugee Convention as the central point of reference in many respects. Their profound impact on migration law will be discussed in the other chapters of this volume.

6

Doctrinal Foundations of the Case Law Good lawyers are sometimes said to be those who can make creative use of the law. There is, no doubt, an important element of truth in this statement, but good lawyers also need to know the law. This becomes particularly relevant on the European continent where doctrinal hermeneutics define the contents of judgments that are delivered on behalf of the collective of judges, in line with the civil law tradition. The judicial reasoning on migration validates the prevalence of interpretation in light of the wording, the general scheme, objectives, or the drafting history; comments on these matters often stretch over several dozen paragraphs. That is why this chapter will present the doctrinal foundations in light of a horizontal inspection of the Court’s output on migration law. Mastery of the doctrinal foundations is much more than a boring exercise of ‘black letter’ law. It is a precondition for practical impact of anyone aiming at influencing the outcome of court proceedings at the domestic or supranational levels. Experts in migration law occasionally ignore the broader outlook of the Court of Justice of the European Union (CJEU), which addresses supranational migration law cases against the backdrop of seven decades of judgments on other subject areas. Our horizontal analysis of the judicial output will considers the judgments on the different migration law instruments as an integral part of a body of laws that is more than the sum of its parts. We shall see the potential but also the limits of any attempt at increasing the coherence and predictability of the case law on the various pieces of legislation discussed throughout this volume, which are defined by an increasing complexity. Our analysis begins with the theoretical claim to legal coherence and autonomy which underlies the concept of ‘legal order’ (6.1). Notwithstanding the trend towards fragmentation, the Court generally supports the uniform and parallel interpretation of the sectoral legislation on migration law (6.2). EU migration law brings a distinct flavour to the constitutional essentials, such as the principle of subsidiarity or the distinction between directives and regulations (6.3). Judgments on migration illustrate the pitfalls of the classic methods of interpretation, which are essential for anyone navigating the steady flow of new rulings (6.4). Individual rights enshrined in secondary legislation are an essential doctrinal tool; they perform similar functions as fundamental rights, although their legal foundations and implications differ (6.5). This chapter will conclude with practical tips for dealing with the supranational case law (6.6).

6.1 ‘Legal Order’ as Doctrinal Self-sufficiency Most experts in EU law employ the term ‘legal order’1 as a neutral description, without being unaware of the underlying aspiration of doctrinal self-sufficiency in foundational

1

See Case 26/62 van Gend en Loos EU:C:1963:1; and Case 6/64 Costa v ENEL EU:C:1964:66.

European Migration Law. Daniel Thym, Oxford University Press. © Daniel Thym 2023. DOI: 10.1093/oso/9780192894274.003.0007

‘Legal Order’ as Doctrinal Self-sufficiency 151 cases. Supranational Union law was decoupled from both international law and domestic settings to establish a world of its own that was expected to provide answers to upcoming questions, if necessary by means of dynamic interpretation to fill any gaps that might exist. To put it differently, a ‘legal order’ is more than the sum of its parts; it is an organic mechanism in search of coherence and perfection. Individual rules and different legislative instruments feed into a consistent whole.2 EU migration law does not exist in isolation but is firmly embedded, in the practice of the CJEU at least, in the holistic aspiration of systemic coherence and conceptual autonomy. Experts with a background in common law are rarely aware of the doctrinal ambitions which the concept of ‘legal order’ encapsulates and which continue to define legal education and judicial practice across continental Europe.3 Among the original Member States, the doctrinal tradition was particularly strong in Italy and Germany, from where the notion of ‘legal order’ was borrowed.4 France contributed unwritten general principles to fill gaps in the legal material. Later-on, other influences coalesced with the continental tradition, including Nordic formalism or Anglo-Saxon common law.5 Union law is a peculiar mix that takes up diverse elements. Chapter 4.4 discussed how this conceptual diversity informs the choice of methodology for the design of research projects on migration in full awareness of the different traditions. In migration law, the ideal of systemic coherence can be difficult to put into practice as a result of the complexity of the legislative material. Nevertheless, overarching themes enlighten the analysis of seemingly unrelated pieces of legislation, for instance via the doctrinal features of individual rights or effet utile. Court judgments complement each other, thus ideally establishing a mosaic rather than a patchwork. There is, so to say, an invisible net of doctrinal concepts and general principles that holds the case law together. Judgments on the Return Directive can have repercussions for the analysis of the Asylum Procedures Directive, in the same way as settled case law on Union citizenship has implications for EU migration law. The degree of specificity and generalisation varies greatly and is not usually reflected upon openly. Yet this obscurity does not undo the theoretical and practical significance of the search for replicable patterns. For academic observers and practitioners alike, such general and overarching features have the benefit of rendering the judicial output more stable and predictable. Predictability is particularly relevant for the CJEU, which depends on domestic courts to apply Union law on a daily basis, including on subject matters on which the Court has not spoken. Unfortunately, the judicial practice does not facilitate the distinction between case-specific arguments and abstract statements. In contrast to the ECtHR, the CJEU does not distinguish between ‘general principles’ and their ‘application to the specific case’. To be sure, formulae like ‘in the specific circumstances of the present case’ indicate a fact-specific outcome, whereas replicable patterns can be identified by the use of generic language or the citation of judgments on different subject matters. Nevertheless, the impact of a judgment is

2 See Loïc Azoulai, ‘“Integration Through Law” and Us’ (2016) 14 ICON 449. 3 See Miguel Poiares Maduro, ‘Interpreting European Law’ (2007) 2/2 Eur J L Stud 1, 4–13; and Rob van Gestel and Hans-Wolfgang Micklitz, ‘Why Methods Matter in European Legal Scholarship’ (2014) 20 ELJ 292. 4 See Giulio Itzcovich, ‘Legal Order, Legal Pluralism, Fundamental Principles’ (2012) 18 ELJ 358, 360–69; and Jacco A Bomhoff, ‘Perfectionism in EU Law’ (2014) 12 Camb Yearb Eur Leg Stud 75. 5 See Frederico Mancini and David Keeling, ‘Language, Culture and Politics in the Life of the European Court of Justice’ (1995) 1 Columbia J Eur L 397.

152 Doctrinal Foundations of the Case Law not usually evident at the point of delivery.6 The Court tends to build the legal order ‘stone by stone’, instead of presenting an overarching plan directly. This incremental and experimental approach allows judges to take on board feedback and criticism from domestic courts and political actors,7 yet it also means that the degree of novelty and generalisability will became apparent only gradually.

6.2 The Promise and Limits of Coherence Anyone reading the directives and regulations that make up EU migration law intuitively understands why the search for coherence can be complex and occasionally futile. Secondary legislation often employs similar terminology, although its meaning may differ (6.2.1). Such fragmentation can undermine the uniform application of Union law if domestic authorities and courts find it difficult to determine the correct meaning. The original idea of an ‘immigration code’ seems to have been abandoned, even though the EU institutions strive to the coordinate the contents of related pieces of draft legislation (6.2.2). Judges at the Court have started exploring horizontal cross-fertilisation by recognising that similar terminology in different migration law instruments should usually be given a coherent meaning (6.2.3).

6.2.1 Legislative fragmentation Coherence is more ambitious than consistency. While ‘consistency’ designates the absence of contradictions, ‘coherence’ hints at an underlying rationality if different parts connect to each other in a sensible way. A classic example of consistent rules lacking coherence is the definition of the ‘third country national’: while most migration law instruments define the term as referring to anyone who is not a Union citizen, the Return Directive and the Schengen Borders Code Regulation exclude those enjoying free movement as family members of Union citizens or on the basis of association agreements.8 There are good reasons for this differentiation, but the discrepancy complicates the work of practitioners and academics, nonetheless. On other occasions, incoherence results from the lack of coordination between legislative instruments. A telling example is the first-entry rule in Article 13(1) Dublin III Regulation (EU) No 604/2013, according to which the Member State whose external border an asylum seeker had ‘irregularly crossed’ shall usually examine the application. In a judgment dealing with movements along the ‘Western Balkans route’ during the winter of 2015/16, the Court rejected the proposal of the Advocate General to interpret the provision in isolation; doing so would have justified not qualifying as ‘irregular crossing’ situations where Member States had tolerated irregular entry. Judges disagreed and upheld the idea of overarching 6 See Marie De Somer, Precedents and Judicial Politics in EU Immigration Law (Routledge 2018) ch 8; and generally Marc A Jacob, Precedents and Case-Based Reasoning in the European Court of Justice (CUP 2014) ch 4. 7 See Joseph HH Weiler, ‘Epilogue: Apology and Critique’ in Maurice Adams and others (eds), Judging Europe’s Judges (Hart Publishing 2013) 235, 247–48. 8 Contrast Long-Term Residents Directive 2003/109/EC art 2(a) and Blue Card Directive (EU) 2021/1883 art 2(1) to Return Directive 2008/115/EC art 3(1) and Schengen Borders Code Regulation (EU) 2016/399 art 2(5).

The Promise and Limits of Coherence 153 coherence by emphasising that the Schengen Borders Code Regulation influences the interpretation of the Dublin III Regulation, even though statutory differences may occasionally result in divergent outcomes.9 The quest for coherence extends to the domestic level. Domestic authorities and courts will habitually orientate themselves at national laws. Their structure and wording may differ from supranational legislation, especially in the traditional destination countries whose migration and asylum legislation predates Europeanisation. Member States are bound to avoid inconsistencies when implementing directives, but they are not obliged positively to promote, as we shall see, coherence in terms of adapting the structure and terminology. That can cause additional frictions. The term ‘expulsion’, for instance, may designate either termination of legal stay or return proceedings.10 Coherence of the multi-level system would be supported if domestic legislatures embraced the supranational terminology. Reliable translation in the various official languages is an important starting point for that endeavour. In addition to the terminology of supranational legislation, the EMN Glossary may support linguistic and conceptual coherence.11

6.2.2 Vision of an ‘immigration code’ EU institutions are aware of the pitfalls of legislative incoherence. People involved in the drafting are called upon, in a ‘joint practical guide’, to ensure formal consistency and substantive coherence within each instrument and with acts already in force, especially in the same field.12 The EU institutions have made an effort to live up to this ideal in recent years. Ongoing debates about asylum reform identified cross-cutting definitions which should be used coherently throughout the legislation on procedures, jurisdiction, qualification, and reception conditions.13 Along similar lines, most instruments on legal migration define the notion of ‘family member’ under reference to Article 4(1) Family Reunification Directive 2003/86/EC, although Article 2(g) Dublin III Regulation (EU) No 604/2013 and Article 2(j) Qualification Directive 2011/95/EU employ distinct terminology in response to political sensitivities and contextual differences. At the same time, terminological consistency is no more than the bottom line of systemic coherence. The civil law jurisdictions on the European continent are proud of the tradition of codification, which may include general principles guiding the interpretation of the provisions on specific subject matters. Moreover, constitutional adjudication supports orientation at higher-ranking norms, including human rights. Against this background, observers complained about the uncontrolled growth of EU migration law early on.14 EU 9 See Case C-646/16 Jafari EU:C:2017:586, para 72; contra AG Eleanor Sharpston, Opinion in ibid, EU:C:2017:443, points 115–41, 155–200. 10 See ch 16.5.1. 11 See https://home-affairs.ec.europa.eu/networks/european-migration-network-emn/emn-asylum-and- migration-glossary_en (accessed 1 March 2023). 12 See ‘Joint Practical Guide for Persons Involved in the Drafting of European Union Legislation’ (European Parliament, Council, and Commission, July 2016), No 6, whose usage of ‘substantive consistency’ corresponds to our notion of ‘coherence’. 13 See ‘Cross-cutting definitions’ (Council doc 8044/1/17 REV 1, 11 April 2017). 14 See Philippe De Bruycker, ‘Legislative Harmonization in European Immigration Policy’ in Ryszard Cholewinski and others (eds), International Migration Law (TMC Asser Press 2007) 329, 331–37; and Jan Hecker, ‘Zur Europäisierung des Ausländerrechts’ [2011] Zeitschrift für Ausländerrecht 46, 48–50.

154 Doctrinal Foundations of the Case Law legislation mirrored the Anglo-Saxon tradition of casuistic law-making, listing one article after another, rather than the European tradition of complex edifices with abstract principles. Unease with the status quo was shared by the European Council, which called for the ‘consolidation of all legislation’ on the basis of amendments ‘to simplify and/or, where necessary, extend the existing provisions and improve their implementation and coherence’.15 That plea for consolidation and simplification was the watered-down version of the ‘immigration code’ the Commission had put on the political agenda.16 The idea of an ‘immigration code’ failed to gain political support, since it was perceived by the Member States as a potential power-grab, by the Commission, to overcome the sectoral approach to economic migration.17 More than a decade after the debate about the immigration code, we may conclude that this ship has sailed. The size of legislative instruments has expanded significantly over the years. While the Family Reunification Directive 2003/ 86/EC and the original Long-Term Residents Directive 2003/109/EC comprised 7 and 10 pages, the Students and Researchers Directive (EU) 2016/801 and the Frontex Regulation (EU) 2019/1896 unfold over no fewer than 32 and 100 pages. Ongoing debate on asylum reform indicates that future legislation will buttress the trend towards seemingly incessant increases in the volume of legislation. The battle for simplification of EU migration law appears to have been lost. Having said this, the objective of consolidation remains valid. It is not intrinsically linked to an expansion of the thematic scope of Union law and does not require abandoning the complex structure of contemporary legislation. Consolidation in search of coherence could be achieved on the basis of existing rules, thus effectively projecting the tradition of codification to the supranational level. EU institutions might wish to adopt sectoral ‘codes’ on different aspects of migration law, or consolidate the existing sectoral legislation.18 Indeed, the idea behind the Schengen Borders Code Regulation and the Visa Code Regulation was exactly to promote legal coherence via thematic ‘codes’. That tradition could be revived, ideally not only by name but in the form of substantive improvements in the structure of the legislation.

6.2.3 Interpretative approximation Legislative complexity entails that there is ample room—and need—for interpretative cross-fertilisation, of which the Court appears to have become increasingly aware. Two judgments emphasised the objective of looking beyond the proverbial nose: the meaning of ‘minor’ in the Return Directive and the Asylum Procedures Directive was aligned ‘[w]‌ith a view to a consistent and uniform application of Union law on asylum and immigration’19; 15 The Stockholm Programme [2010] OJ C115/1, 30. 16 See Commission, ‘Communication: An area of freedom, security and justice serving the citizen’ COM(2009) 262 final 25 for a proposal see Steve Peers, ‘An EU Immigration Code’ (2012) 14 EJML 33. 17 See ch 14.1.4; and Sergio Carrera, ‘The Impact of the Treaty of Lisbon over EU Policies on Migration, Asylum and Borders’ in Elspeth Guild and Paul Minderhoud (eds), The First Decade of EU Migration and Asylum Law (Martinus Nijhoff 2012) 229, 239–43; and Dora Kostakopoulou, Diego Acosta Arcarazo, and Tine Munk, ‘EU Migration Law. The Opportunities and Challenges Ahead’ in Diego Acosta Arcarazo and Cian C Murphy (eds), EU Security and Justice Law (Hart Publishing 2014) 129, 132–34. 18 See European Parliament, ‘Resolution on new avenues for legal labour migration’ (P9_TA(2021)0260, 20 May 2021) Nos 32–33. 19 Case C-441/19 Staatssecretaris van Justitie en Veiligheid EU:C:2021:9, para 38.

Constitutional Essentials in Migration Law 155 the Reception Conditions Directive and the Asylum Procedures Directive were found to ‘belong to the same body of law, namely the Common European Asylum System’, and ‘constitute relevant and necessary contextual elements’ when interpreting the other instrument.20 Other examples followed that line without mentioning the objective interpretive coherence in the abstract, for instance with regard to the meaning of ‘detention’, to be analysed in Chapter 16.7. Mutual dependency, not splendid isolation, is the common thread of the case law. To recognise the objective of coherence is not the same as unconditional synchronisation. Some discrepancies cannot be overcome by means of interpretative hermeneutics—as illustrated by the example of the definition of ‘third country national’ mentioned previously. Such divergences are no specificity of EU migration law. Settled case holds that the notions of ‘worker’ and ‘social assistance’ can have distinct meanings in different single market instruments.21 However, such terminological confusion should ideally be limited to exceptional scenarios if the EU institutions succeed in achieving basic coherence across the sectoral legislation on EU migration law. On that basis, judges may support further approximation by means of coordinating the judgments on related themes, such as fee levels or ‘public policy’, which will take centre stage in our analysis of the general features of secondary legislation in Chapter 10.

6.3 Constitutional Essentials in Migration Law As a supranational legal order, Union law embodies a distinct outlook that contrasts with the international realm and domestic legal orders. While the basic constitutional concepts of direct effect and primacy are not specific to EU migration law, they inform its operation as an everyday practice (6.3.1). The abstract distinction between directives and regulations gives way to gradual variation in the legislation discussed in this book (6.3.2). Articles 77– 80 Treaty on the Functioning of the European Union (TFEU) establish competences on a wide array of themes, provided there is political will to activate them (6.3.3).

6.3.1 Direct effect and primacy as an everyday practice Direct effect and primacy are the hallmarks of supranationalism, as taught in university courses on EU law. Primacy famously requires domestic courts and authorities to ‘set aside any provision of national law which may conflict with’22 Union law, irrespective of whether it was adopted prior or subsequent to the supranational legislation. In an extreme scenario, technical provisions of secondary legislation may prevail over national constitutions. To be sure, some constitutional courts have established national caveats on primacy, which naturally receive much attention. They were raised indirectly by Poland in its challenge of the mandatory relocation of asylum applicants, which the Polish government considered 20 See Joined Cases C- 322/ 19 and C- 385/ 19 The International Protection Appeals Tribunal and others EU:C:2021:11, para 58. 21 See Case C-85/96 Martínez Sala EU:C:1998:217, paras 31–36; Case C-140/12 Brey EU:C:2013:565, paras 46– 58; and Emily Hancox, ‘Judicial Approaches to Norm Overlaps in EU Law’ (2021) 58 CML Rev 1057. 22 Case 106/77 Simmenthal EU:C:1978:49, para 21.

156 Doctrinal Foundations of the Case Law to be incompatible with the Identity of a ‘virtually ethnically homogeneous’ nation, only to be rejected by the Court as being ‘clearly contrary to EU law’, including the Charter.23 National identity equally surfaced in a judgment of the Hungarian constitutional court, albeit without gaining practical relevance.24 A tangible example of the everyday significance of primacy are legal remedies in Article 46(3) Asylum Procedures Directive 2013/32/EU. Domestic courts must disapply any legislation, for instance on mandatory time limits for court proceedings, if they hinder the effectiveness of the judicial oversight function.25 Disapplication is automatic and does not require the Court to say so; domestic courts and authorities do not have to wait for the domestic legislature to change national laws. In the same way, the Court forcefully rejects international law-style reciprocity: a violation of Union law by one country does not justify non-compliance by others.26 Thus, Belgium cannot disapply the Dublin III Regulation (EU) No 604/2013 to people coming from Italy with the argument that Italy does not cooperate in take back procedures. Since disputes about migration law routinely involve individuals challenging state action, the conceptual difference between directives (requiring transposition) and regulations (being directly applicable) is often practically irrelevant. Failure to adopt implementing legislation does not absolve domestic authorities from applying those provisions of a directive that are directly effective. For instance, refugees may invoke domestic treatment in the domain of social assistance under Article 29(1) Qualification Directive 2011/95/EU against regional Austrian legislation which established lower standards.27 Settled case law holds that direct effect presupposes a provision to be precise and unconditional. If these criteria are not meant, domestic courts must interpret national law in conformity ‘so far as possible’.28

6.3.2 No categorical distinction between directives and regulations The basic idea behind directives is to provide Member States with flexibility to adapt the prescriptions of Union law to the domestic legal context. To do so is often an end in itself to give national parliaments some leeway, but directives also allow civil law jurisdictions to integrate supranational legislation into pre-existing codifications at the national level, such as the French Code Napoléon or the German Aufenthaltsgesetz (Residence Act). Settled case law reaffirms that parliaments need not ‘copy and paste’ the wording of a directive into domestic law, provided the national rulebook ensures respect for Union law and renders individual rights visible.29 By contrast, the entire contents of regulations is meant to be directly effective. 23 Joined Cases C-643/15 and C-647/15 Slovakia and Hungary v Council EU:C:2017:631, paras 302–05 refused to apply Article 4(2) TEU. 24 See Hungarian Constitutional Court (Alkotmánybíróság), Case 22/16 (decision of 30 November 2016); and Simone Penasa and Graziella Romeo, ‘Sovereignty-based Arguments and the European Asylum System’ (2020) 22 EJML 11, 13–16. 25 See Case C-406/18 Bevándorlási és Menekültügyi Hivatal EU:C:2020:216, paras 32–34. 26 See Joined Cases 90/63 and 91/63 Commission v Luxembourg and Belgium EU:C:1964:80. 27 See Case C-713/17 Ayubi EU:C:2018:929, paras 36–40. 28 Joined Cases C-397/01 to C-403/01 Pfeiffer EU:C:2004:584, para 113. 29 See Case C-194/01 Commission v Austria EU:C:2004:248, paras 37–39.

Constitutional Essentials in Migration Law 157 In practice, however, the theoretical distinction between directives (requiring transposition) and regulations (being directly applicable) has lost relevance—notwithstanding the claim by the Commission that it would bring about ‘comprehensive harmonisation’30 to replace the Asylum Procedures Directive with a Regulation. To say so may be smart from a public relations perspective, to signal that the Commission does something to overcome discrepancies on the ground, but the practical effects of the switch from a directive to a regulation will be limited. The contents of a provision is more important than the choice of instrument. By way of example, the Proposal for an Asylum Procedures Regulation foresees an obligation to set time-limits to lodge appeals ‘between a minimum of two weeks and a maximum of two months’.31 This is stricter than the vague prescription of ‘reasonable time limits’32 but stays far behind the claim about ‘comprehensive harmonisation’, also considering that many aspects of the court procedure would not be harmonised at all under the Asylum Procedures Regulation. Conversely, directives will not give Member States leeway in all cases. Exclusion from refugee status of Palestinians receiving support of the United Nations Relief and Works Agency for Palestine Refugees (UNRWA), for instance, is fully harmonised, despite being laid down in a directive.33 Regulations, on the contrary, may require implementation, for instance regarding the meaning of ‘risk of absconding’ in Article 2(n) Dublin III Regulation (EU) No 604/2013. The abstract distinction between directives and regulations gives way to gradual variations. Chapter 10.6 on more favourable national legislation will explain how to identify mandatory provisions in the migration law instruments.

6.3.3 Subsidiarity as political preference Articles 77–80 TFEU lay down a broad set of competences for law-making, as explained in Chapter 1.2.4. There are few thematic gaps in the supranational powers, for instance with regard to nationality. Whether to make use of these competences is a matter of political choice. EU institutions decide on the suitability of common action in the ordinary legislative procedure; justice and home affairs establish a ‘shared’ competence.34 In practice, the EU institutions have legislated on many aspects, without, however, occupying all issues with which they might possibly deal. For example, the regularisation of illegal stay, discussed in Chapter 16.8 on return, remains mostly untouched. From a constitutional perspective, the choice for inaction is perfectly valid, irrespective of whether one finds it politically convincing or not. Shared competences require compliance with the principles of subsidiarity and proportionality in accordance with Article 5(3) and (4) TEU. Judges allow political actors much discretion when assessing these principles. To counterbalance this leeway, national parliaments may speak out against new proposals in the early-warning mechanism.35 The 30 Commission, ‘Communication: Towards a reform of the Common European Asylum System and enhancing legal avenues to Europe’ COM(2016) 197 final 10. 31 Amended Proposal for an Asylum Procedures Regulation, COM(2020) 611 final, art 53(7)(b). 32 Asylum Procedures Directive 2013/32/EU, art 46(4). 33 See ch 13.5.6.1; and Case C-585/16 Alheto EU:C:2018:584, paras 98–100. 34 See TFEU, arts 2(2), 4(2)(j). 35 See Protocol [No 2] on the application of the principles of subsidiarity and proportionality [2004] OJ C310/ 207.

158 Doctrinal Foundations of the Case Law procedure serves as a forum for debate with domestic actors who often have a different outlook than politicians in Brussels (or academics specialising in EU migration law). Commission proposals for the mandatory relocation of asylum seekers and a Dublin IV Regulation were among the initiatives criticised most.36 In doing so, national parliaments often expressed political opposition rather than legal concerns.37 Indeed, the Commission repealed both proposals not for considerations of subsidiarity but in response to political stalemate. Note that subsidiarity can only be relied upon to criticise supranational initiatives; it does not oblige EU institutions to become active when joint action might add value. Generally speaking, the argument for common action is particularly strong in the fields of visas, border controls, and asylum, since one country alone cannot address these matters effectively any longer. Moreover, the objective, enshrined in the EU Treaties, of a ‘common’ asylum system supports generous application of the principles of subsidiarity and proportionality.38 By contrast, the argument for Europeanisation is weaker with regard to legal migration and integration.

6.4 Pitfalls of the Interpretative Exercise Montesquieu famously said that judges ‘are only the lips pronouncing the contents of the law’.39 However, following the preference of the legislature is easier said than done, and judges in Luxembourg benefit from a particularly wide leeway as a result of the open- endedness of secondary legislation. Discrepancies between language versions complicate the identification of straightforward outcomes (6.4.1), which can occasionally be unearthed by recourse to the drafting history (6.4.2). By contrast, the case law on migration displays an almost bewildering elasticity of teleological interpretation in light of objectives (6.4.3). Useful doctrinal tools are the autonomy of Union law, which helps to overcome entrenched national practices (6.4.4), and the principle of effet utile, which allows judges to address domestic rules not covered by secondary legislation (6.4.5). Our comments demonstrate why it remains a formidable challenge, for Courts and academics alike, to present an argument in the tradition of doctrinal hermeneutics.

6.4.1 Multilingualism of EU migration law All EU migration law instruments are published in the twenty-four working languages, which have the same authenticity. This pronounced multilingualism can result in frictions and misunderstandings; it also explains the use of clumsy language if consistent translation 36 See Tomasz Jaroszyński, ‘National Parliaments’ Scrutiny of the Principle of Subsidiarity. Reasoned Opinions 2014–2019’ (2020) 16 EuConst 91, 101. 37 See Angela Tacea, ‘From Legal to Political Reasoning. National Parliaments’ Use of Reasoned Opinions in the Area of Freedom, Security and Justice’ (2021) 59 JCMS 1573. 38 See ch 7.1.1; Steve Peers, EU Justice and Home Affairs Law, vol 1 (4th edn, OUP 2016) 328; and Henry Labayle, ‘L’espace de liberté, sécurité et justice dans la Constitution pour l’Europe’ (2005) 41 Revue trimestrielle de droit européen 437, 463. 39 Own translation of ‘ne sont que la bouche qui prononce les paroles de la loi’, from the oeuvre De l’esprit des lois (1748).

Pitfalls of the Interpretative Exercise 159 takes priority over linguistic elegance. While the quality of the translation is usually very good, inconsistencies can occur. By way of example, the Spanish version of the former Schengen Borders Code Regulation suggested that authorities ‘shall’ expel third country nationals under certain conditions, while all other language versions indicated that the authorities ‘may’ do so.40 Other differences are less drastic but can be relevant nonetheless, especially when domestic authorities orientate themselves at ‘their’ language without recognising that other language versions say otherwise. EU migration law has to be interpreted uniformly in such scenarios. Multilingualism may result in confusion, but it does not bring about a babel of divergent meanings. Generally speaking, the Court will follow ‘the great majority of the language versions’,41 if the outlier appears to be a mishap, like in the Spanish example. In some cases, however, a comparison will expose two camps supporting different conclusions. For example, several language versions, including English and French, employ the neutral term ‘resources’ when describing the economic self-sufficiency requirement in Article 5(1)(a) Long-Term Residents Directive 2003/109/EC, while the Dutch and German translation, amongst others, use the term ‘income’, thus indicating that the migrant has to earn the money instead of receiving support from others. Judges resolved this conundrum under recourse to other interpretative standards, including objectives and a comparison with other instruments.42 The bearing of sufficient resources will be examined in Chapter 10.2 on general features of secondary legislation. Of course, there are informal linguistic hierarchies. The English and French version will often have a greater weight in practice than the Finnish or the Bulgarian version, also considering that the intra-and interinstitutional debate often unfolds against the background of the English and French text (remember that the Court’s internal working language is French). Such informal hierarchies do not, however, translate into linguistic hegemony.43 Settled case law holds that ‘where there is divergence between the various language versions of an EU legislative text, the provision in question must be interpreted by reference to the general scheme and the purpose’.44 The same applies to secondary legislation that is based on international conventions of which only some language versions are authentic. The Court did not take up the proposal of Advocate General Sharpston formally to give a greater weight to the English and French version of those provisions of the Qualification Directive that replicate the Refugee Convention.45

6.4.2 Revival of the drafting history In the digital age, information on the legislative process is easily accessible through the EUR-Lex portal. References to preparatory documents can be found in the ‘procedure’

40 See Joined Cases C-261/08 and C-348/08 Zurita García and Choque Cabrera EU:C:2009:648, paras 52–57; the discrepancy was ‘corrected’ on the occasion of a recast. 41 See Case C-231/21 Bundesamt für Fremdenwesen und Asyl EU:C:2022:237, para 48. 42 See Case C-302/18 X EU:C:2019:830, paras 27–43. 43 See TFEU, art 342; and Regulation No 1 determining the languages to be used by the European Economic Community [1958] OJ Spec Ed Vol I/59. 44 Case C-18/19 Stadt Frankfurt am Main EU:C:2020:511, para 33. 45 cf AG Eleanor Sharpston, Case C-364/11 El Karem El Kott and others EU:C:2012:569, points 30–32.

160 Doctrinal Foundations of the Case Law section of the entry on the final instrument in the EUR-lex portal,46 or in academic literature dealing with the drafting history of specific provisions.47 While the Court had attributed limited relevance to the drafting history initially, historic interpretation in light of the ‘origin’48 has gained more importance. In contrast to Article 32 Vienna Convention on the Law of Treaties, the worth of the travaux préparatoires is not supplementary in the EU context. Judgments on EU migration law have mentioned the drafting history regularly, especially if the argument was brought up by one of the parties.49 For instance, the Court found the ‘public policy’ clause in Article 6 Family Reunification Directive 2003/86/EC to require a different interpretation than for the free movement of Union citizens. The preparatory work confirmed that the discrepancy in the wording was a deliberate choice, since the Council had rejected a more generous Commission proposal to the contrary.50 Academic observers and legal practitioners may find recourse to the historic method particularly useful when addressing questions of interpretation that cannot be easily resolved on the basis of the wording and general scheme. Unfortunately, that is often the case as a result of the indeterminacy of the legislative process, which tends to follow the tradition of diplomatic negotiations resulting in open compromise formulae.51 Consulting the drafting history can be particularly useful in such scenarios.

6.4.3 Indeterminacy of teleological interpretation In response to ambiguous drafting, the Court often resolves open questions in light of objectives. This technique was the hallmark of the dynamic early rulings.52 Dynamic interpretation in light of objectives was supported by the binary structure of the policy areas involved: EU Treaties established straightforward objectives, such as factor mobility in the single market. By contrast, Chapter 1.3.1 illustrated that the area of freedom, security, and justice establishes diverse and potentially contradictory objectives for law-making. This indeterminacy of the Treaty regime has a knock-on effect on the interpretation of secondary legislation. Some of the early judgments on EU migration law replicated the tradition of dynamic interpretation in light of seemingly straightforward objectives. The Family Reunification Directive 2003/86/EC, for instance, was deemed to promote family reunification,53 even 46 See https://eur-lex.europa.eu/advanced-search-form.html (accessed 1 March 2023); alternatively, consult the subsection on ‘procedure’ of the COM proposal; see https://eur-lex.europa.eu/collection/eu-law/pre-acts.html (accessed 1 March 2023). 47 See the thorough description of the debate on individual articles in Daniel Thym and Kay Hailbronner (eds), EU Immigration and Asylum Law. Article-by-Article Commentary (3rd edn, CH Beck/Hart Publishing/ Nomos 2022). 48 Case C-647/16 Hassan EU:C:2018:368, para 40. 49 See, amongst others, Case C-285/12 Diakité EU:C:2014:39, paras 20–21, 29; Case C-91/20 Bundesrepublik Deutschland EU:C:2021:898, paras 48– 52; Hassan (n 48) paras 44– 46; and Case C- 670/ 16 Mengesteab EU:C:2017:587, para 90. 50 See ch 10.3.1; and Joined Cases C-381/18 and C-382/18 GS and VG EU:C:2019:1072, para 59. 51 See Eleanor Sharpston, ‘Transparency and Clear Legal Language in the European Union’ (2009/10) 12 Camb Yearb Eur Leg Stud 409, 411–12. 52 See generally Gunnar Beck, The Legal Reasoning of the Court of Justice of the EU (Hart Publishing 2012) 312–13. 53 See Case C-540/03 Parliament v Council EU:C:2006:429, para 69; and Case C- 578/ 08 Chakroun EU:C:2010:117, para 43.

Pitfalls of the Interpretative Exercise 161 though Article 1 states in neutral fashion that it ‘determine[s]‌the conditions for the exercise of the right to family reunification’,54 thus highlighting ‘conditions’ as much as ‘family reunification’. Judges recognised this teleological flexibility in later judgments. The optional minimum age of twenty-one years for spouses was found to serve the prevention of forced marriages,55 while income requirements were considered to minimise burdens on social assistance systems.56 These alternatives rationales behind specific provisions openly compete with the general purpose of promoting family unity.57 One judgment recognised, under recourse to Article 1, the absence of a straightforward telos,58 whereas later rulings repeated the generic priority of the aim of family unity.59 The situation is even more perplexing for the Dublin case law, which displays puzzling elasticity in how the Court handles the teleological argument. In judgments emphasising state prerogatives, judges subscribed to the prevention of ‘secondary movements’ and ‘forum shopping’ as the objective;60 by contrast, swift processing and effective access to asylum procedures took centre stage in more lenient rulings.61 Depending on the outcome, judges would accentuate either of the two legislative objectives without reflecting openly as to why they do so. Such lawyeristic approach to interpretation, with judges activating the formula that fits the outcome, contradicts the ambition of rational reasoning and legal certainty. A step in the right direction would be to recognise, as a recent judgment did, that one should strive for a balance ‘between the different objectives . . . and the competing interests involved’.62 Interpretation in light of objectives need not advance the legal status of migrants, as illustrated by the objective of preventing ‘forum shopping’ in some judgments on the Dublin system. Along similar lines, the Return Directive had originally been found to aim at ‘an effective policy of removal and repatriation’,63 although doing so had the ‘migrant-friendly’ side-effect of prohibiting States from imposing prison sentences for illegal stay that might jeopardise the objective of removal.64 Later judgments mentioned ‘full respect’ for human rights as an additional objective.65 The Schengen Borders Code Regulation was considered to pursue the purpose of ‘combat[ting] illegal immigration’ and security threats,66 whereas 54 On the contrast see Philippe De Bruycker, ‘Chronique de jurisprudence consacrée à l’espace de liberté, de sécurité et de justice’ (2010) 46 Cahiers de droit européen 137, 146; and Daniel Thym, ‘A Bird’s Eye View on ECJ Judgments on Immigration, Asylum and Border Control Cases’ (2019) 21 EJML 166, 186–87. 55 See Case C-338/13 Noorzia EU:C:2014:2092, para 16. 56 See Case C-558/14 Khachab EU:C:2016:285, para 39, after having mentioned the ‘general’ objective in para 26. 57 For similar discrepancies in the case law on the Free Movement Directive 2004/38/EC see Daniel Thym, ‘The Elusive Limits of Solidarity’ (2015) 52 CML Rev 17, 25. 58 See Case C-706/18 Belgische Staat EU:C:2019:993, para 27. 59 See Joined Cases C-133/19, C-136/19 and C-137/19 État belge EU:C:2020:577, para 25. 60 See Case C-616/19 Minister for Justice and Equality EU:C:2020:1010, paras 51–52; Joined Cases C-582/17 and 583/17 H and R EU:C:2019:280, paras 77–79; Joined Cases C-411/10 and C-493/10 NS and others EU:C:2011:865, para 79; and Case C-695/15 PPU Mirza EU:C:2016:188, para 52. 61 See Hassan (n 48) para 56; Case C-163/17 Jawo EU:C:2019:218, paras 58–59; Case C-63/15 Ghezelbash EU:C:2016:409, para 52; Mengesteab (n 49) paras 54, 73, 96; Joined Cases C-245/21 and C-248/21 Bundesrepublik Deutschland EU:C:2022:709, para 56; and Joined Cases C-47/17 and C-48/17 X and X EU:C:2018:900, para 69. 62 Joint Cases C-323/21 and others, para 68. 63 Case C-61/11 PPU El Dridi EU:C:2011:268, para 59; and Case C-290/14 Celaj EU:C:2015:640, paras 21–23. 64 See ch 16.3.3. 65 See Case C-383/13 PPU G and R EU:C:2013:533, para 42; the formula has been taken up ever since, starting with Case C-146/14 Mahdi EU:C:2014:1320, para 38. 66 See Case C-341/18 Staatssecretaris van Justitie en Veiligheid EU:C:2020:76, para 63; and Case C-575/12 Air Baltic Corporation EU:C:2014:2155, para 50.

162 Doctrinal Foundations of the Case Law judgments on the Visa Code Regulation and the former Students Directive emphasised different purposes.67 Such neutral formulae are appropriate to start with. Neither the official recitals nor the drafting history usually sustain the identification of one predominant objective. It is in the nature of the interinstitutional negotiations that the final outcome habitually combines varied and occasionally contradictory political preferences and objectives. Primary law confirms this diverse outlook—in contrast to the freedom-enhancing rationale of the single market. In the field of EU migration law, objectives cannot usually serve as magic words overcoming the built-in teleological indeterminacy of most legislation. To recognise the plurality of intentions supports rational decision-making with judges, practitioners, and academics openly pondering the pros and cons of different solutions, instead of hiding behind seemingly clear-cut objectives.

6.4.4 Autonomous concepts and their limits EU law often regulates state behaviour against the backdrop of entrenched national practices. Such traditions can have a gravitational pull that the Court overcomes by means of the well-established assumption that ‘a provision of EU law which makes no express reference to the law of the Member States for the purpose of determining its meaning and scope must normally be given an autonomous and uniform interpretation throughout the European Union’.68 What might appear self-evident is a decisive tool for the effective and uniform application of EU migration law. While autonomy primarily concerns decoupling from domestic law, it has an external dimension as well. Even if EU migration law employs similar terminology as international agreements, the Court may interpret supranational legislation independently. A pertinent example is the notion of ‘serious harm’, traced in Chapter 13.6 on subsidiary protection, which had been developed against the backdrop of ECtHR case law, from which it was distinguished by the CJEU later. Much more common than decoupling from international law is the internal autonomy. References to ‘public policy’ or ‘unaccompanied minors’ are not open-ended formulae allowing Member States to continue domestic rules, even if they employ the same terminology. Autonomy breaks with national traditions and directs the interpretative exercise towards the supranational level. It covers abstract concepts, such as ‘public policy’, in the same vein as legislation-specific provisions. For example, Chapter 14.3.5 on asylum will explain that the uniform reading of the term ‘unaccompanied minor’ prepared the ground for several judgments holding that family reunification depends on the age at the time of the asylum application, not that of the subsequent administrative decision. Once the Court has established a uniform meaning, this finding will ideally resonate in the practice of domestic courts across Europe. In cases of doubt, they send follow-up references to Luxembourg, thereby allowing the Court to develop the case law. Not all provisions lay down autonomous concepts. A reliable indicator signalling the absence of uniform meaning are references to national laws. For example, Member States are 67 See Case C-544/15 Fahimian EU:C:2017:255, para 34; Case C-84/12 Koushkaki EU:C:2013:862, paras 52–54; and Case C-491/13 Ben Alaya EU:C:2014:2187, paras 28–29. 68 État belge (n 59) para 30.

Individual Rights of Migrants 163 required to lay down objective criteria for the ‘risk of absconding’ for return purposes at the national level, as discussed in Chapter 16.7.2 on detention. However, such leeway should not be misunderstood as unfettered freedom; the principle of effet utile and, possibly, fundamental rights can be activated to counter national excesses. Conversely, autonomy does not imply that Union law necessarily provides a full answer. Judges may conclude that that a provision establishes a mandatory framework of uniform meaning, while leaving domestic authorities with flexibility to adjust certain parameters to the domestic context. A telling example is the case law on the refusal of entry visas for students and tourists. The Court concluded that States have ‘wide discretion’ when determining security threats or the likelihood of overstaying, although both concepts must be interpreted uniformly throughout the Union to start with.69

6.4.5 Effet utile as a useful doctrinal tool The principle of effet utile (French for effective application) means that Union law is to be given an interpretation supporting its effective application. The phrasing was borrowed from international law and has defined many dynamic Court judgments. In migration law, effet utile is regularly employed as an argument amongst other, in support of other means of interpretation. This doctrinal tool will work to the benefit of individuals whenever the provision in question coincides with their interests, but the effet utile is not intrinsically linked to the concept of individual rights. For instance, the argument was applied to dynamically find additional exceptions from the rule that returnees shall be detained in specialised detention facilities.70 As an integral part of the interpretative exercise, the effet utile interacts with the other arguments. On step further, the Court employs the principle effet utile as a free-standing obligation, on the part of the Member States, not to undermine the objectives of secondary legislation. Such independent use of the effet utile can be used to scrutinise domestic rules which are not covered by specific provisions of a directive or regulation. Otherwise put, the effet utile allows judges to address questions they would normally have to stay silent on if they limited their analysis to the provisions which are explicitly set out in secondary legislation. Instructive examples are fees for domestic residence permits in the absence of EU harmonisation, mentioned in Chapter 10.1.3, and a series of judgments limiting the criminalisation of illegal stay, discussed in Chapter 16.3.3 on return. These examples illustrate the added value of effet utile in the doctrinal ‘armoury’, allowing the supranational judiciary to extend legal oversight.

6.5 Individual Rights of Migrants Human rights are essential for the legal analysis, although an analytical focus on human rights will often be supplanted by higher standards enshrined in secondary legislation. In striking contrast to the visibility of human rights in the academic and political discourse,

69 70

See ch 10.3.2; Fahimian (n 67) paras 61–63; and Koushkaki (n 67) paras 60–62. See ch 16.8.4; and Stadt Frankfurt am Main (n 44) para 34.

164 Doctrinal Foundations of the Case Law many Court judgments do not mention the Charter, or refer to it as one argument amongst others. Crucially, such silence need not signal preference for state interests. Judges often employ rely on individual rights enshrined in secondary legislation instead (6.5.1). A comparison with human rights illustrates to what extent individual rights may advance the position of migrants (6.5.2). Their added value explains why many rulings remain vague as to the precise role of human rights (6.5.3). This ambiguity must be abandoned, however, when judges determine, in light of Article 47 of the Charter of Fundamental Rights (CFR), whether EU legislation may restrict legal remedies (6.5.4).

6.5.1 Individual rights enshrined in legislation Individual rights are firmly embedded in the Court’s practice and were recognised as a doctrinal tool that may advance the legal status of migrants early on.71 It was originally developed for the Treaty-based fundamental freedoms but applies to secondary legislation similarly. This book uses the description ‘statutory’ (or ‘legislative’) to distinguish rights enshrined in directives or regulations from ‘constitutional’ guarantees at Treaty level, such as human rights.72 Such individual rights play an important part in the case law.73 They support the effective application of the supranational rulebook and may form the basis for dynamic interpretation. By way of example, judges had interpreted most provisions of the former Dublin II Regulation (EC) No 343/2003 not to serve the interests of asylum applicants, thus rejecting a statutory guarantee to contest take back decisions, unless their complaint was based on human rights. Judges emphasised the prerogative of the legislature to decide on the scope of legal remedies, and they interpreted the Dublin II Regulation to have answered that question to the negative.74 However, the Court changed course when concluding, on the basis of the revised rules in the Dublin III Regulation (EU) No 604/2013, that asylum seekers may invoke the time limits to challenge take back decisions, irrespective of whether the transfer would violate human rights.75 This judicial about-turn is a telling example that an ‘administrative mindset’ focusing on statutory interpretation, instead of human rights, need not result in restrictive outcomes.

6.5.2 Added valued in comparison to human rights From the perspective of domestic authorities, individual rights and human rights have comparable practical effects. They can be relied upon to challenge domestic legislation. 71 See Jürgen Bast, Aufenthaltsrecht und Migrationssteuerung (Mohr Siebeck 2011) 101–11; and Pieter Boeles, ‘What Rights Have Migrating Third Country Nationals?’ in Jaap de Zwaan and Flora Goudappel (eds), Freedom, Security and Justice in the European Union (TMC Asser Press 2006) 151, 161–62. 72 This usage of the term ‘statute/statutory’ refers to the EU institutions as the author and is not meant to transport a specific meaning the term might have in some countries, notably the United Kingdom and Ireland. 73 See Saša Beljin, ‘Rights in EU Law’ in Sacha Prechal and Bert van Roermund (eds), The Coherence of EU Law (OUP 2008) 91; and Thomas Eilmansberger, ‘The Relationship between Rights and Remedies in EC Law’ (2004) 41 CML Rev 1199. 74 See Case C-394/12 Abdullahi EU:C:2013:813, para 60; and Case C-4/11 Puid EU:C:2013:740, paras 25–31. 75 See Ghezelbash (n 61) para 35; Case C-490/16 AS EU:C:2017:585, paras 24–25; Mengesteab (n 49) paras 41– 62; and also Maarten den Heijer, ‘Remedies in the Dublin Regulation’ (2017) 54 CML Rev 859, 862–69.

Individual Rights of Migrants 165 This transformative potential is a legitimate reason why they are occasionally discussed in parallel.76 When we focus on the supranational level, however, conceptual differences between human rights and statutory guarantees stand out. Individual rights reach further than human rights, but their contents remains intimately linked to political choices. Individual rights cannot be used, unlike human rights, to reverse the outcome of the legislative process. The added value of individual rights is explained best by one of the first judgments on the area of freedom, security, and justice. In response to an action for annulment against the Family Reunification Directive, judges held, in line with human rights law, that respect for family life ‘cannot be interpreted as denying States a certain margin of appreciation when they examine applications for family reunification’.77 Nevertheless, Directive 2003/86/EC was interpreted to ‘impose precise positive obligations with corresponding clearly defined individual rights’, which are ‘[g]‌oing beyond’ human rights.78 Otherwise put, the Family Reunification Directive lays down guarantees to receive a residence permit where human rights do not require such an outcome. Chapter 10.1.2 will mention various judgments which found diverse EU migration law instruments to have established individual rights to be authorised entry or residence. Most of these legislative guarantees reach substantially further than human rights, subject to one important caveat: individual rights are conditional upon compliance with the criteria laid down in secondary legislation. This caveat played a minor role initially; early judgments on migration law in the early 2010s emphasised classic doctrinal features to advance the legal status of third country nationals.79 Autonomous and uniform interpretation, predominance of objectives, proportionality, and the ‘strict’ (or ‘narrow’) interpretation of derogations featured prominently and left academic observers with the impression that judges would systematically advance the legal position of migrants over state interests.80 All these doctrinal features remain intact. Nevertheless, Chapter 3.2.3 on the judicial output illustrated that a holistic inspection of the case law mitigates the seemingly clear- cut trajectory in favour of the individual. In the same way as the rulings on the Dublin system enhanced the legal status of asylum seekers by means of statutory interpretation, other judgments accentuated state interests. Chapter 15.4 on integration will explain that the Court found pre-departure language requirements to be compatible with the Family Reunification Directive. Judges insisted on a hardship clause, but they did not reverse a policy choice in favour of integration measures which the EU institutions had endorsed more or less explicitly.

76 See Cathryn Costello, The Human Rights of Migrants and Refugees in European Law (OUP 2016) who, despite the reference to ‘human rights’ in the title, discusses both constitutional guarantees and secondary legislation. 77 Parliament v Council (n 53) para 59. 78 ibid para 60. 79 See Chakroun (n 53) paras 41, 43–45; and Case C-571/10 Kamberaj EU:C:2012:233, paras 77–80, 86, 90. 80 See Diego Acosta Arcarazo and Andrew Geddes, ‘The Development, Application and Implications of an EU Rule of Law in the Area of Migration Policy’ (2013) 51 JCMS 179; Kees Groenendijk, ‘Recent Developments in EU Law on Migration’ (2014) 16 EJML 313, 332–34; and Reinhard Marx, ‘Sprachnachweis und Ehegattennachzug’ [2011] Zeitschrift für Ausländerrecht 15.

166 Doctrinal Foundations of the Case Law

6.5.3 Ambiguous Court practice on the Charter Human rights in the Charter have a higher rank than secondary legislation. They can be relied upon to challenge the validity of legislation, although interpretation in conformity with human rights will usually take precedence.81 Primacy of human rights is independent of whether the recitals say explicitly that an instrument respects the Charter, as contemporary legislation regularly does. Notwithstanding the theoretical potency of fundamental right, the judicial reasoning will not always distinguish neatly whether human rights command a specific outcome or whether they were mentioned as one argument amongst others.

6.5.3.1 Interpretation in conformity with human rights Few judgments state explicitly that the Charter necessitates an outcome which would have been difficult—or impossible—to deduct from secondary legislation without recourse to human rights. A famous example where this was done proves the NS judgment, mentioned in Chapter 13.3.6 on asylum, which required Member States to refrain from the transfer of asylum applicants under the Dublin system whenever there was a real risk of inhumane or degrading treatment contrary to Article 4 CFR. Other examples include the inadequacy of internal administrative appeals in Article 13(1) Return Directive and the obligation not to issue a return decision to unaccompanied minors without a reasonable prospect of removal.82 On other occasions, EU migration law was open enough to accommodate human rights concerns. Nevertheless, judges in Luxembourg instructed domestic courts to respect fundamental rights when interpreting national laws. Discretion, on the part of the States, must not be used to produce outcomes that are incompatible with the Charter. This scenario arose with regard to optional clauses in the Family Reunification Directive and for restrictive court procedures that might possibly contravene the fundamental right to an effective remedy in Article 47 CFR.83 Whenever the Court concludes that the Charter requires a specific outcome, the legislature must respect that position when drafting successor legislation. Amendments violating human rights will be annulled. By contrast, the EU institutions are free to adopt rules overturning earlier judgments whose outcome was based on statutory interpretation. If the legislature supports a different outcome, the Court can be expected to respect that choice, unless human rights require otherwise. The different outlook of the common law, where the legislature operates under the assumption that statutes will be interpreted against the backdrop of the pre-existing judicial precedent,84 cannot be applied in the EU context. 6.5.3.2 Silence or vagueness on the Charter Anyone reading Court rulings on migration will notice that judges often do not discuss human rights at length. While some judgments remain completely silent, others mention the Charter en passant as one argument amongst many, without indicating the extent to which human rights informed the outcome. Human rights seem to serve symbolic or 81 See Joined Cases C-391/16, C-77/17 and C-78/17 M and others EU:C:2019:403, para 77. 82 See chs 3.1.3 and 16.5.5. 83 See ch 7.2.3; and, for Article 47 CFR, Case C-233/19 CPAS de Liège EU:C:2020:757, paras 45, 54–57; Case C- 556/17 Torubarov EU:C:2019:626, paras 54–56; and Case C-662/17 E G EU:C:2018:847, paras 46–48, 69. 84 See https://en.wikipedia.org/wiki/Statute (accessed 1 March 2023).

Individual Rights of Migrants 167 decorative functions.85 However, such instances of judicial silence or ambiguity need not signal disrespect for human rights. Chapter 3.2 explained that the Court serves as a constitutional court and as an administrative tribunal in parallel; jurisdiction covers human rights and legislative interpretation. As a result, the Court need not analyse each case from the human rights viewpoint, as the ECtHR and domestic constitutional courts are bound to do for the simple reason that their jurisdiction hinges on human rights. The CJEU concentrates on human rights when correcting the outcome of the legislative process. By contrast, statutory interpretation will prevail when the minimum level of human rights protection is not at stake. An instructive example is family reunification, in relation to which judges famously found that Directive 2003/86/EC imposes ‘clearly defined individual rights . . . [g]‌oing beyond’ human rights.86 If that is the case, it does not come as a surprise that follow-up rulings focus on the effet utile or the strict interpretation of derogations, which are closely associated with the concept of individual rights. Such interpretation assumes, implicitly at least, that respect for family life under Article 7 CFR ‘cannot be interpreted as depriving the Member States of the margin of appreciation available to them’.87 Judges remain silent on the Charter, or mention it in passing, for the simple reason that legislative guarantees offer a higher level of protection. One factor explaining recourse to fundamental rights is the order of reference of domestic courts. If a preliminary reference invokes the Charter, judges are more likely to discuss human rights, although nothing prevents the Court from raising human rights ex officio.88 An extreme case was the Mahdi judgment, which mentioned the Charter when reproducing the order for reference and when formulating the answer, after having remained silent on the Charter when developing that answer.89 More common are abstract references to human rights as one argument amongst others, without indicating their relative weight. While some judgments should probably be read to indicate a mandatory outcome,90 other rulings remained vague about the degree of constitutional prescription.91 Instructive are judgments on reception conditions for asylum seekers, discussed in Chapter 13.8.3, which fluctuate between abstract invocations of the Charter as one argument amongst others and the potential of constitutional duty. Such vagueness becomes relevant when the EU institutions consider a legislative amendment or whenever practitioners or academics wonder whether the judicial argument can be projected upon other instruments. An instructive example is a judgment on access to the labour market no later than nine months after an asylum application was lodged. Ireland had exempted Dublin cases from that guarantee, and the ECJ found that solution to be incompatible with the Reception Conditions Directive 2013/33/EU. While the reasoning concentrated on profane matters of statutory interpretation, it also mentioned that Recital 11 of the Directive and earlier judgments had conceived of the labour market as contributing to human dignity.92 Does this imply that the EU institutions are prevented from setting a 85 See Diego Acosta Arcarazo, ‘Civic Citizenship Reintroduced? The Long-Term Residence Directive as a Post- National Form of Membership’ (2015) 21 ELJ 200, 210–12; and Jean-Yves Carlier and Luc Lebœuf, ‘Droit européen des migrations’ [2020] Journal de droit européen 132, 146. 86 Parliament v Council (n 77) para 60. 87 Case C-558/14 Khachab EU:C:2016:285, para 28; and Bevándorlási és Menekültügyi Hivatal (n 25) para 65. 88 See Case C-365/02 Lindfors EU:C:2004:449, para 32. 89 See Mahdi (n 65) paras 37, 52, 53. 90 See Case C-473/16 F EU:C:2018:36, para 53. 91 See État belge (n 59) paras 33–44. 92 See The International Protection Appeals Tribunal and others (n 20) paras 69–70.

168 Doctrinal Foundations of the Case Law higher threshold of twelve or fifteen months, since doing so would be incompatible with Article 1 CFR? Probably not, but the example illustrates that ambiguous statements on human rights can contribute to legal uncertainty.

6.5.4 Scope of the right to an effective remedy Article 47 CFR was mentioned in many rulings on the Asylum Procedures Directive, including in scenarios where the Court requested an interpretation of domestic laws in conformity with human rights. Procedural requirements flowing from that provision will be elaborated upon at length in Chapter 7.2.3 on the administrative dimension. A different question concerns the scope of Article 47 CFR ratione materiae. Closer inspection of the reasoning of the judgments on legal remedies under the Dublin system illustrates an important lacuna, which becomes practically relevant in the ongoing debate about asylum policy reform. According to Article 47(1) CFR, the right to an effective remedy applies to ‘[e]‌veryone whose rights and freedoms guaranteed by the law of the Union are violated’. We know for certain that these ‘rights’ need not be fundamental rights; statutory guarantees in secondary legislation are similarly covered. The Charter guarantees, in other words, an effective remedy whenever secondary legislation establishes an individual right. By way of example, legal remedies against the rejection of a Schengen visa are mandatory in light of Article 47(1) CFR if we conclude, in line with case law, that the Visa Code Regulation embraces a guarantee to be issued an entry visa under the conditions put forward by the legislature. But what about a scenario in which the legislature limits the right of appeal on the occasion of a legislative amendment? Is it prevented from doing so even if the right in question transcends human rights? What may sound like a theoretical inquiry is a matter of great practical concern. Remember that the Court interpreted the Dublin III Regulation (EU) No 604/2013 to have established—unlike the predecessor instrument—individual rights to challenge take back requests on procedural and other grounds, even in the absence of a human rights violation. Does that prevent EU institutions from returning to the status quo ante, as the Commission proposes to do? Chapter 13.3.4 will illustrate that the Commission wants to limit appeals to human rights considerations. Any analysis of whether this would be legal or not is complicated by judgments mentioning Article 47 CFR as one argument amongst others as to why legal remedies under the Dublin III Regulation should be interpreted broadly.93 That begs the question whether the legislature may limit judicial oversight, especially when no other human rights are at stake? A similar question comes up regarding resettlement. Chapter 13.10.2 will mention that the Commission proposes exempting the selection process from judicial oversight. Resolution of this inquiry ultimately depends on how we conceive the reference to ‘rights . . . guaranteed by [Union] law’. If Article 47(1) CFR presupposes the existence of an individual right, enshrined in legislation, the abrogation of statutory guarantees can result 93 See eg Joint Cases C-323/21, C-324/21, and C-325/21 Staatssecretaris van Justitie en Veiligheid EU:C:2023:4, para 92; and Case C-194/19 État belge EU:C:2021:270, paras 33–36, 42–43; the argument came up first in Case C- 201/16 Shiri EU:C:2017:805, para 44.

Practical Tips for Dealing with the Case Law 169 in the parallel termination or limitation of legal remedies. The wording supports this interpretation,94 since it does not—in contrast to Article 42(2)(a) CFR on the administrative right to be heard95—apply to any measure negatively affecting an individual.96 The President of the Court has argued in an extra-judicial capacity that the legislation can be changed.97 Such an outcome would mirror case law on Article 6 ECHR.98 If, conversely, access to court under Article 47 CFR transcended than the scope of individual rights, the legislature would have to justify any limitation in light of Article 52(1) CFR.99 Such justification is not warranted, however, if we conclude that an activation of Article 47(1) CFR presupposes negative repercussions for individual rights. A different solution would hold when individuals can rely on human rights. A fine example is take charge requests for unaccompanied minors, for instance an Egyptian minor residing in Greece who requests to be reunited with his uncle living in the Netherlands. Article 27(1) Dublin III Regulation (EU) No 604/2013 does not foresee legal remedies against the refusal of the take charge request, as opposed to the take back request. Nevertheless, the Court concluded that the Dutch authorities must provide for legal redress. To justify this outcome, the Grand Chamber relied on both individual rights enshrined in Article 8(2) Dublin III Regulation in line with previous comments and on Articles 7, 24, and 47 CFR.100 Doing so limited the conclusion to scenarios where individual statutory rights or human rights are at stake; relatives cannot insist on legal remedies, since neither the Dublin III Regulation nor human rights provides them with individual guarantees.101 Along these lines, the Commission’s proposal for the future Dublin system exempts human rights issues from the ban on legal remedies, for take back requests at least.

6.6 Practical Tips for Dealing with the Case Law Court judgments can be found online via the EUR-Lex portal or the Court’s search form.102 The Court’s website is particularly useful whenever one does not know the case number, which is the most reliable marker for identification. Be careful not to misapprehend the meaning of ‘C-187/20’. It does not signify ‘Case 187/20’, as the abbreviation ‘C’ stands for the French Cour de justice (Court of Justice), while ‘T’ denotes the Tribunal, the French name of the General Court. The prefix was introduced when the General Court was established in the late 1980s, and older judgments do not use it as a result. Capital letters behind the case number indicate special procedures, such as ‘PPU’ for the ‘urgent preliminary ruling

94 See AG Michal Bobek, in Case C-403/16 El Hassani EU:C:2017:659, points 74–84. 95 See Case T-429/18 BRF and SHB Comercio e Industria de Alimentos v Commission EU:T:2020:322, para 83; and Case C-141/08 P Foshan Shunde Yongjian Housewares and Hardware v Council EU:C:2009:598, para 74. 96 Contra den Heijer (n 75) 869–70, who proposes interpreting Article 47 CFR in line with Case C-277/11 MM EU:C:2012:2479, paras 83–87, even though that judgment had concerned Article 42 CFR. 97 See Koen Lenaerts, ‘Europarecht und Zuwanderung’ in Julia Iliopoulos- Strangas and others (eds), Migration –Migration –Migrations (Nomos/Stämpfli 2017) 233, 236. 98 See Károly Nagy v Hungary App no 56665/09 (ECtHR [GC], 14 September 2016) §§ 61–62. 99 See Siliva Morgades-Gil, ‘The Right to Benefit from an Effective Remedy against Decisions Implying the Return of Asylum Seekers to European Safe Countries’ (2017) 19 EJML 255, 274–77. 100 See Case C-19/21 Staatssecretaris van Justitie en Veiligheid EU:C:2022:605, paras 33–50. 101 ibid para 50. 102 See https://eur-lex.europa.eu/advanced-search-form.html; and http://curia.europa.eu/juris/recherche. jsf?language=en (both accessed 1 March 2023).

170 Doctrinal Foundations of the Case Law procedure’ (procédure préjudicielle d’urgence) or ‘P’ for ‘appeals’, which are called pourvoi in the working language of the Court. The former Court Reports were discontinued during the 2010s, and the official version of the judgments has been published digitally ever since. ‘ECLI’, the European Case Law Identifier, was introduced around the same time, also retrospectively for older rulings. ECLI introduces a unique number and is increasingly used by domestic courts as well. Note that the Court of Justice does not follow the Anglo-Saxon tradition of citing cases on the basis of the full names of the parties when citing previous judgments that responded to a preliminary reference.103 Instead, the Court refers to one party, which can be identified via the Court’s search form. In order to continue the tradition of case-specific names without abbreviations, which is deeply entrenched in EU legal culture, the Court increasingly refers to the defendant party in the domestic proceedings of a preliminary reference whenever the names of private parties have been anonymised on grounds of data protection concerns.104 If this results in indistinctive case names, such as ‘Staatssecretaris van Justitie en Veiligheid’ or ‘Bundesrepublik Deutschland’ for Dutch or German asylum cases, a fictional name suggested by a computerised automatic name generator will be chosen from 2023 onwards. In order to facilitate distinction, a short thematic description is added in brackets for more recent judgments to respond to the anonymisation of the names of most private parties in response to data protection concerns.

6.7 Summary One may easily feel lost when dealing with European migration law. Secondary legislation is highly complex, and the body of case law is constantly expanding. Idiosyncratic features of secondary legislation amplify the challenge to keep on oversight. EU institutions have occasionally put forward divergent definitions of identical terminology in different legislative acts, such as ‘third country national’, and continue establishing instrument-specific procedures and admission criteria. The idea of an overarching ‘immigration code’ (or several thematic codes) has been abandoned. The vision of simplification and consolidation has given way to sectoral legislation defined by a seemingly unstoppable trend towards idiosyncratic and convolution. In such a situation, it is useful to have a mental map that guides practitioners and academics through the jungle-like landscape. Such guidance allows us to see the proverbial forest for the trees and to develop solutions to questions which have not been decided by courts or examined widely otherwise. Against this background, this chapter set out to explain the doctrinal foundations of the case law of the Court of Justice, which follows the civil law tradition in so far as the concept of ‘legal order’ embodies a claim to doctrinal self-sufficiency. Individual rules are meant to feed into a consistent whole; the legal order is expected to provide answers to new questions, if necessary by means of dynamic interpretation to fill any gaps that might exist. Doctrinal features and unwritten general principles help judges to contribute to the vision of overall coherence. The Court recognises that the different migration law instruments should be interpreted consistently and uniformly, whenever possible. EU migration law is more than

103 104

Both parties are cited for infringement proceedings and actions for annulment. Case C-36/20 PPU was called ‘Ministerio Fiscal’ instead of ‘VL’.

Summary 171 the sum of its parts and establishes a body of laws that supports common answers to similar problems by means of horizontal cross-fertilisation between the multiple instruments discussed in this volume. Constitutional essentials, such as direct effect or the difference between directives and regulations, are known to anyone who has attended a university course on EU law. The perspective of migration law demonstrates that these abstract findings matter in practice. Several judgments empowered domestic courts to set aside domestic legislation. In the institutional practice, the abstract distinction between directives and regulations gives way to gradual variations. Regulations often require implementing legislation, whereas directives embrace provisions which fully harmonise specific subject matters. The contents of a provision is more important than the official title. To replace the Asylum Procedures Directive by a Regulation would not change much in itself. Many judgments on migration law develop classic arguments of legal hermeneutics in the continental tradition: they discuss the wording, the general scheme, objectives, or the drafting history. What may appear as a boring exercise of ‘black letter’ law to some, in fact harbours the potential of practical impact and unexpected outcomes. Anyone trying to understand—or influence—court proceedings at the national or supranational levels will appreciate the potential of doctrinal arguments. Insistence on the autonomous interpretation of Union law challenges entrenched domestic practices. Recourse to effet utile allows judges to scrutinise domestic rules that are not covered by secondary legislation explicitly. At the same time, our inspection of the case law illustrated how fragile and ambiguous the interpretative exercise can be. The multilingual character can complicate the search for straightforward answers, and the judicial practice of teleological interpretation in light of objectives reveals an almost bewildering lack of coherence. EU migration law is cherished by many for having helped to overcome state discretion. EU legislation goes much further than human rights law by laying down clearly defined statutory guarantees on entry and stay. Such individual rights in secondary legislation are an essential doctrinal ‘weapon’ in the armoury of anyone aiming at enhancing the legal position of third country nationals. While the notion of individual rights is firmly established in the case law, judgments on migration law demonstrate built-in ambiguities. On the one hand, they serve as an argument for the generous interpretation of secondary legislation, as the case law on the Dublin III Regulation (EU) No 604/2013 exemplifies. This added value explains why many judgments are silent or vague regarding human rights. On the other hand, the contents and reach of statutory rights is intimately linked to the outcome of the legislative procedure; judges will not usually reverse political choices by means of interpretative acrobatics. Such an outcome requires an interpretation of secondary legislation in conformity with the fundamental rights of the Charter.

7

Administrative Dimension Most experts in EU migration law focus their minds on the output of the EU institutions at the supranational level. To do so is perfectly legitimate, even necessary. At the same time, anyone dealing with everyday practices of administrative authorities across Europe will realise that an exclusive emphasis on the supranational level runs the risk of presenting an incomplete picture. EU rules are being applied by domestic officials and judges in literally thousands of offices and courtrooms in the countryside and capital cities. Officials and judges deciding individual cases are firmly embedded in national administrative traditions; they employ domestic implementing legislation, read, and write in the national language, and refer to rulings of national courts rather than ‘abstract’ supranational legislation or judgments, with whose idiosyncrasies they are not familiar. By focusing on the supranational level, legal experts in EU migration law often take the normative self-sufficiency of legislative prescriptions for granted, as if the publication of a regulation in the Official Journal meant that national authorities respected it quasi- automatically. Empirical studies demonstrate that this is not necessarily the case: the law in books and law in action can be different matters. Anyone trying to understand how Union law ‘reaches’ the individual should pay attention to the rules governing the application of supranational legislation by domestic authorities and courts. Officials and judges at the national level are, so to speak, the arms and legs applying supranational legislation, while domestic rules on administrative procedure and court proceedings are the muscle and bone sustaining the normative organism. Studying the administrative dimension of European migration law is notoriously difficult precisely because it depends on the domestic legal context of the twenty-seven Member States (and the countries associated with the Schengen area). If we conceive of ‘European’ migration law as a composite structure comprising both supranational legislation and domestic rules and practices, we are confronted with more than two dozen different sub-systems. The constitutional framework and interdisciplinary perspectives accentuate these composite and multi-level characteristics (7.1). Union law supports effective implementation by limiting the room for manoeuvre of the Member States in the form of procedural harmonisation, fundamental rights, and unwritten general principles (7.2). A specificity of migration law is transnational legal effects of administrative decisions. Legislative provisions and Court rulings illustrate different options of how to delineate the horizontal authority of the Member States to question the legality of administrative decisions of another country (7.3). Budgetary rules and financial instruments are often ignored by legal analyses, although they can prove to be important tools of governance through incentives (7.4).

European Migration Law. Daniel Thym, Oxford University Press. © Daniel Thym 2023. DOI: 10.1093/oso/9780192894274.003.0008

European Migration Law as a Composite System 173

7.1 European Migration Law as a Composite System Article 291 Treaty on the Functioning of the European Union (TFEU) obliges Member States to ‘adopt all measures of national law necessary to implement legally binding Union acts’. The EU level concentrates on legislative output, whose practical effectiveness depends on domestic authorities. Indirect enforcement is firmly rooted in EU constitutional law and practice, which nevertheless allows for various forms of horizontal and vertical interaction. While agencies will be discussed in Chapter 8, this section demonstrates the conceptual vagueness of the notion of an ‘area’ of freedom, security, and justice and the idea of a common asylum ‘system’ (7.1.1). EU migration law is bound to embrace multiple instruments and actors cooperating in the European administrative space (7.1.2). When analysing justice and home affairs, we should acknowledge their operational character distinguishing migration law from Union activity in other areas (7.1.3). Administrative decision-making requires distinct solutions to establish effective and accountable governance structures, with agencies as one example, amongst others (7.1.4).

7.1.1 Conceptual vagueness of ‘areas’ and ‘systems’ It was not lost upon the founders of today’s European Union that supranational rules would have to be implemented. Closer inspection of the institutional practice reveals greater complexity than the standard rule of indirect implementation suggests. Core supranational policies like competition, state aid, or the former European Coal and Steel Community were based on a high degree of Commission involvement (direct implementation), while the harmonisation of policy fields such as tax law, consumer protection, environmental policy, and social policy relied primarily on domestic authorities (indirect implementation). Other areas developed diverse forms of networks, comitology, and agencies, notably in relation to agriculture, structural funds, and, more recently, banking and telecommunication.1 Such ‘integrated’ or ‘composite’ administration is increasingly common in the field of migration, which adds important innovations. The binding character of supranational legislation should not distract us from the internal plurality of the European project. The structure of EU migration law repudiates simple comparisons with the nation state. The notion of an ‘area’ of freedom, security, and justice does not support a different conclusion, although it may be read, at first reading, to imply territorialisation, given that the term ‘area’ has territorial connotations in the English language. By contrast, however, the French word espace and the German Raum prove linguistically more flexible: they cover both the territorial ‘area’ and the boundless ‘space’.2 The historic context reiterates this point. The designation ‘area of freedom, security and justice’ developed in the late 1990s, mirroring the deliberately indistinct

1 See Paul Craig, EU Administrative Law (3rd edn, OUP 2018) ch 1. 2 Other Roman languages mirror the French, while Germanic languages fluctuate between the German open- endedness and the English ‘area’, which can also designate subjects of activity; Slavic languages were not among the official languages when the term was popularised after the Tampere European Council; see also Michael Schwarz, Grundlinien der Anerkennung im Raums der Freiheit, der Sicherheit und des Rechts (Mohr Siebeck 2016) 217–22.

174 Administrative Dimension usage for other policy projects with uncertain contours, such as the European Economic Area or the European Higher Education Area. It was apparently chosen, at the time of the Treaty of Amsterdam, to evade the notion of ‘policy’, which can be read to imply a higher density of cooperation in the semantic practice of the EU institutions.3 Along similar lines, the ‘Common European Asylum System’, which is nowadays commonly written as an acronym, does not necessarily designate a high level of integration. In the EU context, the term ‘common’ (French: commun) is usually—though not uniformly— employed to designate an intermediate degree of harmonisation, in contrast to a ‘single’ (French: unique) policy.4 The terminology seems to have been a deliberate choice made by heads of state or government at their meeting in Tampere in the autumn of 1999.5 Primary law does not use the formula as a technical term written as an acronym with capital letters— unlike the widely used abbreviation ‘CEAS’. A common asylum ‘system’ may be rationalised linguistically to transcend the focus of other policies on legislative harmonisation by embracing the administrative dimension.6 Visas, border controls, and international cooperation complement the EU’s asylum policy. At a practical level, Article 78(2)(a) TFEU highlights the option of a ‘uniform status of asylum . . . valid throughout the Union’.7 One should be cautious not to read too much into this phrasing. The formula signalled that the Treaty of Lisbon transcended earlier limitations to minimum standards and does not authorise the introduction of a distinct asylum status departing from the Refugee Convention, to which Article 78(1) TFEU refers explicitly.8 Instead, the vision of ‘validity throughout the Union’ may be read to refer to transnational effects: it can mean either favourable free movement for beneficiaries of international protection or the mutual recognition of positive decisions, thereby paving the ground for swift returns in case of secondary movements.9 The example illustrates that the degree of commonality within the CEAS remains ambiguous: one may adequately describe it as a ‘system of [national] systems’.10 In any case, the objective of a common asylum ‘system’ does not involve federal uniformity.

3 See Jörg Monar, ‘The Area of Freedom, Security and Justice’ in Armin von Bogdandy and Jürgen Bast (eds), Principles of European Constitutional Law (2nd edn, Hart Publishing 2009) 551, 556–57; the term ‘policy’ was applied to justice and home affairs by the Treaty of Nice in the early 2000s. 4 Think of the common market (established in 1968) and the later move towards the single market (realised in 1992); similarly, the EU had a common monetary policy before the single currency was introduced; however, we refer to a Common Agricultural Policy or Common Commercial Policy despite a high degree of uniformity. 5 The former Vice-Director General for Justice and Home Affairs, Jean-Louis De Brouwer, recounted at the meeting of the European Migration Network in Helsinki on 26 October 2019 that the former French President Jacques Chirac insisted on the designation of a ‘common’ (not: single) asylum system. 6 See Evangelia (Lilian) Tsourdi, ‘The Emerging Architecture of EU Asylum Policy’ in Francesca Bignami (ed), EU Law in Populist Times. Crises and Prospects (CUP 2020) 191, 200–201; and Cathryn Costello, ‘Administrative Governance and the Europeanisation of Asylum and Immigration Policy’ in Herwig CH Hofmann and Alexander Türk (eds), EU Administrative Governance (Edward Elgar Publishing 2006) 287, 322. 7 TFEU, art 78(2)(b) for subsidiary protection does not use the territorial claim. 8 See Daniel Fröhlich, Das Asylrecht im Rahmen des Unionsrechts (Mohr Siebeck 2011) 162–75; and Daniel Thym, ‘Legal Framework for EU Asylum Policy’ in Daniel Thym and Kay Hailbronner (eds), EU Immigration and Asylum Law. Article-by-Article Commentary (3rd edn, CH Beck/Hart Publishing/Nomos 2022) MN 14–17. 9 See Steve Peers, ‘Institutional Framework’ in Steve Peers and others (eds), EU Immigration and Asylum Law (Text and Commentary), vol 3 (2nd edn, Brill Nijhoff 2015) 7, 13. 10 See https://en.wikipedia.org/wiki/System_of_systems (accessed 1 March 2023); and also AG Yves Bot, Opinion in Case C-213/17 X EU:C:2018:434, point 8.

European Migration Law as a Composite System 175

7.1.2 Multiplicity of actors at different levels Political and legal science started emphasising the ‘multi-level’ character of the European project from the 1990s onwards to highlight interlinkages, mutual dependencies, and conceptual amalgamation.11 This novel description responded to an increasing observation of overlaps in the fields of policy-making, legal effects, and constitutional concepts— in contrast to the earlier focus on separation and distinction in line with the concept of supranational legal autonomy. Multi-level thinking presented itself in both harmonious, coordinated-oriented accounts and confrontational variants accentuating potential differences and disputes. It was applied to the administrative dimension to describe the element of coordination in composite forms of comitology, networks, and agencies, particularly in the work of Italian and German scholars whose rich research output was later taken up by the transnational debate in English. It has become common to refer to an ‘integrated’ or ‘composite’ multi-level administration.12 While this chapter highlights composite features of the administrative dimension, the multi-level character of EU migration law embraces the legislature and the judiciary as well. They are indispensable for adopting implementing legislation and for designing complementary national legislation, such as nationality law, regularisation, and migrant integration. The seemingly clear-cut prescription of domestic treatment means different things across the Union depending on rules for nationals. EU migration law is often significantly less uniform on the ground than the widespread focus on the output of the EU institutions suggests. One may even go as far as saying that it may be necessary to analyse ‘European’ migration law holistically by combining supranational legislation with domestic rules and practices.13 It has been explained in the introduction that many research questions cannot be adequately addressed by discussing the EU level in isolation. It will often be necessary to include the domestic dimension at least exemplarily, for instance by means of comparative analyses. To do so allows for insights regarding the added value of supranational rules; often their impact will differ from country to country and may change over time. It can be a formidable challenge to design such research projects, which have to pay attention to the diversity of national rules and practices without getting lost in the sheer complexity of the legal material in its multiple languages. The term ‘multi-level’ is often used along with the notion of ‘governance’, which designates the spread of informal and non-hierarchical forms of rule-making, including through soft law and the involvement of private actors.14 These softer forms of governance have 11 For political science see Fritz W Scharpf, ‘Community and Autonomy. Multi-Level Policy-Making in the European Union’ (1994) 1 J Eur Politics & Policy 219; for constitutional law see Ingolf Pernice, ‘Multilevel Constitutionalism and the Treaty of Amsterdam’ (1999) 36 CML Rev 703. 12 ‘Composite’ is a rough translation of the German ‘Verwaltungsverbund’; see Eberhard Schmidt-Aßmann, ‘Verfassungsprinzipien für den europäischen Verwaltungsverbund’ in Wolfgang Hoffmann-Riem and others (eds), Grundlagen des Verwaltungsrechts. Band I (2nd edn, CH Beck 2012) § 5; and also Giacinto della Cananea, ‘The European Union’s Mixed Administrative Proceedings’ (2004) 68 L & Contemp Probs 197; and Herwig CH Hofmann and Alexander H Türk, ‘The Development of Integrated Administration in the EU and its Consequences’ (2007) 13 ELJ 253. 13 See generally Armin von Bogdandy, ‘National Legal Scholarship in the European Legal Area’ (2012) 10 ICON 614. 14 See Herwig CH Hofmann, Gerard C Rowe, and Alexander H Türk, Administrative Law and Policy of the European Union (OUP 2011) ch 9.

176 Administrative Dimension traditionally had a lesser weight in the field of migration, which concentrates on classic forms of hierarchical ‘government’. That is not to say that elements of governance have no role to play, as illustrated in Chapter 2.3.4 on soft law and Chapter 4.2.4 for private actors. On the whole, however, their relevance stays short of their involvement in other domains.

7.1.3 Operational character of asylum and border controls Chapter 6 demonstrated that the Court of Justice approaches migration law as an integral part of the supranational legal order; case law on areas other than migration influences the interpretation of legislative instruments—and vice versa. The holistic outlook of the judicial function does not mean, however, that administrative solutions developed for one policy field can easily be transplanted to other domains. It is widely accepted, in research on administrative law, that the structure of the public administration has seen much differentiation in recent decades.15 There are important contextual differences between, for example, the supervision of financial services, the authorisation of new chemicals, the disbursement of farm subsidies, and asylum decisions, in addition to the commonality that all these domains are subject to supranational legislation. When it comes to EU migration law, protracted compliance deficits are palpable in core areas of asylum, border control, and return policies.16 National administrations may lack the political will or the managerial capacities to guarantee swift and fair asylum procedures, adequate reception conditions, and respect for procedural obligations during border controls; individuals often prefer not to comply with mandatory periods of voluntary departure or embark on secondary movements. It does not imply ethical judgment to note these discrepancies between the law in the books and everyday practices. Publication of new rules in the Official Journal does not guarantee success on the ground; it would be insufficient to assume that the doctrine of direct effect and preliminary references alone can guarantee respect for EU law.17 EU institutions are well-advised to consider administrative mechanisms supporting compliance. Generally speaking, there is no uniform expression of EU administrative law but variations with certain elements in common.18 To be successful, administrative rules should reflect the distinct features of the policy field in question; it depends on contextual factors whether regulatory techniques can be transplanted from one policy field to another. If that is correct, it is essential to grasp the specificities of migration law to develop context-sensitive administrative measures supporting practical implementation. Arguably, a decisive element is the operational character of migration law which one might analytically— not pejoratively—describe as an exercise of ‘mass administration’.19 The application of 15 See generally Helmuth Schulze-Fielitz, ‘Grundmodi der Aufgabenwahrnehmung’ in Hoffmann-Riem and others (n 12) § 12; see also Carol Harlow and Richard Rawlings, Law and Administration (3rd edn, CUP 2009) ch 2. 16 That conclusion is no novel phenomenon; see Costello (n 6) 314–22; and Neil Walker, ‘In Search of the Area of Freedom, Security and Justice’ in Neil Walker (ed), Europe’s Area of Freedom, Security and Justice (Hart Publishing 2004) 3, 20–23. 17 See Daniel Thym, ‘The “Refugee Crisis” as a Challenge of Legal Design and Institutional Legitimacy’ (2016) 53 CML Rev 1545, 1554–58. 18 See generally Bruno de Witte, ‘Institutional Variation Across Policy Fields in the Evolution of European Union Law’ (2013) 20 MJECL 465; and Giandomenico Majone, Europe as the Would-be World Power (CUP 2009). 19 See Rainer Wahl, ‘Die Aufgabenabhängigkeit von Verwaltung und Verwaltungsrecht’ in Eberhard Schmidt- Aßmann and others (eds), Reform des Allgemeinen Verwaltungsrechts. Band 1 (Nomos 1993) 177, 192–96, 206–208.

European Migration Law as a Composite System 177 migration law habitually involves the application of supranational rules to numerous individuals by officials with a relatively low degree of formal education (as opposed to, for instance, stakeholders in competition cases). To build an effective and accountable ‘mass administration’ requires a challenging mixture of administrative routine and standardisation to ensure equal treatment and attention to the particularities of the individual case. The sheer number of individual files commands tremendous efforts in terms of administrative resources, organisation, and training. Mutual distrust between migrants and public officials further complicates an effective fulfilment of administrative tasks. These operational challenges are particularly pronounced in the fields of visas, border controls, asylum, and return, while legal migration is usually less confrontational. We shall see that the EU institutions aim to support domestic authorities through financial support and agencies; their structure and output reflects the operational characteristics and distinguishes them from other policy fields and the agencies established on these matters.

7.1.4 Diverse avenues of inter-state cooperation Visibility of Frontex and the Asylum Agency should not leave us with the impression that they are the only or primary element of administrative collaboration. Other forms of cooperation have been developed, including on segments of migration law the agencies do not cover. What is more, neither Frontex nor the Asylum Agency are simple apparatuses of vertical top-down control; their governance structure and operational capacities depend, as we shall see, on loyal cooperation with national authorities, not least for the secondment of personnel. Horizontal cooperation between Member States is a general feature of justice and home affairs, as highlighted by Articles 71–74 TFEU. In a series of judgments on the Dublin system, discussed in Chapter 13.3.6 on asylum, the Court emphasised that inter- state cooperation and mutual trust relate to the ‘raison d’être of the European Union and the creation of an area of freedom, security, and justice’.20 Several explanations illustrate why horizontal cooperation is critical for EU migration law. The need for cooperation becomes obvious when third country nationals cross internal borders, be it as asylum seekers in scenarios of secondary movements, be it as workers engaged in intra-European mobility. It becomes further evident when an administrative decision has transnational legal effects, as in the case of Schengen visas: a visa delivered by an Estonian consulate authorises entry into Finland, Latvia, or any other Schengen country. There are uniform patterns of how inter-state collaboration unfolds. There are variances in the legal rules and the methods employed, which are ultimately defined by the legislative act in question. A bird’s eye view allows us to identify clusters of activity, although these groups do not translate into clear-cut prototypes or distinct models. Not even ‘mutual trust’, highlighted so prominently by the Court, constitutes an overarching constitutional or legislative principle; its characteristics and legal effects depend on the legislation in question.21 At an intermediate level of abstraction, we may identify four patterns of inter-state cooperation. 20 Joined Cases C-411/10 and C-493/10 NS and others EU:C:2011:865, para 83. 21 See Francesco Maiani and Sara Mogliorini, ‘One Principle to Rule Them All! Anatomy of Mutual Trust in the Law of the Area of Freedom, Security and Justice’ (2020) 57 CML Rev 7.

178 Administrative Dimension First, information exchange has been essential ever since the first version of the Schengen Information System (SIS). Recent years have seen upgrades of existing databases and the introduction of new ones, such as the Entry/Exit System (EES) or the European Travel Information and Authorisation System (ETIAS). Although their operation is largely invisible, the practical and conceptual impact is substantial. Databases will take centre stage in Chapter 9. Secondly, we can observe composite procedures and transnational legal effects of administrative decisions, which exist in other segments of Union law as well. Instead of presenting a uniform phenomenon, statutory manifestations of composite procedures and transnational effects differ markedly, both in migration law and beyond.22 They embrace truly transnational decisions when one Member State delivers a Schengen visa on behalf of all the others as well as limited transnational effects in cases of intra-European mobility. By way of example, a second Member State may reject the mobility of researchers and students, admitted by another country, under the conditions laid down in secondary legislation.23 Indirect transnational effects include the rejection of asylum applications as inadmissible if another Member State had already granted protection.24 Composite procedures comprise consultation among national authorities concerning the accuracy of entry bans in the SIS, which will be discussed towards the end of this chapter. Thirdly, EU institutions support convergence through networks among practitioners and financial support. Such networks include formal institutions like the European Migration Network (EMN), which produce valuable comparative reports.25 Such institutionalised networks played a prominent role in the early years of EU migration policy when transnational links had to be established in an area that has traditionally been defined by domestic introspection. Cooperation within Schengen proved essential in establishing an esprit de corps;26 it was later replicated in other areas such as illegal immigration or asylum.27 Working parties within the Council similarly provide an institutional platform for trust building, mutual learning, and policy exchange, including on sensitive areas such as pre-departure language tests as a precondition for family reunification which have spread horizontally between several Member States.28 Finally, there are different forms of operational cooperation ranging from the organisation of joint return flights to cooperation among border guards at the internal borders of the Schengen area.29 Having said this, most operational cooperation in the domains of border

22 See Schwarz (n 2) chs 5–6; and Jürgen Bast, ‘Transnationale Migrationsverwaltung des europäischen Migrationsraums’ (2007) 46 Der Staat 1, 16–32. 23 See Students and Researchers Directive (EU) 2016/801, arts 27–32. 24 See Asylum Procedures Directive 2013/32/EU, art 32(2)(a). 25 See https://home-affairs.ec.europa.eu/networks/european-migration-network-emn (accessed 1 March 2023); the EMN was established by Decision 2008/381/EC establishing a European Migration Network [2008] OJ L131/7, with later amendments. 26 See Ruben Zaiotti, Culture of Border Controls (University of Chicago Press 2011) chs 5–6. 27 See Philippe De Bruycker and Evangelia (Lilian) Tsourdi, ‘Building the Common European Asylum System beyond Legislative Harmonisation’ in Vincent Chetail and others (eds), Reforming the Common European Asylum System (Brill 2016) 473; and Regulation (EU) 2019/1240 on the creation of a European network of immigration liaison officers [2019] OJ L198/88, which replaced the previous Council Regulation (EC) No 377/2004 [2004] OJ L64/1. 28 See ch 14.1.2. 29 See Decision 2004/573/EC on the organisation of joint flights for removals [2004] OJ L261/28; Schengen Borders Code Regulation (EU) 2016/399, arts 17 and 18; and Matthias Laas, Die Entstehung eines europäischen Migrationsverwaltungsraumes (Nomos 2008) 150–212.

National Procedural Autonomy and Its Limits 179 controls and asylum is nowadays channelled through Frontex and the Asylum Agency, which will feature in Chapter 8. In other areas, cooperation often remains bilateral. Many pieces of legislation contain provisions on horizontal cooperation. A telling example are Schengen visas, for which national consulates present in third states establish cooperation mechanisms.30

7.2 National Procedural Autonomy and Its Limits In the absence of harmonisation, Member States apply national rules on administrative and court procedure when deciding individual cases. Such indirect implementation is commonly discussed under the heading of ‘national procedural autonomy’. The degree of freedom varies considerably, depending on the contents of secondary legislation (7.2.1). In the absence of legislative harmonisation, the Court applies unwritten general principles; the principle of effectiveness has gained practical relevance in the case law on migration, especially as a corrective tool for strict time limits (7.2.2). General principles are increasingly buttressed by the procedural safeguards in the Charter of Fundamental Rights. Many judgments rely upon the right to an effective remedy to correct domestic migration laws (7.2.3). Inspection of the case law on procedural deficits under the Asylum Procedures Directive and the Return Directive shows fluctuations in the judicial output, which complicate the search for cross-sectoral coherence (7.2.4).

7.2.1 The many faces of procedural harmonisation Anyone familiar with everyday practices understands how crucial procedural rules are for the resolution of individual cases. They are particularly relevant for visas, border controls, asylum, and return, where domestic authorities routinely assess the situation in foreign countries that are often subject to an enhanced degree of uncertainty. In borderline scenarios where the claim to protection hangs in the balance, procedural rules can be at least as important as questions of substance, such as the definition of refugee status. When discussing the supranational approach to procedures, the EU institutions and the Court are in a precarious situation for the simple reason that domestic rules and administrative traditions differ markedly. This diversity renders it difficult to design common standards and to develop a coherent case law. This dependence of migration law on procedures indicates that the principle of subsidiarity does not generally prevent the EU institutions from harmonising procedural standards despite the general desire, expressed in Article 67(1) TFEU, that justice and home affairs shall respect ‘the different legal systems and traditions of the Member States’. Legal bases in Articles 77(2)–79(2) TFEU must be interpreted to allow for the harmonisation of administrative procedures depending on political choices and the relevance of procedures for the different sub-segments of migration law.31 The level of detail varies considerably. 30 See ch 11.4.5. 31 For asylum see the explicit reference in TFEU, art 77(2)(b); for the other provisions see Daniel Thym, ‘Legal Framework for Entry and Border Controls’ in Thym and Hailbronner (n 8) MN 8, 10a.

180 Administrative Dimension While Article 4 Single Permit Directive 2011/98/EU lays down an abstract procedural skeleton (like most other directives on legal migration), the Asylum Procedures Directive 2013/ 32/EU contains detailed rules. By way of example, it regulates the personal interview in no fewer than four articles, jointly comprising more than 1,300 words. The degree of harmonisation varies greatly between and within the different legal instruments. Chapter 6.3.2 explained that the abstract distinction between directives and regulations gives way to gradual variations in the institutional practice. Even within one instrument, we have to assess the degree of harmonisation carefully. For instance, the Court interprets Article 46(3) Asylum Procedures Directive 2013/32/EU to provide detailed guidance on the density of judicial oversight, which must provide ‘for a full and ex nunc examination of both facts and points of law’. That does not mean, however, that all aspects of the court procedure are fully harmonised. By way of example, Article 46(10) leaves the choice whether to establish time limits for the judicial decision with national parliaments. In such circumstances, the Court will usually respect the decision made at the domestic level, although it may still, as we shall see, scrutinise domestic rules in light of human rights or the unwritten general principles. Even seemingly clear-cut rules like the requirement of ‘a full and ex nunc examination of both facts and points of law’ will not necessarily translate into uniform practices on the ground. Comparative studies portray profound differences between national models in terms of administrative procedure and judicial oversight.32 Those studying the procedural dimension should be aware of these profound national disparities. This dependence on the domestic context has a knock-on effect on the practical effects of legislative harmonisation. Those not familiar with everyday practices may ignore that ostensibly straightforward concepts like the ‘burden of proof ’ and other evidentiary standards can mean different things in different countries: while some countries limit court oversight mostly to the facts and arguments put forward by the parties, others require extensive court scrutiny ex officio.33 Similarly, the notion of executive ‘discretion’ can refer to different legal concepts and judicial powers across the Union.34

7.2.2 Effectiveness as a corrective vehicle In the absence of leges speciales in secondary legislation, states apply national rules to administrative procedures and court proceedings in line with the principle of ‘national procedural autonomy’. Crucially, national procedural autonomy comes along with supranational limitations, which can be activated to restrict the procedural leeway. These unwritten limitations allow judges to fill regulatory gaps within EU legislation on the basis of a twofold caveat. When applying supranational legislation, Member States must ensure, firstly, that the rules for EU-related claims are not less favourable than those governing

32 See Michel Fromont, ‘A Typology of Administrative Law in Europe’ in Armin von Bogdandy and others (eds), The Max Planck Handbooks in European Public Law, vol I (OUP 2017) 579; and the contributions to Giacinto della Cananea and Mauro Bussani (eds), Judicial Review of Administration in Europe (OUP 2021). 33 See Ida Staffans, Evidence in European Asylum Procedures (Martinus Nijhoff 2012); and also Jalek Chlebny, ‘Public Order, National Security and the Rights of Third-Country Nationals in Immigration Cases’ (2018) 20 EJML 115, 120–26. 34 See Takis Tridimas, The General Principles of EC Law (2nd edn, OUP 2007) chs 3–5.

National Procedural Autonomy and Its Limits 181 similar domestic actions (principle of equivalence) and, secondly, that they do not render the exercise of rights conferred by Union law practically impossible or excessively difficult (principle of effectiveness).35

7.2.2.1 Court practice on migration law Effectiveness is far more important than the sister principle of equivalence, whose application the Court often leaves to domestic courts which are better placed to compare national laws.36 In doing so, judges do not insist on categorical parallelism. Instead, they recognise that asylum procedures can be subject to stricter requirements than rules for other areas of domestic law; no equal treatment is required when procedures on two subject matters are not comparable, for instance regarding the suspensive effect of legal remedies.37 The Court found the objective of ‘expeditious’ decision-making to support stricter time-limits for asylum applications in comparison to other areas.38 It also confirmed that distinctions within asylum law between different countries of origin do not constitute unjustified unequal treatment.39 The added value of the effectiveness test concerns domestic procedural standards not subject to harmonisation. On the whole, the Court’s practice is bound to remain sketchy: it concerns selective issues, which do not easily blend into an overarching theory. An underlying reason is the dependence on preliminary references. Domestic judges will rarely consult judges in Luxembourg on procedural questions which are not harmonised, although a first ruling may trigger follow-up references on related issues, thus resulting in a considerable number of judgments on interrelated questions. That is precisely what happened in the field of migration, in particular with regard to time limits. Other subject matters may follow in the years to come. In the majority of cases, judges have found national rules to violate the principle of effectiveness. Nevertheless, such an outcome is not a foregone conclusion and can be rationalised by national courts primarily referring problematic cases. Generally, the Court recognises that it can be legitimate to curb legal remedies, for instance to guarantee that ‘applications for international protection should be dealt with expeditiously’.40 Appeals to a higher court can be limited to matters raised before the lower court.41 Judges also confirmed, in the context of asylum, that legal certainty can outweigh considerations of effectiveness. In doing so, the Court had recourse to judgments on other segments of Union law to confirm that administrative decisions and court rulings, for which time limits for remedies or appeals have elapsed, need not be automatically reversed even though they were based on an incorrect interpretation of Union law.42 Effectiveness proves an essential argument to correct restrictive national rules but not an all-purpose tool.

35 This has been settled case law ever since Case 33/76 Rewe-Zentralfinanz EU:C:1976:188, para 5; see further Koen Lenaerts, Ignace Maselis, and Kathleen Gutman, EU Procedural Law (OUP 2014) ch 4. 36 By way of example see Case C-651/19 Commissaire général aux réfugiés et aux apatrides EU:C:2020:681, paras 39–41, 51–52. 37 See Case C-175/17 Belastingdienst/Toeslagen EU:C:2018:776, paras 41–46. 38 See Commissaire général aux réfugiés et aux apatrides (n 36) para 54. 39 See ch 10.4. 40 Commissaire général aux réfugiés et aux apatrides (n 36) para 54. 41 Case C-161/15 Bensada Benallal EU:C:2016:175, paras 27–28. 42 See Joined Cases C-924/19 PPU and C-925/19 PPU Országos Idegenrendészeti Főigazgatóság Dél-alföldi Regionális Igazgatóság EU:C:2020:367, paras 185–87; and Craig, EU Administrative Law (n 1) 613–18.

182 Administrative Dimension

7.2.2.2 Time limits as a test case The example of time limits illustrates that the operationalisation of the effectiveness test remains notoriously difficult. Why? A comparison between judgments shows that practical consequences depend on the factual and legal specificities of the case at hand.43 In the field of asylum, secondary legislation stipulates in general language that Member States ‘shall provide for reasonable time limits and other necessary rules for the applicant to exercise his or her right to an effective remedy’.44 On this basis, judges held a time limit of 15 days to be acceptable, while striking down an eight-day rule as a matter of principle (subject to a caveat that it might be acceptable in some scenarios).45 Strictly calculated deadlines of 10 days for regular appeals and five days for legal remedies after a subsequent asylum application made during detention were accepted, although the Court highlighted that national judges had to ascertain whether other safeguards, such as access to legal assistance, were practically available.46 While the principle of effectiveness will work to the benefit of third country nationals in many cases, it is not intrinsically linked to such an outcome. For instance, general principles were found to stand in the way of domestic rules obliging authorities to positively respond to an application for a residence permit for the sole reason that time limits for taking a decision had elapsed; authorities must diligently apply the conditions laid down in secondary legislation.47 Conversely, the Court found that a Hungarian rule obliging domestic courts to deliver a ruling within a certain time frame violates the principle of effectiveness, provided that national law prevents judges from performing their oversight function effectively in light of contextual factors such as workload, working conditions, or the difficulty of certain cases.48

7.2.3 Fundamental right to an effective remedy Court judgments on national procedural autonomy increasingly overlap with human rights. On the basis of the general comments in Chapter 5, this section will concentrate on procedural safeguards in the Charter of Fundamental Rights. As a matter of principle, these guarantees hold the potential for surprise outcomes, effectively supplanting ‘shallow’ statutory provisions in secondary legislation by ‘deep’ judicial reshaping of the domestic legal order.49 Procedural guarantees in the Charter can be applied, together with the principle of effectiveness, to domestic rules not subject to supranational harmonisation, in line with the

43 See generally Craig, EU Administrative Law (n 1) 759–87. 44 Asylum Procedure Directive 2013/32/EU, art 46(4). 45 See Case C-69/10 Samba Diouf EU:C:2011:524, paras 66–68; and Case C-564/18 Bevándorlási és Menekültügyi Hivatal (Tompa) EU:C:2020:218, paras 72–73; moreover, Case C-429/15 Danqua EU:C:2016:789, paras 41–48 concerned time limits for submitting an asylum application; see also Marcelle Reneman, ‘Speedy Asylum Procedures in the EU’ (2013) 25 IJRL 717, 730–46. 46 See Case C-755/19 THC. EU:C:2021:108, paras 26–44; and Commissaire général aux réfugiés and aux apatrides (n 36) paras 54–64, with additional comments on the special situation of an applicant residing abroad. 47 See Case C-706/18 Belgische Staat EU:C:2019:993, paras 26, 34–36; and Case C-246/17 Diallo EU:C:2018:499, paras 44–49. 48 See Case C-406/18 Bevándorlási és Menekültügyi Hivatal EU:C:2020:216, paras 26–34. 49 See Jürgen Bast, ‘Of General Principles and Trojan Horses. Procedural Due Process in Immigration Proceedings under EU Law’ (2010) 11 GLJ 1006.

National Procedural Autonomy and Its Limits 183 case law on Article 51(1) CFR defining the scope of the Charter of Fundamental Rights.50 In practice, however, the impact of the Charter depends on domestic courts referring questions of interpretation to the Court in Luxembourg.

7.2.3.1 Availability of legal remedies In the judicial output, fundamental procedural rights have gained relevance for the Asylum Procedures Directive and the Return Directive, in particular. The right to an effective remedy under Article 47 CFR proved particularly relevant, since it transcends existing obligations under human rights law in two respects. First, the Charter requires an effective remedy ‘before a tribunal’, thereby going beyond the combined reading of Articles 6 and 13 ECHR that do not generally require remedies before courts, due to the non-civil and non-criminal character of migration law.51 Chapter 3.1.3 explained that Article 13 Return Directive has to be interpreted in conformity with the Charter. Reference to complaints before an ‘administrative authority’ must be interpreted to require judicial remedies in light of Article 47 CFR. Secondly, the Charter does not, unlike Article 13 ECHR, confine remedies to potential human rights violations.52 The Charter applies whenever ‘rights guaranteed by Union law’ may be violated. In doing so, Article 47 CFR effectively guarantees access to domestic courts whenever we conclude that a directive or regulation lays down an individual right. Chapter 6.5 illustrated that it depends on the outcome of the legislative process whether the instruments discussed in this volume embrace individual rights. Whenever this is the case, Member States are obliged to establish judicial oversight in areas such as refusal of entry or visa procedures which had not traditionally been subject to legal remedies.53 EU migration law played a crucial role in reining in earlier notions of largely unfettered state discretion over the entry and stay of foreigners. 7.2.3.2 Scope of judicial oversight Notwithstanding the wider scope, Article 47 CFR must be interpreted in line with ECtHR case law on Articles 6 and 13 ECHR to start with.54 When analysing the case law, it is not always straightforward to what extent Court rulings on specific subject matters can be generalised, since the judicial reasoning does not always distinguish between constitutional requirements and legislation-specific considerations. A perfect example is the question of whether legal remedies must provide for an ex nunc assessment. While of Article 46(3) Asylum Procedures Directive 2013/32/EU answers that question in the affirmative, Article 27 Dublin III Regulation (EU) No 604/2013 remains silent. Judges interpreted this silence to mean that the legislature ‘has harmonised only some of the procedural rules governing’ legal remedies against take back decisions.55 The principle of effectiveness may be activated 50 See ch 5.4.4; and Case C-403/16 El Hassani EU:C:2017:960, paras 33–34 on the interaction with national procedural autonomy. 51 See MN and others v Belgium App no 3599/18 (ECtHR [GC], 5 May 2020) §§ 137–40; and Maaouia v France App no 39652/98 (ECtHR [GC], 5 October 2000) §§ 32–41. 52 Article 13 ECHR presupposes an arguable complaint that another human right was violated, for instance Article 3 ECHR; see MSS v Belgium and Greece App no 30696/09 (ECtHR [GC], 21 January 2011) § 288. 53 See Schengen Borders Code Regulation (EU) 2016/399, art 14(3); and Visa Code Regulation (EC) No 810/ 2009, art 32(3); see also El Hassani (n 50) paras 39–41; and Case C-949/19 Konsul Rzeczypospolitej Polskiej w N EU:C:2021:186, paras 41–46. 54 See Case C-205/15 Toma and Biroul Executorului Judecătoresc Horaţiu-Vasile Cruduleci EU:C:2016:499, para 40; and Marcelle Reneman, EU Asylum Procedures and the Right to an Effective Remedy (Hart Publishing 2014). 55 See Case C-194/19 État belge EU:C:2021:270, paras 37–41.

184 Administrative Dimension to limit state discretion, but it will not usually require a full ex nunc assessment.56 ECtHR rulings on independent and rigorous scrutiny under Article 13 ECHR may serve as a guideline for legislative instruments which, unlike the Asylum Procedures Directive, do not harmonise the nature of judicial oversight.57

7.2.3.3 Suspensive effect and legal assistance Time is often a matter of great practical importance in border control, asylum, and return procedures, in particular when individuals seek to remain on the territory. In this respect, settled case law foresees that an ‘effective remedy’ does not generally require automatic suspensive effect, let alone a right to be admitted to the territory during judicial proceedings.58 Instead, the ECtHR’s frequently-cited formula that Member States must provide for ‘automatic suspensive effect’ presupposes a real risk of an irreversible human rights violation, in particular with regard to Article 3 ECHR59 (not, however, for private and family life60). Whenever this condition is fulfilled, automatic suspensive effect requires domestic courts to be able to issue a suspension order. The language of automatism refers to an effective guarantee against refoulement; states have leeway on how to organise the procedure.61 The CJEU shares this view. Article 47 CFR does not entail a generic right to remain, unless individuals can demonstrate that there is an arguable claim under Article 4 CFR.62 The exclusion of suspensive effect for take back decisions under the Dublin system and for refusal of entry at external borders can be compatible with fundamental rights.63 In any case, suspensive effect concerns the first instance of judicial oversight only, not, however, legal remedies at a second judicial instance, which is not mandatory under human rights law.64 The abstract requirement, in ECtHR case law, that remedies ‘must be available in practice as well as in law’65 can have repercussions for access to legal assistance and interpretation. A telling example of how important these guarantees can be is the situation in the hotspots, which are often found in the geographic periphery where legal assistance can be difficult to obtain. Articles 19–23 Asylum Procedures Directive 2013/32/EU contain far- reaching statutory guarantees in terms of procedural support, while Article 13(3) Return Directive 2008/115/EC simply refers to the ‘possibility’ of obtaining legal assistance, without, however, obliging Member States to make it factually available. There is complex human rights case law on when legal assistance is mandatory, possibly even free of charge in acute scenarios. Judges do not, however, recognise a generic and unconditional guarantee.66

56 ibid paras 42–47. 57 For a summary see De Souza Ribeiro v France App no 22689/07 (ECtHR [GC], 13 December 2012) §§ 77–83. 58 For a different proposal see Helen Oosterom-Staples, ‘Effective Rights for Third-Country Nationals?’ in Hans Lindahl (ed), A Right to Inclusion and Exclusion? (Hart Publishing 2009) 65, 89–90. 59 See the summary in Khlaifia and others v Italy App no 16483/12 (ECtHR [GC], 12 December 2016) §§ 276–77. 60 See De Souza Ribeiro (n 57) paras 82–83. 61 See Case C-233/19 CPAS de Liège EU:C:2020:757, paras 49–53; and Čonka v Belgium App no 51564/99 (ECtHR, 5 February 2002) § 79. 62 CPAS de Liège (n 61) paras 61–66; Case C-181/16 Gnandi EU:C:2018:465, paras 54–56; Case C-239/14 Tall EU:C:2015:824, paras 56–58; and Case C-562/13 Abdida EU:C:2014:2453, paras 44–53. 63 On Dublin see ch 6.5.4; the compatibility of Schengen Borders Code Regulation (EU) 2016/399, art 14(3) with human rights is practically guaranteed by the option to apply for asylum; see also Reneman (n 54) ch 7. 64 See Belastingdienst/Toeslagen (n 37) paras 22–24, 32–36. 65 MSS v Belgium and Greece (n 52) § 290. 66 See Case C-279/09 DEB EU:C:2010:811, paras 37–52; and Case C-249/13 Boudjlida EU:C:2014:2431, paras 64–70.

National Procedural Autonomy and Its Limits 185 Experts in migration law should understand that to find otherwise would have far-reaching implications for other segments of the law.

7.2.4 Good administration The fundamental right to good administration in Articles 41 CFR is an innovative feature. While the provision directly binds the institutions and agencies of the Union, the Court has recognised that it can be applied domestically as an unwritten general principle.67 Practical relevance depends on the domestic context. Member States with sophisticated procedural guarantees will be less affected than other countries. The right to be heard and the obligation to give reasons have features in the supranational case law on migration so far.

7.2.4.1 Giving reasons and right to be heard Judgments on the right to good administrative exhibit caution. The Court found that the standard form for rejecting visa applications meets the obligation to state reasons under Article 42(2)(c) CFR, although the set of pre-defined answers are highly abstract, mirroring a multiple choice test more than elaborate explanations.68 It also concluded that reasons must be given for ordering or prolonging detention under Article 15(3) Return Directive, not however for the review of an existing detention order.69 Conversely, national authorities cannot generally refrain from giving reasons on public security grounds, as the Charter requires them to always convey the essence of their reasoning in a manner which takes due account of the necessary confidentiality; in doing so, they have some flexibility in how they structure digital documents, nor is there an obligation to use handwritten signatures.70 To oblige domestic authorities to justify their behaviour, albeit at a high level of abstraction, allows individuals to decide whether to launch legal remedies. The right to be heard guarantees that domestic authorities take the position of the third country national into account. It applies as an integral part of the rights of the defence71 and covers, in accordance with Article 42(2)(a) CFR, any individual measure ‘which would affect him or her adversely’. As a consequence, it is not limited—unlike Article 47 CFR—to scenarios where individual rights in secondary legislation are at stake. The multifaceted judicial output leaves us with mixed messages about the consequences of procedural deficits. In a number of rulings, the Court avoided an outcome in which non-compliance with procedural safeguards resulted in an annulment of the administrative decision.

67 See Case C-604/12 HN EU:C:2014:302, paras 49–51; and Case C-166/13 Mukarubega EU:C:2014:2336, paras 42–45. 68 See ch 11.4.3; and Joined Cases C-225/19 and C-226/19 Minister van Buitenlandse Zaken EU:C:2020:951, paras 44–47; as well as, for refusal of entry, Case C-584/18 Blue Air—Airline Management Solutions EU:C:2020:324, paras 82–86. 69 See Case C-146/14 Mahdi EU:C:2014:1320, paras 41–52. 70 See Case C-159/21 Országos Idegenrendeszeti Főigazgatóság and others EU:C:2022:708, paras 44–59; Case C- 300/11 ZZ EU:C:2013:363 on Union citizenship; and Case C-564/22 Bundesrepublik Deutschland EU:C:2022:951, paras 33–63. 71 See Case C-560/14 M EU:C:2017:101, para 25.

186 Administrative Dimension

7.2.4.2 Consequences of procedural deficits Judgments on the right to be heard remind us that fundamental rights can be limited in accordance with Article 52(1) CFR. ‘Balancing’ private and public interests is a defining feature of human rights law. Judges are strict when it comes to personal interviews in asylum proceedings, since secondary legislation reinforces the significance of the right to be heard.72 Absence of an administrative hearing will regularly result in a verdict of illegality, also considering that the statutory prescriptions on interviews in the Asylum Procedures Directive transcend the human right to be heard under Article 43(2)(a) CFR.73 The higher level of protection for asylum was confirmed by judgments on the Return Directive. In the absence of detailed statutory prescriptions, the lack of a hearing does not generally result in the illegality of the administrative decision.74 Academic observers have met these findings with disappointment.75 Yet, the case law can be rationalised by the contents and context of the instruments under analysis. Even in the field of asylum, the right to be heard remains subject to limitations. Judges did not submit judicial proceedings to the same procedural rigidity as the administrative procedure. Domestic judges may deliver their rulings on the basis of the information in the administrative file, without being required to perform a second personal interview or organise a hearing.76 Judges also gave states leeway as to whether the domestic judiciary may substitute the administrative decision with their own. Some countries do not foresee such substitution powers, even in situations where judges conclude that a specific outcome was mandatory. Instead, the file may be referred back to the authorities to take a new decision in light of the findings of the domestic court.77 Judges confirmed that neither the Asylum Procedures Directive nor human rights or unwritten general principles mandated substitution powers on the part of judges.78 From a theoretical perspective, the case law illustrates that the Court of Justice often replicates, in the field of migration, the position it defends—in varying degrees—in other segments on Union law, notably with regard to the administrative dimension of competition cases.79 Experts in migration law are well-advised, therefore, to relate their analysis to the administrative dimension to Union law at an intermediate level of abstraction, so to better understand the outcome but also to identify new arguments. In any case, they should beware of the simple transfer of domestic practices to the EU level. Union law often develops a solution of its own, combining the tradition of several domestic legal orders. Judgments on the consequences of procedural deficits are a good example. While common law countries 72 See Case C-517/17 Addis EU:C:2020:579, paras 64–71. 73 See M (n 71) para 54. 74 See Mukarubega (n 67) paras 46–53; and Case C-383/13 PPU G and R EU:C:2013:533, paras 32–40; see also Boudjlida (n 66) paras 55–56; and Valeria Ilareva, ‘The Right to be Heard. The Underestimated Condition for Effective Returns and Human Rights Consideration’ in Madalina Moraru and others (eds), Law and Judicial Dialogue on the Return of Irregular Migrants from the European Union (Hart Publishing 2020) 351. 75 See Piotr Sadowski, ‘A Safe Harbour or a Sinking Ship? On the Protection of Fundamental Rights of Asylum Seekers in Recent CJEU Judgments’ (2019) 11 EJLS 29, 52–53; and Philippe De Bruycker and Sergo Mananashvili, ‘Audi alteram partem in Immigration Detention Procedures between the ECJ, the ECtHR and Member States’ (2015) 52 CML Rev 569, 582–87. 76 See Case C-348/16 Sacko EU:C:2017:591, paras 33–40, 45–48. 77 See Marcelle Reneman, ‘Asylum and Article 47 of the Charter’ in Andrea Crescenzi and others (eds), Asylum and the EU Charter of Fundamental Rights (Editoriale Scientifica 2018) 59, 67–71. 78 See Case C-556/17 Torubarov EU:C:2019:626, paras 54–59; and Case C-585/16 Alheto EU:C:2018:584, paras 145–48. 79 See Tridimas (n 34) ch 8.

Transnational Effects of Horizontal Cooperation 187 tend to highlight the self-supporting character of procedural guarantees, civil law jurisdictions emphasise their supportive function and do not, as a result, generally sanction non- compliance autonomously.80

7.3 Transnational Effects of Horizontal Cooperation Widespread focus on the vertical dimension of EU migration law, regarding agency involvement or the limits of national procedural autonomy, should not detract from the significance of horizontal cooperation between the Member States. In the area of freedom, security, and justice, administrative decisions taken by one country often have transnational legal effects. A variety of different scenarios gives rise to a common question: whether, and if so under which conditions, may other Member States, or individuals living there, challenge the legality of a decision delivered by a first Member State? The answer to that question depends on the legislative instrument applicable, but the underlying challenge is the same. EU institutions, including the Court, have to strike a balance between countervailing claims to legal certainty and the rule of law. As a matter of principle, justice and home affairs rely—like the single market—on loyal cooperation. Member States have to respect decisions made elsewhere under usual circumstances. The idea is firmly embedded in the collective memory of the European project with its preference for ‘mutual trust’ and ‘mutual recognition’—two concepts that do not embody an overarching general principle.81 Instructive examples beyond the domain of migration are judgments on the mutual recognition of drivers’ licences or regarding sanctions for transnational television broadcasts.82 They illustrate that the conditions and limits of transnational legal effects ultimately depend on the rules in question. What is valid for driving licences, need not be appropriate for human rights limits to Dublin transfers, which will feature in Chapter 13.3.6 on asylum. Comments hereafter concentrate on transnational effects as a matter of administrative law, not human rights standards. In EU migration law, we can distinguish at least three scenarios. Distinct rules exist for visas and databases where secondary legislation authorises retroactive administrative control (7.3.1). Some instruments establish inter-state consultation requirements, without stating what the legal effects of a communication failure shall be (7.3.2). The example of ‘golden visas’ indicates that general principles may be relied upon in the absence of secondary legislation (7.3.3).

80 See Saša Beljin, ‘Rights in EU Law’ in Sacha Prechal and Bert van Roermund (eds), The Coherence of EU Law (OUP 2008) 91, 100–105; and Michael Fehling, ‘Eigenwert des Verfahrens im Verwaltungsrecht’ (2011) 70 Veröffentlichungen der Vereinigung Deutscher Staatsrechtslehrer 278. 81 See Maiani and Mogliorini (n 21) 7–44; Schwarz (n 2) chs 4–5; and Stephen Weatherill, ‘Why there is No “Principle of Mutual Recognition” in EU Law’ in Kai Purnhagen and Peter Rott (eds), Varieties of European Economic Law and Regulation (Springer 2014) 401. 82 See Joined Cases C-329/06 and C-343/06 Wiedemann and Funk EU:C:2008:366; and Joined Cases C-244/10 and C-245/10 Mesopotamia Broadcast EU:C:2011:607.

188 Administrative Dimension

7.3.1 Visas and databases: rules in secondary legislation Schengen visas are an evident example of transnational effects: they are issued by the consulate of one country and authorise entry and stay in the Schengen area as a whole. Article 34 Visa Code Regulation (EC) No 810/2009 states that a visa shall be revoked ex nunc or annulled ex tunc where there are serious grounds for believing that the conditions for issuing the visa are no longer met or had never been fulfilled, notably as a result of fraud. Revocation or annulment ‘shall’ usually be decided by the Member State which had issued the visa, although other countries ‘may’ do so as well, as highlighted by the Court in a complex case concerning criminal proceedings for illegal entry.83 Neither the legislation nor the judgment required the second Member State to consult the country that had issued the visa beforehand, instead requiring retroactive information. States shall be able to respond forcefully to alleged fraudulent behaviour, subject to legal remedies without suspensive effects.84 Schengen visas have far-reaching transnational effects, but Member States are in a strong position in terms of annulment and revocation. Inaccurate entries in the Schengen Information System (SIS) and the Visa Information System (VIS) can be a nuisance for third country nationals. It is important, therefore, that they are able to challenge wrong data. On paper, the SIS Regulation gives individuals much influence: they can access and, if need be, ask for the correction or deletion of personal data in the country of their choice. However, the authorities where the complaint was submitted cannot decide autonomously, as in the case of Schengen visas; rather, they shall consult with the country which had entered the information and whose position they are bound to follow.85 In essence, the second Member State serves as a ‘front office’ transmitting the complaint to the country that had issued the alert. In case of legal remedies, inter-state communication unfolds before the domestic court where the complaint had been filed; the issuing state undertakes to enforce the judgment, even if it was delivered abroad.86 The VIS Regulation differs insofar as individuals shall always address the state that entered the data; requests made elsewhere are forwarded to the country responsible.87 The solution defended by the legislature with regard to databases contrasts with a judgment on visa procedures. A Dutch consulate had rejected an application for a Schengen visa, based on information by Hungary and Germany that the applicant posed a public security threat.88 Judges in Luxembourg held that the Dutch court should limit oversight to formal aspects of the prior consultation procedure, while the substantive legality of the public policy alert had to be challenged in the country that had issued the alert.89 If several states did so, the third country national apparently has to file several lawsuits. This finding, on the part of the Court, was perplexing, since the Visa Code Regulation states explicitly that appeals ‘shall be conducted against the Member State that has taken the final decision’90. One might possibly rationalise the judgment through the specificities of public policy threats, which are often based on confidential information. Having said this, the ruling should be

83 84 85 86 87 88 89 90

See Case C-83/12 PPU Minh Khao Vo EU:C:2012:202, paras 38–40. See Visa Code Regulation (EC) No 810/2009, art 34(7). See SIS Border Checks Regulation (EU) 2018/1861, art 53(2). ibid art 54(2); in line with the former SIS II Regulation (EC) No 1987/2006, art 43. See VIS Regulation (EC) No 767/2008, arts 38–40. See Visa Code Regulation (EC) No 810/2009, arts 22, 32(1)(a)(vi). See Joined Cases C-225/19 and C-226/19 Minister van Buitenlandse Zaken (n 68) paras 50–52. Visa Code Regulation (EC) No 810/2009, art 32(3).

Transnational Effects of Horizontal Cooperation 189 criticised for contradicting supranational legislation and for ignoring what the Court had said on other scenarios of transnational legal effects. Roughly 100 countries are ‘blacklisted’ in Annex I of the Visa List Regulation (EU) 2018/ 1806, meaning that their nationals must have a visa before entering the Schengen area. However, many Member States do not have consulates in all regions, and it is common, therefore, to sign representation arrangements for specific countries. Thus, the Netherlands authorised the Swiss consulate in the Sri Lankan capital of Colombo to process applications for Schengen visas for which the Netherlands would normally be responsible.91 Article 8 Visa Code Regulation, in the version applicable at the time, foresaw two options for such bilateral representation arrangements. The Swiss consulate may act as a front office collecting documents to be forwarded to the Dutch authorities, or it may, alternatively, decide autonomously as a trustee (as it did in Colombo).92 To cut a long story short: the Swiss rejected the application for lack of means of subsistence,93 and the Court found that Swiss— not Dutch—courts were responsible for legal remedies.94 Seizing Swiss courts without any personal relations to the country will usually prove difficult for the applicants. One might possibly have defended a different outcome by arguing that Switzerland took the decision ‘on behalf of ’95 the Netherlands, although such an outcome would sit uncomfortably with the explicit legislative prescription that appeals shall be conducted against the country ‘that has taken the final decision on the application’.96 Complexity is an inevitable side effect of horizontal cooperation. Even so, the legislative provisions and Court judgments leave us with mixed results. There is a noticeable trend towards a ‘unitary model’ of legal remedies. Composite procedures do not translate into a ‘separation model’ that distinguishes neatly between the contributions of each country which individuals have to challenge separately (a solution the Court defended for entry alerts on public security grounds).97 Instead, transnational cooperation takes place behind the curtain, and individuals challenge the final outcome in the country whose authorities took the decision. Domestic courts are responsible for incidentally checking the input of other Member States. That preference for a unitary model holds important lessons for the legal oversight of the input of Frontex and the Asylum Agency in national decision-making, to be analysed in Chapter 8.

7.3.2 Entry bans: failure of inter-state consultation Procedural rules on inter-state consultation are rarely an inspiring read. Nevertheless, they can have profound implications for the rights of third country nationals, as the example of the Dublin III Regulation illustrates. Chapter 6.5.2 explained that the Court interpreted 91 ibid art 5(1)(a), with later amendments. 92 ibid art 8, which was changed by Regulation (EU) 2019/1155 amending the Visa Code [2019] OJ L188/25. 93 ibid art 32(1)(a)(iii). 94 See Case C-680/17 Vethanayagam EU:C:2019:62, paras 58–77, with additional comments on the ‘reliability’ of the Swiss as an associated Schengen country; see also Jean-Yves Carlier and Luc Lebœuf, ‘Droit européen des migrations’ [2020] Journal de droit européen 132, 134. 95 Visa Code Regulation (EC) No 810/2009,, art 8(1)(a), both the original text and the revised version. 96 ibid art 32(3). 97 See also Wolfgang Kahl, ‘Rechts-und Sachkontrolle in grenzüberschreitenden Sachverhalten’ in Josef Isensee and Paul Kirchhof (eds), Handbuch des Staatsrechts. Band XI (3rd edn, CF Müller 2013) § 253.

190 Administrative Dimension the time limits in the Dublin III Regulation to contain directly applicable individual rights, which asylum seekers can rely upon to challenge transfer to another country on formal grounds. A comparison with the case law on the European Arrest Warrant illustrates that a different outcome would have been possible: the expiry of time limits does not relieve states from carrying out the execution procedure.98 This leaves us with the question about the implications of the provisions on inter-state consultation in other migration law instruments. The issue is particularly relevant for entry bans. Chapter 16.4 on return will demonstrate that they oblige domestic authorities from across the Schengen area to refrain from allowing entry or issuing residence permits within the territory, thus effectively excluding the legalisation of irregular stay. In order to avoid conflicting decisions, an administrative authority that considers granting a residence permit shall consult with the country that had entered an alert into the SIS—vice versa if an entry ban is entered for someone holding a residence permit elsewhere. In these consultations, states shall coordinate their views to ensure that third country nationals are either subject to an entry ban or receive a residence permit.99 That is a reasonable solution, but the consultation requirement begs the question of what happens in cases of silence, when a country does not comply with the strict time limits for inter-state consultation under the new SIS Regulations. Intuitively, one might say that silence should be held against the uncooperative country, and this is exactly what the SIS Regulation stipulates in cases of regularisation whenever one country wants to override an entry ban. Silence does not prevent granting a residence permit.100 Does the same apply in the reverse scenario if, for example, a Nigerian national with a Spanish residence permit is to be banned from Finland after drug-related criminal convictions; provided the Spanish authorities do not respond to the consultation request by the Finnish authorities? The Court answered that question to the negative on the basis of Article 25(2) Convention Implementing the Schengen Agreement, which had stated expressly that silence should not be held against the third country national.101 However, this provision does not feature in Articles 38–30 SIS Border Checks Regulation (EU) 2018/1861 any longer. It is not self- evident, therefore, that the same conclusion would be confirmed in follow-up judgments, which might alternatively interpret silence as implicit consent. Irrespective of the answer judges would give in follow-up judgments, the case allows us to highlight an important element. The Court held, in this specific constellation at least, that transnational consultation requirements are directly applicable.102 That meant that the Nigerian national could rely on the communication failure before Finnish courts to object the return decision. Direct applicability had tangible benefits for individuals—a finding that cannot necessarily be generalised. We have to assess on a case-by-case basis whether an article embraces an individual right or not, as illustrated in Chapter 6.5 on the doctrinal foundations. By way of example, the Irish Supreme Court came to a negative conclusion for the Dublin III Regulation.103 For our purposes, the example of entry bans and the direct

98 See Case C-237/15 PPU Lanigan EU:C:2015:474, paras 34–42. 99 See SIS Border Checks Regulation (EU) 2018/1861, arts 27–30; SIS Return Regulation (EU) 2018/1860, arts 9–12; and Case C-193/19 Migrationsverket EU:C:2021:168, paras 30–37. 100 SIS Border Checks Regulation (EU) 2018/1861, art 27(c). 101 See Case C-240/17 E EU:C:2018:8, paras 38, 53–54. 102 ibid paras 57–59. 103 See Niovi Vavoula, ‘Information Sharing in the Dublin System’ (2021) 22 GLJ 391, 409–13.

Money as a Means of Governance 191 invocability of inter-state coordination demonstrates that seemingly dull procedural questions can have significant repercussions.

7.3.3 Intra-EU mobility: ‘golden visas’ as an abusive practice? In situations of intra-EU mobility the legislature thought it to be sufficient to lay down grounds for rejecting mobility, instead of dealing with transnational effects explicitly, like in the case of visas and databases. Member States where someone intends to relocate can apply these conditions, to be presented in Chapter 14.4.5 on legal migration, but they cannot usually check whether another country complied with Union law when delivering a residence permit that gives rise to mobility. What if national authorities willingly apply secondary legislation negligently? Chapter 15.7.3 on settlement will explain that national ‘golden visa’ and ‘golden passport’ schemes raise this question, if rich foreigners effectively purchase residence statuses. To be sure, mobility rights under the Long-Term Residents Directive 2003/109/EC come with strings attached: they require five years of legal and continuous residence. Few Chinese industrialists may want to live that long in Malta or Cyprus. That is why the Commission worries that Member States might apply the residence criterion negligently and proposes introducing a monitoring system at the national level to ensure compliance.104 Can other Member States refuse mobility on the basis of golden visas in the absence of such mechanism? The Long-Term Residents Directive is silent, apart from an abstract reference to mutual trust.105 This silence need not necessarily indicate that other Member States are generally bound to respect administrative decisions made elsewhere. That obligation may possibly be overcome under recourse to the general principle to prevent abuse, which will be discussed in Chapter 10.8 on overarching features. That is precisely what the Court did in the domain of social security coordination to give Member States some leeway, despite an explicit rule stating that certificates issued elsewhere must generally be recognised.106 This line of reasoning can possibly be applied to migration law, especially when the legislative framework for inter-state cooperation is loosely knit.

7.4 Money as a Means of Governance The European model of executive federalism entails that the costs of indirect implementation lie with the Member States. By way of example, the German taxpayer financed almost €10 billion for the reception of one million asylum applicants during 2015/16; the annual budget of the federal police, who deal with border controls, stands close to €5 billion. These

104 See Commission, ‘Report on Investor Citizenship and Residence Schemes in the EU’ COM(2019) 12 final, 10. 105 See Long-Term Residents Directive 2003/109/EC, recital 17; and Johannes Eichenhofer, ‘Das Recht auf Daueraufenthalt’ in Ferdinand Wollenschläger (ed), Enzyklopädie Europarecht. Band X (2nd edn, Nomos 2021) § 25 MN 35. 106 See Case C-359/16 Altun and others EU:C:2018:63, paras 46– 49; and Case C- 370/ 17 CRPNPAC EU:C:2020:260, paras 48–86.

192 Administrative Dimension exemplary figures illustrate that the indirect implementation of EU migration law can entail substantial costs at the national level (not only for Germany, of course). They also put into perspective the remarkable increase of the EU’s justice and home affairs budget. Funds for migration are scheduled to increase by almost 80 per cent during the current Multiannual Financial Framework (MFF) to reach €4.5 billion in 2027—not much less than the annual budget of the German federal police. EU funds can add value if they are more than a simple contribution to national budgets. They may ideally support forward-thinking investment in new technologies, human rights compliance, transnational cooperation, pilot projects for migrant integration, and inter-state solidarity. In doing so, EU funding may trigger innovation and change through investment in other domains than salaries for civil servants, which are often the biggest expenditure at the national level. That is precisely what the institutions have tried to do over the past twenty years—mirroring experiences with other policy areas. Not surprisingly, events of 2015/16 resulted in a substantial increase and readjustment of priorities. The situation at the external borders became the new top priority. This emphasis was an end in itself and promoted the solidarity with ‘frontline’ Member States, echoing the reference to ‘financial implications’ in Article 80 TFEU. To get a quick overview of justice and home affairs spending is comparatively easy, on the basis of the spending ceiling for continuous seven year periods. Negotiations on the MFF 2021–27 were characterised by the usual inter-state bargaining on the size of the overall budget, spending priorities, and national allotments. The final compromise required unanimity in the Council and the consent of the Parliament, which managed to negotiate a financial top-up for border management and Frontex, after the European Council had reduced—by more than 25 per cent—the amount for justice and home affairs in comparison to the Commission Proposal.107 On the basis of the MFF, the Council and the Parliament will agree on the annual budgets, for which the degree of flexibility is limited, since the legislative instruments for specific EU funds foresee distinct allocation criteria agreed upon in advance as an integral part of the overall compromise package. It is in the nature of such negotiations that governments ensure that they get a substantial piece of the ‘cake’. In doing so, they will have regard to the budget as a whole. Member States may, for instance, agree to receive less money for home affairs in return for higher spending on agriculture. Migration-related funding was concentrated under a single budget heading on ‘migration and border management’ on the occasion of the MFF 2021–27. Additional funds can be found in the headings for external relations and social cohesion. While the Asylum, Migration, and Integration Fund (AMIF) covers short-term needs upon arrival, the European Regional Development Fund and the European Social Fund+support migrant integration more generally.108 Such mainstreaming suits the cross-sectional character of integration policies, discussed in Chapter 15.5, provided that states actually use part of the money for migration-related projects.

107 See TFEU, arts 311 and 312; Commission, ‘Communication: A Modern Budget for a Union that Protects, Empowers and Defends’ COM(2018) 321 final, 14–15; European Council, ‘Conclusions’ (EUCO 10/20, 21 July 2020) Nos 101–08; and ‘EU’s Next Long-Term Budget & NextGenerationEU: Key Facts and Figures’ (Commission Factsheet, 11 November 2020). 108 Migrant integration— both short- term and long- term— had been covered primarily by the Asylum, Migration and Integration Fund (AMIF) and the European Integration Fund (EIF) under the MFF 2007–13 and the MFF 2014–20.

Money as a Means of Governance 193 Since 2015 the institutions have had recourse to the Civil Protection Mechanism, humanitarian aid funding, and a flexibility reserve to make extra funds available to support states like Greece.109 By 2020, almost €3 billion has been granted to the Greek government.110 In response to the war of aggression against Ukraine, Member States were authorised to channel several billion euros for regional development and migration which had not been spent under the previous MFF into refugee protection and integration.111 Additional resources were made available under the Civil Protection Mechanism. In addition to short-term needs, migration-related projects can be financed from the research budget for Horizon Europe within the programme line on ‘civil security for society’.112 Research may support, for instance, integrated surveillance and security technology used in border management, as mentioned in the chapters on agencies and databases in this volume. The external dimension is more difficult to grasp. It covers spending within the regular budget as well as special instruments like the Facility for Refugees in Turkey and the EU Emergency Trust Fund (EUTF) for Africa.113 Financial contributions to the Facility for Refugees in Turkey came from the EU budget, together with direct payments of the Member States, who contributed two-thirds of the overall amount of €6 billion for the period from 2016 to 2020.114 EUTF for Africa was set up after a summit in La Valetta in December 2015 and similarly rested on a complex legal construct with migration management as one of the priorities, which received substantial funds.115 While the bulk of the resources came from the EU budget, the Member States contributed additional money. Considerable amounts were channelled into neighbouring countries like Libya, Morocco, and Mali in the framework of the EUTF for Africa, which can be difficult to track for external observers. Similar extra-budgetary arrangements will be used in the future, for instance in the form of the Ukraine Solidarity Trust Fund for reconstruction. From 2022 onwards, the former trust funds have been integrated into the Neighbourhood, Development and International Cooperation Instrument (NDICI), which embraces more than €70 billion and describes migration as a new priority area; 10 per cent of the funds have been earmarked for migration-related projects.116 A separate Instrument for Pre-Accession Assistance (IPA III) finances projects in the Western Balkans and other accession candidates.117 Examples of what kind of international projects are funded can be found in Chapter 18 on the external dimension. Readers should recognise that it requires 109 See Zsolt Darvas, EU Funds for Migration, Asylum and Integration Policies (Study for the European Parliament, PE 603.828, 11 April 2018) 14–16, 47–52. 110 See ‘EU Financial Support to Greece’ (Commission Factsheet, November 2020). 111 See Regulation (EU) 2022/562 as regards Cohesion’s Action for Refugees in Europe (CARE) [2022] OJ L109/ 1; and Commission, ‘Communication: Temporary protection for those fleeing Russia’s war of aggression against Ukraine’ COM(2023) 140 final, 19–21. 112 See Regulation (EU) 2021/695 establishing Horizon Europe [2021] OJ L170/1, Annex IV, arts 4(1)(b)(iii), 12(2)(b)(iii), No 13. 113 See Leonhard den Hertog, ‘Money Talks: Mapping the Funding for EU External Migration Policy’ (CEPS Papers in Liberty and Security in Europe No 95, 2016); the European Development Fund (EDF) was integrated into the general budget from 2021 onwards. 114 See the contents of and the predecessor instruments to Commission Decision of 14 March 2018 on the Facility for Refugees in Turkey [2018] OJ C106/5; and the ‘Updated Common Understanding establishing a Governance and Conditionality Framework for the Refugee Facility for Turkey’ (Council doc 11329/18, 20 July 2018). 115 For a snapshot see Commission, ‘Progress Report on the Implementation of the European Agenda on Migration’ COM(2019) 481 final, 9–17. 116 See NDICI Regulation (EU) 2021/947, Annex III, recitals 50–51, arts 3(2)(c)(iv), 6. 117 See Regulation (EU) 2021/1529 establishing the Instrument for Pre-Accession assistance (IPA III) [2021] OJ L330/1, arts 3(2)(a), (3)(b), 4.

194 Administrative Dimension an advanced knowledge of budgetary rules and practice to identify how much is being spent for specific projects.118 The budget heading on ‘migration and border management’ under the MFF 2021–27 comprises the budgets of the migration agencies, Frontex and EUAA, as well as two distinct instruments called the Asylum, Migration, and Integration Fund (AMIF) and the Instrument for Border Management and Visa (BMVI). The latter vehicle is by far the biggest element of the Integrated Border Management Fund (IBMF), which covers customs cooperation as well. The AMIF and the BMVI have been allocated almost €10 billion and €7.3 billion respectively for the 2021–27 period. Funding is scheduled to increase annually. In 2027, the ceiling will reach almost €2 billion for the AMIF and €1.3 billion for the IMBF, on top of no less than €1.5 billion for the agencies, notably Frontex. In line with previous comments, the small print will be determined on an annual basis during budget negotiations among the institutions, which will translate the MFF baselines into funding for specific projects in accordance with the regulations on the funds. Spending targets for justice and home affairs may be modest in comparison to domestic budgets, but the growth is astonishing, nonetheless. When the former European Refugee Fund and the ARGO programme on administrative cooperation were first set up in the early 2000s, they covered €50 million for refugees and €5 million for administrative cooperation per annum.119 This translates into a staggering overall increase of the annual EU budget for migration of no less than 5,600 per cent between 2005 and 2027. This highly impressive surge started with four distinct funds for the period from 2007 to 2013 (altogether, €4 billion).120 Modest growth had initially been foreseen for two funds during 2014 and 2020 (jointly €6 billion),121 which gathered momentum through budgetary reallocation in the aftermath of 2015 (altogether, more than €11 billion).122 The legislative instruments establishing the funds anticipate the allocation between different envelopes and sub-programmes based on complex distribution keys of multiple variables.123 To do so is standard practice in budgetary negotiations. Nevertheless, EU institutions have started reserving more funds for flexible distribution as a lesson learned from the policy crisis of 2015/16. Part of the funds are foreseen for ‘thematic facilities’ instead of national portions. They are often administered by the Commission, while other resources

118 See also Thomas Spijkerboer and Elies Steyger, ‘European External Migration Funds and Public Procurement Law’ (2019) 4 European Papers 493. 119 See Council Decision 2000/596/EC establishing a European Refugee Fund [2000] OJ L252/12, art 2(1); Council Decision 2004/904/EG establishing the European Refugee Fund for the period 2005 to 2010 [2004] OJ L381/52, art 2(1); and Council Decision 2002/463/EC adopting an ARGO programme [2002] OJ L161/11, art 11(1). 120 The sum of €628 million was allocated under Decision 2007/573/EC establishing the European Refugee Fund [2007] OJ L144/1, art 12(1); €1.82 billion was foreseen by Decision 2007/574/EC establishing the External Borders Fund [2007] OJ L144/22, art 13(1); €€676 million was the financial foundation of Decision 2007/575/EC establishing the European Return Fund [2007] OJ L144/45, art 13(1); the sum of €825 million was mentioned in Council Decision 2007/435/EC establishing the European Fund for the Integration of Third-Country Nationals [2007] OJ L168/18, art 13(1). 121 €2.76 billion had originally been foreseen by Regulation (EU) No 515/2014 establishing, art 5(1), as part of the Internal Security Fund, the instrument for financial support for external borders and visas [2014] OJ L150/143; €3.137 billion was to be allocated under Regulation (EU) No 516/2014 establishing the Asylum, Migration and Integration Fund [2014] OJ L150/168, art 14(1). 122 See Darvas (n 109) 15. 123 See especially AMIF Regulation (EU) 2021/1147, art 10, Annex I; and BMVI Regulation (EU) 2021/1148, art 7, Annex I; for criticism see UNHCR/ECRE, ‘The Way Forward: A Reflection Paper on the New Proposals for EU Funds on Asylum, Migration and Integration 2021–2027’ (November 2018).

Summary 195 are usually handed out under shared management. Member States allocate them to specific projects on the basis of the common rulebook.124 In light of unfortunate experiences with fraud and mismanagement, the EU has put in place multiple safeguards to ensure compliance, including cumbersome reporting requirements. Most projects are subject to a co-financing rate of 25 per cent to prevent funding for the proverbial white elephants which lack local ownership and are not sustainable. Co- financing requirements can be loosened or renounced for specific projects, notably emergency assistance.125 Anyone trying to find the way through the EU’s budgetary rulebook should have the stamina to digest highly technical rules, which, nonetheless, are of great practical importance.

7.5 Summary Inspection of the administrative dimension highlights the bottom-up perspective, whose significance experts in EU law occasionally underestimate. The administrative dimension is not limited to the agencies, discussed in the next chapter, since the overwhelming majority of decisions is taken by domestic authorities without the involvement of any supranational actor. This is in line with the European tradition of executive federalism: Member States implement supranational legislation. Our comments revealed that seemingly ambitious concepts like the ‘area’ of freedom, security, and justice or a common asylum ‘system’ depend on national contributions. At the same time, administrative cooperation transcends the simple application of supranational rules by domestic authorities. Multiple actors serve as nodes in a network with multiple horizontal and vertical linkages. Articles 71–74 TFEU are testament to the significance of inter-state cooperation, which ranges from databases over operational cooperation to composite procedures. Secondary legislation often embraces procedural rules, notably on asylum. Yet, even complex rules leave states some leeway as to how to organise administrative procedures or court proceedings. Such leeway does not, however, mean unfettered freedom. The Court of Justice has recourse to the unwritten limits of procedural autonomy to exercise basic oversight of domestic practices. An in-depth analysis of the case law on migration demonstrated that the principle of effectiveness has gained relevance, for instance with regard to time limits. Effectiveness is increasingly supplanted by the fundamental right to an effective remedy in Article 47 CFR, which judges interpret in line with ECtHR case law, while adding additional guarantees. A specificity of Union law is the right to good administration under Article 41 CFR in relation to which judges are strict on matters concerning asylum, often on the basis of detailed secondary legislation; by contrast, assorted judgments on return illustrate that the Court accentuates countervailing public policy objectives that may justify limitations on the exercise of human rights. Transnational legal effects of administrative decisions receive little attention. They are governed by rules in secondary legislation in the case databases and visa applications, for instance when the domestic authorities or courts of one country decide on the legality of alerts entered into the SIS by another Member State. Judgments demonstrate that the Court

124 125

See generally Hofmann, Rowe, and Türk (n 14) ch 10. See AMIF Regulation (EU) 2021/1147, art 15; and BMVI Regulation (EU) 2021/1148, art 12.

196 Administrative Dimension favours a unitary approach, supporting the joint assessment of the legality of the different contributions, even though the complexity of both the legislation and case law preclude clear-cut conclusions. The same applies to the legal consequences of transnational communication failures on entry bans, which exemplify the importance of highly technical procedural rules for the legal status of migrants. Intra-European mobility demonstrated that even seemingly straightforward rules on the recognition of administrative decisions could possibly be subject to limitations when a first Member State issues a ‘golden visa’ or long-term residence permit in disregard of EU law.

8

Agencies (Frontex and Asylum Agency) While the brunt of the responsibilities for implementing migration law rests with domestic authorities, the supranational level increasingly contributes administrative resources of its own. Frontex and the Asylum Agency have witnessed exponential growth in terms of stronger mandates, operational capacities, and practical reach, thus raising formidable challenges of political accountability and legal oversight. To find solutions can be challenging for the simple reason that the agencies transcend traditional models of EU integration by means of legislative harmonisation and horizontal inter-state cooperation. Frontex and the Asylum Agency are building up human and technical resources of their own, to be exercised together with domestic authorities. The relative weight of the supranational actors can be difficult to disentangle. To do so requires a high degree of practical expertise in everyday practices on the ground. Agencies are not simple instruments of top-down enforcement. Their organisational set-up, operational powers, and practical functioning transcends simple trajectories of centralisation. The migration agencies mirror the characteristics of the European administrative space as a conglomerate of diverse actors at multiple levels, as described in Chapter 7. Frontex and today’s Asylum Agency have traditionally served as a clearing house for inter- state cooperation by means of seconded national personnel, which will be complemented by a standing corps of supranational officers both agencies have started hiring. These impressive operational capacities and involvement in national decision-making raise administrative cooperation to a new level, to support Member States under pressure. At the same time, the role of the agencies remains intricately connected to the model of decentralised implementation. We shall see that this composite character is the main reason why a satisfactory answer has not yet been found for how to guarantee effective legal oversight by courts or through other accountability mechanisms. In political and academic debates, Frontex and the Asylum Agency often assume a prominent position. Such prominence is fully justified, provided that we do not mistake stronger agencies as straightforward centralisation. Unequivocal federalisation is supported by neither the policy debate (8.1) nor the EU Treaties (8.2). For many years, the agencies had comparatively meagre operational powers, until the crisis of asylum policy during 2015/16 led to several upgrades of the mandates, capacities, and practical tasks. Frontex and today’s Asylum Agency turned into powerful actors whose precise role on the ground is notoriously difficult to grasp. While the Frontex Regulation has been amended several times (8.3), the official transformation of the former Asylum Support Office into the Asylum Agency was delayed. The resulting mismatch between statutory powers and operational activities was overcome when the new mandate came into force in 2022 (8.4). The composite character of administrative decision-making complicates the design of accountability mechanisms at the domestic and supranational levels (8.5).

European Migration Law. Daniel Thym, Oxford University Press. © Daniel Thym 2023. DOI: 10.1093/oso/9780192894274.003.0009

198 Agencies (Frontex and Asylum Agency)

8.1 Theory and Policy Design Political support for stronger agencies is a recurrent feature in the policy debate in Brussels and in national capitals. This political support was essential for the continuous promotion of both Frontex and the Asylum Agency, which, nevertheless, should not be equated with the simple centralisation of decision-making (8.1.1). A reminder of the comparatively modest motivation behind the creation of the border agency and the former Asylum Office (8.1.2) illustrates how impressive the increase in tasks and functions has been in recent years (8.1.3). Both Frontex and the Asylum Agency are witnessing a period of exponential growth, which accelerated after the asylum policy crisis of 2015/16 and raises challenges of managerial adaptation and human rights compliance (8.1.4).

8.1.1 Elusive vision of administrative centralisation In the English language, the term ‘federal’ is usually associated with centralisation. This mirrors the American tradition where federal legislation is being applied by federal authorities and courts. In the United States, the regions will not usually apply federal laws. By contrast, the EU replicates the German model where the term ‘federal’ can equally accentuate regional differences. Institutions at the central level concentrate on law-making and the creation of apex courts for uniform interpretation, whereas administrative implementation and judicial oversight by courts of first and second instance are left to the regions. Regional authorities and courts must apply common rules, and the central institutions will establish administrative offices throughout the territory in exceptional cases only. The EU follows this model of ‘cooperative’ or ‘executive’ federalism.1 Political actors and academic observers occasionally call for an act of liberation from the complexities and weaknesses of decentralised implementation. Agencies have always enthused abstract aspirations of administrative ‘centralisation’ and ‘federalisation’, especially among those who are generally enthusiastic about the European project. Administrative capacities at the supranational level are a symbol of further integration, as indicated by ‘Europol’ or the ‘European Public Prosecutor’. Their official designation hints at a high level of integration, despite the absence of far-reaching powers. We shall see that the situation is similar for Frontex and the Asylum Agency; powers on the ground often lack behind abstract declarations of intent. The centralisation of asylum decision-making was officially put on the agenda as an option by the Commission in 2016, although this idea was not actively pursued further in the years thereafter.2 Frontex has been a projection sphere for Europhile calls for administrative centralisation ever since the early days of the area of freedom, security, and justice. The abstract vision of a ‘European Border Police’3 came up in the policy debate surrounding the Constitutional Treaty, which paved the way for the introduction of today’s Article 77(2)(d) Treaty on the 1 See Robert Schütze, From Dual to Cooperative Federalism (OUP 2009) chs 1–2. 2 See Commission, ‘Communication: Towards a Reform of the Common European Asylum System and enhancing legal avenues to Europe’ COM(2016) 197 final, 8–9; see also Harald Dörig and Christine Langenfeld, ‘Vollharmonisierung des Flüchtlingsrechts in Europa’ [2016] Neue Juristische Wochenschrift 1. 3 See ‘Feasibility Study for the Setting up of a European Border Police’ (Final Report, 30 May 2002) https://cmr. jur.ru.nl/cmr/docs/61.pdf (accessed 1 March 2023).

Theory and Policy Design 199 Functioning of the European Union (TFEU).4 The related notion of a ‘European Border Corps’ was introduced in the 2010s, as shorthand for different models of administrative integration ranging from mixed administrative settings to the ‘full integration’ of a uniform institution with regional offices and local branches throughout the Union.5 The Commission took up these long-standing reform options during the policy crisis of 2015/16. A stronger mandate for Frontex was adopted within a couple of months during 2016 to signal political determination at a time when the structural weaknesses of EU asylum policy had become apparent. The degree of reform initially hardly justified the grand new name ‘European Border and Coast Guard’ (EBCG), which hints at a high level of centralisation.6 Notwithstanding this initial limitation, the ongoing introduction of supranational administrative capacities is gradually turning Frontex into an executive actor in its own right. Moreover, the official designation is less straightforward than one might think intuitively. The official definition of the EBCG comprises both the supranational agency Frontex and national authorities responsible for border management.7 The ECBG is based, in other words, on a composite structure and does not signal the objective of supranational border guards replacing national officers. We should distinguish the integrated overall structure from the specific powers of the border agency. That is one reason why it is preferable to continue using the acronym ‘Frontex’ (for French frontières extérieures), in line with the practice of the institution itself.8 To do so ensures the retention of a ‘brand name’ critics prefer as well. It also counters the impression that the Warsaw-based agency brings about fully fledged administrative centralisation, as the official title might be read to suggest in an exercise of semantic overstretch. Important reasons as to why the political vision of administrative centralisation is bound to remain a pipe dream are the operational character of EU migration law and the composite structure of the European administrative space described in the previous chapter. Frontex and the Asylum Agency may increase collaboration with domestic authorities and could possibly even take certain decisions themselves. By contrast, it would take years if not decades to organise a truly federal border police or asylum agency deciding individual cases across the Union. Legal oversight would have to be guaranteed by the General Court, or specialised tribunals, possibly through decentralised chambers hearing and deciding about legal remedies at the external borders, as described in Chapter 3.1.3 on the judicial architecture. Such centralised decision-making on literally tens of thousands of individual cases would be tremendously challenging to realise in practice. The EU is compelled, therefore, to continue the precarious and complex path of a composite administrative setting involving domestic authorities and supranational agencies.

4 See Daniel Thym, ‘Legal Framework for Entry and Border Controls’ in Daniel Thym and Kay Hailbronner (eds), EU Immigration and Asylum Law. Article-by-Article Commentary (3rd edn, CH Beck/Hart Publishing/ Nomos 2022) MN 19–21. 5 See Unisys, ‘Study on the Feasibility of the Creation of a European System of Border Guards to Control the External Borders of the Union’ (16 July 2014) 21–38. 6 On the initial mismatch see Philippe De Bruycker, ‘The European Border and Coast Guard: A New Model Built on an Old Logic’ (2016) 1 European Papers 559. 7 See Frontex Regulation (EU) 2019/1896, art 4. 8 ibid recital 3 mentions the informal name.

200 Agencies (Frontex and Asylum Agency)

8.1.2 The choice for Frontex and the Asylum Office Articles 71–74 TFEU bear witness of how important transnational operational cooperation is for justice and home affairs. Inter-state coordination initially took place in the framework of the existing legislation, inter-state cooperation, and financial support.9 These activities prepared the ground for Frontex and the former European Asylum Support Office (EASO). Their creation was an integral part of the ‘agencification’ of administrative law, beyond the earlier focus on indirect implementation and horizontal cooperation.10 Europol was established in 1998 as the first justice and home affairs agency, Frontex began its work in 2005, and EASO followed in 2010. The decentralised location reiterates the focus on transnational coordination: Europol has its seat in The Hague, Frontex in Warsaw, and the Asylum Agency in Valletta. There were various reasons for the creation of the agencies. To start with, they were symbols of European integration and signalled, to sceptical public opinion, that the EU institutions were determined to counter practical deficits.11 Critics and supporters mutually stabilised the impression that Frontex, in particular, was a powerful actor, notwithstanding the initial mismatch between the ambitious designation and the reality of meagre operational powers. Over the years, however, the mandate, technical infrastructure, and operational tasks saw unprecedented growth, thus turning Frontex and the Asylum Agency into veritable operational powerhouses and fascinating objects of academic study and political scrutiny. The case for Frontex had always been strong. External border controls are carried out in the collective interest.12 Frontex was a potential solution to the ‘Schengen dilemma’ of mutual dependency, with its built-in risk of reciprocal accusations: inbound Member States lament the negligent control practices of outbound states, which, by contrast, complain about a solidarity deficit in the face of a common task: ‘efficient monitoring of the crossing of external borders’.13 Germany was among the first to call for a supranational border police force as early as 1989 to give governments in the middle of the continent indirect control over the continent’s external borders.14 Italy and other countries at the external borders supported the project in order to share this (costly) task. Italy commissioned the feasibility study which prepared the ground for the institutional design of Frontex, thus confirming our earlier finding that justice and home affairs thrived thanks to governmental support.15 The initial Frontex Regulation (EC) No 2007/2004 was negotiated in less than a year to be adopted immediately after the enlargement of 2004. Again, the motivation of indirect

9 See Ferruccio Pastore, ‘Visas, Borders, Immigration’ in Neil Walker (ed), Europe’s Area of Freedom, Security and Justice (OUP 2004) 89, 120–27. 10 See Eduardo Chiti, ‘Decentralisation and Integration into the Community Administrations’ (2004) 10 ELJ 402. 11 See ch 4.2.3; and Jorrit J Rijpma, ‘Hybrid Agencification in the Area of Freedom, Security and Justice and Its Inherent Tensions’ in Madalina Busuioc and others (eds), The Agency Phenomenon in the European Union (Manchester UP 2012) 84, 90–91. 12 See Schengen Borders Code Regulation (EU) 2016/399, recital 6. 13 TFEU, art 77(1)(b). 14 See Mechthild Baumann, Der deutsche Fingerabdruck (Nomos 2006) 124. 15 See ch 2.1; and ‘Feasibility Study’ (n 3); besides Member States, the Commission supported the initiative and developed it further.

Theory and Policy Design 201 control played a role, since the border agency was meant to assist the transformation of the border police of the new members to perform their future tasks as ‘frontline’ countries.16 The choice for today’s Asylum Agency followed more conventional rationales, mirroring the experience with other agencies. The Asylum ‘Support’ Office was primarily a source of information gathering, technical expertise, training, and quality improvement for the decentralised application of asylum laws.17 EASO was supposed to add value by serving as a catalyst for mutual learning and better quality in the collective interest.18 Stark discrepancies in the recognition rates for different countries of origin, mentioned in Chapter 13.5.1 on asylum, demonstrated that supranational legislation was not enough to achieve administrative convergence on the ground. Moreover, today’s Asylum Agency was an instrument of support and gentle tutoring in the setting up of functioning asylum systems in states in southern and eastern Europe with little or no experience as asylum destination countries. In recent years, the function as an instrument of solidarity with countries at the external borders has gained momentum.

8.1.3 Formal expansion of the mandates The choice for a network approach, instead of administrative centralisation, meant that Frontex initially served as a facilitator of transnational cooperation. Officers from different countries with diverse equipment and administrative traditions cannot simply be instructed to cooperate effectively. Doing so requires the coordination of technical standards, training, contact points, and the like. In the first years, Frontex was very much preoccupied with establishing such network structure, which prepared the ground for operational cooperation and was supported financially by EU funds.19 Information exchange and risk analysis also featured among the tasks, although they played a lesser role than for EASO.20 The original tasks of Frontex were a far cry from the operational powers and budgetary means of today’s border agency. An important legal change came in 2007 when the Council authorised the formation of Rapid Border Intervention Teams, which became known under the misplaced acronym RABITS, to support states facing exceptional pressure.21 They marked the beginning of the dynamic evolution of legal rules and practical standards for joint operations in the years to come.22 A first amended was adopted in 2011 to establish some preliminary powers in the areas of return and international cooperation, together with tentative human rights

16 See Marek Adamczyk, ‘Balancing Openness with Control’ in Joanna Apap (ed), Justice and Home Affairs Law in the EU (Edward Elgar Publishing 2004) 281; and David Fernández Rojo, EU Migration Agencies (Edward Elgar Publishing 2020) 27–30. 17 See generally Merijn Chamon, EU Agencies (OUP 2016) ch 3. 18 See Elspeth Guild, ‘Does the EU Need a European Migration and Protection Agency?’ (2016) 28 IJRL 585; and Rijpma (n 11) 87–90. 19 See ‘Plan for the Management of the External Borders of the Member States’ (Council doc 10019/02, 13 June 2002) 10–19; and Commission, ‘Report on the evaluation and future development of the Frontex Agency’ COM(2008) 67 final, 3–5. 20 See the first Frontex Regulation (EC) No 2007/2004, art 4. 21 See Regulation (EC) No 863/2007 establishing a mechanism for the creation of Rapid Border Intervention Teams [2008] OJ L199/30. 22 For a reliable summary see Fernández Rojo (n 16) 36–38, 49–57, 64–76, 93–96; and Melanie Fink, Frontex and Human Rights (OUP 2018) 35–79.

202 Agencies (Frontex and Asylum Agency) safeguards.23 A major reform came with the revised Frontex Regulation (EU) 2016/1624, which expanded the operational tasks across the board and introduced supervisory powers vis-à-vis the Member States. Today’s Frontex Regulation (EU) 2019/1896 went further, by initiating the formation of a standing corps of supranational officials and by enhancing the powers in the areas of border controls, return, and international cooperation. Human rights supervision was also strengthened. In contrast to Frontex, the former EASO Regulation (EU) No 439/2010 was not changed initially. Nevertheless, the agency saw substantial growth after 2015. The Commission had proposed to transform the Support Office into an Asylum Agency as early as 2016. There was broad agreement that this should be done, and the agency was given substantial more powers de facto, meaning that its operational activities were occasionally out of sync with the legal framework.24 An interinstitutional agreement on the new mandate was partially agreed upon in the trilogue format in 2017 but fell victim to the political stalemate over asylum policy.25 A second and amended proposal was tabled by the Commission one year later. Again, there was a consensus that the agency should be strengthened, although the adoption of the new EUAA Regulation (EU) 2021/2303 was postponed as a result of the ‘package approach’ to asylum reform, presented in Chapter 13.1.5. The revised mandate finally entered into force in January 2022.

8.1.4 Exponential growth in practice Both Frontex and the Asylum Agency are witnessing a period of exponential growth, which accelerated after the asylum policy crisis of 2015/16. The reform measures built upon the initial design and added new functions. A defining feature of both Frontex and the Asylum Agency is the operational character of many activities, in contrast to the regulatory focus of other EU agencies.26 Comments throughout this chapter will highlight how supranational staff and seconded national personnel participate in national decision-making and cooperate with third states. These enhanced operational powers have gathered momentum over the years. Formal expansion of the mandates was supported by a substantial increase in funding. The agencies’ growth was dynamic and at times exponential, thus posing formidable challenges of managerial adaptation and absorption. Basic numbers outline the level of ambition. The budget for Frontex was almost trebled from €140 to €350 million between 2015 and 2020; it is expected to treble again to reach almost €1.4 billion per annum in 2027. EASO’s budget increased exponentially from €15 to more than €100 million until 2020 and is scheduled to stand at roughly €160 million per annum in 2027.27 These figures are 23 See Regulation (EU) No 1168/2011 amending the initial Frontex Regulation (EC) No 2007/2004 [2011] OJ L304/1. 24 For an early description see Sergio Carrera, Leonhard den Hertog, and J Parkin, ‘The Peculiar Nature of EU Home Affairs Agencies in Migration Control’ (2013) 15 EJML 337, 341–44. 25 See ‘State of play and guidance for further work’ (Council doc 10555/17, 27 June 2017). 26 See Thomas Groß, ‘Die Kooperation zwischen europäischen Agenturen und nationalen Behörden’ (2005) 40 Europarecht 54, 56–63; and Fernández Rojo (n 16) ch 1. 27 See ‘Spending Ceiling Multiannual Financial Framework 2021–2027 (in commitments) –Current prices’ https://ec.europa.eu/info/publications/multiannual-financial-framework-2021-2027-commitments_en (accessed 1 March 2023).

Constitutional Foundations 203 impressive but indicate underlying risks as well. Such a rapid build-up requires human resources, managerial skills, and operational expertise that money alone cannot buy.28 Some of the difficulties may be the result of institutional overstretch trying to achieve too much too fast. ‘Hotspots’ on the Greek islands have become an epitome of operational and structural challenges. The lacklustre performance and partial failure of the hotspot approach reiterates the significance of managerial adaptation. To develop prototypes for operational cooperation ‘learning by doing’ in the geographic periphery would never have been easy, but the administrative challenge was multiplied by institutional stress, political pressure, and humanitarian imperatives accompanying the venture. Pushback allegations against national border guards in the context of Frontex operations and accusations of mismanagement reinforce the message that political scrutiny and legal oversight will play an essential role in the years to come.

8.2 Constitutional Foundations The dynamic evolution of operational powers de facto and de jure requires thorough inspection of the constitutional limits for the transfer of further competences, especially with regard to the involvement of Frontex and the Asylum Agency in deciding individual cases (8.2.1). The Meroni doctrine is an important constitutional yardstick for the regulatory powers of EU agencies; however, the rationale behind the case law turns out to be largely irrelevant for agencies with a predominantly operational mandate (8.2.2). Finally, readers are reminded of the status of Ireland, Denmark, and associated Schengen members (8.2.3).

8.2.1 Involvement in administrative decision-making Several provisions in the EU Treaties confirm that the supranational level should concentrate on legislative harmonisation, while decisions affecting individuals are normally taken by domestic authorities.29 Treaty rules for the area of freedom, security, and justice similarly endorse national prerogatives: residence permits and visas shall be ‘issue[d]‌by Member States’ and one country shall be designated to have jurisdiction for considering asylum applications.30 Chapter 7.1.1 explained that neither the idea of a common asylum ‘system’ nor reference to an asylum status ‘valid throughout the Union’ requires the abolition of national asylum systems. Articles 71–74 TFEU explicitly call for effective horizontal cooperation among the Member States, thus presupposing the co-existence of distinct national administrations. Such constitutional preference for indirect implementation need not, however, exclude a greater role of the agencies. To start with, the Court interprets the term ‘measure’, used in Articles 77(2), 78(2), and 79(2) TFEU, to cover the creation of agencies. It even allows for the conferral of

28 For a snapshot see Frontex, ‘State of Play of the Implementation of the EBCG 2.0 Regulation in View of Current Challenges’ (Council doc 7607/20, 28 April 2020). 29 See TFEU, art 291; and TEU, arts 4(2), 5(1), 19(1)(2). 30 See TFEU, arts 79(2)(a) and 78(2)(e).

204 Agencies (Frontex and Asylum Agency) administrative decision-making powers to supranational agents, at least when states do not fulfil certain functions in a satisfactory manner.31 Article 80 TFEU supports such an interpretation in so far as the activities of the agencies contribute to inter-state solidarity. Recourse to Articles 77–80 TFEU has the advantage that the ordinary legislative procedure applies; there is no need to activate Article 74 TFEU as an additional legal basis, mirroring the example of the first Frontex Regulation (EC) No 2007/2004 and the former EASO Regulation (EU) No 439/2010. To recognise that the EU Treaties authorise the creation of the migration agencies does not, however, end our inquiry. When considering the detail, we have to distinguish between the legal bases for border controls and asylum. With regard to the former, Article 77(2)(d) TFEU authorises ‘the gradual establishment of an integrated management system for external borders’. This open-ended formulation takes up the idea of ‘integrated border management’, which had been discussed widely in the early 2000s. The precise meaning of integrated border management fluctuated between structural cooperation across policy areas (eg security, migration, customs) and institutional measures.32 One step further, the drafting history of Article 77(2)(d) TFEU confirms that the provision concerns institutional cooperation revolving around the border agency Frontex.33 The ambitious reference to ‘any measure necessary’ does not contain a precise vision but signals a certain flexibility. It was the desire of the European Convention to lay the foundations for the progressive development of Frontex, without stating clearly what this might entail. Against this background, we may conclude that legislation adopted on the basis of Article 77(2)(d) TFEU may move beyond the original policy design. The wording and drafting history, with its explicit reference to Frontex, argue against a direct comparison with the legal bases for Europol and Eurojust.34 Precise limits are difficult to draw in the abstract, but it seems to me that a realistic solution would be to recognise that Frontex can be given autonomous decision-making powers on selected subject matters or with regard to specific situations, while a genuine federal border guard substituting national units would require Treaty change.35 Below that threshold, there is room for further integration transcending the status quo. By contrast, the Asylum Agency does not rest on an authorisation in the Treaties. Its legal basis is the abstract call for ‘measures’ on common procedures, asylum jurisdiction, and reception conditions in Article 78(2)(d)–(f) TFEU, possibly read in conjunction with Articles 74 and 80 TFEU.36 Taken together, these provisions undoubtedly authorise the former Support Office and of the new Asylum Agency. They may even be interpreted as to approve involvement in administrative decisions on individual cases at the national level, or the exceptional direct enforcement of the supranational legislation in line with the case law 31 See Case C-270/12 United Kingdom v Parliament and Council EU:C:2014:18, paras 97–115; and Roman Lehner, ‘Rechtliche Möglichkeiten zur Schaffung einer EU-Asylbehörde’ in Roman Lehner and Friederike Wapler (eds), Die herausgeforderte Rechtsordnung (BWV 2018) 183, 190–96. 32 See Violeta Moreno-Lax, Accessing Asylum in Europe (OUP 2017) 28–40; and Herbert Rosenfeldt, Frontex im Zentrum der Europäischen Grenz-und Küstenwache (Mohr Siebeck 2021) 46–50. 33 See Jörg Monar, ‘Die Vertragsreformen von Lissabon in den Bereichen Inneres und Justiz’ [2008] Integration 379, 385; against Roberta Mungianu, Frontex and Non-Refoulement (CUP 2016) 22–24. 34 Contra Jorrit J Rijpma, ‘Frontex and the European System of Border Guards’ in Maria Fletcher and others (eds), The European Union as an Area of Freedom, Security and Justice (Routledge 2017) 217, 236–37. 35 See Thym (n 4) paras 19–20a. 36 These latter may be relevant for information exchange and training (supported by Article 74 TFEU) and involvement in decision-making at the external borders (in line with Article 80 TFEU).

Constitutional Foundations 205 mentioned previously. That being said, the argument in favour of supranational decision- making remains an uphill struggle, since primary law is less ambitious than for Frontex. The legislature benefits from a certain leeway to develop a robust asylum agency going beyond the status quo, although the level of ambition is constitutionally bound to stay behind the powers of Frontex.37

8.2.2 ‘Meroni’ doctrine: no decisive hurdle Constitutional limits for operational powers embrace the Meroni doctrine, which should be distinguished from the scope of Union competences. Competences serve as the first hurdle, since measures not covered by primary law cannot be adopted by the EU institutions. The Meroni doctrine, named after an early Court judgment, enters the equation once the first hurdle has been cleared. It limits the power of the EU institutions to delegate certain matters to agencies. The position of the Court of Justice on the delegation of decision-making power has changed over the decades in response to the evolving characteristics of the supranational agencies and in light of the Treaty of Lisbon.38 In essence, the contemporary Meroni doctrine allows for the adoption of administrative decisions affecting individuals whenever the powers delegated to an agency are precisely delineated and amenable to judicial review.39 Application of secondary legislation on the basis of technical expertise can usually be delegated, as long as delegation does not result in what one might call ‘political’ discretion.40 Such ‘technical’ application of the law cannot be distinguished from ‘political’ choices easily, also considering that the discussion so far has focused on regulatory agencies, not operational powers. On the whole, the operational functions entrusted on the Asylum Agency and Frontex are probably compatible with the Meroni doctrine with regard to participation in individual decisions. A different conclusion might possibly be defended for decisions about whether to launch operations and perform risk assessments,41 even though the agencies’ governance structure reduces the freedom of manoeuvre on such politically sensitive matters. In the case of both Frontex and the Asylum Agency, the Management Board, the Council, the Commission, and the European Parliament participate in programming and in decisions having budgetary implications.42 That involvement renders it difficult to conclude that there is ‘political’ leeway. When it comes to the participation in administrative decision-making, it would not enough, in light of the case law, to find that the agencies have factual leeway (or ‘discretion’43), 37 On different options see Evangelia (Lilian) Tsourdi, ‘The Emerging Architecture of EU Asylum Policy’ in Francesca Bignami (ed), EU Law in Populist Times (CUP 2020) 191, 212–17; and Lehner (n 31) 197–204. 38 See generally Paul Craig, EU Administrative Law (3rd edn, OUP 2018) 168–71; and Chamon (n 17) ch 4. 39 See United Kingdom v Council and Parliament (n 31) paras 44–54, 63–68; which modernised the earlier findings in Case 9/56 Meroni EU:C:1958:7; and Case 98/80 Romano EU:C:1981:104. 40 See Carl Fredrik Bergström, ‘Shaping the New System for Delegation of Powers to EU Agencies’ (2015) 52 CML Rev 219. 41 See Rosenfeldt (n 32) 142–74; and Rijpma (n 11) 93–94. 42 See Frontex Regulation (EU) 2019/ 1896, arts 102, 115– 16; and EUAA Regulation (EU) 2021/ 2303, arts 42, 52–53. 43 On the polysemy of the notion of discretion see Jonas Bornemann, The Transformation of Discretionary Power in European Migration Law (PhD Thesis, University of Konstanz 2022); and Marta Simoncini, Administrative Regulation beyond the Non-Delegation Doctrine (Bloomsbury/Hart Publishing 2018) ch 3.

206 Agencies (Frontex and Asylum Agency) Table 8.1 Participation in the most relevant entry and border control instruments Instrument

Ireland /UK (before Brexit)

Denmark

Schengen? Accession (NO, IS, CH, LI)

Frontex Regulation (EU) 2019/1896

no

no (int. law)

yes

yes

EUAA Regulation (EU) 2021/2303

no

no (int. law)

yes

yes

for instance when agency staff prepares an inadmissibility decision in response to an asylum application or whenever Frontex personnel refuses the entry of a third country national in line with comments that will follow. One would have to demonstrate, rather, that doing so involves policy choices which go beyond the application of secondary legislation.44 Such a conclusion is almost impossible to defend given the level of detail of the legislative rulebook and the guarantee of legal remedies before domestic courts.45 One would have to develop, instead, a new model of constitutional limitations for operational activities which no-one has seriously tried to put forward so far. Involvement in individual decisions raises complex questions of legal oversight, but it does not involve political flexibility on the basic layout of EU migration law.

8.2.3 Territorial scope Participation in Frontex presupposes membership of the Schengen area. Table 8.1 illustrates that associated countries participate fully, including the right to vote in meetings of the management board. Chapter 2.4.2 explained that the Court rejected a British attempt to join without abolishing internal border controls. Ireland does not, as a result, participate either, although it may cooperate with Frontex.46 The special status of Denmark reflects its half-hearted Schengen membership, as described in Chapter 2.4.1 on the opt-outs. New Member States participate immediately after accession, to prepare national border guards for full Schengen membership.47 All Member States without an opt-out fully participate in the Asylum Agency. Associated countries can join as observers and participate in the operational activities, provided they contribute to the budget.48

44 See Édouard Dubout, ‘Les enjeux constitutionnels du pouvoir de substitution de l’agence Frontex’ [2017] Revue trimestrielle de droit européen 457, 469–73. 45 See Fernández Rojo (n 16) 171–74. 46 See Frontex Regulation (EU) 2019/1896, art 70. 47 See ch 12.3.1; and, by way of example, Act concerning the conditions of accession of the Republic of Croatia [2012] OJ L112/21, Annex II, art 4(1), No 8. 48 See EUAA Regulation (EU) 2021/2303 recital 65, art 34.

Frontex: Institutional Muscle Formation 207

8.3 Frontex: Institutional Muscle Formation Frontex often assumes a prominent position in public debates. The border agency is, much more than the sister agency for asylum, a powerful symbol of the move towards ‘fortress Europe’ for some and evidence of ‘a Europe that protects’ for others.49 Operational powers are the hallmark of Frontex activities, ranging from joint operations at the external borders to involvement in individual decisions (8.3.1). Politically sensitive are supervisory functions, including the option of compulsory intervention in Member States under pressure (8.3.2). Cooperation with third states is set to become a central arena in the future (8.3.3), together with the purchase of technical equipment and advanced technological border surveillance (8.3.4).

8.3.1 Operational powers Operations are widely discussed. The first sea operation ‘Hera’ started off the Canary Islands in 2006 where many ships with migrants from Western Africa arrived during that period. The central Mediterranean became the focal point thereafter, with the Frontex missions ‘Triton’ and, later, ‘Themis’ superseding the Italian operation called ‘Mare Nostrum’ from 2014 onwards.50 These operations made headline news for their involvement in search and rescue, which brought several hundred thousand refugees to Europe and did not prevent more than 10,000 deaths; the central Mediterranean earned the dubious reputation as the deadliest migration route in the world. Some additional information on these controversial Frontex missions can be found in Chapter 18.3.4 on externalisation; international and human rights standards will be discussed in Chapter 12.2 on border controls. Smaller operations took place at land borders in parallel, for instance between Greece and Turkey on the banks of the river Evros. At the time of writing, three main operations were ongoing: ‘Themis’ in the Central Mediterranean; ‘Poseidon’ in the Aegean Sea; ‘Minerva’ and ‘Indalo’ near the Spanish mainland and the Canary Islands; as well as several smaller missions, for instance in Lithuania. Generally speaking, the new mandate distinguishes between ‘joint operations’ for the deployment of coastguard ships from several Member States, ‘rapid border intervention’ involving seconded national staff or supranational personnel in administrative decisions, and ‘migration management support teams’ in scenarios of mixed migratory flows.51 While Frontex operations may last for longer periods, rapid border intervention and support teams are confined, in theory at least, to exceptional scenarios of disproportionate challenges. From an institutional perspective, Frontex missions are governed by an operational plan (OPLAN), which is confidential and drawn up by the executive director, in cooperation with the host country and the fundamental rights officer.52 Much less visible than the deployment of equipment or personnel at the external borders are the activities in the field of return. Agency powers have increased significantly over

49

On the public discourse see ch 1.2.6.

51

See Frontex Regulation (EU) 2019/1896, arts 36–40. ibid arts 38, 109(2)(e).

50 See https://en.wikipedia.org/wiki/Operation_Mare_Nostrum (accessed 1 March 2023). 52

208 Agencies (Frontex and Asylum Agency) the years, both in terms of regulatory leverage and operational capabilities. While the decision whether someone should be returned remains the prerogative of the Member States, Frontex may support voluntary and forced return in practice. Contributions encompass, amongst others, the organisation and coordination of return operations, in particular flights, the exchange of best practices, financial support, the acquisition of travel documents, and the provision and training of return monitors.53 We can expect return activities to grow further precisely because they are less visible. Originally, Frontex served primarily as a clearing house and facilitator for national contributions. In the future, the agency will deploy many resources of its own. It purchases equipment and has started setting up a ‘standing corps’. The standing corps is scheduled to comprise 10,000 staff in the year 2027, with 3,000 supranational officers, employed by Frontex directly, and 7,000 national staff for short-term and long-term secondment.54 This marks an important innovation, with Frontex hiring officials to be deployed, as federal border guards, alongside national personnel. Members of the standing corps will act under the political and legal responsibility of the host state.55 This ensures that the corps stays short of fully fledged administrative centralisation substituting national units, which might require Treaty change. Statutory provisions on the involvement in administrative procedures will be presented in the context of legal oversight. It is not a foregone conclusion that Member States will wait in line to welcome Frontex. For example, Spain has never accepted Frontex involvement in Ceuta or Melilla, and Poland rebuffed a deployment at the border with Belarus during the stand-off about the instrumentalisation of migration in 2021. Considerations of ‘pride’ are one explanation, with national border guards considering Frontex as a sort of ‘neo-colonial’ actor which signals their own incapacity. Frontex also gathers transnational media attention and public scrutiny, thus putting national practices in the spotlight. If Member States were unwilling to host Frontex missions, the standing corps and other supranational personnel might be used to reinforce cooperation with third states.

8.3.2 Supervision of the Member States A legally tricky and politically controversial move is the authority to intervene on the ground, if needs be against the will of the Member State concerned. These emergency powers were a response to the policy crisis of 2015/16 when some countries deemed it appropriate to have far-reaching tools to force ‘negligent’ partners to take border controls seriously. They are problematic from a constitutional perspective but are probably compatible with the EU Treaties. In situations of extreme urgency, the principles of solidarity and loyal cooperation, enshrined in Article 4(3) TEU and Article 80 TFEU, arguably outweigh national prerogatives for self-government and the maintenance of law and order, which are protected under Article 4(2) TEU and Article 72 TFEU.56 Such compulsory federal jurisdiction would hardly succeed in practice without the cooperation of the authorities of the state 53 See ibid arts 48–53; and Gkliati, Mariana, ‘The EU Returns Agency’ (2022) 24 EJML 545. 54 ibid Annexes I–IV. 55 ibid arts 54–62, Annex V. 56 See Rosenfeldt (n 32) 174–88; Dubout (n 44) 473–76; and Jorrit Rijpma, The Proposal for a European Border and Coast Guard (Study for the European Parliament, PE 556.934, March 2016) 18.

Frontex: Institutional Muscle Formation 209 where the operation shall take place.57 Support is particularly important whenever agency staff is involved in administrative decision-making or depends on the host country for accommodating asylum applicants and other migrants. Procedurally, the mandatory deployment of Frontex presupposes persisting deficits in certain segments of the external border, to be identified on the basis of a ‘risk analysis’ (with regard to migratory movements) and a ‘vulnerability assessment’ (regarding the preparedness of the Member State).58 Frontex would play a critical role in the risk analysis and vulnerability assessment, while the authority to mandate the compulsory intervention would rest with the Council, acting by a qualified majority.59 In light of the need for practical cooperation, the option of compulsory intervention will primarily serve dissuasive functions—to use a metaphor that has become popular in EU-circles in recent years: it is the ‘nuclear option’ no one wants to employ. This ultimate instrument had not been activated at the time of writing, in contrast to the ‘softer’ Schengen evaluation mechanism, mentioned in Chapter 12.3.2 on border controls.

8.3.3 Cooperation with third states Cooperation with third states is a domain where Frontex can add value, precisely because it can be useful to pool resources. Indeed, international activities feature prominently in the new mandate. Cooperation may range from information exchange, for instance via liaison officers, over technical and operational assistance to cooperation on return.60 At the time of writing, the EU institutions were considering the future deployment of Frontex personnel in important transit countries in the Sahel region, Morocco, and Iraq. These missions further afield would complement ongoing activities in the Western Balkans, which have seen and are still experiencing a significant expansion. The tasks and job description of agency staff depend on the context and can change over time. A ‘model working arrangement’ indicates the breadth of possible actions.61 Frontex would work alongside other Union activities, such as financial and other support for capacity building and migration management, which will be presented in Chapter 18.3.2 and 18.4.2 on the external dimension. Working arrangements establish a political framework for practical cooperation without being legally binding.62 Nevertheless, the international activities of Frontex raise complex legal questions from an institutional perspective under both Union law and public international law.63 Data protection may possibly prove to be a vehicle to hold the border agency to account for international cooperation.64 Additional concerns exist with respect to compliance with human rights and non-refoulement obligations, to be discussed in Chapter 12.2 on border controls. A far-reaching form of cooperation are status agreements, which 57 Frontex Regulation (EU) 2019/1896, art 42(5) recognises the need for cooperation insofar as the OPLAN requires the consent of the host state. 58 ibid arts 29, 32–35. 59 ibid arts 32(10), 42(1); and TEU, art 16(4). 60 Frontex Regulation (EU) 2019/1896, arts 70–78; and see Florin Coman-Kund, European Union Agencies as Global Actors (Routledge 2018) 178–92. 61 See Commission, ‘Communication: Model working arrangement’ COM(2021) 830 final. 62 ibid Annex I, No 13; and Coman-Kund (n 60) 193–202. 63 See Merjin Chamon, ‘A Constitutional Twilight Zone’ (2019) 56 CML Rev 1509. 64 See Frontex Regulation (EU) 2019/1896, art 86.

210 Agencies (Frontex and Asylum Agency) authorise Frontex staff to exercise executive powers. Such status agreements are legally binding and are concluded in accordance with procedure set out in Article 218 TFEU.65 A first set of status agreements were signed with Albania and Macedonia in 2018,66 thereby effectively projecting the rationale behind joint operations and rapid border intervention to the territory of third states. From a constitutional perspective, there is little doubt that Frontex can be entrusted with international cooperation. Article 77(2)(b) and (d) TFEU do not specify the geographic location of border controls, which need not be restricted, as a result, to the immediate environment of the borderline. The Treaty objective of ‘efficient monitoring’ ‘at all stages . . . of migration flows’67 calls for some flexibility of the legislature regarding pre-arrival measures on the high seas or in third countries.68 If need be, additional legal bases can be activated, such as Article 78(2)(g) TFEU for cooperation on asylum or Article 79(2)(c) TFEU for return.69 That is precisely what the Council did when concluding the status agreement with Albania.70 Therefore, it is not problematic per se that the former limitation of executive powers to neighbouring states was discontinued in the latest version of the Frontex Regulation.71 Arguably, Frontex could even be authorised to support return to another third state, as the Commission had originally proposed.72 The powers of Frontex are not universal, however; they require a functional linkage with movements towards Europe.

8.3.4 Eurosur and procurement of equipment An important area of Frontex activities is often overlooked by legal observers and in the policy debate. Frontex has started buying equipment and manages the European Border Surveillance System (EUROSUR), which builds upon previous Spanish practices.73 While operational equipment is mostly acquired to support national border guards or joint operations (also as an instrument of financial solidarity), EUROSUR coordinates the deployment of drones, reconnaissance aircrafts, and satellites to track migratory movements; the material is partly owned by Frontex. EUROSUR transforms border controls via modern technology, including the future use of artificial intelligence.74 To perform these tasks, the agency cooperates with industrial partners. Substantial parts of the budget will be spent on lucrative contracts to operate existing systems and develop pioneering surveillance

65 ibid art 78(3); and ch 2.3.5. 66 See eg Status Agreement between the European Union and the Republic of Albania on actions carried out by the European Border and Coast Guard Agency (adopted 5 October 2018, entered into force 1 May 2019) [2019] OJ L46/3. 67 TFEU, arts 77(1), 79(1). 68 See ch 12.2.1. 69 See ch 18.2.1. 70 See Council Decision (EU) 2019/267 concerning the conclusion of the Status Agreement between the EU and Albania [2019] OJ L46/3. 71 Contrast Frontex Regulation (EU) 2019/1896, art 73(3) with the former Frontex Regulation (EU) 2016/1624, art 54(3). 72 See Proposal for a Frontex Regulation, COM(2018) 631 final, art 75(4); and Thym (n 4) MN 20b. 73 See Frontex Regulation (EU) 2019/1896, arts 18–23; and ch 18.3.3. 74 See Communication: ‘Commission establishing multiannual strategic policy European integrated border management’ COM(2023) 146 final, Annex 1; and Frontex, ‘Artificial Intelligence-Based Capabilities for the European Border and Coast Guard’ (Final Report, March 2021).

Asylum Agency: Latecomer with Subtle Influence 211 apparatuses.75 Scrutinising these activities from a legal perspective requires an advanced knowledge of public procurement law and budgetary procedures. The operation of remote high-tech border controls raises important political and legal questions.76 A telling example are contracts Frontex has signed with Airbus and Israeli companies to monitor movements in the Mediterranean.77 Apparently, some information obtained from drones or conventional reconnaissance aircraft was forwarded to the Libyan coastguard on at least one occasion in 2021. Doing so provided the Libyan authorities with the information they needed to ‘rescue’ migrants in distress at sea and to take them back to a country where plenty of human rights abuses take place.78 In contrast to the information contained in databases, which will be discussed in Chapter 9, information on border surveillance will only be subject to data protection laws when it qualifies as ‘personal data’. The high quality of satellite imagery and drone supervision may possibly cross that hurdle in specific circumstances. By contrast, remote control by means of avant-garde technology will not usually amount to an exercise of state jurisdiction, as a precondition for the application of human rights.79

8.4 Asylum Agency: Latecomer with Subtle Influence Today’s Asylum Agency has always been, much more than Frontex, a forum for information gathering and training to improve the quality and uniformity of asylum decisions and reception facilities. These tasks should not be discarded as less relevant; anyone familiar with everyday asylum practices knows how terribly important reliable and constantly updated information is for decision-making (8.4.1). Information gathering continues to be an essential task, notwithstanding the focus of much of the political and legal debate on the operational activities in hotspots and elsewhere (8.4.2). Like Frontex, the Asylum Agency will increasingly become active in supervising Member States and in cooperating with third states (8.4.3).

8.4.1 Information gathering and practical guidance Chapter 13.5 on asylum will illustrate that refugee status determination requires a forward- looking prognosis as to whether someone would be persecuted in the home state. Doing so involves an assessment of the general situation and of the personal circumstances in distant places. Manifest discrepancies in recognition rates bear testimony to the importance of up-to-date ‘country of origin information’ (COI). Collecting and updating information on 75 See Jacqueline Andres, EU-Grenzregime. Profiteure von Entmenschlichung und mythologisierten Technologien (The Left in the European Parliament 2021); and Statewatch, ‘Funds for Fortress Europe’ (Analysis, 29 January 2022). 76 See Luisa Marin and Kamila Krajčíková, ‘Deploying Drones in Policing Southern European Borders’ in Aleš Završnik (ed), Drones and Unmanned Aerial Systems (Springer 2015) 101. 77 See Jasper Jolly, ‘Airbus to Operate Drones Searching for Migrants Crossing the Mediterranean’ The Guardian (20 October 2020). 78 See Sara Creta and others, ‘Wie Frontex hilft, Migranten in libysche Folterlager zurückzuschleppen’ Der Spiegel (30 April 2021). 79 See ch 12.2.2.

212 Agencies (Frontex and Asylum Agency) countries such as Syria, Nigeria, or Afghanistan represents a critical task, which the Agency performs together with national authorities.80 Indeed, major asylum destination countries have long pursued centralised COI collection.81 To move towards supranational information gathering may benefit from economies of scale and improve quality. EUAA input shall be taken into account by administrative authorities and domestic courts when deciding individual cases. COI ‘reports’ (regarding the factual situation) and ‘guidance’ (including advice on how to assess individual cases) gained in quality and quantity during the early 2020s.82 We can expect their visibility to increase further. While the practical influence of COI reports and guidance is palpable, the Asylum Agency also assists—and gently tutors—national authorities in other respects. Support and training tools are meant to improve the quality and coherence of decision-making. By way of example, asylum officers can benefit from practical guidelines on evidence assessment and on how to conduct the personal interview, either in person or remotely; similarly, national authorities are advised on how to structure administrative processes.83 Training materials are available for legal aspects of refugee status determination, including for judges at domestic courts (notwithstanding full independence in the exercise of their function).84 Readers may take advantage of the ‘EASO Case Law Database’ to find judgments from the CJEU, the ECtHR, and national courts.85 Other useful sources of information are the ‘European Database of Asylum Law’ (EDAL) and the ‘Asylum Information Database’ (AIDA), which are coordinated by ECRE.86

8.4.2 Operational support in ‘hotpots’ and beyond From the beginning, the former EASO Regulation foresaw asylum support teams to be deployed where Member States face problems with asylum procedures or reception facilities.87 Support teams assist domestic authorities with practical needs, as defined in the OPLAN; deployment of interpreters, presumably for personal interviews, and access to databases were mentioned as an example.88 These explicit authorisations in the original mandate must be read in conjunction with an abstract limitation of the agency’s powers.89 Against this background, some of the powers which asylum support teams performed de facto probably went beyond the original mandate, especially autonomous preparation of asylum decisions and carrying out personal interviews.90 Article 16(2) EUAA Regulation (EU) 2021/2303 remedied this shortcoming in the form of a detailed list of multiple operational tasks the 80 See EUAA Regulation (EU) 2021/2303, arts 9–11. 81 See Michael Jandl and Brigitte Suter, Study on Country of Origin Systems in Ten European Countries and the Potential for Further Improvement of COI Co-Operation (ICMPD 2006). 82 See https://euaa.europa.eu/asylum-knowledge/country-guidance (accessed 1 March 2023). 83 See EUAA Regulation (EU) 2021/2303, art 13; and https://euaa.europa.eu/practical-tools-and-guides (accessed 1 March 2023). 84 ibid art 8; and https://euaa.europa.eu/asylum-knowledge/courts-and-tribunals (accessed 1 March 2023). 85 ibid art 6(3); and https://caselaw.euaa.europa.eu (accessed 1 March 2023). 86 See https://www.asylumlawdatabase.eu and https://asylumineurope.org (both accessed 1 March 2023). 87 See the former EASO Regulation (EU) No 439/2010, arts 10, 13–23. 88 ibid arts 15, 18(1)(b). 89 ibid art 2(6), recital 14. 90 See Ombudsman, ‘EASO involvement in the decision-making process in the Greek Hotspots’ (Decision 735/ 2017/MDC, 5 July 2018) paras 33–34; and Evangelia (Lilian) Tsourdi, ‘Holding the European Asylum Support Office Accountable for its Role in Asylum Decision-Making: Mission Impossible?’ (2020) 21 GLJ 506, 511–20.

Asylum Agency: Latecomer with Subtle Influence 213 agency may perform with regard to identification and registration, asylum procedures, and reception conditions. At the same time, the new legal basis is more than the codification of previous administrative practices. Some formulations, for instance regarding ‘the necessary assistance in the [asylum] procedure’,91 are so broad that they may cover almost anything. Along similar lines, assistance with ‘the provision of reception facilities . . . in particular emergency accommodation, transport, and medical assistance’, effectively authorises EUAA to operate reception centres, under recourse to equipment purchased or leased by the Agency.92 A memorandum of understanding with Greece on a joint reception and identification centre in Lesvos may serve as a model for the collaborative provision of reception facilities in the future.93 Practical details are set out in the OPLAN and additional standard operation procedures, to be drawn up the executive director in cooperation with the host country; they are not publicly available.94 ‘Asylum support teams’ may include agency staff, national officials, and other experts (such as sub-contracted interpreters and personnel); in scenarios of mixed flows, they may form ‘migration management support teams’ together with Frontex and Europol.95 Member States shall make available a ‘reserve pool’ of 500 officers to be deployed at short notice.96 In contrast to Frontex, EUAA will not set up a supranational standing corps, although the agency has the power and the resources to hire its own staff to perform specific tasks. Presumably, a formal standing corps is not warranted as asylum officers and other experts usually have diverse educational backgrounds, depending on the tasks they are supposed to perform. Much will depend, therefore, on what the Asylum Agency will do with the enhanced mandate in practice. The new legislation builds upon years of experience with ‘hotspots’ on the Greek islands and in Southern Italy, which had served as laboratories for operational cooperation. The ‘hotspot’ concept emerged during discussions about the future of Frontex and generally stands for greater agency involvement in emergency situations.97 Tasks depend on the context and can change over time. Thus, the job description for the Italian hotspots primarily concerned assistance for purposes of disembarkation, registration, and relocation; in Greece, the design of the hotspots changed fundamentally after the EU–Turkey Statement, with unprecedented involvement in decision-making. Asylum support teams performed vulnerability assessments (with important consequences under Greek law) and prepared decisions on the admissibility and the merits of asylum applications, which seem to have been rubber-stamped by the Greek authorities for the most part.98 Moreover, Frontex was involved in identification, screening, and return.99 Future missions will build on these models, notwithstanding concerns about accountability. 91 EUAA Regulation (EU) 2021/2303, art 16(2)(c). 92 ibid arts 16(2)(f), 23. 93 See Annex Decision approving the Memorandum of Understanding on a Joint Pilot for the establishment of a new Multi-Purpose Reception and Identification Centre, C(2020) 8657 final. 94 See EUAA Regulation (EU) 2021/2303, arts 17–19; and Catharina Ziebritzki, ‘The Integrated EU Hotspot Administration and the Question of the EU’s Liability’ in Markus Kotzur and others (eds), The External Dimension of EU Migration and Asylum Policies (Nomos 2020) 253, 259–66. 95 EUAA Regulation (EU) 2021/2303, arts 20–21. 96 ibid art 19(6), Annex I. 97 See Unisys (n 5) 24; and Commission, ‘Communication: A European Agenda on Migration’ COM(2015) 240 final 6, Annex II. 98 See Fernández Rojo (n 16) 104–12, 133–59; and Ziebritzki (n 94) 259–77. 99 See Frontex Regulation (EU) 2019/1896, arts 38(3)(d), 48(1)(b), (2), 82.

214 Agencies (Frontex and Asylum Agency)

8.4.3 Supervision and international cooperation EUAA will receive, like Frontex, enhanced monitoring functions to identify structural weaknesses and to respond to deficiencies. Ideally, these powers will prevent crisis situations from occurring, including through training, operational guidance, and standard setting. An informative example is guidance on reception conditions, which provides Member States with a template of how to implement the abstract legislative obligation to guarantee an ‘adequate standard of living’ for asylum applicants.100 One step further, intervention in response to pressing deficits mirrors the powers of Frontex: the Commission will adopt non-binding recommendations, to be followed by a Council decision adopted by qualified majority, if needs be against the will of the country concerned. These compulsory measures will apply once the current Dublin III Regulation is replaced by a successor document.101 The introduction of this ‘sunrise clause’ was the political price for abandoning the ‘package approach’ to asylum reform. As in the case of Frontex, compulsory federal intervention will hardly succeed in practice without the cooperation of domestic authorities. International cooperation may concern standard setting, assistance to capacity building, and support for setting up asylum systems in third states. EUAA staff may be involved in the implementation of projects funded under the EU budget for refugee protection in neighbouring countries and the global South, which will be described in Chapter 18.4.3 on the external dimension. In doing so, the Agency will have to cooperate with international organisations and private actors. Working arrangements will have an administrative character, dealing with practical aspects of international assignments.102 Unlike Frontex, the Asylum Agency does not have formal powers to perform executive functions in third states, with the exception of resettlement where agency staff may support Member States in organising the process.103

8.5 Accountability for Wrongdoing Administrative law generally performs a dual function: it supports the effective realisation of public tasks, while, at the same time, ensuring respect for the rule of law. The tension between empowerment and control is not specific to Frontex and the Asylum Agency, yet the challenge of developing effective accountability mechanisms is particularly pressing. The hotspot approach has been accused of recurring illegalities in terms of reception conditions and asylum procedure.104 Reports about alleged pushback practices by Greek, Croatian, Hungarian, Polish, and Lithuanian border guards gathered widespread public attention during the period from 2020 to 2022. These developments nourished the impression that agency involvement, which was often ad hoc and non-transparent, undermined the rule of law through slippage into legal loopholes.

100 See ch 13.8.2; and EASO Guidance on Reception Conditions. Operational Standards and Indicators (EASO 2016). 101 See EUAA Regulation (EU) 2021/2303, arts 15(4)–(8), 22, 73(2). 102 ibid arts 35–38. 103 ibid art 35(6); and ch 13.10. 104 By way of example see ECRE, ‘The Implementation of the Hotspots in Italy and Greece: A Study’ (December 2016).

Accountability for Wrongdoing 215 Instead of assessing specific instances of wrongdoing, our analysis will concentrate on the institutional and legal design of accountability mechanisms. The starting point shall be inspection of the legislative provisions on composite procedures in the mandates of Frontex and the Asylum Agency, which are entrusted with supporting domestic decision-making by the national authorities of the host country (8.5.1). Institutional oversight of these activities encompasses non-judicial accountability instruments, including internal governance structures (8.5.2). Such political and administrative measures complement legal oversight by courts. Procedural requirements emphasise the primary responsibility of domestic courts for the indirect review of the behaviour of Frontex and EUAA in the context of domestic proceedings (8.5.3).

8.5.1 Legislative provisions on agency involvement To describe agency involvement in administrative decisions in hotspots, experts in migration law often employ the term ‘joint processing’, which had been introduced by a feasibility study on the future Asylum Agency.105 On that basis, Tsourdi proposed a valuable distinction between simple assistance, joint processing, and the potential of centralised decision- making.106 Such an ideal-typical taxonomy serves as an heuristic device to distinguish different levels of agency involvement and does not constitute distinct legal categories. The powers of the agencies are defined in the mandate which can lay down intermediate solutions. Moreover, the academic literature and Court judgments on other domains of administrative law hold important insights into the role of the migration agencies. For this reason, this book uses the generic term ‘composite procedures’, instead of the asylum-specific notion of ‘joint processing’, in order semantically to connect research on EU migration law to the debate about administrative law generally. Doing so can be particularly helpful when identifying competent accountability fora. Their jurisdiction depends, as we shall see, on the relative weight of an administrative actor in a ‘joint’ decision. In other areas of Union law, composite procedures are more common. In the field of product placement, for instance, the admission of certain hazardous products to the single market is based on the administrative decision of one Member State, taken after feedback from across the Union; admission of medicines, by contrast, relies on transnational input into a final decision at the supranational level.107 We may, in other words, distinguish between different forms of agency involvement in composite procedures: decisions taken by national authorities or an EU institution after preparatory input from an agency or, alternatively, agency decisions based on contributions by domestic authorities.108 With regard to Frontex, we can observe both situations: whereas the border agency may adopt 105 See Ramboll/EurAsylum, ‘Study on the Feasibility and legal and practical implications of establishing a mechanism for the joint processing of asylum applications on the territory of the EU’ (HOME/2011/ERFX/FW/ 04, 13 February 2011). 106 See Evangelia (Lilian) Tsourdi, ‘Bottom- up Salvation? From Practical Cooperation towards Joint Implementation through the European Asylum Support Office’ (2016) 1 European Papers 997. 107 See Herwig CH Hofmann, Gerard C Rowe, and Alexander H Türk, Administrative Law and Policy of the European Union (OUP 2011) 405–10. 108 For an analytical taxonomy see Mariavittoria Catanzariti and Alexander H Türk, ‘EU Agencies and the Rise of a Mixed Administration in the EU Multi-Jurisdictional Setting’ in Miroslava Scholten and Alex Brenninkmeijer (eds), Controlling EU Agencies (Edward Elgar Publishing 2020) 18, 22–30.

216 Agencies (Frontex and Asylum Agency) recommendations, after input from the Commission and Member States, to remedy existing shortcomings, Frontex assists the Council in deciding whether to oblige states to accept mandatory agency support against their will.109 Legal remedies will have to be directed against the body that formally took the decision as a matter of principle. With regard to involvement in individual decisions, the mandate distinguishes two scenarios. First, Frontex shall ‘provide technical and operational assistance’ to the domestic authorities in the areas of search and rescue, border controls, and return in accordance with the OPLAN.110 The ordinary and legal meaning of the term ‘assistance’ indicates help or support given to someone else; accordingly, the border agency will not act autonomously but under the direction and authority of the host state. Secondly, the new mandate goes one important step further as Member States may ‘authorise members of the [support] teams to act on its behalf ’, by adopting administrative decisions on refusal entry or in other aspects.111 Members of the standard corps and seconded national personnel may also use force and carry weapons in accordance with the laws of the host state.112 To act ‘on behalf of ’ someone involves enhanced operational autonomy, while articulating, both in common and civil law, legal attribution of the agent’s conduct to the principal. Again, the Frontex mandate emphasises host state responsibility. Powers of the Asylum Agency are limited to assistance. Such assistance may be far- reaching, mirroring previous practices in the hotspots, but stays short of the acting on behalf of the host state. The new mandate says unambiguously that the assistance shall facilitate ‘examination by the competent national authorities’ and is ‘without prejudice to the competence of Member States to decide on individual applications’.113 The practical difference to the initial proposal to ‘carry out the entire procedure or parts of the procedure for international protection at the administrative stage’114 are not immediately clear, besides the semantic emphasis on host state responsibility. In any case, the provisions in the mandate on the involvement of agency staff in composite procedures remain modest in comparison to agency powers in other domains. In the terminology of international rules on state responsibility, the conduct of Frontex and the Asylum Agency is attributable to the host state at whose disposal they have been placed.115 Some academic analyses intuitively rely on these well-known international standards when discussing the accountability of the agencies under Union law. We should be careful, however, not to overstate their significance. They concern the responsibility of the Member States for misconduct of agency staff under international law. For the purposes of Union law, by contrast, the Court has developed distinct rules on attribution that define the admissibility of court proceedings. Attribution is a normative category, independent of factual causality.

109 See Frontex Regulation (EU) 2019/1896, arts 32(7), 42(1). 110 ibid arts 36(2)(e), 37(1), 38(3)(d), 48(1)(b), (2). 111 ibid art 82(4), (11). 112 ibid arts 55(7)(c), 82(8), Annex V. 113 EUAA Regulation (EU) 2021/2303, arts 16(2)(c), 18(2)(j). 114 Proposal for an EUAA Regulation, COM(2018) 633 final, art 16a(1)(a). 115 See UNGA Resolution 56/83, ‘Articles on the Responsibility of States for Internationally Wrongful Acts’ (12 December 2001) art 7 ‘Draft Articles on the Responsibility of International Organizations’ (UN doc A/CN.4/ L.778, UN, 30 May 2011), art 6 lays down an additional ‘effective control’ criterion; for comments see Mungianu (n 33) ch 4.

Accountability for Wrongdoing 217 Against the backdrop of the legislative provisions, we have to distinguish between the weight of agency involvement de facto and the responsibility of the Member States de jure. The practical opacity of the situation on the ground contrasts with a comparatively straightforward allocation of responsibilities to the host state in the design of secondary legislation. This solution mirrors the structure of composite procedures in other areas, which similarly distinguish between preparatory input and final decisions. We shall see that the Court generally follows these rules. The responsibility of host states for agency involvement in administrative decision-making is reaffirmed by the contents of the Return Directive and the Asylum Procedures Directive, which both guarantee legal remedies before domestic courts.

8.5.2 Institutional governance Chapter 3.4 described why accountability outside courts can be critical to hold the public administration accountable. Some areas, such as cooperation with third states, can be difficult to address by legal means. Other domains are rarely dealt with by judges, as stark discrepancies in the number of preliminary references illustrate. Hotspots and border controls are prime examples of such fields where accountability mechanisms can be useful. We have seen in Chapter 3.4.2 that own-initiatives reports by the Ombudsman paved the way for a complaints mechanism within Frontex; the Court of Auditors criticised mismanagement in the hotspots in Greece and Italy; and the European Parliament set up an extraordinary ‘Frontex scrutiny group’ to investigate pushbacks allegations. These forms of ex post control complement judicial oversight on the basis of legal standards. They are further enhanced by the investigative powers of human rights treaty bodies and political fora at the international level mentioned elsewhere.116 Mechanisms of ex post control are preceded by avenues of ex ante institutional governance, which may ideally prevent wrongdoing from occurring in the first place. Such institutional governance revolves around appointment, reporting, and hierarchical subordination, in relation to which agency-specific solutions must be found.117 Frontex and the Asylum Agency are no simple subsidiaries of the Commission but independent bodies. There is a tension between autonomy and control, which similarly defines the position of the Commission and other executive bodies at the national or international levels. Inter- institutional checks and balances ideally guarantee that autonomy does not translate into unaccountable sovereignty.118 For Frontex and EUAA, the management boards are the essential steering bodies; they decide by majority with each participating state casting one vote, besides the Commission which has two votes (UNHCR participates in EASO as an observer).119 This governance structure supports smooth cooperation with domestic authorities. At the same time, however, the dominant position of the Member States means that neither the Commission nor

116 See ch 5.5; and Cathryn Costello and Itamar Mann, ‘Border Justice’ (2020) 21 German LJ 311, 325–32. 117 See ‘Common Approach on decentralised agencies’ (Joint Statement of the European Parliament, the Council, and the Commission, 2012). 118 We have to distinguish between de jure autonomy and de facto influence by means of political clout or managerial expertise; see Bjorn Kleizen and Koen Verhoest, ‘Opportunities and Threats of Agency Autonomy in EU Governance’ in Scholten and Brenninkmeijer (n 107) 39–59. 119 See Frontex Regulation (EU) 2019/1896, arts 101, 105; and EUAA Regulation (EU) 2021/2303, arts 40, 45.

218 Agencies (Frontex and Asylum Agency) the European Parliament can instruct the agencies how to proceed. Voting rules may result in a stalemate if members cannot agree on a common approach. The resignation of Frontex Executive Director Fabrice Leggeri in April 2022 came after months of criticism, which was buttressed by investigative journalism, parliamentary scrutiny, and an investigation of the anti-fraud office OLAF.120 The episode was a telling reminder of how important complementary accountability fora can be. The position of a fundamental rights officer (FRO) is meant to create awareness for human rights within the agencies, including through participation in operational planning and cooperation with third states. Doing so cannot replace ex post accountability by courts or other fora, but it may be a useful tool to change the agencies from within; a consultative forum serves as an interface with civil society and public debates.121 The implementation of the enhanced mandate of the FRO within Frontex was delayed by practical difficulties in hiring staff and developing operating procedures. Once the system has been set up, however, fundamental rights monitors will constantly assess operational activities of the border agency.

8.5.3 Legal remedies for composite procedures Frontex and the Asylum Agency are supranational bodies and are, therefore, subject to the jurisdiction of the CJEU. Primary law recognises that decisions and other conduct of the agencies can be challenged.122 These provisions reiterate the basic position of the Court that neither the Member States nor the EU institutions ‘can avoid a review of the question whether the measures adopted by them are in conformity with the basic constitutional charter, the Treaty’.123 Operational activities of Frontex and the Asylum Agency can be challenged, but in which court? The answer to that question is less straightforward than one might think intuitively and has not been settled definitely so far. Several observations may elucidate the debate. First, actions for annulment by individuals before the General Court may be directed against measures ‘intended to produce legal effects’ in the form of any behaviour capable of changing the legal position of the claimant.124 That broad definition may cover operational activities, provided we conclude on a case-by-case basis that it brings about a ‘distinct change in the legal position’ even in the absence of a formal decision. Having said this, agency input into formal decision-making by host states will rarely meet this threshold. Individuals have to challenge the final decision before domestic courts, which may make a preliminary reference to the CJEU. Similarly, direct actions against an OPLAN before the General Court would probably be declared inadmissible. Settled case law on Article 263(4) TFEU gives priority to indirect challenges against implementing measures.125 A singular

120 See ‘Frontex Boss Leggeri Resigns from EU Border Control Agency’ Le Monde (29 April 2022). 121 See Frontex Regulation (EU) 2019/1896, arts 108–11; EUAA Regulation (EU) 2021/2303, arts 49–50; as well as Frontex Consultative Forum, Eighth Annual Report 2020 (Frontex 2021). 122 Frontex Regulation (EU) 2019/1896, art 98, which is declaratory in character. 123 Case 294/83 Les Verts EU:C:1986:166, para 23. 124 See Joined Cases C-105/15 P to C-109/15 P Mallis and Malli v Commission and ECB EU:C:2016:702, para 51. 125 See ch 3.3.1; reaffirmed for human rights violations by Case C-565/19 P Carvalho and others v Parliament and Council EU:C:2021:252, paras 77–78.

Accountability for Wrongdoing 219 judgment of the Court intervening in the suspension of the governor of the Latvian Central Bank cannot be generalised easily.126 Secondly, domestic courts do not have authority to control the behaviour of Frontex and the Asylum Agency directly, as the agencies enjoy jurisdictional immunity.127 Crucially, this immunity does not entail, however, that individuals are left without legal protection. Rather, the wording and general scheme of the mandates underscores that the conduct of Frontex and EUAA staff in composite procedures shall be attributed to the host state as a matter of principle. The same holds for national personnel seconded to another country in the context of rapid border intervention or asylum support teams.128 These legislative provisions are a strong indication that third-country nationals have to complain with domestic courts about asylum decisions or refusals of entry of the host state prepared by supranational staff or seconded national personnel. Case law on other segments of Union law confirms this finding. Judges have held repeatedly that preparatory acts are controlled at a subsequent stage, when reviewing the legality of the final outcome of the procedure.129 This was confirmed explicitly for an agency opinion preparing a Commission decision on medical products.130 On step further, several judgments have developed distinct reasoning for composite procedures involving agencies and national officials (or horizontal inter-state cooperation). In essence, the Court reaffirmed the contents of the Frontex and EUAA Regulation: preparatory measures are controlled in the context of legal remedies against the final outcome; whenever necessary, national laws must be interpreted in light of Article 47 CFR and Article 19(1)(2) TEU to ensure effective judicial oversight of preparatory acts in the context of legal remedies against the final decision.131 Judges even went so far as to prohibit autonomous legal remedies against preparatory acts to prevent the duplication of legal oversight.132 For our purposes, the case law reaffirms the mandates that domestic courts of the host state are responsible for reviewing the legality of preparatory acts and other assistance by the agencies in the context of legal remedies against the final decision; Court judgments do not require an identification of the precise role of EUAA and Frontex.133 Experts in migration law may prefer otherwise but preference for indirect legal remedies is firmly embedded in Union law, as mentioned in Chapter 3.3.1 on the judicial architecture. We cannot expect the Court to reverse seminal rulings for the sole reason that Greek courts are, unfortunately, unreliable partners. The Court would certainly deliver clearly worded judgments

126 See Case C-202/18 Rimšēvičs v Latvia EU:C:2019:139, paras 69–74. 127 See Frontex Regulation (EU) 2019/1896, art 96; EUAA Regulation (EU) 2021/2303, art 61; TFEU, art 343; and Protocol [No 7] on the Privileges and Immunities of the European Union [2008] OJ L115/266, art 11(a). 128 The English-language literature often ignores that most personnel deployed in Frontex or EUAA missions is not supranational staff; see Matthias Lehnert, Frontex und operative Maßnahmen an den europäischen Außengrenzen (Nomos 2014) ch 11. 129 See Case C-69/10 Samba Diouf EU:C:2011:524, paras 54– 55; in line with Case C- 13/ 01 Safalero EU:C:2003:447, paras 54–56. 130 See Case T-326/99 Olivieri v Commission and EMEA EU:T:2003:351, para 53. 131 See Case C-219/17 Berlusconi and Fininvest EU:C:2018:1023, paras 42–51; Case C-682/15 Berlioz Investment Fund EU:C:2017:373, paras 75–101; and Paolo Mazzotti and Mariolina Eliantonio, ‘Transnational Judicial Review in Horizontal Composite Procedures’ (2020) 5 European Papers 41. 132 This can create loopholes for final decisions taken at the EU level with regard to compliance of preparatory acts with national laws; in the reverse scenario, however, the gap does not exist, as domestic courts must apply Union law; see also Filipe Brito Bastos, ‘An Administrative Crack in the EU’s Rule of Law’ (2020) 16 EuConst 63. 133 See, however, Gaia Lisi and Mariolina Eliantonio, ‘The Gaps in Judicial Accountability of EASO in the Processing of Asylum Requests in Hotspots’ (2019) 4 European Papers 589, 599.

220 Agencies (Frontex and Asylum Agency) emphasising the need for effective oversight at the domestic level.134 It will not, however, step in to take their place. Thirdly, one may possibly defend a different outcome in special circumstances. For example, one could try to argue that Frontex staff taking decisions about refusal of entry are to be treated similar to binding preparatory acts the Court considers to be amendable to independent legal remedies.135 Having said this, the odds are against the success of the argument considering the rules on attribution in the mandates.136 More promising may be attempts at exploring a conceptual lacuna in legal remedies against operational conduct not resulting in a formal decision. Case law on Article 263 TFEU could possibly be developed dynamically, in light of Article 47 CFR, to embrace factual conduct having negative consequences on the legal position of an individual.137 Pushbacks or other ill-treatment by agency staff may possibly pass that hurdle, provided these measures do not involve or prepare (oral) decisions on behalf of the host state, which would have to be challenged before domestic courts of the host state. If Article 263 TFEU was extended to factual conduct which has severe consequences, there would be no need to demonstrate a ‘sufficiently serious’ breach of EU law, as is necessary for non-contractual liability. Time limits under Article 263(6) TFEU would have to be respected, whereas the absence of ongoing (quasi)legal effects of the factual conduct need not prevent admissibility.138 Finally, the academic literature often focuses on non-contractual liability, possibly reflecting the visibility of tort law in common law countries.139 Be it as it may, actions for damages are certainly no panacea. The mandate of Frontex and the Asylum Agency confirm that the CJEU shall have jurisdiction for non-contractual liability, although it remains uncertain how these provisions relate to the explicit conferral of ‘civil liability’ to the host state.140 In a next step, attribution for the purposes of Article 340(4) TFEU would have to be assessed; case law on this matter is complex and would have to be related to the legislative provisions on agency staff ‘assisting’ domestic authorities, or ‘acting on their behalf ’.141 In any case, rules on attribution for non-contractual liability differs from the standards for actions for annulment.142 The most difficult hurdle will be to demonstrate a ‘sufficiently serious’ breach of Union law (serious fundamental rights violations may regularly pass that test), together with a causal link between the conduct of agency officials and the damage.143 That can be tricky. Think of an asylum decision drafted by an asylum support team, which Greek authorities

134 See Case C-583/11 P Inuit Tapiriit Kanatami and others v Parliament and Council EU:C:2013:625, paras 89– 107; and Berlioz Investment Fund (n 131) paras 75–101. 135 See Case C-97/91 Oleificio Borelli v Commission EU:C:1992:491, paras 9–15; and Case C-225/19 and C-226/ 19 Minister van Buitenlandse Zaken EU:C:2020:951, paras 50–52. 136 See also TFEU, art 263(5). 137 Mallis and Malli v Commission and ECB (n 123) para 51; see the overview by Koen Lenaerts, Ignace Maselis, and Kathleen Gutman, EU Procedural Law (OUP 2014) ch 7.13–7.26; and the proposal by Thomas Groß, ‘Defizite des Grundrechtsschutzes bei Frontex-Einsätzen’ [2020] Zeitschrift für Ausländerrecht 51, 54. 138 See Case T-229/05 PKK v Council EU:T:2008:87, para 49. 139 In German administrative law, by contrast, one would intuitively prioritise a dynamic interpretation of Article 263 TFEU in line with comments above. 140 See Frontex Regulation (EU) 2019/1896, arts 84, 97(4)–(5); and EUAA Regulation (EU) 2021/2303, arts 26, 66(3)–(4). 141 See generally Lenaerts and others (n 137) ch 11.25–11.30; and, for a dynamic proposal, Fink (n 22) 232–316. 142 The former may be permissible even if the latter would fail; see Joined Cases C-8/15 P–C-10/15 P Ledra Advertising v Commission and ECB EU:C:2016:701, paras 49–61. 143 See Case T-528/20 Kočner v Europol EU:T:2021:631, paras 58–69.

Summary 221 are not legally obliged to follow; or a Slovenian vessel handing over a rubber boot with potential asylum applicants to the Greek coast guard, which later tows the boat into Turkish waters. Arguments have been put forward that the causality yardstick need not present an unsurmountable hurdle.144 Notwithstanding these difficulties, non-contractual liability provides a procedural safety net that could possibly be made use of. Test balloons were pending before the General Court at the time of writing.145

8.6 Summary Frontex and the Asylum Agency have, for good reasons, received much scholarly and political attention. This was no foregone conclusion, since the initial design of the agencies was modest. They served as instruments of information gathering, promoted operational cooperation, allowed national governments in the interior to influence external border control practices, and fostered inter-state solidarity. The latter components gathered momentum after the policy crisis of 2015/16 and explain the striking upgrade of the agencies’ mandates. Frontex and EUAA have become veritable powerhouses of operational action, thus raising formidable political and legal questions. Answers are complicated by the composite character of the agencies, which support domestic authorities. Occasional calls for the ‘centralisation’ or ‘federalisation’ of decision-making in the form of supranational staff substituting national authorities contradict the European tradition of executive federalism and would presumably require Treaty change. The Frontex mandate was reinforced on several occasions, thus gradually turning the Warsaw-based body into the largest EU agency with impressive financial and technical resources. A standing corps of supranational policemen, to be set up during the 2020s, will be the hallmark of operational sway, to be deployed alongside national personnel in future missions. Operational powers are complemented by supervisory functions, including the controversial option of compulsory Frontex missions to remedy shortcomings on the ground. We can expect this ‘nuclear option’ primarily to have dissuasive effects; it cannot be performed effectively without the support of domestic authorities anyway. By contrast, cooperation with third states may turn out to be dynamic in the years to come. Agency staff may be deployed in countries of origin or transit to support capacity building, return, migration management, and operational activities. Procurement of equipment and technology-based border surveillance deserve more attention, precisely because the use of drones, for example, remains largely free from legal oversight. EUAA Regulation (EU) 2021/2303 finally became operational in 2022, thus ending several years of legal uncertainties during which the Asylum Agency had performed functions de facto which transcended the scope of the old mandate. Involvement in domestic decision-making and the provision of reception conditions are epitomised by the ‘hotspot’ approach on the Greek islands, which has been widely criticised for managerial deficiencies and possible illegalities. Information gathering and practical guidance are essential functions as well. High-quality country of origin information, in particular, may prove a decisive tool to support the convergence of recognition rates across Europe. Monitoring of

144 145

See Fink (n 22) chs 4–5; and Ziebritzki (n 94) 286–93. See Case T-600/21 and Case T-136/22 (pending).

222 Agencies (Frontex and Asylum Agency) the Member States follows the model of Frontex, even though the compulsory support is conditional upon the reform of the Dublin system. Enhanced powers should be accompanied by fortified accountability, which, unfortunately, lag behinds the increase in institutional prerogatives. The case of Frontex reiterates our earlier finding that non-judicial instruments are crucial to hold the external border agency accountable, through political supervision, financial auditing, or internal administrative measures such as fundamental rights officers with practical sway. These extra- judicial accountability mechanisms supplement retroactive judicial oversight by courts. While there is no doubt that the behaviour of the migration agencies must be amenable to legal remedies, the identification of the correct forum remains difficult. Actions for annulment before the CJEU against agency involvement in administrative decision-making will probably fail in light of legislative provisions stating explicitly that Frontex and EUAA staff shall ‘assist’ national authorities or ‘act on their behalf ’. Case law on composite procedures in other segments of Union law confirms that preparatory measures must be challenged together with the final outcome before domestic courts. A different solution may possibly be defended for ill-treatment that is not related to administrative procedures, whereas the legal standards for non-contractual liability establish high hurdles for success.

9

Databases ‘Big brother’ has become a synonym for dystopian visions of mass surveillance and totalitarian repression. The label emanates from the literary classic 1984, written by George Orwell, and became a household name because of the title of a popular reality TV series. It is an appropriate starting point for the analysis of a highly dynamic area of European migration law which receives little academic and political attention: the impressive build-up of databases and their forthcoming interoperability. To be sure, these developments will not turn the EU in a quasi-totalitarian regime. Databases are regulated by a complex body of rules and subject to data protection standards. In spite of that, their thematic breadth and technical complexity call for heightened scrutiny. To do so presupposes a basic understanding of underlying policy concepts, operational features, legal rules, and accountability mechanisms. The aim of this chapter is to convey that knowledge. One reason why databases receive little attention is their limited visibility. They serve as a technical infrastructure operating in the background. This does not mean, however, that they perform a supportive function only. Article 8 Charter of Fundamental Rights (CFR) reiterates that the collection and processing of personal data require distinct legal bases and are subject to human rights constraints. Human rights recognise, in other words, the self-sustaining character of modern technology, which holds the potential to augment the knowledge base considerably and, hence, the operational leverage of the administration. The collection of more and more information through an increasing number of justice and home affairs databases and their forthcoming interoperability underlines the pertinence of this finding. None of this is Orwellian in the sense of creating an omniscient totalitarian superstate. Rather, the legal construction and practical functioning of the databases has to be embedded firmly into supranational standards of administrative law and human rights. Studying databases is complicated by a sectoral approach, with multiple pieces of legislation. Nevertheless, we can identify general features in terms of theoretical implications, practical functioning, and ongoing expansion (9.1). Migration databases rest on the Treaty competence for migration, and their technical infrastructure reflects the multi-level characteristics of the European administrative space (9.2). Notwithstanding the emphasis on entry and border controls, functions are broader and are schedule to increase substantially in the near future (9.3). Notwithstanding practical limitations of the existing databases, the prospect of interoperability and algorithm-based processing under the European Travel Information and Authorisation System (ETIAS) and Visa Information System (VIS) Regulations holds the potential to transform migration control (9.4). Interoperability and the advent of artificial intelligence reinforce the need for refined data protection standards and effective control mechanisms (9.5).

European Migration Law. Daniel Thym, Oxford University Press. © Daniel Thym 2023. DOI: 10.1093/oso/9780192894274.003.0010

224 Databases

9.1 Theory and Policy Design State authorities are not usually pioneers of new technology. They often lag behind the private sector and require the submission of paper documents at a time when most people have become accustomed to organising many aspects of their daily lives online. Databases on migration are an exception in the sense that they were developed early on as a flanking measure compensating Member States for the loss of control occurring at the internal borders of the Schengen area. Digital capacities often disappear from view in the policy debate precisely because they are—unlike earlier forms of biopolitics—an invisible manifestation of post-modern governmentality (9.1.1). A defining feature of EU developments has been the constant ‘function creep’, through the creation of more and more databases and the extension of access rights (9.1.2).

9.1.1 Information technology as ‘invisible’ infrastructure While information exchange had characterised EU migration law early on, the ‘war on terror’ declared by the US government after 9/11 marked the beginning of a new era. Under the broader heading of ‘homeland security’, the US administration relied heavily on digital methods to transform entry and border controls. Passenger name records, an entry/exit system, and prior travel authorisation were spearheaded across the Atlantic, thus advancing an organisational and functional overlap of migration control, criminal matters, and intelligence services.1 These developments in the US served as a model for EU initiatives discussed below, as did the advanced British e-borders programme in parallel to EU activities as a result of the British opt-out.2 Overlap of migration control and law enforcement has defined the EU’s justice and home affairs databases from the beginning. Alerts of ‘wanted’ criminal offenders featured in the Schengen Information System (SIS), Eurodac emulated the use of fingerprints as a device for identification, and law enforcement authorities were given access to the migration databases. The interdisciplinary ‘securitisation’ hypothesis described in Chapter 4.3 perceives databases to be a post-modern expression of Foucauldian governmentality, thus complementing classic symbols of state power through physical coercion.3 In daily practice, databases tend to seep into the background and disappear from view. They are largely invisible and politically less contentious as a result.4 Widespread invisibility need not, of course, entail practical irrelevance. Infrastructure generally presents itself as a depoliticised technical apparatus, although its formation and maintenance embody policy choices and consume substantial resources.5 EU migration law is a perfect example to illustrate how the creation of databases sustains political dynamics to collect ever more data and to use them for wider purposes. The prospect of interoperability and the advent of artificial 1 See Dana Priest and William Arkin, Top Secret America (Little, Brown & Company 2011). 2 See Public Accounts Committee of the House of Commons, ‘E-Borders and Successor Programmes’ (27th Report of Session 2015/16, February 2016). 3 See Denis Broeders, Breaking Down Anonymity: Digital Surveillance of Illegal Migrants in Germany and the Netherlands (Amsterdam UP 2009) ch 2. 4 See Thierry Balzacq, ‘The Policy Tools of Securitization’ (2008) 46 JCMS 75. 5 See generally Jörg Niewöhner, ‘Infrastructures of Society, Anthropology of ’ in James D Wright (ed), International Encyclopedia of the Social & Behavioral Sciences, vol 12 (2nd edn, Elsevier 2015) 119.

Theory and Policy Design 225 intelligence in the design of ETIAS and the VIS indicate that the challenge will increase in future. Fingerprints and iris scans are increasingly common in migration law and law enforcement, as well as in everyday lives. Many readers will use them to unlock smartphones or to authorise payments. Building upon this expansion, the International Civil Aviation Organization (ICAO) promotes best practices for biometric traveller identification.6 Union law partakes in the spread of biometric identifiers, including through e-gates for registered travellers under the Entry/Exit System (EES), which are at the forefront of the use of face recognition for migration control purposes.7 Digital instruments are a crucial component of the stratification of contemporary border controls. Chapter 11.1 will describe that Union citizens may use them to cross borders swiftly, while governments employ similar methods to obstruct irregular movements.

9.1.2 ‘Function creep’ over three decades The ‘war on terror’ accelerated the formation of databases which had been on the political agenda ever since the intergovernmental Schengen cooperation had established the SIS. Chapter 12.1.3 on border controls will explain that interior ministries used the Schengen cooperated to introduce new migration control and law enforcement instruments. The SIS was the forerunner of technology-based migration control in the early years of the digital age. Biometric identifiers, especially fingerprints, soon complemented the initial reliance on alphanumeric information such as the name, date of birth, or the description of physical characteristics. Entry bans for third country nationals featured among the data to be entered in the original SIS, together with information that can be useful for police cooperation and criminal matters. The migration component thrived in the slipstream of the terrorist attacks in New York and incidents in Europe, such as the train bombings in Madrid and London in 2004/05 or the attacks in Paris and Brussels in 2015/16. They created a sense of threat to which EU institutions responded, amongst others, by accelerating the creation of transnational databases.8 Developments were and are defined by a threefold extension: new databases have been created, ever more information is being collected, and access rights have been extended. The first aspect of gradual extension was the creation of new databases, as well as linkages between them. While the SIS, Eurodac, and the VIS had originally been conceived as sectoral databases for the distinct purposes, the idea of connecting information stored in different systems was pursued under the heading of ‘interoperability’ early on.9 The interoperability agenda encompassed ETIAS, EES, and EUROSUR, which were publicised by

6 See ‘TRIP Guide on Evidence of Identity’ (Version 5.4, ICAO, May 2018); see also Rebekah Thomas, ‘Biometrics, International Migrants and Human Rights’ (2005) 7 EJML 377, 378–88. 7 See generally Javier Galbally Herrerra and others, Study on Face Identification Technology for its Implementation in the Schengen Information System (Commission Joint Research Centre 2019). 8 See Anneliese Baldaccini, ‘Counter-Terrorism and the EU Strategy for Border Security’ (2008) 10 EJML 31, 44–48; and Niovi Vavoula, Immigration and Privacy in the Law of the European Union (Brill 2022). 9 See Commission, ‘Communication on improved effectiveness, enhanced interoperability and synergies among European databases’ COM(2005) 597 final.

226 Databases the Commission under the programmatic title of ‘smart borders’.10 At the time of writing, the latest SIS and VIS upgrade, EES, ETIAS, and the interoperability package were in the process of implementation after years of political debate and technical preparation. They are currently scheduled to be phased in until the end of 2023, thus heralding the beginning of a new phase of information-based migration control. A second component of ‘function creep’ concerns the collection of more and more personal data. Progress in information technology made it possible to complement alphanumeric information with biometric identifiers, such as fingerprints or facial images.11 Databases were extended step by step, with regard to the categories of people covered and in terms of the additional information. The example of Eurodac illustrates what function creep meant in practice: it originally comprised fingerprints of asylum seekers for 10 years and of those entering irregularly for 2 years, with all data being erased after a positive asylum decision;12 this exception for beneficiaries of international protection was deleted later, meaning that their data was retained.13 More recently, the Commission has proposed, as we shall see, a major extension including longer storage periods, compulsory entries on those staying irregularly, and the transformation of Eurodac into a veritable migration management database. Future transformation may possibly include the use of biometric facial images for the additional purpose of automated face recognition.14 The third aspect of ‘function creep’ is easier access, including law enforcement. The ‘war on terror’ saw a series of legislative amendments which gradually extended access rights of national security authorities, Europol, and Eurojust to data stored in the SIS, VIS, and Eurodac.15 This were legally problematic, since any extension of the purpose for which data can be used must be justified in its own right.16 The factual availability of data does not automatically translate into an expectation that it can be used for any legitimate purpose. To do so requires a political decision that can be assessed as to whether it complies with human rights. Complex rules govern the conditions under which law enforcement agencies can consult data collected for migration control purposes; access is typically confined to the prevention, detection, or investigation of terrorist offences or other serious criminal offences.17 Access for law enforcement is usually not automatic but involves basic procedural safeguards to ensure compliance with statutory requirements.18 In the aftermath of 2015/16, the policy debate focused on improving migration control. For instance, national migration authorities will be able to access information in the EES and the VIS to identify those staying irregularly within the territory. One step further,

10 See Commission, ‘Communication: Smart borders: options and the way ahead’ COM(2011) 680 final; and the initial idea presented in Commission, ‘Communication: Preparing the next steps in border management in the European Union’ COM(2008) 69 final. 11 See Evelien Brouwer, Digital Borders and Real Rights (Martinus Nijhoff 2008) 135–40; and Kristrún Gunnarsdóttir and Kjetil Rommetveit, ‘The Biometric Imaginary’ (2017) 26 Public Understanding of Science 195. 12 See Regulation (EC) No 2725/2000 concerning the establishment of ‘Eurodac’ [2000] OJ L316/1, arts 4, 6, 8, 10, 12. 13 See Eurodac Regulation (EU) No 603/2013, art 18. 14 See Niovi Vavoula, ‘Artificial Intelligence (AI) at Schengen Borders’ (2021) 23 EJML 457, 474–82. 15 By way of example see Decision 2008/633/JHA concerning access for consultation of the VIS [2008] OJ L218/129. 16 See EDPS, ‘Opinion concerning access for consultation of the VIS by the authorities of Member States and by Europol’ [2006] OJ C97/6. 17 See Niovi Vavoula, ‘Consultation of EU Immigration Databases for Law Enforcement Purposes’ (2020) 22 EJML 139. 18 By way of example see EES Regulation (EU) 2017/2226, arts 31–33.

Constitutional Foundations 227 return may be facilitated if information about travel documents, used during previous visa applications, can be accessed after entry.19 In 2020, the Commission proposed including scanned colour copies of travel documents into Eurodac, to be used for identification at a later point; the information may even be transferred to third states for return purposes.20 To facilitate comparison of the Eurodac data with the information on the part of third countries, EU institutions simultaneously support third states in setting up national biometric identity systems.21 Such cross-fertilisation is a perfect example how ‘function creep’ may support effective migration control.

9.2 Constitutional Foundations Primary law authorises the creation of the databases discussed in this chapter (9.2.1). Notwithstanding the primary responsibility of the EU institutions for designing the legal framework, the technical infrastructure mirrors the multi-level character of the European administrative space. The supranational component, managed by eu-LISA, is complemented by national systems through which domestic authorities enter and retrieve data (9.2.2). This decentralised structure has important consequences for legal remedies, which will be discussed, together with human rights, towards the end of this chapter. The migration databases cover all members of the Schengen area as a matter of principle, whereas Eurodac has a wider scope (9.2.3).

9.2.1 Supranational competences Databases are not mentioned in the Treaties, yet there is no doubt that primary law authorises their creation. The SIS was the prototype of a justice and home affairs database compensating states for the loss of control options at the internal borders.22 Rules governing the SIS belonged to the Schengen acquis integrated in the EU framework after the Treaty of Amsterdam on the basis of the regular competences for migration and criminal matters; information gathering also informed the policy debate in the European Convention drafting today’s Treaty articles in the broader context of the ‘war on terror’.23 Primary law calls for enhanced inter-state cooperation, amongst others through the introduction of Articles 71– 74 TFEU. There are some uncertainties as to the interaction of Article 74 TFEU on inter-state cooperation and the competences for border controls, visas, asylum, and return in Article 77(2)(a), (b), 78(2)(e), 79(2)(c) TFEU. The former provision was often cited as the legal basis in early legislation but has disappeared in most regulations adopted recently. A pragmatic 19 ibid arts 16(1)(b), 26–27; VIS Regulation (EC) No 767/2008, arts 9(4)(b), 20, as amended by EES Regulation (EU) 2017/2226, art 61; and also Broeders (n 3) chs 2–6. 20 See Amended Proposal for a Eurodac Regulation, COM(2020) 614 final, art 12(1)(j); and Proposal for a Eurodac Regulation, COM(2016) 272 final, art 38. 21 See ch 18.3.3; and Commission, ‘Report: Assessment of third countries’ level of cooperation on readmission in 2019’ COM(2021) 55 final, which is confidential but can be found online. 22 See Case C-378/97 Wijsenbeek EU:C:1999:439, para 40. 23 See generally European Convention, ‘Final Report of Working Group X “Freedom, Security and Justice” ’ (CONV 426/02, 2 December 2002) 16.

228 Databases demarcation line might be to use the substantive competences for the collection and storage of data concerning individuals, while inter-state cooperation and the creation of the databases is supported by Article 74 TFEU as an additional legal basis.24 In practice, the Parliament will prefer Articles 77–80 TFEU which require co-decision. Access for purposes not related to migration may require recourse to other Treaty provisions, especially criminal matters. The principle of subsidiarity will not usually stand in the way of supranational measures for the simple reason that inter-state information exchange cannot be achieved effectively by Member States acting alone.

9.2.2 Decentralised network structure Notwithstanding the responsibilities of the European Union Agency for the Operational Management of Large-Scale IT Systems in the Area of Freedom, Security and Justice (eu- LISA), the technical infrastructure and practical functioning rests on decentralised cooperation among national authorities. The mandate of eu-LISA essentially concerns the provision of the central components and the guarantee of high-quality services.25 The agency has its seat in Tallinn, the capital of Estonia, which is proud of its advanced e- government strategy, whilst the operational centre is in Strasbourg, where the central unit of the SIS had always been located. Research on new IT solutions can be financed by the Instrument for Border Management and Visa and under the Horizon Europe programme, mentioned in Chapter 7.4 on funding. The collection of data, storage periods, and access rights follow the model of indirect implementation, in accordance with the comments in Chapter 7.1 on the administrative dimension. National authorities enter the data to be retrieved by officials across the Schengen area. Eu-LISA has no direct role in handling personal data besides providing the technical infrastructure and serving as a clearing house. As a result, legal responsibility for wrongdoing primarily rests with the Member State entering or using the data. Such transnational cooperation continues to define the migration databases, although we shall see that the advent of artificial intelligence may possibly transform their functioning in the medium term.

9.2.3 Territorial scope Legislation on databases is subject to country-specific opt-outs for Ireland and Denmark, whose distinct layout was presented in Chapter 2.4. Generally speaking, the territorial scope, summarized in Table 9.1, hinges on whether the instrument in question is classified as building on the Schengen acquis, in which the associated countries and Denmark participate in accordance with the comments in Chapters 2.4.1 and 17.2.2. With regard to Eurodac, association with the Dublin rulebook entails that the associated countries and Denmark participate as well, whereas Ireland applies the former legislation (it may decide 24 See Steve Peers, EU Justice and Home Affairs Law, vol I (4th edn, OUP 2016) 81; and Jürgen Bast, Aufenthaltsrecht und Migrationssteuerung (Mohr Siebeck 2011) 157–58. 25 See Regulation (EU) 2018/1726 on the European Union Agency for the Operational Management of Large-Scale IT Systems in the Area of Freedom, Security and Justice (eu-LISA) [2018] OJ L295/99, with later amendments.

Constitutional Foundations 229 Table 9.1 Participation in the most relevant entry and border control instruments Instrument

Ireland / UK (before Brexit)

VIS Regulation (EC) no No 767/2008, with later amendments

Denmark

Schengen/ Dublin? (NO, IS, CH, LI)

Accession

no (int. Law)

yes

read-onlya (≠ Cyprus)

Eurodac Regulation (EU) No 603/2013

yesb

no (int. Law)

yes

yes

EES Regulation (EU) 2017/2226

no

no (int. Law)

yes

read-onlyc (≠ Cyprus)

ETIAS Regulation (EU) no 2018/1240

no (int. Law)

yes

no

SIS Return Regulation (EU) 2018/1860

no

no (partly)

no (partly)d

yese (≠ Cyprus)

SIS Border Checks Regulation (EU) 2018/1861

yes (partly)f

no (int. Law)

yes

yes (≠ Cyprus)

SIS Criminal Matters Regulation (EU) 2018/1862

yes

no (int. Law)

yes

yes (≠ Cyprus)

Borders and Visa Interoperability Regulation (EU) 2019/817

no

no (int. Law)

yes

no

Asylum and Migration Interoperability Regulation (EU) 2019/818

no

no (int. Law)

yes

no

aSee Decision (EU) 2017/1908 on the putting into effect of certain provisions relating to the VIS in the

Republic of Bulgaria and Romania [2017] OJ L269/39. bIreland applies the former legal framework; see recital 20 of the initial Eurodac Regulation (EC) No 2725/2000 (n 12); and Eurodac Regulation (EU) No 603/2013, recital 53. cSee Decision 2010/365/EU on the application of the provisions relating to the SIS in the Republic of Bulgaria and Romania [2010] OJ L166/17, art 1(4); and the identical Decision (EU) 2017/733 [2018] OJ L165/37 on Croatia. dOn the scope of participation see Decision 1999/437/EC on certain arrangements for the application of the Agreement [1999] OJ L176/31, Pt 1(C). ePrevious restrictions were lifted by Decision (EU) 2018/934 on the putting into effect of the remaining provisions of the SIS in the Republic of Bulgaria and Romania [2018] OJ L165/37; we can expect a similar decision on Croatia in the near future. fSee Decision 2002/192 concerning Ireland’s request to take part in some of the provisions of the Schengen acquis [2002] OJ L64/20, arts 1(a)(ii), 5(1)(d)–(e), which seem not to have been repealed when the new legal framework was adopted.

230 Databases to upgrade its status on the occasion of the ongoing revision). New Member States waiting to join the Schengen area are usually given passive access to Schengen-related databases as a first step, subject to gradual extension. The complex legal construction of the new SIS framework was motivated, in part at least, by the distinct territorial scope of the three different regulations adopted in 2018.

9.3 An Overview of Migration Databases Five databases are the backbone of information gathering for migration control purposes: the traditional SIS comprising entry bans and other information (9.3.1); Eurodac on the fingerprints of asylum seekers, which the Commission proposes transforming into a genuine migration management system (9.3.2); VIS on visa applications and, in future, domestic residence permits to enhance migration management within the territory (9.3.3); ETIAS establishing entry authorisation for third country nationals who are exempted from visa requirements (9.3.4); and data gathered in the EES on the occasion of border crossings (9.3.5). Secondary legislation can be complex. By way of example, the EES Regulation (EU) 2017/2226 extends over no fewer than sixty-three pages of single-spaced text in the Official Journal. That is why our description will describe the above-mentioned databases at an intermediate level of abstraction. Additional datasets can be relevant for migration control. They include ECRIS-TCN on previous criminal convictions, which forms part of the interoperability package,26 and the Passenger Name Record (PNR) for air-bound arrivals from third states, which is communicated prior to landing.27 Their primary objectives are the fight against terrorism and the prevention of serious criminal offences. Moreover, the European Border Surveillance System (EUROSUR), mentioned in the previous chapter on Frontex, collects information that will not necessarily involve personal data, as defined in data protection laws. Finally, biometrics and security features in passports and ID cards of Union citizens and residence permits of third country nationals support migration control; they will feature in Chapter 12.7 on document security.

9.3.1 Schengen Information System (SIS) The SIS served as the prototype of a migration database and was established in the framework of the original Schengen cooperation. It started operating in March 1995 when the founding members of the Schengen area discontinued border controls among themselves. The SIS contains data collected for the diverse purposes of migration control, criminal matters, and police cooperation.28 In purely quantitative terms, the number of entries on stolen 26 See Regulation (EU) 2019/816 establishing a centralised system for the identification of Member States holding conviction information on third-country nationals (ECRIS-TCN) [2019] OJ L135/1; Asylum and Migration Interoperability Regulation (EU) 2019/818, art 3; and Vavoula (n 8) ch 7. 27 See Directive 2004/82/EC on the obligation of carriers to communicate passenger data [2004] OJ L261/ 24; and Elif Mendos Kuşkonmaz, Privacy and Border Controls in the Fight against Terrorism (Brill Nijhoff 2021) chs 3–5. 28 See Convention Implementing the Schengen Agreement, arts 94(3), 95–98, 100; and, nowadays, SIS Border Checks Regulation (EU) 2018/1861, art 24; and SIS Criminal Matters Regulation (EU) 2018/1862, arts 26, 32, 34, 36, 38, 40.

An Overview of Migration Databases 231 cars and other objects has always outnumbered the amount of data on people. With regard to the latter, however, entry bans are by far the most relevant category. Numbers of third country nationals subject to entry bans in the SIS have been relatively stable throughout the years, with between 500,000 and 700,000 entry bans across Europe—of almost 100 million entries altogether, mainly on documents and vehicles.29 Comments hereinafter will concentrate on the migration component. Enlargement of the Schengen area after the millennium change required a technical upgrade of the SIS. Moreover, EU institutions agreed in the context of the war on terror to include new categories of data, notably biometric identifiers such as fingerprints, and additional users, in particular for law enforcement purposes. The project became known under the heading of ‘SIS II’, designating the second generation. The SIS II was put into practice gradually on the basis of several smaller legislative amendments; a major overhaul finally became operational in 2013, after years of delay.30 The SIRENE network provides a platform for the bilateral and non-automatic exchange of supplementary information among national authorities.31 Another major reform, adopted in 2018, abandoned the terminology of successive generations by simply referring to the ‘SIS’ again. It became operational in March 2023. Once again, more data and users were added to the database, thus reaffirming the trend towards ‘function creep’. Changes include access for Frontex and national staff in migration management teams and an obligation to enter return decisions to verify compliance.32 The legal framework is difficult to understand intuitively, since it consists of three regulations adopted in parallel: one on border checks, one on return, and one on criminal matters. This convoluted construction adds an additional lawyer of complexity to an already sophisticated rulebook. Readers should be careful not to confuse the separate instruments, whose contents differs in important respects. For instance, DNA profiles can be consulted for criminal matters only, and the use of facial images is limited to regular border crossings; they are not meant to support the identification of third country nationals within the territory.33 Access by Frontex is defined vaguely ‘insofar it is necessary for the performance of their task’, with additional specification to be contained in operational plans.34 The European Data Protection Supervisor stressed the supporting role of the border agency for whose activities host states are ultimately responsible,35 in accordance with our findings on the legal accountability for operational action in Chapter 8. For the purposes of migration control, refusals of entry and stay are the most important feature of the SIS. They will be discussed in Chapter 16.4 on return.

29 See ‘SIS II –2021 Statistics’ (eu-LISA, March 2022) 10–16; and Brouwer (n 11) 65–70; note that nationals of new Member States stopped being third country nationals with accession, thus reducing comparability over time. 30 See Brouwer (n 11) ch 4; and Regulation (EC) No 1987/2006 on the establishment, operation and use of the second generation SIS [2006] OJ L381/4, with later amendments. 31 See SIS Border Checks Regulation (EU) 2018/1861, arts 3(2), 7–8. 32 ibid arts 36–37; SIS Return Regulation (EU) 2018/1860, art 3; and Commission Implementing Decision (EU) 2023/201 setting the date on which operations of the SIS start [2023] OJ L27/29. 33 SIS Border Checks Regulation (EU) 2018/1861, arts 20(2)(3), 33(4); and SIS Criminal Matters Regulation (EU) 2018/1862, art 20(3)(x). 34 SIS Border Checks Regulation (EU) 2018/1861, art 36(1). 35 See EDPS, ‘Opinion on the new legal basis of the SIS’ (May 2017) paras 24–28.

232 Databases

9.3.2 Eurodac Most readers will be familiar with the acronym ‘Eurodac’ for the storage of fingerprints of asylum seekers, although few will know that ‘dac’ stands for ‘dactylographic data’, which essentially relates to fingerprints. It was the first supranational database to use biometric identifiers, in addition to classic alphanumeric information such as names. Eurodac became operational in 2003 and was the first piece of supranational legislation on asylum adopted under the Treaty of Amsterdam, thus realising a project Member States had agreed upon politically in an intergovernmental context to render the original Dublin Convention more effective.36 The rationale behind Eurodac is simple: many asylum seekers do not reveal the travel route, nor do they have passports. As a result, it can be challenging, if not impossible, to identify and prove the Member State responsible under the Dublin system. Fingerprints are meant to overcome these difficulties, since they are considered to be ‘an important element in establishing the exact identity’37. Categories of persons covered by Eurodac were extended, as we have seen, in several steps up until the adoption of Eurodac Regulation (EU) No 603/2013 to cover asylum applicants, beneficiaries of international protection, and those apprehended in the context of irregular border crossings.38 The number of entries varies over the years depending on the number of asylum applications as well as the practices of national authorities, which were not always complying with fingerprinting obligations. Asylum applicants make up roughly 80 per cent of the data; the rest concerns irregular border crossings.39 The database is indispensable for the functioning of the Dublin system, since the overwhelming majority of take back requests is based upon ‘hits’ in Eurodac. Proposals have been put forward, by the Commission, to turn Eurodac into a fully- fledged asylum management database. They consist of two lawyers to be read jointly: the original proposal of 2016 and an amendment introducing additional elements of 2020. In the summer of 2022, the Council was moving towards a common position on Eurodac, as an integral part of the ‘graded approach’ towards asylum reform with ‘mini deals’ on selected topics. Interinstitutional negotiations focused on the new elements of the 2020 proposal, thus building on a partial political agreement on the 2016 Proposal in the trilogue format. At the time of writing, it appeared likely that the reform would be adopted before the European elects in 2024. The changes would effectively transform Eurodac into an asylum management database.40 Additional information such as facial images, date and place of birth, nationality, and scanned copies of travel documents would buttress migration control across the board.41 The minimum age for data collection shall be reduced from 14 to six years, and storage periods in case of irregular entry or stay would be extended to five years.42 Illegal stay would bring about an obligation to enter personal information, thus extending the scope beyond asylum applicants and irregular border crossings.43 Additional information

36 37 38 39 40 41 42 43

See Proposal for a Eurodac Regulation, COM(1999) 260 final 3–4. Eurodac Regulation (EU) No 603/2013, recital 5. ibid arts 9, 14, 18. See ‘Eurodac: 2020 Statistics’ (eu-LISA, March 2021). See Vavoula (n 8) ch 4.6. See Eurodac Proposal of 2016 (n 20) art 12; and Amended Eurodac Proposal of 2020 (n 20) art 12(j). Eurodac Proposal of 2016 (n 20) arts 17–19. ibid art 14.

An Overview of Migration Databases 233 on disembarkation, relocation, resettlement, asylum decisions, and return are meant to allow domestic authorities across the Union to ‘track’ migration statuses over time.44 Doing so would give us a better picture of secondary movements and put an end to inflated numbers of asylum applications as a result of double counting.

9.3.3 Visa Information System (VIS) Schengen visas embody the most enhanced form of transnational effects when, by way of example, a visa issued by a Slovenian consulate grants temporary mobility throughout the Schengen area. These transnational effects explain why the initial introduction the VIS was comparatively straightforward: it allowed issuing states to take into account the views of other countries. At the same time, the VIS enhanced the knowledge base by requiring the collection of biometric data, notably fingerprints and facial images, and the type and number of travel documents.45 Access for law enforcement was extended several times, amongst others to Europol.46 The original VIS was gradually set up from 2004 onwards and became operational in 2011.47 In practical terms, the VIS is a large database. About 17 million new visa applications were entered in the database during 2019 before the decrease in international traffic during the Covid-19 pandemic; altogether, almost 69 million sets of fingerprints were registered at the time, which are kept for five years after the visa application.48 The number of biometric searches was much lower and stood at 7 million during 2019; the average processing time for fingerprint verification was an impressive 1.87 seconds. Consulates abroad and border guards at crossing points were the main users.49 We can expect that the forthcoming interoperability and the extension of access rights will result in many more searches for other purposes. A particularity of the VIS is that it comprises data on Union citizens serving as invitees or guarantors of visa applicants.50 The original idea behind the VIS was to permit consulates to check whether someone had applied for a Schengen visa previously. Moreover, national authorities, including border guards, may use the database to identify forged or revoked documents. Amendments agreed upon during 2021 make a decisive step to extend the VIS beyond visa applications, thus turning the database into a generic instrument of migration management (mirroring the proposed new functions of Eurodac for asylum). Readers should use a consolidated version of the VIS Regulation (EC) No 767/2008, which includes later amendments and can be found in the EUR-Lex portal.51 Changes agreed upon during 2021 were integrated into the original text, which, moreover, had been amended several times in between. The new legislation is scheduled to become operational during 2024, once the interoperability package has become effective. 44 See Amended Eurodac Proposal of 2020 (n 20) arts 11, 12, 14a, 14b. 45 See VIS Regulation (EC) No 767/2008, art 9(4)–(7), as amended by Regulation (EU) 2021/1134. 46 ibid art 3(1). 47 ibid; and Decision 2004/512/EC establishing the Visa Information System (VIS) [2004] OJ L213/5. 48 VIS Regulation (EC) No 767/2008, art 23(1). 49 See ‘Report on the Technical Functioning of the VIS’ (eu-LISA, August 2020). 50 See VIS Regulation (EC) No 767/2008, art 9(4)(f), with later amendments; and Visa Code Regulation (EC) No 810/2009, art 21(3)(b), (5), Annex II, with later amendments. 51 See ch 6.6.

234 Databases The main thrust of the latest changes reaffirms the tendency for ‘function creep’. Additional data include facial images, information on travel documents, and a lower minimum age for fingerprinting of six instead of twelve years (presented as a contribution to fight trafficking and other abuse involving children52). Information on travel documents can be forwarded to third states for return purposes, and Frontex will be given access.53 Moreover, national authorities will be obliged to enter data on long-term visas and on any residence permit delivered within the territory for diverse purposes such as family reunification, studies, or work.54 Member States may have similar systems domestically already, but this extension entails nevertheless that VIS is gradually being transformed— notwithstanding its name—into a generic visa and residence permit repository. A major transformation will be the introduction of automated screening following the model of ETIAS, to be discussed in the context of interoperability.

9.3.4 Travel authorisation (ETIAS) Regulation (EU) 2018/1240 introduced the European equivalent of the US-style travel ESTA scheme. It will be known under the acronym ETIAS, for European Travel Information and Authorisation System, and is currently scheduled to become operational by the end of 2023. Political agreement was reached comparatively quickly after the terrorist attacks in Paris and Brussels, at a time when the return of foreign fighters who had joined the terror regime of the ‘Islamic State’ was a major concern. Five years earlier, the Commission had still concluded that prior authorisation for visa-exempt travellers should not be pursued, since the added value for national security ‘would neither justify the collection of personal data at such a scale nor the financial cost’55. Two additional instruments were adopted in 2021, which are scheduled to become effective, together with the remainder of the interoperability package, at the end of 2023, although such deadlines have proven tentative.56 Applications for travel authorisation can be made any time through a website or an app available in the EU’s official languages; basic information will be translated in at least one official language of each state on the ‘white list’ for visa-free travel under Annex II Visa List Regulation (EU) 2018/1806. Those holding residence permits are exempt from the authorisation requirement (as are third country nationals subject to visa requirements). Applications will be processed quickly in the overwhelming majority of cases through computer-based processing, and the travel authorisation will be issued automatically by the ETIAS Central Unit, or Member States in case of manual follow-up. It will be valid for three years and cost €7 to be paid online, cheaper than the USD 21 charged by the United States.57 In practical terms, filling in the form will be a mere ‘nuisance’ for most visitors, while a small percentage will be rejected even before their arrival at border crossing points. In case of refusal, applicants may lodge a legal remedy with the domestic courts of the country 52 See Proposal for a Regulation amending Regulation (EC) No 767/2008, COM(2018) 302 final, 6. 53 See VIS Regulation (EC) No 767/2008, arts 31(2), 45d–45e, as amended by Regulation (EU) 2021/1134. 54 ibid arts 4(18), (19), 22a–22j. 55 Commission (n 10) 11; taking up the criticism of EP, ‘Resolution: Next Steps in Border Management in the EU’ (P6_TA(2009)0085, 10 March 2009) No 11. 56 See Regulation (EU) 2021/1151 and Regulation (EU) 2021/1152 as regards the establishment of the conditions for accessing other EU information systems for the purposes of the ETIAS [2021] OJ L249/7 and 15. 57 See ETIAS Regulation (EU) 2018/1240, arts 2, 15–16, 18, 20, 20–22, 36, 38(1).

An Overview of Migration Databases 235 that issued the refusal. Grounds of refusal will be listed, like in the case of visas, in a short standard form without detailed comments.58 Applicants are expected to enter a wealth of personal data concerning both personal identity (name, sex, date of birth, nationality, etc.) and travel-related information (such as type, number, and date of expiry of the passport). Similar information is contained in other databases and can be exchanged on the basis of the rules on interoperability, which authorise cross-checks between ETIAS, the VIS, the SIS, and ECRIS-TCN.59 Applicants for travel authorisation are expected to enter additional data regarding education, an email address, and the current job group chosen from a pre-determined list; they will also have to respond to basic queries about recent stays in conflict zones or conviction of serious criminal offences over the past 10 years.60 These data are collected for travel authorisation purposes only and play an important role in the automated processing to filter out those who might pose a risk for internal security, in terms of illegal immigration, or regarding public health. Such algorithm-based decision-making is a novelty and raises important technical and legal questions, which will be addressed on the pages that follow. At a practical level, only affirmative travel authorisation will be issued automatically by computers. In case the algorithm identifies a potential risk, the file will be forwarded to the national unit responsible for manual processing. National authorities are expected to decide within no more than four days after submission; they may require additional information or invite the applicant to an interview, with a final decision being required within two to four days thereafter.61 Grounds of refusal reiterate the entry conditions all visa-exempt travellers must fulfil. These entry requirements remain valid after an affirmative travel authorisation,62 meaning that border guards may still refuse entry.63 At the time of writing, the practical effects of automated processing were impossible to predict, since much would depend on the implementing regulations and the design of the algorithm. The ETIAS Regulation foresees a transitional period of at least six months, during which automated processing remains optional, to be followed by a grace period when applicants can be allowed to enter without authorisation. Thereafter, prior authorisation will become a mandatory condition for external borders crossings.64 Compulsory travel authorisation weakens the guarantee of visa-free travel the EU has given to several third states in visa-facilitation agreements. Considerations of good faith and effet utile support such a critique. At the same time, however, much shorter processing times, procedural differences, lower costs, longer validity periods, different names, and the absence of formal visa stickers support the conclusion that the need for prior travel authorisation is compatible with the visa facilitation agreements,65 also considering that these bilateral agreements define the term ‘visa’ narrowly.66

58 ibid arts 37(3), 38(2), (3); and ch 11.4.3. 59 ibid art 11(2)–(6), as amended by Regulation (EU) 2021/1152 (n 56). 60 ibid art 17(2)–(4) and Annex; by contrast, Proposal for an ETIAS Regulation, COM(2016) 731 final, art 15(4) (c) had foreseen a question whether one had ‘ever’ been convicted of ‘any criminal offence in any country’. 61 ibid arts 21(2), 25–30, 32. 62 ibid arts 3(5)–(8), 37(1)(b)–(d). 63 ibid art 36(6); and Schengen Borders Code Regulation (EU) 2016/399, art 14. 64 ETIAS Regulation (EU) 2018/1240, arts 80, 83. 65 See Julius Buckler, ‘Schengen-Grenzmanagement 2022’ [2020] Die Öffentliche Verwaltung 749, 753. 66 By way of example see Agreement between the European Union and Georgia on the facilitation of the issuance of visas (adopted 22 November 2010, entered into force 1 March 2011) [2011] OJ L52/34, art 3(d).

236 Databases

9.3.5 Entry/Exit System (EES) Visas and entry stamps in passports are epitomes of the age of paper-based bureaucracy, which is about to come to an end in the Schengen area. The EES will catapult border controls into the digital age, by replacing manual stamping with the digital registration of arrivals and departures of anyone not holding a residence permit already.67 The primary aim is the prevention of illegal entry and stay. At present, departures are not registered, and national authorities do not always know, as a result, whether someone has left the Schengen area or not. The new database is scheduled to become operational during 2022. The EES does not, unlike ETIAS, autonomously prepare administrative decisions, although national authorities will use the system when assessing individual cases. For example, it can help to identify ‘overstaying’ or multiple uses of the same visa.68 Moreover, the database can be accessed for law enforcement purposes and will have to be consulted by carriers involved in cross-border transport.69 Data to be stored concerns personal identification (name, date of birth, number of passport, etc.), travel-related information (time and place of entry or exit), and a facial image. Visa-exempt travellers must also provide fingerprints, which others have given during visa applications already. Those refused entry are registered as well. Data is kept for three years.70 The Commission was eager to present digitalisation as a mechanism making life easier for frequent travellers through self-service gates and e-gates, in the same way as many readers provide companies with sensitive personal data to facilitate online shopping and other digital services. Amendments to the Schengen Borders Code Regulation will become effective in parallel to the Entry/Exit System.71 The idea had originally been presented under the name ‘registered traveller programme’,72 thus reaffirming the general finding in Chapter 11.1 that migration controls sustain highly stratified cross-border flows combining easy access for the rich with multiple hurdles for others. The number of personal data contained in the EES will be substantial, once national authorities have started registering every entry and exit.

9.4 Prospect of Interoperability and Artificial Intelligence The EU’s migration databases will expand significantly in the years to come, once the recent amendments have been put into practice. The new digital architecture will result in an almost seamless surveillance of entries, exits, and migration statuses of third country nationals: asylum seekers and those or staying entering irregularly will be registered in Eurodac, whereas the EES records data on regular arrivals and departures at border crossing points. Information on pre-arrival travel authorisation and visa applications will feature in ETIAS and the VIS, together with residence permits issued post-entry. Eurodac

67 See EES Regulation (EU) 2017/2226, art 2. 68 ibid arts 12, 21–27. 69 ibid arts 13(4), 29–33. 70 ibid arts 14–18, 54(1). 71 See Schengen Borders Code Regulation (EU) 2016/399, arts 8a–8d, as amended by Regulation (EU) 2017/ 2225 as regards the use of the Entry/Exit System [2017] OJ L327/28. 72 See Proposal for a Regulation amending the Schengen Borders Code Regulation, COM(2013) 96 final.

Prospect of Interoperability and Artificial Intelligence 237 is proposed to collect asylum-related decisions, and the SIS gathers refusals of entry or stay and return decisions. Against this background, this section assesses horizontal questions to be addressed in the future. ETIAS and the VIS foresee algorithm-based automated processing which may use artificial intelligence (9.4.1). More profane quality concerns hamper the effectiveness and credibility of cutting-edge technology (9.4.2), while ‘interoperability’ is used by the EU institutions as a magic formula to increase the leverage of the databases, with potentially far-reaching practical effects (9.4.3).

9.4.1 From support to automated processing Information exchange on cross-border movements is neither surprising nor problematic per se. However, the methods and instruments have changed considerably over the years. They used to be reactive and ad hoc in the sense of direct communication among national authorities on specific cases; an officer would call a colleague working elsewhere to ask her to provide him with information. In the field of migration law, such individualised communication can realistically only take place selectively.73 The sheer number of people crossing external borders requires modern technology to make transnational communication feasible in practice. Without the SIS, border guards at the Finnish-Russian land border could not realistically learn about entry bans issued by a Slovak authority. Proactive data upload makes such information available to any national authority. This basic rationale of the migration databases is unproblematic. That being said, information gathering by means of proactive upload soon developed a momentum of its own. Secondary legislation foresees, as we have seen, the collection of more and more information for ever wider purposes. Collection of fingerprints in the EES, for instance, does not require individuals to pose a security threat or illegal migration risk, unlike in the case of entry bans; the overwhelming majority of people crossing the external borders are bone fide travellers. Nevertheless, their data will be collected, since it could possibly be used at a later point to prevent abuse or to identify overstayers. Data collection was extended from tangible threats to abstract risks which might possibly emerge in the future.74 Interoperability will multiply such indirect linkages. For example, fingerprints stored in the VIS on the occasion of an unsuccessful visa application can be used on a later occasion to convince home states to accept return. Such multiplication of purposes can be problematic from a human rights perspective. The example of ETIAS and the VIS demonstrates another twist. Unlike fingerprints, information on education levels, the home address, or the job of a person is not directly relevant to achieve legitimate aims. The very idea of screening under ETIAS and the VIS is that a plethora of information, which is not directly related to migration control purposes, may become relevant indirectly if connected to other data submitted by the individual, or available otherwise. That is usually referred to as ‘profiling’ in information science if algorithms discover patterns of correlations through computerised data analysis to identify risk factors

73 By way of example see SIS Border Checks Regulation (EU) 2018/1861, arts 27–30. 74 See Jens-Peter Schneider, ‘Basic Structures of Information Management in the European Administrative Union’ (2014) 20 Eur Public L 89, 98– 105; and Kristina Heussner, Informationssysteme im Europäischen Verwaltungsverbund (Mohr Siebeck 2007) 174–82.

238 Databases that individual pieces of information do not reveal.75 Google and other companies make a great deal of money from personalised advertising based on user profiles; intelligence services use the same method. ETIAS and the VIS mean that algorithm-based processing will have its debut in EU migration law. This raises difficult technical and legal questions for future research and policy debates. At the time of writing, the practical rules for screening had yet to be defined. The EU institutions, the ETIAS Central Unit within Frontex, and eu-LISA will fine-tune the methods employed in computerised risk assessments in the years to come, which—besides human rights concerns—are technologically challenging and cost-intensive.76 Screening rules will be developed in two steps. First, the Commission shall define what kind of security, illegal migration, and high pandemic risks are relevant for the automated assessment in the form of a delegated and an implementing act.77 These acts will be published in the Official Journal and will be reviewed on a regular basis.78 A crucial question for the practical functioning of automated processing concerns the level of detail in the definition of risks in the forthcoming delegated and implementing acts. Secondly, the ETIAS Central Unit will define, together with screening boards comprising national experts, ‘specific risk indicators’, which will inform the automated assessment of travel authorisation and visa applications on the basis of algorithms developed for that purpose, possibly through the use of artificial intelligence.79 The author admits that he is unsure whether the algorithm and artificial intelligence will be used primarily for developing the list of ‘specific risk indicators’ at an intermediate level of abstraction, or whether they will similarly be applied to process individual applications in light of all the available information.80 In the latter case, automated processing would reach further, as it commonly does in other domains. Specific risk indicators are to be compiled from among a list of potential indicators, although the legislation is open to interpretation whether this list is enumerative or not.81 Publication of the risk indicators is not foreseen, and the legislature may not have had a very precise idea of their appearance. In short, much will depend on how the rules are put into effect. An assessment of automated processing in light of human rights law will have to consider a landmark ruling on the use of passenger name records within the Schengen area. The Grand Chamber laid down standards for artificial intelligence. State authorities must be able to influence the relevant standards and to perform an individual review by non-automated means whenever a decision negatively affects a person.82 The ETIAS and VIS Regulation comply with these judicial standards as a matter of principle, although future cases may develop additional requirements or translate these findings into detailed prescriptions. 75 See https://en.wikipedia.org/wiki/Profiling_(information_science) (accessed 1 March 2023). 76 See ‘Artificial Intelligence in the Operational Management of Large-Scale IT Systems’ (eu-LISA, July 2020) 15–23. 77 See ch 2.4.3; ETIAS Regulation (EU) 2018/1240, art 33(2), (3); and VIS Regulation (EC) No 767/2008, art 9j(2), (3), as amended by Regulation (EU) 2021/1134. 78 As required by Opinion 1/15 Draft Agreement between Canada and the EU EU:C:2017:592, paras 168–74. 79 See ETIAS Regulation (EU) 2018/1240, arts 7, 9, 33(4), (5); and VIS Regulation (EC) No 767/2008, art 9j(4), (5), 9k, as amended by Regulation (EU) 2021/1134. 80 See also Susie Alegre, Julien Jeandesboz, and Niovi Vavoula, ‘European Travel Information and Authorisation System (ETIAS)’ (Study for the European Parliament, PE 583.148, April 2017) 23–27. 81 See ETIAS Regulation (EU) 2018/1240, art 33(4); and VIS Regulation (EC) No 767/2008, art 9j(4), as amended by Regulation (EU) 2021/1134. 82 See Case C-817/19 Ligue des droits humains EU:C:2022:491, paras 179–210.

Prospect of Interoperability and Artificial Intelligence 239 An obvious danger of screening is reliance on group-specific considerations, which define the very idea of profiling. The legislature foresaw this challenge and laid down an abstract obligation that risk indicators must not be based ‘solely’ on a person’s sex, although sex may be an element amongst other when defining indicators.83 Thus, a risk indicator such as ‘young men of X-Y years working in certain job categories in countries A and B (or parts thereof)’ appears not fall foul to the prohibition of profiling ‘solely’ on grounds of sex. Legislation seems to authorise, in other words, indirect differentiation on grounds of sex, while forbidding direct discrimination.84 It will have to be tested whether such indirect discrimination on grounds mentioned in Article 21 CFR can be justified in light of overriding public policy objectives.85 The additional requirement that the risk assessment ‘shall in no circumstances be based on information revealing’, amongst others, a person’s colour, race, language, religion, or membership of a national minority similarly contrasts with the explicit permission to use the city of residence and, putatively, place of birth.86 After all, residence may often indicate the language, religion, and the like. Such proxies are apparently permitted to feature among the specific risk indicators.87 Moreover, those with technical expertise may try to decipher the effects of the algorithm for the identification of the risk indicators; there is ample room for hidden and indirect forms of reliance on questionable standards in the design of risk indicators.88 We can expect the EU to explore the enhanced and innovative use of artificial intelligence in the years to come, building upon lessons learned and the tradition of ‘function creep’. Agencies promote forward-looking thinking; they have put the vision of a ‘European System for Traveller Screening’ (ESTS) on the agenda, with advanced profiling using data from diverse sources.89 Moreover, negotiations on the EU’s future Artificial Intelligence Act will have repercussions for automated processing for diverse purposes.90

9.4.2 Quality concerns and implementation deficits Technological solutions are often hailed, by state authorities, as a panacea to increase effectiveness, while critics dread the danger of omniscient ‘big brother’. Both are right, yet digital technologies do not always live up to the promise—or menace—for the simple reason that they depend on the quality of the dataset. Moreover, it requires considerable managerial skills and financial resources to set up the supranational central unit, to establish the 83 ETIAS Regulation (EU) 2018/1240, art 33(4)(a), (5); VIS Regulation (EC) No 767/2008, art 9j(4)(a), (5); and Vavoula (n 14) 470–74. 84 See also FRA, ‘The Impact on Fundamental Rights of the Proposed Regulation on the ETIAS’ (Opinion 2/ 2017, June 2017) 22–24. 85 See ch 10.4.1; CFR, art 52(1); Case C-220/17 Planta Tabak EU:C:2019:76, para 43; and Case C-236/09 Association belge des Consommateurs Test-Achats ASBL and others EU:C:2011:100, paras 28–24. 86 See ETIAS Regulation (EU) 2018/1240, art 33(4)(b), (5); and VIS Regulation (EC) No 767/2008, art 9j(4) (b), (5), as amended by Regulation (EU) 2021/1134; the place of birth is collected under Article 17(2)(a) ETIAS Regulation but not listed as a potential risk indicator among the reference list in Article 33(4). 87 See ch 10.4.4; and Draft Agreement between Canada and the EU (n 78) paras 164–67. 88 See generally Carsten Orwat, ‘Risks of Discrimination through the Use of Algorithms’ (Study Compiled with a Grant from the Federal Anti-Discrimination Agency, September 2019). 89 See ‘Final Report Future Group on Travel Intelligence and Border Management’ (Council doc 6767/22, 3 March 2022). 90 See Proposal for a Regulation laying down harmonised rules on artificial intelligence, COM(2021) 206 final.

240 Databases technical interface with national units, and to connect domestic authorities and consulates across Europe and worldwide. The example of the SIS II illustrates that these challenges can be difficult to live up to: the Commission delivered the second generation six years late, failed to supervise subcontractors, and did not foresee cost overruns from €23 million to €189 million.91 Costs for later projects were calculated more realistically. ETIAS, for instance, was predicted to cost €212 million from the beginning, while the price tag for the EES was calculated to exceed €1 billion as a result of major technical adjustments at the national level.92 Besides the initial costs of setting up a database, annual costs will incur for operation and maintenance. States will receive some financial support from the funds mentioned in Chapter 7.4. At the same time, digitalisation might help to save some money by replacing personnel or allow it to work more efficiently when tasks which had been performed manually beforehand are taken over by machines. To gather managerial and technical expertise for establishing and operating the databases, eu-LISA was created. Delays in implementing the broader interoperability package, including ETIAS and the EES, were real but less dramatic than for the SIS II. Practical effects depend on the amount and quality of information and on the willingness of national authorities to use it. In that respect, major discrepancies have always existed. Initially, the (illegal) German and Dutch practice of systematic entry bans for all unsuccessful asylum applicants caused major discrepancies between the numbers of SIS entries per country,93 and the VIS was not used extensively by consulates abroad in the first years of operation.94 Today, the number of searches shows major differences if, by way of example, Croatia and Romania consult the SIS far more often than other countries of a similar size and geographic location.95 For many years, the effectiveness of Eurodac was hampered by the cold boycott on the part of countries at the external borders. One of the original ideas behind the ‘hotspot’ concept, devised in the spring of 2015, was to ensure systematic fingerprinting.96 On the whole, officials seem to use the databases more and more.97 Interoperability will boost consultation further if automatic cross-checks increase the chance of a ‘hit’ that may be used by domestic authorities. Wrong or incomplete entries have plagued the databases from the beginning. Prominent alphanumeric deficits include spelling errors, incorrect transcription of names into the Latin alphabet, or the recording of birth dates when the precise date is unknown.98 Biometric identifiers are equally not immune to mistakes, in the form of either ‘false negatives’ to the benefit of individuals, when the database does not recognise a similarity, or ‘false positives’ when it wrongly reports a ‘hit’.99 To be sure, manual processing and legal remedies allow for 91 See Court of Auditors, ‘Development of the second generation SIS II’ (Special Report 03/2014, 14 May 2014). 92 See Proposal for an ETIAS Regulation (n 60) 21–22; and Julien Jeandesboz and others, ‘The Commission’s Legislative Proposals on Smart Borders. Their Feasibility and Costs’ (Study for the European Parliament, PE 493.026, October 2013). 93 See ch 16.4.1; and Brouwer (n 11) chs 12–14. 94 See Kay Hailbronner and Imelda Higgins, ‘General Rapporteur Report’ in Imelda Higgins (ed), Migration and Asylum Law and Policy in the European Union: FIDE 2004 National Reports (CUP 2004) 455, 456–61. 95 See ‘2020 Statistics’ (n 29) 7, for manual searches. 96 See Commission, ‘Communication: A European Agenda on Migration’ COM(2015) 240 final, 6, 13. 97 See Court of Auditors, ‘EU information systems supporting border control’ (Special Report 20/2019, 11 November 2019). 98 ibid 29–32. 99 See Jillyanne Redpath, ‘Biometrics and International Migration’ in Ryszard Cholewinski and others (eds), International Migration Law (TMC Asser Press 2007) 427, 441–43.

Prospect of Interoperability and Artificial Intelligence 241 the correction of such mistakes, but the danger of practical disadvantage or harm remains. Legislative instruments oblige domestic authorities to submit quality data, and eu-LISA performs a basic quality check, which could possibly be rendered more effective through artificial intelligence.100 Having said this, the prospect of improvement is cold comfort for those affected by inaccurate data.

9.4.3 The spectre of ‘interoperability’ Michel Foucault popularised the ideal-typical prison architecture ‘panopticon’ as an allegory for administrative techniques to control the population in a disciplinary society.101 The panopticon is—like the ‘big brother’ formula—frequently referred to in debates about mass surveillance in the digital age. In the EU context, critics often employ the language of ‘fortress Europe’ for physical coercion at the external borders, whereas the panopticon terminology stands for databases.102 These comparisons are powerful metaphors, even though the element of invisibility of information gathering contrasts with the original usage of the panopticon allegory for the physical prison architecture in the Foucauldian oeuvre. Notwithstanding this caveat, the panopticon concept is particularly well-suited for discussing interoperability. The technical linkage of several databases corresponds to the architectural outline of a panopticon where state agents have an overview in multiple directions from a single viewpoint. There are no blind spots if the information is combined. The Commission emphasised the objective of seamless data collection when it proposed ETIAS and the EES, together with the prospect of interoperability. It presented a graphic of a jigsaw puzzle with missing pieces closing potential ‘information gaps’.103 Critics will argue that this positive framing disguises the controversial character of full supervision, which underlies the ‘gap’ rhetoric. Interoperability has been on the political agenda for decades to optimise the use of the existing databases; they were presented as ‘silos’ to be connected in the name of efficiency and effectiveness.104 The policy debate gathered momentum after the terrorist attacks and the asylum policy crisis in 2015/16.105 Two regulations were adopted in 2018 to create the technical infrastructure for connecting the different databases.106 They have to be read in combination with database-specific rules indicating for which purposes the information stored in a database can be accessed. Indeed, interoperability does not automatically extend access rights, which, rather, depend on the small print of the legislation. By way of example, information on travel documents submitted during visa procedures can be accessed for return, while job categories collected under the ETIAS Regulation cannot usually be retrieved

100 By way of example see SIS Border Checks Regulation (EU) 2018/1861, arts 13–14, 15(4), (5); see also ‘Artificial Intelligence’ (n 76) 26–28. 101 See Michel Foucault, Surveiller et punir. Naissance de la prison (Gallimard 1975). 102 See Broeders (n 3) 28–34. 103 See Commission Proposal of 2018 (n 52) 7–8. 104 See Commission, ‘Enhanced Interoperability’ (n 9); and Brouwer (n 11) 132–35. 105 See Vavoula (n 8) ch 8. 106 See Borders and Visa Interoperability Regulation (EU) 2019/817; and Asylum and Migration Interoperability Regulation (EU) 2019/818.

242 Databases for other purposes.107 EU institutions were right when pointing out that interoperability does not create new access options as such. Nevertheless, interoperability will support a substantial increase in data usage for diverse purposes once it starts being implemented at the end of 2023, according to the latest schedule. First, access rights had already been extended on earlier occasions, and the prospect of interoperability fed the appetite further.108 ETIAS and the EES are a case in point, since a political justification for setting them up was the vision of interoperability under the ‘smart borders’ initiative. Their creation was motivated, in part at least, by the desire to prepare the ground for holistic datasets. Technical linkage under the interoperability package is the proverbial cherry on top, completing a process that had been in the making for years. Secondly, interoperability may have self-enhancing effects from a practical perspective, since it makes it easier for officials to access the information stored in the different databases. The ‘European Search Portal’, created by the interoperability package, facilitates ‘the fast, seamless, efficient, [and] systematic’109 access by domestic authorities and EU agencies to several datasets simultaneously. It replaces, in other words, the need for multiple search entries in separate databases with the option of synchronised access to the EES, VIS, ETIAS, Eurodac, SIS, and ECRIS-TCN. Doing so reduces transaction costs and increases the chance of a ‘hit’ on the basis of a single search query, thus rendering it more likely that national authorities and EU agencies use the databases in practice—both as information recipients and in terms of entering data. Interoperability may herald the arrival of a turning point when digital services overtake physical and manual operations as the hallmark of migration control activities. Thirdly, interoperability is more than a technical linkage of existing databases. A ‘Common Identity Repository’ will be established as a central infrastructure. It will contain an individual file on each third country national with core data such as names, age, nationality, fingerprints, photos, or travel document details. This information will be gathered from the different databases and will be stored for as long as the corresponding data are stored in at least one of the databases.110 Together with a ‘shared biometric matching service’ and a ‘multiple identity detector’, the Common Identity Repository will help to detect multiple identities and identify third country nationals withholding their identities.111 It remains to be seen whether the centralised storage of templates of the information contained in the individual databases entails that the central unit controls the personal data, thus requiring the introduction of procedural safeguards at the supranational level.112 EU institutions opted against that solution for the time being; they entrusted Member States with ensuring compliance with data protection standards.113 Such decentralised legal oversight is a general attribute of the justice and home affairs databases. 107 cf ETIAS Regulation (EU) 2018/1240, art 11(2)–(6), as amended by Regulation (EU) 2021/1152 (n 56); and VIS Regulation (EC) No 767/2008, arts 9(4)(l), 22g–22j, 22o(5), 22r(5), as amended by Regulation (EU) 2021/ 1134. 108 See Michiel Besters and Frans WA Brom, ‘Greedy Information Technology’ (2010) 12 EJML 455, 461–63. 109 Borders and Visa Interoperability Regulation (EU) 2019/ 817, art 6(1); and Asylum and Migration Interoperability Regulation (EU) 2019/818, art 6(1). 110 ibid art 23(2); and EDPS, ‘Opinion 4/2018 on the framework for interoperability’ (April 2018) 19–21. 111 Borders and Visa Interoperability Regulation (EU) 2019/817, arts 12–26; and Asylum and Migration Interoperability Regulation (EU) 2019/818, arts 12–16. 112 See EDPS (n 110) 24–25. 113 See Borders and Visa Interoperability Regulation (EU) 2019/817, arts 47–48; and Asylum and Migration Interoperability Regulation (EU) 2019/818, arts 47–48.

Respect for Data Protection Standards 243

9.5 Respect for Data Protection Standards An essential feature of data protection laws is their self-sustaining character. Collecting and processing of personal data amounts to an interference with Article 8 CFR and has to be justified autonomously. Third country nationals can complain about storage and usage independent of the administrative decision the information is meant to prepare. Data protection standards can be found in several instruments: abstract human rights norms, the detailed prescriptions of the General Data Protection Regulation, and the legislation on the justice and home affairs databases (9.5.1). Selected questions illustrate what kind of legal problems may surface in the debate about the migration databases (9.5.2). There are several accountability fora promoting compliance with data protection laws. While administrative oversight by independent bodies functions reasonably well, legal remedies play a marginal role in practice (9.5.3).

9.5.1 Principles of data protection For our purposes, the Charter is the central point of reference. Judges in Luxembourg have interpreted the guarantee of data protection and respect for private life under Articles 7 and 8 CFR dynamically, thus replacing the ECtHR as the judicial epicentre of pan-European data protection standards.114 Judgments declaring illegal the retention of communication data and reaffirming a ‘right to be forgotten’ with regard to sensitive information on the internet are symbols of the dynamism of the CJEU.115 Nevertheless, the case law of the ECtHR in Strasbourg remains relevant as a benchmark for general principles guiding the assessment of individual cases.116 These human rights requirements establish a minimum level of protection. Human rights are complemented by statutory requirements in the General Data Protection Regulation (EU) 2016/679 (GDPR) and the legislation on the databases.117 The relationship between the GDPR and the specific instruments is governed by a mix of deference and displacement. The regulations establishing the SIS, ETIAS, and the EES say explicitly that the rules in the GDPR shall apply,118 although individual provisions of the GDPR may still be supplanted by leges speciales in the legislation on databases. At a practical level, one should check the contents of the more specific legislation first to see whether it contains detailed prescriptions, which must be related to the contents of the GDPR in a second step. A good example is the transfer of personal data to third states for identification purposes in

114 The ECJ does not usually distinguish between Articles 7 and 8 CFR. 115 See Joined Cases C-293/12 and C-594/12 Digital Rights Ireland EU:C:2014:238; and Case C-131/12 Google Spain SL and Google Inc EU:C:2014:317. 116 See the overview by Vavoula (n 8) ch 1; and Stephen Kabera Karanja, The Schengen Information System and Border Control Co-operation (Martinus Nijhoff 2008) ch 4. 117 The Convention for the Protection of Individuals with Regard to Automatic Processing of Personal Data (adopted 20 January 1981, entered into force 1 October 1985) ETS No 108 is an additional source of inspiration, although secondary legislation is often more generous and despite the EU not being bound directly, as explained in ch 18.2.1. 118 See SIS Border Checks Regulation (EU) 2018/1861, art 51; ETIAS Regulation (EU) 2018/1240, art 56; and EES Regulation (EU) 2017/2226, art 49; as well as the reference to the former Directive 95/46/EC in Eurodac Regulation (EU) No 603/2013, recital 38; and VIS Regulation (EC) No 767/2008, recital 17, with later amendments.

244 Databases the context of return, which requires a holistic analysis of the general rules in the GDPR and the database-specific regulations.119 Information stored in the databases qualifies as ‘personal data’ and is, therefore, subject to data protection laws.120 By contrast, the situation may be different for data collected by drones during border surveillance in the context of EUROSUR; such situational data need not amount to ‘personal data’ subject to enhanced legal safeguards. In line with human rights, any collection, storage, or access of personal data amounts to an interference that has to be provided for by law and be proportionate.121 As a result, the legitimacy of data storage for a specific purpose does not imply that it can be accessed for any purpose whatsoever. Rather, access by means of consultation, retrieval, or disclosure must be justified as an interference in its own right and has to be provided for by law for each purpose separately. The almost bewildering complexity of the secondary legislation partly results from that obligation. What is meant to support effective data protection can have the opposite effect in practice if the intricacy of the data protection rulebook renders it difficult for newcomers to grasp the legal framework. Article 5 GDPR lays down six principles, which are derived from human rights law and govern the collection and processing of personal data: purpose limitation; data minimisation; accuracy; lawfulness, fairness, and transparency; storage limitation; as well as integrity and confidentiality.122 By way of example, the principles of lawfulness and purpose limitation imply that access rights are laid down in detailed provisions, while accuracy aims at preventing the upload of incorrect information. Secondary legislation reinforces the accuracy principle through an explicit obligations of domestic authorities to ensure data quality and to rectify or delete mistakes.123 Even so, quality concerns have accompanied, as we have seen, the migration databases from the beginning. By contrast, there have been no major security incidents where a database was accessed or even falsified by third parties, although such data breaches remain a threat. Interoperability increases that risk, since more data could be accessed simultaneously.

9.5.2 Core elements of the human rights assessment Any holistic assessment of the justice and home affairs databases in light of data protection standards remains a Herculean task, precisely because the principles of purpose limitation and data minimisation require careful distinction between the personal data concerned and the diversity of access rights. The storage of fingerprints of asylum seekers for identification in response to tangible problems in applying the Dublin III Regulation is not the same as the use of the same fingerprints during return procedures, let alone the indiscriminate collection of fingerprints of tourists and other bona fide travellers. Similarly, information on travel documents should not be measured against the same yardstick as entries

119 See eg GDPR, arts 44–46; and the mostly concurring leges speciales in EES Regulation (EU) 2017/2226, art 41; and Eurodac Proposal of 2016 (n 20) arts 37–38. 120 See GDPR, art 4(1); and Kabera Karanja, Schengen Information System (n 116) 137–42. 121 See CFR, art 52(1); ECHR, art 8(2); S and Marper v United Kingdom App nos 30562 and 30566/04 (ECtHR [GC], 4 December 2008) §§ 95–104; and Draft Agreement between Canada and the EU (n 78) paras 133–41. 122 See also Brouwer (n 11) 204–21; Kabera Karanja (n 116) 144–77; and Heussner (n 74) ch 4. 123 See eg SIS Border Checks Regulation (EU) 2018/1861, arts 6–17, 44, 53.

Respect for Data Protection Standards 245 on education levels or job categories. Proportionality requires us to assess these situations independently, thus supporting differentiated outcomes. For our purposes, a couple of observations may help readers to acquaint themselves with pertinent questions. Most data is meant to support the identification of individuals (name, place of birth, travel document details, fingerprints) or concerns migration-specific evidence (date of border crossing, asylum applications, migration status). Both the CJEU and the ECtHR are comparatively generous in terms of allowing the collection and storage of such information. Document security, identification, travel-related evidence, and, more generally, the correct application of migration law and the prevention of irregular entry and stay were recognised as legitimate public policy objectives.124 That is certainly no carte blanche, but it indicates that the creation of the databases to facilitate identification and to support the effectiveness of EU migration law can be subject to more generous standards than data collection for internal purposes. Emphasis on identification and migration-specific purposes marks an important difference to legislation that allowed state authorities to design personality profiles. The Court considered illegal the retention of communication data, since they provided an intimate insight into the private life and targeted the whole population indiscriminately.125 By contrast, fingerprints, facial images, migration status, or travel data are undoubtedly sensitive personal data, but their abstract character distinguishes them, in the eyes of judges, from the intimate nature of other information.126 Judges in Luxembourg distinguish between databases concerning ‘practically the entire European population’ and information specifically on third country nationals.127 Interdisciplinary analyses may criticise such distinction as the securitarian construction of the ‘risky foreigner’ posing a generic threat,128 but this does not reverse the outcome of Court judgments allowing the collection of migration-related information independent of individualised risk assessments. Academic observers have highlighted repeatedly that we lack reliable statistics to what extent the justice and home affairs database contribute to the realisation of public policy objectives, thus questioning compliance with the principles of data minimisation and proportionality.129 One may try to argue, moreover, that the end of the compartmentalisation as a result of interoperability supports stricter standards.130 Doing so remains an uphill struggle, however, since the Common Identity Repository, in particular, will essentially store information in support of identification, which judges have deemed unproblematic so far—as opposed to information on education levels or job categories in ETIAS and the VIS, which are of a more general character. Closer inspection of the ETIAS and VIS Regulation illustrates that their contents has been modelled on a previous ruling on ‘passenger name 124 See Case C-70/18 A and others EU:C:2019:823, paras 47–49; Case C-101/13 U EU:C:2014:2249, paras 41–47; Case C-524/06 Huber EU:C:2008:724, paras 58–62; Draft Agreement between Canada and the EU (n 78) para 150; and Ligue des droits humains (n 82) para 161. 125 See Digital Rights Ireland (n 115) paras 26–27, 56–59. 126 See A and others (n 124) paras 58–59; Ligue des droits humains (n 82) para 120; S and Marper v United Kingdom (n 121) §§ 68, 78–86; and Julius Buckler, ‘Auf dem Weg zu einer digitalen europäischen Grenzkontrollarchitektur?’ [2018] Bayerische Verwaltungsblätter 73, 76–77. 127 Contrast Digital Rights Ireland (n 115) para 56 to the relative generosity of Draft Agreement between Canada and the EU (n 78) paras 186–89, 197–99; A and others (n 124) paras 47–49, 58–62; and Huber (n 124) paras 58–62. 128 See Vavoula (n 17) 160; and Thomas Gammeltoft-Hansen, ‘Filtering Out the Risky Migrant. Migration Control, Risk Theory and the EU’ (AMID Working Paper 52, 2006). 129 See Brouwer (n 11) 148–52; and Vavoula (n 17) 156–60. 130 See Vavoula (n 8) ch 8.5.

246 Databases records’.131 This illustrates that the success of future legal challenges against automated processing is not a foregone conclusion. Arguably, purpose limitation is the most relevant protection standard when assessing the justice and home affairs databases. ‘Function creep’ has been a defining feature of legislative developments over the past two decades. Ever more authorities were granted access for diverse law enforcement and migration control objectives. Purpose limitation implies that each extension of access rights must be justified independently: the collection of information for entry clearance is not the same as usage of the same data for migration control within the territory, or storage after departure to prevent irregular migration threats in case of re-entry.132 Future analyses may wish to concentrate on these extended access rights. While judges insist on strict procedural safeguards when allowing access for law enforcement,133 they seem to be more lenient when it comes to migration control. The Court generally accepted the storage of data of Union citizens and the collection of biometric identifiers of third country nationals to prevent identity fraud and irregular entry or stay. It did not require an individualised assessment of risks, let alone procedural safeguards mirroring judgments on criminal matters.134 Nevertheless, one may question, for instance, whether the indiscriminate storage of data of bona fide travellers complies with human rights. Information in the EES, VIS, and ETIAS will be deleted after three or five years, even if the third country national leaves the Schengen area.135 The outcome of this proportionality assessment is difficult to predict. On the whole, the comparatively weak constitutional status of third country nationals, described in Chapter 1.3, appears to inform the case law. A series of early judgments had emphasised the special status of Union citizens and their family members; entry bans in the SIS were required not to undermine free movement.136 Judges similarly rejected the use of national databases with regard to Union citizens for law enforcement in light of Article 18 TFEU, unless similar systems exist for nationals.137 Such privileged treatment was not, however, extended to third country nationals, thus reinforcing the judicial message about graded non-discrimination standards, which will be presented in Chapter 10.4.1 on overarching features of secondary legislation. The ECtHR rejected a complaint against an entry ban which had prevented a highly qualified third country national from entering the Schengen area, since states ‘benefit from a principled margin of appreciation when it comes to entry into the territory’.138

131 See Draft Agreement between Canada and the EU (n 78). 132 See generally ibid paras 197–201, 204–206. 133 See ibid para 202; and Joined Cases C-203/15 and C-698/15 Tele2 Sverige, Watson and others EU:C:2016:970, paras 118–20. 134 See ch 10.5.2; Huber (n 124) paras 58–62; and A and others (n 124) paras 47–57; both cases concerned national databases not subject to EU harmonisation. 135 See EDPS, ‘Opinion 6/2016 on the second EU smart borders package’ (September 2016) 9–11. 136 See Case C-503/03 Commission v Spain EU:C:2006:74; and Tim Eicke, ‘Paradise Lost? Exclusion and Expulsion from the EU’ in Kees Groenendijk and others (eds), In Search of Europe’s Borders (Kluwer 2003) 147, 155–64; databases discussed in this chapter usually exempt Union citizens and their family members ratione personae. 137 See Huber (n 124) paras 69–81; and Kay Hailbronner, ‘Die Speicherung personenbezogener Daten im AZR’ [2009] Zeitschrift für Ausländerrecht 178–82. 138 Dalea v France App no 964/07 (ECtHR, 2 February 2010) § 2 (own translation; only available in French); see also Buckler (n 126) 76.

Respect for Data Protection Standards 247

9.5.3 Preventive and reactive supervision Databases share the basic characteristics of the administrative dimension as a composite system described in Chapter 7.1. They are set up on the basis of supranational legislation, but their daily operation relies on domestic authorities entering and retrieving data. Decentralised implementation also defines the technical architecture, which combines a central system, operated by eu-LISA, with national systems organised by the Member States.139 In contrast to what some might intuitively think, the role of the central system is essentially limited to provide the technical infrastructure for information exchange and to secure basic data quality and security, while the responsibility for entering and consulting the data rests with the Member States. Domestic authorities are either explicitly defined as ‘controller’ in the meaning of the GDPR or implicitly assigned that function.140 That is relevant in so far as the controller assumes the legal responsibility to inform individuals, grant them access to personal data, rectify, or delete information, and provide for legal remedies.141 The only exception is ETIAS in relation to which Frontex operates the central unit and serves as the controller of information submitted by individuals for travel authorisation purposes. Frontex is responsible, as a result, for dealing with requests for access, rectification, or erasure.142 The legislation remains vague as to whether individuals have a legal remedy before the General Court in case Frontex does not adequately respond to their requests.143 That being said, the legal responsibility of the central unit under the ETIAS Regulation concerns the collection and storage of personal data only. By contrast, national authorities will reject travel authorisation in case the algorithm-based screening reports a ‘hit’; national courts have jurisdiction for appeals in these cases.144 Actions for annulment against the list of specific risk indicators for automated processing under the ETIAS and VIS Regulation are likely to be declared inadmissible in accordance with Article 263(4) TFEU, as individuals are expected to challenge the implementing measure in the form of refusal of travel authorisation. Domestic courts may consult the CJEU to determine whether the list of specific risk indicators complies with the Charter. With regard to other databases, the legislation states unequivocally that the responsibility for data protection lies with domestic authorities. They have to respond to requests for access, rectification, or erasure in accordance with the GDPR and the legal speciales on the justice and home affairs databases.145 Complex situations of transnational legal effects are inherent in decentralised implementation if, by way of example, a French consulate rejects a visa application on the basis of an entry ban issued by a Spanish authority. Corresponding 139 By way of example see EES Regulation (EU) 2017/2226, art 7; and SIS Border Checks Regulation (EU) 2018/ 1861, arts 4, 6. 140 See the explicit definition in EES Regulation (EU) 2017/2226, art 57; Borders and Visa Interoperability Regulation (EU) 2019/817, art 40; Asylum and Migration Interoperability Regulation (EU) 2019/818, art 40; as well as the responsibilities of national authorities in Eurodac Regulation (EU) No 603/2013, art 23; VIS Regulation (EC) No 767/2008, arts 37–40, with later amendments; SIS Border Checks Regulation (EU) 2018/1861, arts 52–54; and SIS Return Regulation (EU) 2018/1860, art 19. 141 See GDPR, arts 4(7), 12–18; and the leges speciales on databases. 142 See ETIAS Regulation (EU) 2018/1240, arts 7, 57, 63–64, which partly establish a parallel jurisdiction of national authorities. 143 ibid art 64(4) mentions legal remedies ‘where relevant’; see further ch 6.5.4. 144 ibid arts 25–26, 37–38. 145 See GDPR, arts 12–18; and eg EES Regulation (EU) 2017/2226, arts 50–54.

248 Databases scenarios of transnational judicial oversight were discussed in Chapter 7.3, which demonstrated myriad challenges. There had always been a mismatch between the limited number of court proceedings and the ever-increasing size of the databases.146 One reason may be that many disputes are resolved whenever an individual complains with the competent authorities.147 Nevertheless, the relative scarcity of judicial proceedings stands out. Individual rights are comparatively strong on paper but rarely exercised in practice.148 Lack of expertise among legal experts may be a factor explaining why legal remedies are not used more widely. Moreover, third country nationals are primarily interested in challenging administrative decisions the databases prepare; they do not seem to be concerned much with data storage in the absence of tangible effects in practice. In that respect, the justice and home affairs databases benefit from a similar negligence as massive data collection by private companies such as Google or Facebook. The practical weakness of judicial ex post oversight explains why preventive control by the European Data Protection Supervisor (EDPS) and corresponding national authorities is crucial. They serve as instruments of ex ante governance creating public awareness about human rights impediments and promoting risk minimisation. Throughout this chapter, we have referred to several opinions of the EDPS, which generally play an important role during the legislative process.149 It will also supervise the everyday functioning of the databases in the years to come, jointly with national data protection officers.150 Their reports and recommendations are an important source of inspiration for observers and decision- makers alike.

9.6 Summary Databases are a dynamic segment of European migration law which should receive more attention. They have featured on the political agenda ever since Member States started recording entry bans in the SIS. Databases have expanded significantly over the years. Entry bans in the SIS and fingerprints in Eurodac were the beginning of ‘function creep’ with the creation of additional databases, the collection of more information, and an extension of access rights. Alphanumeric records such as the date of birth were complemented by biometric identifies like fingerprints or scanned copies of passports, thus facilitating the identification of third country nationals and their potential return. The ‘fight against terrorism’ was an important impetus for policy change, before better migration control became an end in itself from 2015 onwards. Access rights were extended to law enforcement authorities first and have been buttressed for migration control purposes more recently. Databases

146 See Helen Staples, ‘Adjudicating the External Schengen Border’ in Groenendijk and others (n 136) 215, 225–35. 147 See SIS II Supervision Coordination Group, ‘Report on the Exercise of the Rights of the Data Subject in the SIS’ (October 2014). 148 See Evelien Brouwer, ‘Interoperability of Databases and Interstate Trust. A Perilous Combination for Fundamental Rights’ Verfassungsblog (25 May 2019). 149 For mandatory consultation see Regulation (EU) 2018/1725 on the protection of natural persons with regard to the processing of personal data by the Union institutions [2018] OJ L295/39, art 42(2). 150 See eg Eurodac Regulation (EU) No 603/2013, arts 31–32.

Summary 249 operate in the background as an invisible infrastructure; as a result, their impressive build- up receives little political and academic scrutiny. ETIAS and the Entry/Exit System are scheduled to complement the SIS, Eurodac, and the VIS at the end of 2023. The new digital architecture will be reinforced by legislative amendments, which have been adopted for the SIS and VIS and are still under discussion for Eurodac. The end result will be an almost seamless surveillance of entries, exits, and migration statuses. Regular and irregular border crossings of asylum applicants and other third country nationals will be monitored by the Entry/Exit System and Eurodac. Pre-arrival and post-entry information on visa applicants, visa-exempt foreigners, and residence permits issued domestically will be contained in ETIAS and the VIS. Eurodac is foreseen to collect asylum-related decisions, and the SIS will gather entry bans and return decisions. The information contained in the databases can be accessed for diverse aspects of migration law, ranging from refusal of entry and determination of asylum jurisdiction over the identification of illegal stay and security threats to improving return. An important motivation for legislative change was the prospect of interoperability, which is currently schedule to be put into practice by the end of 2023 after almost two decades of political debate and technical preparation. The interoperability package establishes an overarching technical infrastructure connecting the information stored in the different databases. Interoperability may have self-enhancing effects from a practical perspective, by making it easier for officials to access different databases simultaneously on the basis of a single search query. One step further, digital processing will start defining migration control besides physical and manual operations. A critical novelty is the automated risk assessment. ETIAS and VIS will use algorithm-based specific risk indicators for the computer-based screening of applications for travel authorisation and Schengen visas. Practical and legal effects will depend on implementing rules, which the EU institutions will fine-tune over the years. Automated processing requires close scrutiny. Human rights law necessitates detailed provisions on the kind of personal data that may be collected, storage periods, and access rights. The ensuing complexity of the legislation can complicate critical scrutiny for the simple reason of limited user-friendliness. Several rulings of the CJEU and the ECtHR indicate that judges are comparatively generous when it comes to the collection of personal data facilitating the identification of third country nationals and supporting the application of migration law; they distinguished migration- specific information from other sensitive personal data. That does not imply, however, that the considerable extension of access rights cannot be challenged on human rights grounds. Purpose limitation is an essential data protection standard to test the limits of EU legislation. The legitimacy of data storage for one purpose does not imply that it can be accessed for other objectives as well. Rather, the usage of personal data for additional purposes is an interference in itself which requires justification. Court oversight has played a marginal role so far, thus rendering the control function of the European Data Protection Supervisor and national officers essential for effective supervision.

PART II

SEC TOR A L L E GI SL AT ION A ND P OL IC I E S

10

General Features of EU Legislation EU migration law comprises roughly twenty pieces of core legislation, several dozen smaller instruments, and more than 200 judgments of the Court in Luxembourg. As a result, it is notoriously difficult to keep an oversight. The increasing complexity of directives and regulations complicates this venture further. The size of the Dublin Regulation more than doubled from ten to twenty-three single-spaced pages in the Official Journal between 2003 and 2013. Veritable legislative ‘monsters’ are the Students and Researchers Directive (EU) 2016/801, with thirty-two pages and the Frontex Regulation (EU) 2019/1896, with 100 pages, excluding annexes—contrasting with the lean and comparatively easy to read seven and ten pages of the Family Reunification Directive 2003/86/EC and the original Long- Term Residents Directive 2003/109/EC. Even those with years of experience will occasionally get lost in a legislative jungle of highly complex provisions, which can be exacerbated by the lack of coordination between the instruments. Against this backdrop, this introductory chapter to the second part will present elements that feature in several pieces of legislation. Identifying overarching themes has several advantages. First, it allows us to draw parallels between sector-specific rules, thus improving our understanding of the legislation by means of horizontal cross-fertilisation: judicial practices and academic research on instrument A can be useful for the interpretation of instrument B. Secondly, a horizontal analysis supports the systemic coherence of Union law as a self-sufficient ‘legal order’, in accordance with the comments in Chapter 6.1. Thirdly, an emphasis on general features can be a matter of self-interest. The sheer complexity of the legal material means that experts in European migration law, both established and aspiring, should use scarce resources effectively. Focusing on overarching themes allows us to cope better with the intricate patterns and convoluted character of the supranational legislation. In contrast to Union citizenship, third country nationals require residence permits, although they may benefit from individual rights to be issued with one if they fulfil the conditions laid down in secondary legislation (10.1). Recurring elements of secondary legislation are the ‘sufficient resources’ requirement (10.2) and the ‘public policy’ exception (10.3). Judges in Luxembourg and Strasbourg tread carefully when applying the human right to equality before the law, thus complicating the contestation of unequal treatment (10.4). Closer inspection of the judicial practice on proportionality shows built-in ambiguities as to whether abstract legislative rules are permissible or not (10.5). Along similar lines, the implications of the provisions on more favourable domestic rules differ in substance and effects (10.6). Member States may exceptionally deviate from supranational legislation by invoking the internal security exception in Article 72 TFEU (10.7), or by relying on the general principle of ‘abuse’ (10.8). This chapter will conclude with practical tips for dealing with secondary legislation (10.9).

European Migration Law. Daniel Thym, Oxford University Press. © Daniel Thym 2023. DOI: 10.1093/oso/9780192894274.003.0011

254 General Features of EU Legislation

10.1 Acquisition of Residence Permits Member State authorisation is a conditio sine qua non for legal residence for most third country nationals. Administrative decisions by domestic authorities on entry and stay are usually called ‘residence permits’ or ‘visas’ in the legislation discussed in this volume, notwithstanding terminological ambiguities (10.1.1). A comparison with Union citizenship demonstrates that it was not self-evident that the Court of Justice would recognise the constitutive character of state authorisation as a prerequisite for legal stay (10.1.2). General principles of Union law can be relied upon to challenge excessive fee levels, in accordance with the case law of the Court of Justice (10.1.3).

10.1.1 Of ‘residence permits’ and ‘visas’ Third country nationals who do not possess a valid residence permit or visa are ‘illegally staying’.1 Only a handful of scenarios are exempted from the obligation to present a residence permit or entry visa: asylum seekers, nationals of countries on the ‘white list’ for visa- free travel, family members of Union citizens, and beneficiaries of free movement under association agreements. All other third country nationals will usually require prior state authorisation. EU legislation discussed in this book harmonises many purposes for which states may hand out residence permits. Prominent examples include family reunification, Schengen visas, international protection, or seasonal work. These supranational provisions are complemented by national statuses for other purposes. The harmonisation of these additional national residence permits is essentially limited to a uniform format and security features, thus supporting domestic authorities of other Member States in identifying the authenticity of documents issued elsewhere.2 While the term ‘residence permit’ is used as an overarching category comprising visas in some instruments,3 others distinguish between ‘residence permits’ and ‘visas’: the former are handed out by domestic authorities within the country, whereas the latter are issued by consulates abroad.4 Visas come in the form of ‘regular’ Schengen visas for short stays or ‘national’ visas for other purposes, although that usage is not uniform. Thus, one may refer to a ‘family reunification visa’ for entry authorisation issued by a consulate in a third state, even though the Family Reunification Directive does not distinguish between applications made from abroad or within the territory. This contribution routinely employs the expression ‘visa’ for entry-specific authorisation and uses the term ‘residence permit’ as a generic term, irrespective of the place of delivery.

1

See Return Directive 2008/115/EC, art 3(2). See Residence Permit Format Regulation (EC) No 1030/2002, with later amendments; and ch 12.7. 3 ibid art 1(2). 4 See Schengen Borders Code Regulation (EU) 2016/399, art 6(1)(b). 2

Acquisition of Residence Permits 255

10.1.2 Constitutive state authorisation Treaty rules on cross-border movements are defined by a basic distinction, which was described in Chapter 1.3. While Union citizens benefit from a constitutional promise of free movement, third country nationals can rely on human rights, which rarely oblige states to issue an entry visa or residence permit. That distinction does not imply, however, that Member States retain unfettered discretion to regulate entry and stay. EU legislation lays down a tightly knit network of statutory prescriptions, which often comprise individual rights to receive a residence permit. Applicants can go to court in case of non-compliance. Such individual rights have been recognised by the Court for the admission of asylum seekers,5 family reunification,6 students,7 long-term residents,8 and beneficiaries of international protection.9 The situation under the Visa Code Regulation remains controversial but the practical effects of this uncertainty are limited, as the legislature foresaw legal remedies against refusal anyway.10 Protection under most migration law instruments stretches far beyond that of human rights. A basic assumption of EU migration law comes to the fore through a comparison with Union citizenship. In a foundational judgment on the free movement of workers, the Court found national residence permits to be declaratory. The right to entry and residence of Union citizens is ‘conferred directly by the Treaty’ and is, therefore, ‘acquired independently of the issue of a residence permit’.11 Polish workers may, in other words, take up employment in Ireland without having to ask for prior authorisation with the domestic authorities.12 Non-compliance with an ‘administrative formality’ cannot be sanctioned in a disproportionate manner.13 This interpretation was extended to the EU-Turkey association agreement,14 and some observers had hoped that it would similarly define the judicial approach to the area of freedom, security, and justice.15 Things turned out differently. Judges found application procedures and the ensuing administrative decision to be constitutive in character. To be sure, third country nationals can enforce legislative guarantees to be issued residence permits via domestic courts, but this does not alter the conclusion that the ‘acquisition of that status is not however automatic’.16 Judges did not elaborate on underlying reasons, but there is little doubt that the complex provisions on application procedures and administrative decisions exhibit the choice, on the part of the legislature, that state authorisation shall be constitutive.17 The horizontal

5 See Case C-808/18 Commission v Hungary EU:C:2020:1029, paras 87–103. 6 See Case C-540/03 Parliament v Council EU:C:2006:429, paras 59–60. 7 See Case C-491/13 Ben Alaya EU:C:2014:2187, paras 23–27 with regard to the former Students Directive 2014/114/EC. 8 See Case C-508/10 Commission v Netherlands EU:C:2012:243, para 68. 9 See Case C-373/13 T EU:C:2015:413, para 63. 10 See Case C-84/12 Koushkaki EU:C:2013:862, para 55; and further AG Priit Pikamäe, Opinion in Joined Cases C-225/19 and C-226/19 Minister van Buitenlandse Zaken EU:C:2020:679, points 53–57. 11 Case 48/75 Royer EU:C:1976:57, paras 31–32. 12 Free Movement Directive 2004/38/EC, art 8 approves of registration requirements. 13 See Case C-215/03 Oulane EU:C:2005:95, para 24; and further Gareth Davies, ‘Bureaucracy and Free Movement’ [2003] Nederlands Tijdschrift voor Europees Recht 81. 14 See Case C-351/95 Kadiman EU:C:1997:205, para 51. 15 See Steve Peers, ‘The Court of Justice Lays the Foundations for the Long-Term Residents Directive’ (2013) 50 CML Rev 529, 547–48. 16 Case C-557/17 YZ and others EU:C:2019:203, para 59. 17 By way of example see ICT Directive 2014/66/EU, arts 8, 12–16.

256 General Features of EU Legislation definition of entry conditions and ‘illegal stay’ in the Schengen Borders Code and the Return Directive confirms this finding.18 Third country nationals are ‘illegally staying’ if they are not in possession of a residence permit or visa.

10.1.3 Fee levels In the absence of supranational harmonisation, Member States determine fee levels for resident permits. The Visa Code Regulation (EC) No 810/2009 is unique in setting a uniform amount of EUR 80, which may be increased or lowered depending on whether home states cooperate on return, as we shall see in Chapter 18.3 on the external dimension. All other instruments contain no or vaguely formulated provisions. That did not, however, prevent judges from reviewing Dutch fees for long-term resident status: high fee levels that might deter third country nationals from submitting an application fall foul of the effet utile of EU legislation, even in the absence of an explicit provision.19 That argument can be applied to other instruments. Judges emphasised that Member States have a ‘margin of discretion’,20 subject to two caveats. First, the Court found Dutch practices on long-term resident status to be illegal, since amounts charged were up to seven times higher than for Dutch identity cards (judges were careful not to require a strict parallelism, as domestic ID cards were a relative comparator).21 The message was clear: punitively high charges with dissuasive effects are forbidden. Whether domestic fee levels may have such effects has to be determined under due regard to national specificities. Secondly, the Court declared that an Italian rule was disproportionate, since application fees were calculated so as to include potential costs domestic authorities may have to shoulder beyond the administrative activity necessary to verify whether the conditions for issuing long-term resident status were satisfied.22 Again, judges seem not to have required strict financial equivalence between domestic fee levels and administrative costs. Italy had acted disproportionately, however, by earmarking half of the funds for return purposes.23 Such linkage transgressed state discretion. To complicate things, legislation adopted in recent years contains vaguely formulated provisions on fee levels using differing formulae: while the Single Permit Directive can be read to have established a strict linkage to the actual cost of the administrative service, other instruments mention the principle of proportionality without specifying what it entails.24 We may extract from the case law mentioned previously that the proportionality requirement does not undo a certain leeway. Moreover, the proportionality test remains abstract in line with comments hereinafter. Abstract proportionality does not require—unlike in the 18 See Schengen Borders Code Regulation (EU) 2016/399, art 6(1)(b); and Return Directive 2008/115/EC, art 3(2). 19 See ch 6.4.5; and Commission v Netherlands (n 8) paras 65, 69–70. 20 Commission v Netherlands (n 8) para 64; and Case C-309/14 CGIL and INCA EU:C:2015:523, para 22. 21 Commission v Netherlands (n 8) paras 77–78. 22 See CGIL and INCA (n 20) paras 26–33. 23 ibid paras 24–25, 29–30 relied on proportionality and dissuasive effects as the abstract principle and mentioned the linkage to actual costs with regard to the Italian legislation only. 24 Contrast Single Permit Directive 2011/98/EU, art 10; to Blue Card Directive (EU) 2021/1883, art 12(1); Students and Researchers Directive (EU) 2016/801, art 36; and Seasonal Workers Directive 2014/36/EC, art 19.

Sufficient Resources Requirement 257 case of sufficient resources—individualised outcomes taking into account the specific situation of each applicant.

10.2 Sufficient Resources Requirement Economic self-sufficiency is a critical gateway. Both the Family Reunification Directive and the Long-Term Residents Directive require applicants to demonstrate ‘stable and regular resources’,25 while other instruments insist on ‘sufficient resources’.26 Experts familiar with everyday administrative practices know how terribly relevant the income hurdle can be in real life: it is a common ground for rejection and deters others from applying. Despite the high degree of practical salience, economic self-sufficiency receives comparatively little attention.27 Non-compliance is generally considered to justify the rejection of family reunification to ensure the ‘economic well-being of the country’ in accordance with Article 8(2) ECHR,28 while entry for other purposes is not usually subject to significant human rights constraints in the first place. Residence permits for beneficiaries of international and temporary protection are a rare instance which does not require economic self-sufficiency. Sectoral legislation has in common that it does not define the meaning of ‘regular and stable’ or ‘sufficient resources’, although some instruments allow States to lay down ‘reference amounts’ or to ‘take into account’ minimum wages.29 Judgments delivered so far indicate that the Court strives for a horizontal understanding of these diverse provisions, which are interpreted to pursue the objective of preventing recourse to social assistance.30 As a starting point, the Court recognised that the notions of ‘regular and stable’ and ‘sufficient resources’ are autonomous concepts of Union law that should be interpreted narrowly.31 This entails that Member States cannot define income levels independently, as explained in Chapter 6.4.4 on the doctrinal foundations, although the open texture of EU legislation endows them with a certain leeway. With regard to the small print, the Court required domestic authorities to perform an individualised assessment. Reference amounts can be applied, provided that they do not prevent a different amount from being considered appropriate in light of individual circumstances.32 Article 7(3) Students and Researchers Directive (EU) 2016/801 takes up this judicial insistence on an individualised assessment. With regard to family reunification, the use of the present tense when stating that applicants must ‘have’ stable and sufficient resources was interpreted to refer to the next year or so as the focal point of the forward-looking- prognosis; reliance on the situation prior the application as a proxy was also considered 25 See Family Reunification Directive 2003/86/EC, art 7(1)(c); and Long-Term Residents Directive 2003/109/ EC, art 5(1)(a). 26 See ICT Directive 2014/66/EU, art 6(3); Students and Researchers Directive (EU) 2016/801, art 7(1)(e); Seasonal Workers Directive 2014/36/EC, art 6(3); and Blue Card Directive (EU) 2021/1883, art 8(2)(c). 27 See, however, Eva Hilbrink, Adjudicating the Public Interest (PhD Thesis, Vrije Universiteit Amsterdam 2017). 28 ibid 42–59; Konstatinov v Netherlands App no 16351/03 (ECtHR, 26 April 2013) § 59; and App nos 12510/18 and 57303/18 (pending). 29 By way of example see Students and Researchers Directive (EU) 2016/801, art 7(3); and Family Reunification Directive 2003/86/EC, art 5(1)(a). 30 See Case C-558/14 Khachab EU:C:2016:285, para 46. 31 See Case C-578/08 Chakroun EU:C:2010:117, paras 43, 46; and Case C-302/18 X EU:C:2019:830, para 26. 32 Chakroun (n 31) para 48; and Case C-519/18 Bevándorlási és Menekültügyi Hivatal EU:C:2019:1070, paras 73–75.

258 General Features of EU Legislation acceptable.33 These arguments can be projected upon other instruments using similar language. To answer follow-up references, the Court had recourse to case law on Article 7(1)(b) Free Movement Directive 2004/38/EC as a matter of principle, although the wording and context of migration law instruments may—like in the case of public policy—support sector-specific solutions. By way of example, the notion of ‘resources’, which is semantically coined as ‘income’ in some language versions, allows third country nationals to rely on ‘stable and regular’ support by third parties, such as siblings, in addition or instead of their own revenue.34 By contrast, judges distinguished the notion of financial ‘dependence’ on which family reunification may hinge in special scenarios from free movement case law to take account of the distinct context of applications for family reunification made by refugees. Unlike Union citizens, refugees do not have to demonstrate that a sibling had already received financial support from the refugee while residing abroad; a genuine situation of future dependency was considered sufficient.35 There are weighty reasons that the notion of ‘means of subsistence’ in the Schengen Borders Code Regulation and the Visa Code Regulation should not be aligned with the trend towards horizontal cross-fertilisation of the ‘regular and stable’ or ‘sufficient resources’ requirement in the supranational case law. The wording, the distinct context, and the general scheme of both instruments can be interpreted to support differentiation, since the legislature authorised Member States to set what seem to be precise financial thresholds on the basis of average costs for board and accommodation, to be multiplied by the number of days of the intended stay.36 Thus, the need for individualised assessments could possibly yield to fixed subsistence levels when it comes to rejection at the border and Schengen visas issued abroad.

10.3 Public Policy Exception The ‘public policy’ exception is a core component of EU migration law. It features prominently in most legislative instruments and authorises Member States to refuse or withdraw a residence permit even if the other requirements have been met. These provisions did not come out of the blue. The ‘public policy’ (or: ‘public order’37) exception had been well established long before the EU institutions adopted the legislation discussed in this volume. There is settled case law on the meaning of ‘public policy’ in the single market and for Union citizens. Judges confirmed that this case law can be projected upon migration law as a matter of principle (10.3.1), while insisting on sector-specific outcomes reflecting the contents and context of the migration law instruments (10.3.2).

33 See Khachab (n 30) paras 29–47. 34 See X (n 31) paras 40–43, with regard to the Dutch inkomsten or the German Einkünfte. 35 See Bevándorlási és Menekültügyi Hivatal (n 32) paras 46–53, with regard to Family Reunification Directive 2003/86/EC, art 10(2). 36 See Schengen Borders Code Regulation (EU) 2016/399, arts 6(1)(c), (4), 34(1)(c); and Visa Code Regulation (EC) No 810/2009, art 21(5), with later amendments. 37 The French notion of ordre public is not uniformly translated in Union law.

Public Policy Exception 259

10.3.1 Overlap with Union citizenship In essence, the Court defends a mixed solution. Early judgments had referred to free movement cases to highlight that the ‘public policy’ caveat has a uniform meaning in the supranational legal order.38 On one occasion, the Court stated generically that ‘the extent of the protection a society intends to afford to its fundamental interests cannot vary depending on the legal status of the person that undermines those interests’.39 Closer inspection demonstrates, however, that implications were less straightforward than the idea of parallel interpretation suggests. The underlying argument is explained best under recourse to the Ziebell judgment, which will be mentioned in Chapter 17.1.3 on association agreements: judgments on the Free Movement Directive 2004/38/EC were found not to be applicable to similar provisions regarding Turkish nationals, since Union citizenship transcended the purely economic rationale of the association agreement.40 In other words, even identical terminology can be given a distinct interpretation in light of other criteria. Judges developed general standards on how to identify the degree of differentiation gradually, thereby overcoming the fuzziness of the initial case law. An early ruling recognised that the rationale of the public security exception in the Qualification Directive differed from the Free Movement Directive, without deducting any consequences from that distinction.41 In three judgments on the Return Directive, the CJEU appeared to advance widespread parallelism by insisting on an individualised assessment, while indicating, without further explanation, that rules for Union citizens may be stricter.42 This line of imprecise verdicts was extended for the withdrawal of long-term resident status. One judgment referred to the Ziebell ruling, which had distinguished the status of Turkish nationals from citizens’ rights, without stating as to how the public policy formula should be given a specific meaning for long-term residents.43 In a remarkable step, the Grand Chamber granted authorities a wide discretion when rejecting student visas in situations which might involve terrorism. The Fahimian judgment defined ‘public security’ under recourse to citizenship case law and commanded an individualised assessment, while stating—in direct contrast to Article 27(2) Free Movement Directive 2004/38/EC—that potential threats were sufficient to justify refusal and that these threats need not be based solely on the personal conduct of the applicant.44 Another Grand Chamber ruling elevated the obligation to consider the personal conduct to the rank of primary law for the purposes of Union citizenship.45 Migration law towards third country nationals differs. In December 2019, judges finally adopted a generic position on how to combine a uniform meaning of the ‘public policy’ exception with sector-specific outcomes in two

38 See Case C-601/15 PPU N EU:C:2016:84, paras 66–67; Case C-554/13 Zh and O EU:C:2015:377, para 50; Case C-544/15 Fahimian EU:C:2017:255, para 39; and Case C-240/17 E EU:C:2018:8, paras 48–49. 39 T (n 9) paras 77–78. 40 See Case C-371/08 Ziebell EU:C:2011:809, paras 60–73. 41 See T (n 9) paras 77–79. 42 See Case C-18/19 Stadt Frankfurt am Main EU:C:2020:511, paras 42–44; E (n 38) paras 48–49; and Case C- 82/16 KA and others EU:C:2018:308, paras 90–94. 43 See Case C-636/16 López Pastuzano EU:C:2017:949, paras 26–27. 44 See Fahimian (n 38) paras 40, 44–46. 45 See Case C-165/14 Rendón Marín EU:C:2016:675, paras 81–87.

260 General Features of EU Legislation cases concerning the Schengen Borders Code and the Family Reunification Directive.46 Mirroring the Ziebell judgment, judges emphasised that the meaning of ‘public policy’ shall have a parallel meaning for Union citizens and third country nationals to start with, even though the interpretative exercise may support differentiated outcomes.47 In order to determine the legislation-specific meaning, it is ‘necessary . . . to take into account the wording of th[e]‌provision, its context and the objectives pursued by the legislation of which it forms part’.48 We need to assess, in other words, the abstract concept of ‘public policy’ in light of the interpretative criteria described in Chapter 6.4.

10.3.2 Sector-specific outcomes Settled case law on the meaning of ‘public policy’ and the sister principle of ‘public security’ in the single market serves as the starting point of the interpretative exercise. ‘Public policy’ is an autonomous concept of Union law that does not—unlike the similarly worded formula in private international law49—allow Member States to define public policy autonomously. It is generally understood to require a ‘genuine and sufficiently serious threat . . . affecting one of the fundamental interests of society’.50 This definition serves as an umbrella notion giving Member States a certain leeway to define country-specific standards within the confines of Union law.51 Instead of requiring uniform outcomes across the Union, ‘the concept of public policy may vary from one Member State to another and from one era to another’.52 On the basis of these abstract findings, the EU migration law instruments have to be analysed. In everyday practices, the ‘public policy’ exception is much more relevant than the ‘public security’ caveat, which has traditionally been interpreted narrowly. Public security relates to internal or external threats which call into question ‘a country’s existence since . . . its institutions, its essential public services and even the survival of its inhabitants depend upon [the interest at stake]’.53 Classic examples are terrorism and organised crime, although judges have broadened their outlook by emphasising that ‘a direct threat to the calm and physical security of the population’ can constitute a public security risk, for instance sexual abuse of children or drug trafficking.54 The distinction between public policy and public security has little practical bearing in EU migration law, since Articles 17(2)(d) and 21(2)(a) Qualification Directive 2011/95/EU are the only two provisions which do not embrace the broader notion of public policy at the same time. 46 They were delivered less than a year after the publication of Daniel Thym, ‘A Bird’s Eye View on ECJ Judgments on Immigration, Asylum and Border Control Cases’ (2019) 21 EJML 166, 179–83, which had made a similar proposal of how to rationalise the diffuse case law. 47 See Case C-380/18 EP EU:C:2019:1071, paras 29–32. 48 ibid para 33; and also Joined Cases C-381/18 and C-382/18 GS and VG EU:C:2019:1072, paras 53–55. 49 See Daniel Thym, ‘The Constitutional Dimension of Public Policy Justifications’ in Panos Koutrakos and others (eds), Exceptions from EU Free Movement Law (Hart Publishing 2016) 171, 179–83. 50 The formula was first used by Case 30/77 Bouchereau EU:C:1977:172, para 35; and, for EU migration law, by Fahimian (n 38) para 40. 51 See Case 41/74 Van Duyn EU:C:1974:133, para 18 for EU citizens; and T (n 9) para 77 for migration law. 52 See Case C-208/09 Sayn-Wittgenstein EU:C:2010:806, para 87. 53 Case 72/83 Campus Oil EU:C:1984:256, para 34. 54 See Case C-348/09 I EU:C:2012:300, para 28; Case C-145/09 Tsakouridis EU:C:2010:708, paras 45–50; and Loïc Azoulai and Stephen Coutts, ‘Restricting Union Citizens’ Residence Rights on Grounds of Public Security’ (2013) 50 CML Rev 553.

Public Policy Exception 261 Specific solutions for third country nationals ultimately depend on the contents and context of the legislation in line with previous comments. Having said this, a couple of similarities between the judgments on migration law stand out.55 A classic position of Article 27(2) (1) Free Movement Directive 2004/38/EC is that restrictions ‘shall be based exclusively on the personal conduct of the individual concerned. Previous criminal convictions shall not in themselves constitute grounds for taking such measures’.56 Not a single on EU migration law instrument contains such an explicit guideline, although the Commission had proposed—unsuccessfully—to use similar language for the Long-Term Residents Directive and the Family Reunification Directive.57 Moreover, free movement law requires ‘a genuine, present and sufficiently serious threat’,58 thus excluding latent and potential threats which are less likely to materialise. A common thread of the judgments on migration law is to give states more leeway. The Court found that criminal convictions may amount to a public policy threat in themselves, without the need to consider the personal conduct thereafter. It justified this conclusion by the distinct wording, drafting history, general scheme, and objectives of the Schengen Borders Code Regulation and the Family Reunification Directive.59 Along similar lines, the Court found a ‘potential’ threat to justify the non-admission of students and went as far as obliging domestic courts to limit judicial review to manifest errors.60 That position was reaffirmed regarding the relocation of asylum seekers.61 Member States may consider criminal behaviour that had not yet resulted in a verdict,62 and membership in a terrorist organisation can meet the ‘public policy’ threshold independent of the behaviour of the individual.63 Asylum authorities are not obliged to follow the opinion of the secret service when assessing whether someone poses a security threat.64 A crucial restriction remains. To find that Member States may rely on a threat to public policy without considering the personal conduct does not automatically justify refusal, withdrawal, or non-renewal of residence permits. Settled case law requires a proportionality test in a second step. General principles underlying the proportionality test will be discussed elsewhere in this chapter. We shall see that the level of protection depends on the specificities of the case at hand. As a rule of thumb, individual interests weigh less whenever someone applies for a residence permit from abroad for purposes not related to human rights,65 while having a greater impact when family reunification or the withdrawal of

55 See also Pieter Boeles and others, Public Policy Restrictions in EU Free Movement and Migration Law (Meijers Committee 2021) 33–53. 56 This definition was introduced by the legislature in Article 3 Directive 64/221/EEC on the co-ordination of special measures concerning the movement and residence of foreign nationals [1963/64] OJ Special Ed I/117, long before the Court started interpreting primary law along similar lines. 57 See Commission Proposal for a Long-Term Residents Directive, COM(2001) 127 final, recital 11, art 7(1)(2); and Commission Proposal for a Family Reunification Directive, COM(1999) 638 final, art 8(2). 58 Free Movement Directive 2004/38/EC, art 27(2)(1). 59 See EP (n 47) paras 33–46; and GS and VG (n 48) paras 58–60. 60 See Fahimian (n 38) paras 40, 44–46, in line with the former Students Directive 2004/114/EC, recital 14; see also Katharina Eisele, ‘Public Security and Admission to the EU of Foreign Students: Fahimian’ (2018) 55 CML Rev 279, 288–93. 61 See Joined Cases C-715/17, C-718/17 and C-719/17 Commission v Poland and others EU:C:2020:257, paras 148–53. 62 See E (n 38) paras 48–49. 63 See T (n 9) paras 77–79 under reference to a legislative recital. 64 See Case C-159/21 Országos Idegenrendeszeti Főigazgatóság and others EU:C:2022:708, paras 72–81. 65 See EP (n 47) paras 36–37; and Fahimian (n 38) paras 41–43; referring to Koushkaki (n 10) paras 60–62.

262 General Features of EU Legislation long-term resident status are at stake.66 Distinguishing third country nationals from Union citizenship does not result in uniform outcomes. Article 12 Long-Term Residents Directive 2003/109/EC on status withdrawal is unique among the EU migration law instruments insofar as it requires an ‘actual’ threat, thus prohibiting expulsion based on a previous criminal conviction alone—a conclusion that does not extend to the initial refusal of long-term residence.67 Moreover, the legislature laid down distinct criteria for the proportionality analysis. Recital 16 and Article 12(2) mirror case law on Article 8 ECHR, which will be discussed in Chapter 15.2.2 on integration. The example reiterates that the meaning of public policy is defined by a unique blend of general definitions and legislation- specific considerations.

10.4 Equal Treatment as a Constitutional Guarantee For more than a century, inequality had been the hallmark of migration law. Chapter 5 demonstrated that states were generally allowed to treat foreigners differently before the arrival of universal human rights. In this overall context, the potential of equal treatment lies in its horizontal character: it can be applied in cases where other human rights do not mandate specific outcomes. States are, by way of example, generally free to define the level of social benefits, but this does not prevent advocates of migrants’ rights from challenging differentiation on grounds of nationality or migration status. Success is far from guaranteed, however. Judges distinguish the equal treatment of migrants from the level of protection for Union citizens (10.4.1). A bird’s eye view on the case law presents a multi-faceted picture of different scrutiny levels (10.4.2). Of particular relevance are distinctions between different migration statuses (10.4.3) and criteria which might serve as proxies for race or ethnic origin (10.4.4).

10.4.1 Human rights instead of Union citizenship Equal treatment on grounds of nationality informed much of the CJEU’s dynamic free movement case law. Against this backdrop, observers have highlighted that the open wording of Article 18 TFEU can be interpreted to embrace third country nationals.68 That position was not, however, taken up by the Court. Settled case law holds that Article 18 TFEU ‘is not intended to apply to cases of a possible difference in treatment between nationals of Member States and nationals of non-member countries’.69 The same applies to Article 21(2) Charter of Fundamental Rights (CFR), which replicates Article 18 TFEU.70 66 See GS and VG (n 48) paras 64–68; and López Pastuzano (n 43) paras 26–27. 67 Directive 2003/109/EC, art 6(1) uses a different wording; see Daniel Thym, ‘Long-Term Residents Directive 2003/109/EC’ in Daniel Thym and Kay Hailbronner (eds), EU Immigration and Asylum Law. Article-by-Article Commentary (3rd edn, CH Beck/Hart Publishing/Nomos 2022) Article 6 MN 11–12, Article 12 MN 1–10. 68 By way of example see Chloé Hublet, ‘The Scope of Article 12 of the Treaty of the European Communities vis-à-vis Third-Country Nationals’ (2009) 15 ELJ 757, 761–74; and Dáire McCormack-George, ‘Equal Treatment of Third-Country Nationals in the European Union’ (2019) 21 EJML 53, 57–60. 69 Joined Cases C-22/08 and C-23/08 Vatsouras and Koupatantze EU:C:2009:344, para 52; and Opinion 1/17 EU-Canada CETA-Agreement EU:C:2019:341, paras 168–70; see also Joined Cases C-95/99–C-98/99 and C-180/ 99 Khalil EU:C:2001:532, para 40. 70 See Case C-930/19 État belge EU:C:2021:657, paras 50–51, referring to the official ‘Explanations relating to the Charter’ [2007] OJ C303/17, 24.

Equal Treatment as a Constitutional Guarantee 263 Equal treatment on grounds of nationality is closely linked to the single market and Union citizenship. Chapter 1.3 explained that judges cannot be expected to project these standards upon third country nationals, who can invoke human rights instead. Articles 20 and 21(1) CFR confirm that the EU institutions must respect the human right to equality before the law, which embraces, as leges speciales, guarantees against discrimination on grounds of sex, race, colour, ethnic or social origin, genetic features, and several other characteristics. As freestanding equal treatment guarantees, Articles 20 and 21 CFR do not depend on the parallel application of other human rights—in contrast to Article 14 ECHR.71 Within this wider field of application, the interpretation of Articles 20 and 21 CFR shall be ‘based on’ the case law of the European Court of Human Rights (ECtHR), nonetheless.72 States may justify unequal treatment on the basis of legitimate reasons and subject to the principle of proportionality. In contrast to widespread preoccupation of the theoretical debate on migration with equal treatment, the judicial practice remains patchy. Judges did not replicate the dynamism of the citizenship model and have occasionally signalled a reluctance to challenge migration law—despite the prediction by academic commentators that the human right to equal treatment would turn into a crucial constitutional yardstick.73 International human rights law may serve as an additional source of inspiration, although it will rarely result in precise guidelines for EU migration law.74 Human rights in the Charter apply to both EU legislation and national measures implementing Union law, in line with previous comments in Chapter 5.4. Academic and political debates about non-discrimination transcend the domain of migration. Experts in constitutional law, for instance, emphasise that equal treatment guarantees present a slippery slope of constitutional adjudication if judges intervene in salient political matters requiring democratic choices. It may be a virtue, therefore, to give parliaments breathing space.75 On the other side of the spectrum, feminist and postcolonial scholars highlight the significance of equal treatment in support of the subaltern; they point out that robust judicial intervention should move beyond formal equality in terms of equal application of the law, focusing on substantive inequalities in the form of unequal outcomes instead.76 Such theoretical arguments can be projected on migration law to develop progressive proposals.77

71 It is irrelevant, therefore, that many Member States have not ratified Additional Protocol No 12 to the ECHR. 72 See Joined Cases C-199/12–C-201/12 X, Y and Z EU:C:2013:720, para 54. 73 See McCormack-George (n 68) 68–81; and Sara Iglesias Sánchez, ‘Fundamental Rights Protection for Third Country Nationals and Citizens of the Union’ (2013) 15 EJML 137. 74 See ch 5.5; and Wouter Vandenhole, ‘Migration and Discrimination’ in Vincent Chetail and Céline Bauloz (eds), Research Handbook on International Law and Migration (Edward Elgar Publishing 2014) 216. 75 See Elise Muir, EU Equality Law (OUP 2018) ch 4; and Johanna Croon-Gestefeld, Reconceptualising European Equality Law (Hart Publishing 2017). 76 See Sandra Fredman, ‘Substantive Equality Revisited’ (2016) 14 ICON 712; and Aileen McColgan, Discrimination, Equality and the Law (Hart Publishing 2016) ch 1. 77 By way of example see Karin de Vries, ‘The Non-National as “The Other”’ in Moritz Jesse (ed), European Societies, Migration, and the Law (CUP 2020) 192.

264 General Features of EU Legislation

10.4.2 Different scrutiny levels in the judicial practice Comparative constitutional law illustrates that judges often apply different levels of scrutiny. Justices at the US Supreme Court, in particular, have identified ‘suspect grounds’ warranting stricter scrutiny, thus rendering it more difficult to justify unequal treatment. Neither the CJEU nor the ECtHR distinguishes different scrutiny levels in the abstract. Nevertheless, the judicial practice mirrors the American model insofar as judges are comparatively strict on some subject matters, while leaving political institutions leeway in other domains. For instance, the finding that Article 18 TFEU cannot be extended to third country nationals indicates that the Court will not generally treat nationality as a suspect ground when it comes to migration law. Unfortunately, it can be difficult to distil clear patterns from the case law.78 Judgments range from intense scrutiny with regard to social security to generic statements that the legal status of migrants cannot be compared to the situation of Union citizens. Thus, the Court in Luxembourg found on several occasions that unequal treatment of nationals and third country nationals did not require any justification at all, since their situation was not comparable in the first place.79 To say so mirrors the Aristotelian formula that unequal things should not be treated equally, which the Court employed in other domains as well.80 Along similar lines, judges in Strasbourg recognised that Union citizenship cannot be compared to the migration status of third country nationals, since ‘the Union forms a special legal order, which has, moreover, established its own citizenship’.81 Such findings of incomparability curtail the significance of equal treatment. Although nationality is not listed as a suspect ground in Article 14 ECHR, judges maintain that only ‘very weighty reasons’ can justify unequal treatment based on nationality.82 Practical effects of this generic finding were watered down, however, in many follow-up rulings which have led to a patchwork of sector-specific outcomes.83 An example of strict scrutiny are judgments censuring domestic legislation excluding legally staying migrants from social benefits. Strict scrutiny remained limited, however, to social security, including family benefits.84 Doing so largely coincides with the equality provisions in numerous migration law directives which habitually require equal treatment in the different branches of social security. Conversely, judges found that states ‘may have legitimate reasons for curtailing the use of resource‑hungry public services . . . by short‑term and illegal immigrants, who, as a rule, do not contribute to their funding’.85 That margin concerns health care, social housing, or 78 For an overview see Marie-Bénédicte Dembour, When Humans Become Migrants (OUP 2015) ch 8; and Paul Minderhoud, ‘Social Security Rights of Third Country Nationals’ (2010) 17 J Social Sec L 227. 79 See Case C-579/13 P and S EU:C:2015:369, paras 39–43; Joined Cases C-443/14 and C-444/14, Alo and Osso EU:C:2016:127, paras 54, 59; État belge (n 70) paras 57–76; and Case C-718/19 Ordre des barreaux francophones et germanophone and others EU:C:2021:505, para 65. 80 See Paul Craig, EU Administrative Law (3rd edn, OUP 2018) 577–78. 81 Ponomaryovi and others v Bulgaria App no 5335/05 (ECtHR, 21 June 2011) § 54; reiterating Moustaquim v Belgium App no 12313/86 (ECtHR, 18 February 1991) § 49. 82 See Gaygusuz v Austria App no 17371/90 (ECtHR, 16 September 1996) § 42. 83 See also Natalia Caicedo Camacho, ‘Social Rights and Migrants before the European Court of Human Rights’ in David Moya and Georgios Milios (eds), Aliens before the European Court of Human Rights (Brill 2021) 191; and Evelyn Brouwer and Karin de Vries, ‘Third-Country Nationals and Discrimination on the Ground of Nationality’ in Marjolein van den Brink and others (eds), Equality and Human Rights (SIM 2015) 123. 84 See Gaygusuz v Austria (n 82) §§ 43–52; and Dhabi v Italy App no 17120/09 (ECtHR, 8 April 2013) §§ 45–54. 85 Ponomaryovi and others v Bulgaria (n 81) § 54.

Equal Treatment as a Constitutional Guarantee 265 study benefits and reminds us of the position among constitutional theorists that judges may struggle to equal treatment to redistributive matters. The Grand Chamber even recognised the direct discrimination of long-term residents.86 Chapter 15.3.3 on migrant integration will demonstrate in how far secondary legislation goes beyond the level of protection under human rights law. On the whole, judges often accentuate legal residence and the length of stay as criteria supporting enhanced protection. Against this background, one may conclude, as a rule of thumb, that settled migrants, including long-term residents, can ask for more equality.87 By contrast, the ECtHR found that states retain a margin of appreciation when dealing with migrants with a temporary or precarious residence status.88 In a series of decisions, judges even highlighted that those staying illegally can generally be excluded from social benefits.89 These distinctions reiterate that the judicial practice fluctuates between strict scrutiny in situations of legal stay and settlement to restraint in scenarios of irregular or temporary presence.

10.4.3 Distinctions based on migration status In a judgment of principle, the ECtHR recognised that ‘immigration status is not an inherent or immutable personal characteristic such as sex or race, but is subject to an element of choice’,90 thus indicating a lower level of scrutiny. That being said, the judicial finding does not give states a carte blanche to do as they please. A lower level of scrutiny does not absolve states from demonstrating legitimate reasons and from complying with the principle of proportionality. The need for justification extends to the uneven treatment of different migration statuses. To challenge distinctions within migration law constitutes an alternative to the direct comparison with nationals, which judges are hesitant, as we have seen, to declare unlawful. Comparing migration statuses is established in international practice and comparative constitutional law.91 Again, it remains difficult to distil clear patterns from the case law. On the one hand, the ECtHR and the CJEU have rejected justification for differentiation between refugee status and subsidiary protection, as discussed in Chapter 15.3.5 on integration. On the other hand, the Court in Strasbourg stated generically ‘that there are in general persuasive social reasons for giving special treatment to those who have strong ties with a country, whether stemming from birth within it or from being a national or a long-term resident’.92 The Grand Chamber also held that inter-state reciprocity may justify unequal access to 86 See Savickis and others v Latvia App no 49270/11 (ECtHR [GC], 9 June 2022). 87 See Vandenhole (n 74) 225–26, 229. 88 See Koua Poirrez v France App no 40892/98 (ECtHR, 30 September 2003) §§ 46–49; and Ponomaryovi and others v Bulgaria (n 81) §§ 52–55, after having reaffirmed the ‘weighty reasons’ formula. 89 See Yeshtla v Netherlands App no 37115/11 (ECtHR, 15 January 2019); the parallel decisions against the Netherlands in Said App no 34299/14; Heerawi App no 36558/14; Aghmadi App nos 70475/14 and 70530/14; and AD App nos 71815/14 and 71827/14; and Mel Cousins, ‘The European Convention on Human Rights and Residence Requirement for the Purposes of Social Assistance Benefits’ (2019) 21 EJML 541, 544–47. 90 Bah v United Kingdom App no 56328/07 (ECtHR, 27 November 2011) § 47. 91 See Francesca Ippolito, ‘Le migrant confronté aux discriminations selon l’origine’ in Guy S Goodwin-Gill and Philippe Weckel (eds), Migration & Refugee Protection in the 21st Century (Martinus Nijhoff 2015) 245. 92 Biao v Denmark App no 38590/10 (ECtHR, 25 March 2013) § 94, which was overturned on other grounds by the Grand Chamber a few years later.

266 General Features of EU Legislation social benefits.93 Similarly, the Court in Luxembourg concluded ‘that the nationality of the applicant for asylum is an element which may be taken into consideration to justify the prioritised or accelerated processing of an asylum application’.94 Equal treatment is no magic potion that helps to overcome differences.

10.4.4 Racial and ethnic discrimination Strict scrutiny standards would apply if applicants succeeded in branding distinctions among status groups as (in)direct race discrimination. The argument has occasionally been brought forward, taking up the rich American tradition.95 A classic example are ‘white’ and ‘black’ lists for visa requirements, to be discussed in the following chapter. Having said this, neither the ECtHR nor the CJEU have taken up the criticism of ‘passport apartheid’. Differences among countries of origin do not constitute discrimination on grounds of race or ethnic origin.96 Similarly, the International Court of Justice found travel restrictions for certain nationalities not to amount to direct or indirect discrimination on ground of ethnic origin or race, thus rejecting the non-binding views of the Committee on the Elimination of All Forms of Racial Discrimination.97 Judges in Europe seem to prefer the general equal treatment guarantee which does not use the morally charged notion of ‘race’, or the category of ‘ethnic origin’,98 possibly reflecting the different historic outlook.99 In doctrinal terms, ‘indirect discrimination’ serves as a critical criterion for restraining— or extending—the leeway of the legislature. Indirect discrimination has traditionally been interpreted broadly in the single market and for the equal treatment of women and men, as well as in rulings concerning Roma.100 By contrast, the CJEU interpreted indirect discrimination in the Racial Equality Directive 2000/43/EC and Article 21 CFR to cover only measures that concern one particular ethnic or racial group. Domestic legislation relying on the place of birth, language requirements, or university education abroad as the door-opener for social benefits, financial services, or study grants were found not to amount to indirect racial or ethnic discrimination.101 Critical scholars will be disappointed by such outcomes, which re-orientate indirect discrimination from substantive notions of equality to the motivation behind the differentiation.

93 See Carson and others v United Kingdom App no 42184/05 (ECtHR [GC], 16 March 2010) § 88. 94 Case C-175/11 D and A EU:C:2013:45, paras 71–73; see also AG Juliane Kokott, Opinion in Case C-153/14 K and A EU:C:2015:186, point 20. 95 See generally E Tendayi Achiume, ‘Race, Refugees, and International Law’ in Cathryn Costello and others (eds), The Oxford Handbook of International Refugee Law (OUP 2021) 43. 96 See Abdulaziz, Cabakes and Balkandali v United Kingdom App nos 9214/80, 9473/81, and 9474/81 (ECtHR, 28 May 1985) §§ 85–86; Case C-571/10 Kamberaj EU:C:2012:233, paras 48–50; and the multiple other judgments mentioned previously which did not even consider the argument. 97 See Application of the International Convention on the Elimination of All Forms of Racial Discrimination (Qatar v United Arab Emirates) (ICJ Judgment No 172, 4 February 2021) §§ 71–105, 109–13. 98 AG Nils Wahl, Opinion in Case C-668/15 Jyske Finans EU:C:2016:914, points 31, 34–35 considered the ‘race’ to be a cub-category of ‘ethnic origin’. 99 See Mathias Möschel, Law, Lawyers and Race (Routledge 2014) 81–109. 100 See DH and others v the Czech Republic App no 57325/00 (ECtHR [GC], 13 November 2007); and Case C-83/ 14 CHEZ Razpredelenie Bulgaria EU:C:2015:480, paras 92–109. 101 See Case C-94/20 Land Oberösterreich EU:C:2021:477, paras 55–56, 63; Case C-668/15 Jyske Finans EU:C:2017:278, paras 17–21; and Case C-457/17 Maniero EU:C:2018:912, paras 46–51.

Proportionality 267 Our conclusion does not contrast with the Biao judgment, in which the ECtHR concluded that more generous family reunification for nationals and those having resided legally for twenty-eight years amounted to discrimination on grounds of ethnic origin, since the government had relied on ‘speculative arguments’ and ‘prevailing social prejudice’.102 That verdict was representative insofar as judges concentrated on the intention behind unequal treatment, rather than practical effects of domestic rules, as highlighted by the postcolonial and feminist critique. Nationality, migration status, or other indicators such as language skills or place of birth are not generally treated as immutable (racial) characteristics which warrant strict scrutiny.

10.5 Proportionality Proportionality has a dual relevance in Union law to determine the scope of supranational competences under Article 5(4) TEU and to fine-tune the legality of restrictions on individual rights. Article 52(1) CFR reiterates that any interference with human rights is ‘[s]‌ubject to the principle of proportionality’. Proportionality was borrowed from the German legal order and requires compliance with a multi-step test: the state measure must pursue a legitimate aim; it needs to be suitable for achieving the objective; necessity concerns the availability of less onerous ways of achieving the objective; proportionality sensu stricto requires balancing of countervailing interests—judges often merge necessity and balancing. This four-step test provides abstract criteria for assessing individual cases. In doing so, the degree of judicial scrutiny will vary widely (10.5.1), and judges fluctuate between abstract and individual solutions (10.5.2).

10.5.1 Context-specific outcomes It does not come as a surprise that proportionality is a uniform principle whose application is context-specific. Balancing is by definition about the relative weight of the various factors, in light of circumstances. Individual interests may have a different weight: to withdraw long-term resident status after a criminal conviction is not the same as the rejection of a Schengen visa for touristic purposes. Moreover, the degree of judicial scrutiny will depend on the subject area under consideration. Pre-departure language tests as a precondition for family reunification may be treated differently from police checks near internal Schengen borders, or time limits for accelerated asylum procedures. At an intermediate level of abstraction, we can distinguish at least five context-specific reasons which influence the application of the proportionality principle, although it remains notoriously difficult to predict as to when and why judges opt for a higher or lower level of scrutiny.103 The case law provides little abstract guidance.

102 Biao v Denmark App no 38590/10 (ECtHR [GC], 24 May 2016) §§ 125–26; for comments see Karin de Vries, ‘Rewriting Abdulaziz’ (2016) 18 EJML 467. 103 See Jonas Bornemann, ‘The Guises of and Guidance to Administrative Discretion in the Interpretation of EU Immigration Law’ (2019) 12 Rev Eur Admin L 97; and Jan Zglinski, ‘The Rise of Deference’ (2018) 55 CML Rev 1341.

268 General Features of EU Legislation First, judges tend to leave the legislature more leeway where policy choices do not interfere with individual rights, such as the decision whether mandatory relocation contributes to achieving the Treaty objectives.104 Similarly, the Court did not exercise strict scrutiny when determining whether police checks near the internal Schengen borders comply with the Schengen Borders Code Regulation.105 Lesser scrutiny of basic political choices extends to national legislation,106 in line with the position of some constitutional theorists that judicial restraint may be warranted with regard to politically or culturally sensitive subject matters.107 Secondly, absence of detailed guidance, on the part of the Court, will not always signal deference towards political preferences. The preliminary reference procedure rests on a ‘clear separation of functions between the national courts and the Court of Justice’.108 While the Court interprets Union law, domestic courts will apply these standards to the individual case. Since proportionality is in essence about weighing countervailing interests, it is conceptually sound if the Court leaves the assessment to domestic courts. Inspection of the judicial practice in Chapter 3.3 illustrated that rulings fluctuate between vagueness and micromanagement. Thirdly, stricter scrutiny is generally warranted when judgments affect the rights of migrants. Their individual rights may emanate from human rights in the Charter or from guarantees in secondary legislation. When it comes to human rights, the Court of Justice will not usually have to reinvent the wheel, since it can often refer to ECtHR case law. At the same time, judges may have recourse to the proportionality test in scenarios not subject to individual rights. It may be applied to fee levels or non-admission at the border in cases not involving a real risk of illegal refoulement.109 In such situations, where individual rights do not take centre stage, judges will—as a rule of thumb—engage in less intense scrutiny. Fourthly, the relative weight of individual rights may differ. The example of family reunification illustrates that their significance depends on the legal and factual context. Settled human rights case law holds that Article 8 ECHR does not usually involve an obligation, on the part of states, to grant entry visas, whereas the Family Reunification Directive 2003/ 86/EC goes further by establishing clearly defined statutory rights.110 As a result, Member States may find it more difficult to justify restrictions under the Directive than under human rights law. Chapter 15.4.3 on pre-departure language tests will explain that the Court required Member States to foresee exemptions, which human rights alone would not necessarily have supported. Finally, context-specificity takes centre-stage when it comes to proportionality in a strict sense. Assessing the relative weight of countervailing interests requires thorough identification of individual interests and policy objectives. By way of example, judges concluded that pre-removal detention of third country nationals may last eight months or more, although 104 See Joined Cases C-643/15 and C-647/15 Slovakia and Hungary v Council EU:C:2017:631, paras 206–24. 105 See Case C-9/16 A EU:C:2017:483, paras 60–61, applying an ‘equivalent effect’ test, not the principle of proportionality as such. 106 See generally Craig (n 80) chs 19–20. 107 See Mark Dawson, The Governance of EU Fundamental Rights (CUP 2017) ch 3; and Janneke Gerards, ‘Pluralism, Deference and the Margin of Appreciation Doctrine’ (2011) 17 ELJ 80. 108 Case C-235/95 Dumon and Froment EU:C:1998:365, para 25. 109 See Johan Rochel, ‘Working in Tandem. Proportionality and Procedural Guarantees in EU Immigration Law’ (2019) 20 German LJ 89, 99–105. 110 See chs 6.5.2, 14.2.3.

Proportionality 269 the same time frame would be illegal for Union citizens, whose special status carries much weight.111 The application of the proportionality test may result in a spectrum of possible solutions whose application to the case at hand ultimately depends on the context.

10.5.2 Abstract or individual assessment? A horizontal survey unearths an astonishing discrepancy that is rarely discussed. Can proportionality be used to challenge legislative provisions that lay down exact thresholds, without leaving authorities leeway to consider the individual case? Several judgments on migration law can be read to thwart abstract rules. Examples include pre-departure language requirements as a precondition for family reunification and the definition of stable and regular resources, both mentioned previously. Proportionality served as a door-opener for individualised assessments instead of legislative categorisation. A classic example for the latter are traffic rules: we are not allowed to cross the street while the traffic light is red, even if the factual situation leaves no doubt that doing so would not involve any risk whatsoever. Some rulings on migration law accepted abstract legislative standards as being proportionate, without considering personal circumstances. Examples include fee levels for residence permits, discussed previously, and a time limit of fifteen days for legal remedies.112 The Noorzia judgment stated generally that Member States may introduce a minimum age of twenty-one years for family reunification,113 and judges interpreted the exclusion clause for applications to family reunification with refugees strictly.114 Chapter 9.5.2 on databases explained that the Court is comparatively generous when assessing the proportionality of data collection and storage for migration-specific purposes. Numerous other abstract rules were applied by the Court, without even considering the idea of individualised fine-tuning. Think of five-year residency requirement for long-term residence or the six-month time limit for the transfer of jurisdiction in Dublin cases. Abstract standards also featured in free movement cases, which had traditionally put much emphasis on individualised assessments. Judges accepted the generic exclusion of students from study grants during a five-year period and of job-seekers from social assistance, since the outcome of the legislative process ‘itself takes into consideration various factors characterising the individual situation of each applicant’, thus substituting the need for an individual assessment.115 German judicial practice, from which the proportionality test was borrowed, similarly recognises the option of categorisation, especially when the legislature defines abstract criteria to accommodate countervailing interests.116 Along these lines, the German judiciary interpreted the Court’s proportionality requirement in Rottmann on the loss of nationality 111 See Ordre des barreaux francophones et germanophone and others (n 79) paras 65–72. 112 See Case C-69/10 Samba Diouf EU:C:2011:524, paras 66–69, also highlighting that an exemption from the strict rule may exceptionally be necessary. 113 See Case C-338/13 Noorzia EU:C:2014:2092, paras 15–19; and ch 14.3.2.2 on silence on human rights. 114 See Case C-380/17 K and B EU:C:2018:877, paras 43–50. 115 See Case C-67/14 Alimanovic EU:C:2015:597, paras 59–60; and Case C-158/07 Förster EU:C:2008:630, paras 51–60. 116 See Federal Constitutional Court (Bundesverfassungsgericht), Case 1 BvL 6/10 (decision of 17 December 2013) paras 84– 85; and Liesa Plappert, ‘Der unionsrechtliche Verhältnismäßigkeitsvorbehalt’ [2020] Europarecht 364.

270 General Features of EU Legislation to require abstract balancing.117 If proportionality may support both solutions, we are left with the question as to when and why an abstract appraisal should give way to administrative fine-tuning. At present, the question often comes up indirectly, since it is mentioned by one of the parties or an Advocate General.118 Judges, by contrast, hardly ever explain their preference. Future analysis may rely on the above-mentioned criteria for context-specific outcomes. Whenever they support strict scrutiny, the argument for individualised assessments is stronger. In this overall context, human rights do not, however, always command administrative fine-tuning. A telling example is an ECtHR judgment on family reunification with beneficiaries of subsidiary or temporary protection which found an unconditional waiting period of two years to be acceptable (abstract proportionality), whereas the situation of the applicant had to considered thereafter (individual assessment).119 The ECtHR assumes that general measures are legitimate when a case-by-case examination ‘would give rise to a risk of significant uncertainty, of litigation, expense and delay, as well as of discrimination and arbitrariness’.120

10.6 More Favourable Domestic Rules Most directives on EU migration law contain a provision stating that Member States may introduce or retain more favourable domestic rules. They are generally understood to permit rules in favour of third country nationals. Nevertheless, the precise scope of the leeway remains unclear, especially when EU legislation limits favourable treatment ‘insofar as [national rules] are compatible with this Directive’.121 Restrictive tendencies embodied in such formulae were reinforced by a change in the Treaty framework. Member States are not free to ignore provisions that are mandatory (10.6.1). One step further, the Court found domestic residence schemes deviating from mandatory provisions to fall outside the scope of Union law (10.6.2).

10.6.1 Identification of mandatory provisions Before the Treaty of Lisbon, the legal basis for asylum allowed for the adoption of ‘minimum standards’ only.122 This limitation was widely understood, at the time, as generally permitting more generous domestic rules.123 It is no longer relevant whether this position was correct, since the Lisbon Treaty discontinued that rule. On this basis, the second generation of asylum legislation was adopted, which generally foresees more favourable national rules 117 See (German) Federal Administrative Court (Bundesverwaltungsgericht), Case 1 C 1.17 (judgment of 19 April 2018) para 61; and ch 15.7.2. 118 See AG Paolo Mengozzi, Opinion in Case C-221/17 Tjebbes and others EU:C:2018:572, points 51–118. 119 See MA v Denmark App no 6697/18 (ECtHR [GC], 9 July 2021) §§ 162, 147–50. 120 Animal Defenders International v United Kingdom App no 48876/08 (ECtHR [GC], 22 April 2013) § 108. 121 Return Directive 2008/115/EC, art 4(3). 122 See EC Treaty, art 63(1)(a)–(c), (2), as amended by the Treaty of Nice (adopted 26 February 2001, entered into force 1 February 2003) [2006] OJ C321E/37. 123 See Hugo Storey, ‘EU Refugee Qualification Directive’ (2008) 20 IJRL 1, 16–22; and Steve Peers and others, EU Immigration and Asylum Law (Text and Commentary), vol 3 (2nd edn, Brill Nijhoff 2015) 11.

More Favourable Domestic Rules 271 ‘insofar as’ they are compatible with the instrument in question.124 This double change of the constitutional context and the legislative design indicates that states have less flexibility. A crucial yardstick is what is usually called full—or complete—harmonisation pre- empting national choices. Proponents of migrants’ rights will often defend the finding of full harmonisation, which excludes state discretion. However, there is an another side to this argument. Provisions limiting migrants’ rights can similarly embody complete harmonisation. By way of example, judges have found the five-year qualification period for long-term residence and exclusion from refugee status to be mandatory,125 thus preventing Member States from laying down more favourable rules domestically. One has to establish on a case-by-case basis, in light of the interpretative standards discussed in Chapter 6.4, whether a provision results in full harmonisation or not. In doing so, we should recognise that the Treaty objective of a ‘common European asylum system’ supports greater uniformity,126 mirroring similar conclusions, on the part of the CJEU, for uniform standards in the Schengen Borders Code and the Visa Code Regulation.127 In some instances, interpretation will be straightforward. The term ‘shall’ designates mandatory rules, while the word ‘may’ indicates a certain level of flexibility.128 At the same time, ‘may’ should not be misunderstood as unfettered leeway, since the article in question may limit the room for manoeuvre. One may conclude, for instance, that Member States ‘may’ choose between solutions A or B, while solution C would be incompatible with the directive. The degree of flexibility depends, in other words, on the contents of each legislative instrument.

10.6.2 Different patterns in secondary legislation Provisions on more favourable domestic treatment follow different patterns. A holistic inspection of EU migration law allows for the identification of four different appearances. First, many directives permit national solutions only ‘insofar as [they] are compatible with this Directive’.129 Such provisions require an assessment on a case-by-case basis as to whether articles are mandatory or not in line with previous comments. By way of example, Member States may not grant international protection to those covered by the exclusion clause.130 Secondly, several directives on economic migration say explicitly that Member States may deviate from some provisions only, thus preventing more favourable domestic rules with regard to other aspects.131 Such provisions promise legal certainty and are a model for the future.

124 See Asylum Qualification Directive 2011/95/EU, art 3; Asylum Procedure Directive 2013/32/EU, art 5; and Asylum Reception Conditions Directive 2013/33/EU, art 4. 125 See Case C-469/13 Tahir EU:C:2014:2094, paras 26–37; and Joined Cases C-57/09 and 101/09 B and D EU:C:2010:661, para 115. 126 See ch 7.1.1; and Steve Peers, EU Justice and Home Affairs Law, vol I (4th edn, OUP 2016) 242. 127 See Case C-575/12 Air Baltic Corporation EU:C:2014:2155, paras 65–68; and Koushkaki (n 10) paras 48–50. 128 Interpretation can exceptionally indicate that ‘may’ means ‘shall’; see the opinion of the Council’s legal service referred to in ‘Articles 1 and 4’ (Council doc 14348/02, 15 November 2002). 129 See n 124; Return Directive 2008/115/EC, art 4(3); and Employer Sanctions Directive 2009/52/EC, art 15. 130 See Case C-91/20 Bundesrepublik Deutschland EU:C:2021:898, paras 38–40, 46. 131 See Seasonal Workers Directive 2014/36/EU, art 4(2); ICT Directive 2014/66/EU, art 4(2); and Blue Card Directive (EU) 2021/1883, art 4(2).

272 General Features of EU Legislation Thirdly, some directives stipulate that Member States remain free ‘to adopt or maintain provisions that are more favourable to the persons to whom it applies’, without indicating, like in the first scenario, that domestic rules must comply with Union law.132 Member States have more leeway as a result, although judges limited that freedom. Domestic rules must be linked to the rationale of instrument in question, in the sense of being thematically connected.133 This rationale applies to family reunification and temporary protection, in particular. Moreover, domestic schemes will not be covered by Union law in line with the comments that follow. Note that the Court will answer preliminary references if Member States voluntarily extend the rules adopted for the transposition of a directive to other scenarios, which are not covered by EU legislation.134 Finally, the Long-Term Residents Directive generally permits more favourable permanent residence schemes, while stating that they do not confer intra-European mobility rights.135 The revised Blue Card Directive follows the same logic, while also requiring domestic legislation to treat those with domestic permits similar to blue card holders in some respect, thus limiting the ability of national parliaments to undermine the effectiveness of the blue card scheme by introducing more attractive national programmes.136 Follow-up problems arise when national residence permits are issued under conditions that are less strict than EU legislation. By way of example, Member States may grant domestic permanent residence permits after four years, instead of the mandatory 5-year requirement for the EU status. Such domestic schemes are permitted as a form of more favourable treatment, but beneficiaries ‘cannot in any circumstances be long-term residents . . . within the meaning of that directive’.137 A similar interpretation was put forward for domestic permits for family reunification deviating from mandatory provisions of Directive 2003/86/EC ‘on the basis of their national law alone’.138 Parallel national schemes are not covered by supranational legislation. A similar argument was defended—albeit with an additional twist—by the Court with regard to ‘complementary’ protection schemes for asylum seekers on the basis of domestic laws under conditions which are more generous than the ones in Qualification Directive 2011/95/EU. Such domestic schemes are permissible, provided that they do not undermine the effet utile of supranational legislation, which, in practice, requires Member States to draw ‘a clear distinction . . . between national protection and protection under the Directive’.139 It must be possible, in other words, to distinguish clearly whether a status is governed by Union law or not. Member States act within the scope of retained powers when introducing national schemes.140 This has important doctrinal repercussions, since CJEU has no jurisdiction and the Charter cannot be relied upon.

132 See Family Reunification Directive 2003/86/EC, art 3(5); Students and Researchers Directive (EU) 2016/801, art 4(2); and Single Permit Directive 2011/98/EU, art 13(2). 133 See Case C-652/16 Ahmedbekova EU:C:2018:801, paras 69–71. 134 See GS and VG (n 48) paras 41–47. 135 See Long-Term Residents Directive 2003/109/EC, art 13. 136 See Blue Card Directive (EU) 2021/1883, arts 3(3), 11(6), 12(2), 13(5), 15(6), 16(7), 17(10). 137 Tahir (n 125) para 43; see also Case C-312/12 Ajdini EU:C:2013:103, para 25; and, beforehand, Achilles Skordas, ‘Immigration and the Market. The Long Term Residence Directive’ (2006) 13 Columbia J Eur L 201, 207. 138 Bevándorlási és Menekültügyi Hivatal (n 32) para 43. 139 See B and D (n 125) para 120; and Bevándorlási és Menekültügyi Hivatal (n 32) paras 42–43; and, on effet utile, ch 6.4.5. 140 See Case C-542/13 M’Bodj EU:C:2014:2452, paras 42–46.

Exceptional Non-Compliance (Article 72 TFEU) 273

10.7 Exceptional Non-Compliance (Article 72 TFEU) Article 72 TFEU maintains that the area of freedom, security, and justice ‘shall not affect the exercise of the responsibilities incumbent upon Member States with regard to the maintenance of law and order and the safeguarding of internal security’. The exact meaning of that proviso is subject to some controversy, which can be traced back to the predecessor provisions in the Treaties of Maastricht and Amsterdam.141 Three positions on how to interpret Article 72 TFEU can be distinguished. Some commentators maintained that the caveat mirrors the public policy exception and may, as a result, justify non-compliance with EU legislation.142 Others primarily saw it as a statement about competences which does not constrain the doctrinal force of Union law.143 An intermediate position argued that Article 72 TFEU cannot usually justify disrespect for secondary legislation, even though it requires the EU institutions to leave Member States breathing space, for instance for the rejection of visas on public policy grounds. Article 72 TFEU informs the interpretation of such provisions and may result—in extreme and highly exceptional scenarios144—in the annulment of legislation disregarding national security interests.145 ‘According to settled case-law’,146 the Court supports the intermediate position in regular circumstances.147 At the same time, the Grand Chamber found that Article 72 TFEU may— in line with the first position—exceptionally justify non-compliance with secondary legislation, since the provision can be compared with the derogations under Articles 36, 45, 52, 65, 346, and 347 TFEU, ‘which deal with exceptional and clearly defined cases’.148 When invoking such a scenario, Member States must prove underlying reasons; national governments are not free to determine unilaterally whether to introduce a derogation;149 failure to bring forward any reasons will entail that the Court will not consider applying the exception.150 To date, the Court has not been recognised the arguments brought forward by national governments in a single case. The meaning of the terms ‘maintenance of law and order’ and ‘safeguarding of internal security’ is not immediately clear, since they do not follow established EU terminology, in the English language at least. A comparison with other language versions, such as the French (ordre public) and the German (öffentliche Ordnung), shows that Article 72 TFEU emulates the ‘public policy’ yardstick. That indicates that the provision should be given a similar meaning to start with, referring to a fundamental interest of society in line with

141 See EU Treaty, as amended by the Treaty of Maastricht, art K.2; and EC Treaty, arts 64(1), 68(2), as amended by the Treaty of Amsterdam. 142 See Hemme Battjes, European Asylum Law and International Law (Martinus Nijhoff 2006) 157; and Marcus ter Steeg, Das Einwanderungskonzept der EU (Nomos 2006) 150–58. 143 See Peers (n 126) 84–85; and Stephan Breitenmoser and Robert Weyeneth, ‘Artikel 72 AEUV’ in Hans von der Groeben and others (eds), Europäisches Unionsrecht. Band 2 (7th edn, Nomos 2015) MN 5. 144 The option was mentioned by AG Priit Pikamäe, Opinion in Case C-808/18 Commission v Hungary EU:C:2020:493, point 105. 145 See Kay Hailbronner, Immigration and Asylum Law and Policy of the European Union (Kluwer 2000) 102; and Kay Hailbronner and Daniel Thym, ‘Grenzenloses Asylrecht? Die Flüchtlingskrise als Problem europäischer Rechtsintegration’ [2016] Juristenzeitung 753, 761–63. 146 Stadt Frankfurt am Main (n 42) para 28. 147 Commission v Poland and others (n 61) paras 144–45. 148 ibid para 143; contra AG Eleanor Sharpston, Opinion in ibid, EU:C:2019:917, points 208–21. 149 ibid paras 146–47; emphasis on the need for justification can be read as an implicit rejection of AG Francis Jacobs, Opinion in Case C-120/94 Commission v Greece EU:C:1995:109, points 50–51 regarding TFEU, art 347. 150 See Országos Idegenrendeszeti Főigazgatóság and others (n 64), paras 84–85.

274 General Features of EU Legislation previous comments. While ‘internal security’ usually means internal events, there may be exceptional scenarios where internal and external threats coincide, also considering that entry, border controls, and asylum are often intricately linked to international developments. There is no clear-cut distinction between Article 72 TFEU and Article 347 TFEU on international tensions constituting a threat of war, apart from both provisions specifying the bearing of Article 4(2) TEU for distinct scenarios. An activation of Article 72 TFEU was discussed heatedly in Germany and Austria during the asylum policy crisis of 2015/16 as a potential justification for the suspension of asylum applications at internal Schengen borders.151 The issue was never resolved, and the Jafari judgment cannot be read as an authoritative statement to the contrary, since no government had maintained, neither before the Court nor in its practice, that Article 72 TFEU should have been activated.152 One step further, the provision was invoked unsuccessfully by Hungary in infringements proceedings the Commission had brought against draconian domestic asylum legislation,153 as well as in proceedings regarding the reintroduction of internal border controls. One reason why recourse to Article 72 TFEU will be limited to exceptional scenarios are leges speciales in secondary legislation. Any activation presupposes that the arguments put forward in favour of derogation cannot be adequately addressed on the basis of secondary legislation.154 Along these lines, the Court found provisions on the reintroduction of internal border controls,155 the public policy exception in the relocation decisions,156 and the asylum border procedures to accommodate security concerns.157 Advocates General supported the same view on the rejection of family reunification and student visas; it was also deemed possible to extend the time limits for border procedures in response to overwhelmed reception capacities.158 In doing so, EU legislation may be interpreted in light of Article 72 TFEU, in line with the intermediate position. Moreover, EU institutions can agree on new measures to alleviate concerns of the Member States, including provisional support under Article 78(3) TFEU. Relocation of asylum seekers, financial support, negotiations with third states, and are examples of such compensatory measures. Foreign policy concerns may give rise to similar arguments. In early 2020, the Greek government invoked to Article 72 TFEU, and the sister guarantee of Article 347 TFEU, as a potential justification for the closure of the border with Turkey and the temporary suspension of domestic asylum laws when the Turkish president encouraged foreigners to leave

151 See Daniel Thym, ‘Judicial Maintenance of the Sputtering Dublin System on Asylum Jurisdiction’ (2018) 55 CML Rev 549, 556–58; an activation of Article 72 TFEU was supported by Walter Obwexer, ‘Gutachten für die österreichische Bundesregierung’ (2016); see also Academic Service of the German Parliament (Wissenschaftliche Dienste), ‘Obergrenzen für Asylsuchende und Bürgerkriegsflüchtlinge im Lichte des EU-Rechts’ (PE 6–3000–153/ 15, 16 December 2015) 36–42. 152 See Case C-646/16 Jafari EU:C:2017:586, paras 93–96. 153 See Commission v Hungary (n 5) paras 212–26 in line with AG Pikamäe, points 98–108. 154 This solution mirrors the integration of Article 4(2) TEU into the ‘regular’ analysis of the fundamental freedoms by eg Sayn-Wittgenstein (n 52) paras 83–84. 155 See ch 12.4.2; and Joined Cases C- 368/ 20 and C- 369/ 20 Landespolizeidirektion Steiermark and Bezirkshauptmannschaft Leibnitz EU:C:2022:298, paras 87–90; contra AG Henrik Saugmandsgaard Øe. 156 See Commission v Poland and others (n 61) paras 148–53. 157 See Commission v Hungary (n 5) paras 222–24. 158 See AG Juliane Kokott, Opinion in Case C-540/03 Parliament v Council EU:C:2005:517, points 34–42; AG Maciej Szpunar, Opinion in Case C-544/15 Fahimian EU:C:2016:908, point 71; and AG Richard de la Tour, Opinion in Case C-72/22 PPU Valstybės sienos apsaugos tarnyba EU:C:2022:431, points 126–27.

Abusive Practices 275 the country—a ‘hybrid aggression’ in the eyes of the Greek government.159 That argument was taken up by Lithuania and Poland when the Belarusian dictator threatened to ‘flood’ Europe with migrants in response to economic sanctions during the summer of 2021.160 Similarly, Spain raised the idea when the Moroccan authorities stopped preventing several thousand migrants from entering Ceuta on 17 and 18 May 2021.161 In the case of Belarus, the ‘hybrid attack’ hypothesis was even sanctioned politically by the European Council and the Commission President.162 Judges rejected that argument in a case involving secondary movements between Poland and Lithuania, as opposed to entry directly from Belarus. Lithuania had failed to substantiate its claim and the Asylum Procedures Directive comprised special procedures to respond to security threats.163 Even so, the argument remains on the table as a matter of principle, since the judgment did not overturn the idea of non-compliance. Article 72 TFEU provides national governments with legal ammunition to justify restrictive practices. Settled case law advises caution, but it does not categorically prevent recourse to such arguments. Even if Article 72 TFEU was activated, doing so would not introduce a state of emergency in the theoretical sense of disrespecting legal obligations to save the legal order from collapse.164 The Court has made clear that any recourse to the derogation is subject to judicial oversight. Furthermore, the duty of loyal cooperation under Article 4(3) TEU requires Member States to coordinate among themselves and with EU institutions to minimise negative effects. Derogations will usually have to be limited in time to comply with the principle of proportionality. They are bound to respect, moreover, fundamental principles of Union law, including human rights and the prohibition of refoulement, which are not subject to emergency caveats under the Charter or international law.165 Article 72 TFEU does not allow for an exit from the EU’s constitutional framework in which it is integrated, rather, as an exceptional flexibility measure.

10.8 Abusive Practices The Court has developed unwritten general principles to fill gaps in primary law and secondary legislation. Such general principles are not simply projected onto migration law, which, rather, partakes in developing them further.166 A telling illustration is the notion 159 See Achilles Skordas, ‘The Twenty-Day Greek-Turkish Border Crisis (Part I & II)’ EU Immigration and Asylum Law (5 and 8 May 2020). 160 See Seimas (Lithuanian parliament), ‘Resolution on Countering Hybrid Aggression’ (No XIV-505, 13 July 2021) https://www.lrs.lt/sip/portal.show?p_r=35403&p_k=2&p_t=277512 (accessed 1 March 2023); and the legal amendments in Poland according to the ECRE Weekly Bulletin (15 October 2021) https://mailchi.mp/ecre/ecre- weekly-bulletin-15102021 (accessed 1 March 2023). 161 See José Ignacio Torreblanca, ‘This Time Is Different. Spain, Morocco, and Weaponised Migration’ European Council on Foreign Relations (26 May 2021). 162 Ursula von der Leyen, ‘Strengthening the Soul of our Union’ (State of the Union Address, 15 September 2021); and European Council, ‘Conclusions’ (ECUO 17/21, 22 October 2021) Nos 19–21. 163 See Case C-72/22 PPU Valstybės sienos apsaugos tarnyba EU:C:2022:505, paras 72–74, 92. 164 cf Carl Schmitt, Politische Theologie (Duncker & Humblot 1922). 165 See AG de la Tour (n 158) points 118–19; Hailbronner and Thym (n 145) 762–63; and ECRE, ‘Derogating from EU Asylum Law in the Name of “Emergencies” ’ (Legal Note No 6/2020) 6–7. 166 See Henry Labayle and Jean-Sylvestre Bergé, ‘Les principes de l’Espace de liberté, de sécurité et de justice’ [2016] Revue trimestrielle de droit européen 589, 601–602.

276 General Features of EU Legislation of ‘abuse’ which is frequently used in political debates about asylum by those favouring restrictions. By way of example, the Commission aims to discourage secondary movements as a form of ‘abuse’.167 Such broad political usage should not be confused with the unwritten general principle bearing the same name. The latter was developed by the Court in tax law and with regard to economic market regulation, building on wider notions of ‘abuse’ in many civil law jurisdictions; it can be compared to common law-style estoppel.168 As a legal principle, abuse permits authorities to sanction behaviour which formally complies with secondary legislation. The underlying rationale is simple, although notoriously difficult to operationalise: anyone abusing a right is barred from invoking a guarantee that would normally apply. In the Court’s case law, this requires, first, a combination of objective circumstances in which, despite formal observance of the [law], the purpose of those rules has not been achieved, and, secondly, a subjective element consisting in the intention to obtain an advantage from the European Union rules by artificially creating the conditions laid down for obtaining it.169

Practical difficulties in applying the ‘abuse’ benchmark stem from the inherent vagueness of what kind of behaviour contradicts the objective of a rule whose requirements have been formally met. This implies the concept of abuse cannot be relied upon—in contrast to the political usage—in an abstract manner without an individual examination.170 ‘Abuse’ cannot generally be held against those using an option under Union law. With regard to secondary movements, the Court found that it is not legally abusive to file multiple asylum claims, to delay an application, or to appeal against a negative decision even if the probability of success is small.171 Domestic authorities can sanction non-cooperation, reject applications, or initiate take back requests under the Dublin III Regulation, but that does not render the behaviour abusive in a legal sense. Similarly, advising and counselling persons whose asylum application is bound to be rejected in accordance with Articles 8(2) and 22(1) Asylum Procedures Directive 2013/32/EU cannot be criminalised as an abusive practice.172 A rare scenario where the general principle might be relevant are ‘golden visa’ schemes, discussed in Chapter 7.3.3, when states hand out residence permits in disrespect of secondary legislation, effectively to sell transnational mobility rights.

10.9 Practical Tips for Dealing with Legislation EU migration law is comprised of more than twenty important pieces of secondary legislation as well as multiple other instruments of a more technical character, which jointly fill 167 Commission, ‘Communication on a New Pact on Migration and Asylum’ COM(2020) 609 final 6. 168 See Stefan Vogenauer, ‘The Prohibition of Abuse of Law’ in Rita de la Feria and Stefan Vogenauer (eds), Prohibition of Abuse of Law (Hart Publishing 2011) 521. 169 Case C-456/12 O and B EU:C:2014:135, para 58 on Union citizenship. 170 See Case C-202/13 McCarthy and others EU:C:2014:2450, paras 52–57. 171 See Joined Cases C- 322/ 19 and C- 385/ 19 The International Protection Appeals Tribunal and others EU:C:2021:11, paras 86–93. 172 See Case C-821/19 Commission v Hungary EU:C:2021:930, paras 111–33.

Practical Tips for Dealing with Legislation 277 hundreds of small font size, single-spaced pages in the Official Journal. It is challenging to keep an oversight. Practical knowledge about the formalities and the contours of supranational legislation can be extremely useful to prevent confusion. To start with, secondary legislation can be easily identified on the basis of the document number via the EUR-Lex portal.173 By way of example, the Visa List Regulation (EU) 2018/1806 can be found by entering the year ‘2018’ and the number ‘1806’. Purists will notice that the official titles changed over the years. Regulations used to begin with the number (instead of the year) and included the abbreviation ‘No’. By contrast, directives had always mentioned the year first, never comprised the abbreviation ‘No’, and concluded with the abbreviation ‘EU’ or ‘EC’. The abbreviation ‘EC’ was discontinued when the Treaty of Lisbon entered into force, and uniform document references were introduced in 2015. Nevertheless, it is correct to cite older instruments with the title published in the Official Journal.174 Official names of directives and regulations can be long and complex. For that reason, this contribution uses ‘nicknames’ for important instruments, such as ‘Students and Researchers Directive (EU) 2016/801’. An overview of the nicknames can be found in the list of abbreviations. While it is straightforward to find legislation via the document number, readers should be careful not use the original version unreservedly. Amendments habitually come in the form of new instruments with a separate document number. For instance, the Visa Code Regulation (EC) No 810/2009 was fundamentally altered—not replaced—by several follow- up regulations. These instruments must be read in parallel as a result. Furthermore, instruments on a specific topic may amend other pieces of legislation. By way of example, the Visa Code Regulation was amended by Article 6 former Schengen Borders Code Regulation (EU) No 610/2013. Information on amendments can be found via the EUR-Lex portal in the section on ‘document information’. There, you may also find informal ‘consolidated versions’ summarising existing amendments, for instance for the Visa Code Regulation. The EU institutions want to adopt ‘recast’ legislation more frequently to promote user- friendliness by means of codified versions.175 The Interinstitutional Style Guide requires articles to be cited before the name of the legislative act; moreover, it foresees capital letters when referring to a specific provision, eg ‘Article 7(1)’, and does not use abbreviations, such as ‘art/s’.176 Judgments of the CJEU and the ECtHR follow that line as well, as do the domestic practices in most continental civil law jurisdictions.177 This contribution continues this tradition in the main text, while obeying— grudgingly—the Anglo-Saxon model of mentioning subdivisions behind the legal instrument and in an abbreviated form. Doing so has advantages if it is not immediately clear to which instrument an article belongs. By contrast, the downside is equally apparent: it parts from the practice of the EU institutions in a book which concentrates on their output.

173 See https://eur-lex.europa.eu/advanced-search-form.html (accessed 1 March 2023). 174 Note that there are different versions: legislation can be found in the ‘L’ edition (for French législation), while non-binding documents are published in the ‘C’ edition (for French communications). 175 See Interinstitutional Agreement between the European Parliament, the Council of the European Union and the European Commission on better law-making [2016] OJ L123/1, No 46. 176 See Interinstitutional Style Guide, section 3.3.2 https://publications.europa.eu/code/en/en-000100.htm (accessed 1 March 2023). 177 I am grateful to colleauges throughout Europe who have responded to my little query regarding the citation practice.

278 General Features of EU Legislation

10.10 Summary European migration law is complex. Literally thousands of provisions are spread across several dozen pieces of legislation. To keep an oversight is notoriously difficult, for beginners and experts alike. Having said this, there are overarching patterns which constitute a sort of ‘general part’ of horizontal features characterising several instruments discussed in this volume. It adds value in terms of using scarce resources efficiently and supporting cross- fertilisation to analyse these questions together. Comments in this chapter concentrated on such general characteristics. One of them comes to the fore when we compare migration law to the single market. In contrast to the free movement of Union citizens, residence permits for third country nationals are constitutive in character; beneficiaries can have an individual right to receive a permit, but they are staying irregularly without it. State authorisation comes in the form of ‘visas’, delivered by consulates abroad, and ‘residence permits’, which can be issued within or outside the territory. Judgments on fee levels illustrate why the horizontal analysis is beneficial: findings on long-term resident status can be projected on permits issued for other purposes. The Court prohibits excessive fee levels, which are much higher than the actual administrative costs and might have dissuasive effects on potential applicants. The quest for interpretative coherence, with judges exploring linkages between the migration law instruments, is particularly pronounced with respect to numerous judgments on the meaning of the economic self-sufficiency and the public policy exception. Under recourse to free movement case law, the Court interpreted the ‘stable and regular’ and ‘sufficient resources’ benchmarks, used in the legislation on legal migration, to require an individualised assessment that takes into account personal circumstances—in contrast to the notion of ‘means of subsistence’ in the Schengen Borders Code Regulation and the Visa Code Regulation, for which Member States may apparently lay down mandatory reference amounts. ‘Public policy’ and ‘public security’ are essential building blocks of EU migration law: they authorise the refusal or withdrawal of residence permits even if the other conditions have been met. Interpretation of these benchmarks follows free movement standards to start with. On that basis, judges will inspect the wording, general scheme, drafting history, and objectives of the instrument concerned to determine the level of protection for each migration law instrument separately. The outcome of the interpretative exercise will often differ from Union citizenship. As a rule of thumb, family reunification and long-term residence benefit from higher standards, while applications by students and for Schengen visas can be rejected on the basis of an abstract risk assessment, considering other factors than the personal conduct of the applicant. Context-specific solutions similarly define the judicial practice on the proportionality test. Inspection of the case law unearthed an important divergence that is rarely discussed. Whereas many judgments emphasise the need for administrative fine-tuning to take into the account the specificities of the individual case, other rulings accept abstract criteria as an outcome of the weighing exercise, which do not leave room for the consideration of personal circumstances. Our analysis identified several parameters which help to identify situations where judges require an individualised assessment. The human right to equality before the law can be a powerful argument to call into question unequal treatment on the basis of migration status, or between nationals and foreigners. Much depends on the level of scrutiny judges apply when examining justification. A bird’s

Summary 279 eye view reveals mixed messages from the CJEU and the ECtHR. While better standards for nationals or Union citizens are often accepted, differences between third country nationals are sometimes found to be illegal. Unfortunately, it remains difficult to distil clear patterns from the case law: judges insist that only ‘weighty reasons’ can justify unequal treatment on grounds of nationality but do not qualify migration status as an immutable personal characteristic warranting strict scrutiny; states are given leeway for redistributive policies and in situations of irregular presence; and the legal category of indirect ethnic of racial discrimination is much narrower than in the interdisciplinary debate. Additional protection can be available under equal treatment clauses in secondary legislation, which will be discussed in Chapter 15.3 on integration. EU legislation often allows for more favourable domestic rules, although the degree of flexibility depends on the design of secondary legislation. Settled case law obliges Member States not to deviate from mandatory prescriptions which amount to full harmonisation and requires parallel domestic schemes to be distinct from residence permits governed by EU legislation; the Court holds no jurisdiction as a result. Article 72 TFEU concerns the opposite scenario, since it may exceptionally be relied upon to justify non-compliance with Union law for safeguarding internal security. Judges insist, however, that the derogation must be interpreted strictly and that it does not typically apply when EU legislation has introduced common standards on how to address an internal security threat. Nevertheless, Article 72 TFEU provides national governments with legal ammunition to counter allegations of illegality in extreme scenarios, such as the ‘instrumentalisation’ of migration by neighbouring states. Along similar lines, the implications of the legal concept of ‘abuse’ are much stricter than the widespread political usage of the term; it does not usually justify non-compliance with written guarantees of EU migration law.

11

Visa Policy Passports and visas have long been the ‘first line of defence against the entry of undesirables’.1 On the European continent, they became widespread after the First World War, replacing the earlier focus on expulsion after entry, as described in Chapter 1.1.2 on the historic evolution. In recent years, migration control has become a central motivation for visa requirements, complementing the earlier focus on foreign affairs, security, and trade. States check compliance with admission criteria before individuals show up at the external borders. Doing so permits thorough inspection of the individual case. Most importantly, however, pre-entry clearance replaces retroactive enforcement by preventive action. Frontloading avoids the legal and factual difficulties in enforcing migration law once a person is physically present. Europeanisation played a critical role in the proliferation of such pre-arrival measures. Reliance on ‘remote control’2—or the ‘border abroad’3—has significant repercussions for refugees. Visa requirements are, together with carrier sanctions, the single most effective instrument states employ to ‘manage’ movements of refugees and other potential asylum applicants. Afghans will be prohibited from boarding a flight from Teheran or Istanbul to a destination in the Schengen area unless they are in possession of an entry visa. Remote control practices are an essential component of the contemporary regime of stratified mobility: barriers for some contrast with privileges for tourists and other bona fide travellers (11.1). EU institutions benefit from a broad range of competences for entry policies among the members of the Schengen area (11.2). Human rights and cooperation with third states will feature in Chapters 12 and 18. EU visa policy principally concerns short-term travel for up to ninety days. Residence permits handed out by consulates abroad for longer periods are subject to supranational legislation or domestic laws discussed elsewhere, for instance with regard to family unity and labour migration. After years of cumbersome debate, the European Union finally achieved a high degree of harmonisation for ‘Schengen visas’ regarding short stays. States whose nationals must be in possession of visas are contained in the Visa List Regulation (11.3), whereas the Visa Code Regulation harmonises admission criteria and application procedures (11.4). Humanitarian visas are a bone of contention in legal and political circles (11.5). Carrier sanctions prevent refugees and other migrants from reaching the external borders where they might apply for asylum (11.6). The Visa Information System (VIS) and the European Travel Information and Authorisation System (ETIAS) were presented previously, in Chapter 9 on databases. 1 John Torpey, The Invention of the Passport (2nd edn, CUP 2018) 224. 2 Aristide R Zolberg, ‘The Great Wall against China’ in Jan Lucassen and Leo Lucassen (eds), Migration, Migration History, and History (Peter Lang 1997) 291, 308. 3 Elspeth Guild, ‘The Border Abroad: Visas and Border Controls’ in Kees Groenendijk and others (eds), In Search of Europe’s Borders (Kluwer 2003) 87. European Migration Law. Daniel Thym, Oxford University Press. © Daniel Thym 2023. DOI: 10.1093/oso/9780192894274.003.0012

Theory and Policy Design 281

11.1 Theory and Policy Design Cosmopolitan internationalists are proud of the rich collection of visas and entry or exit stamps they receive when visiting exotic holiday destinations. Such trophies of globetrotting contrast with the lived experience of persons from the Global South who find it difficult to obtain Schengen visas. International travel is defined by profound asymmetries, which effectively establish two distinct worlds of stratified movement. The sociologist Zygmunt Bauman coined the phrase ‘tourists and vagabonds’ to describe the bifurcated character of (non-)privileged mobility.4 Iconic manifestations of stratification on the European continent are images of beaches in Greece or on the Canary Islands showing both sunbathing tourists and refugees leaving rubber boats. Experts in migration law occasionally overlook the immense regulatory and technical resources states invest in facilitating the movements of some; they concentrate on the obstruction of mobility.5 While visa requirements are obvious instruments of stratification, other measures are less visible. Technical infrastructure includes airports, as well as coastguard vessels and drones financed by the Member States and Frontex. Chapter 12.7 on document security will mention the harmonisation of technical standards for machine readable travel documents and biometric identifiers within the EU and the International Civil Aviation Organization (ICAO). Self-service systems and e-gates will be introduced at the external borders for registered travellers when the Entry/Exit System (EES) becomes operational in 2024.6 Validity periods for multiple-entry visas were extended in 2019.7 EU migration law enables the speedy and contactless travel of bona fide travellers, while simultaneously obstructing irregular movements. The numbers of persons involved are impressive and illustrate why most governments are eager to support stratified travel. Tourism, in particular, can be an important source of economic development. International statistics show that almost 1.5 billion visitors cross inter-state borders each year for business, leisure, or holidays, visits to friends or family, or other purposes such as medical treatment.8 Europe is one of the most popular destinations, even if we exclude cross-border travel across its internal borders.9 The sheer number of arrivals means that states are unable to check each person systematically at crossing points (think of busy terminals at airports like Paris Charles de Gaulle).10 Stratification allows for the swift movement of the privileged, while obstructing—or even forestalling—the arrival of ‘unwanted’ migrants. Interdisciplinary perspectives will be discussed in Chapter 13.1 on asylum and Chapter 18.1 on the external dimension.

4 See Zygmunt Bauman, Globalization: The Human Consequences (Columbia UP 1998) ch 4. 5 See Steffen Mau and others, Liberal States and the Freedom of Movement (Palgrave Macmillan 2012); and Thomas Spijkerboer, ‘The Global Mobility Infrastructure’ (2018) 20 EJML 452. 6 See Schengen Borders Code Regulation (EU) 2016/399, arts 8a–8d, as amended by Regulation (EU) 2017/ 2225 as regards the use of the Entry/Exit System [2017] OJ L327/28. 7 See Visa Code Regulation (EC) No 810/2009, art 24(2), as amended by Regulation (EU) 2019/1155 amending Regulation (EC) No 810/2009 [2019] OJ L188/25. 8 See World Tourism Organisation (UNWTO), World Tourism Barometer: Statistic Annex (Vol 19/3, May 2021), for the period before the Covid-19 pandemic. 9 During 2019, intra-EU travel accounted for 540 out of roughly 750 million arrivals to a European country, half of the almost 1.5 billion worldwide. 10 See also Rey Koslowski, ‘International Travel’ in Alexander Betts (ed), Global Migration Governance (OUP 2011) 109, 115–22.

282 Visa Policy

11.2 Constitutional Foundations Visa policy is—unlike border controls—subject to few human rights constraints as a result of the ‘primarily territorial’ conceptualisation of state jurisdiction by the European Court of Human Rights (ECtHR). Core features of the human rights case law will be presented in Chapter 12.2.2 on border controls. Comments that follow will focus on the broad range of supranational competences (11.2.1), as well as the degree and scope of Member State participation (11.2.2).

11.2.1 Supranational competences Certain aspects of visa policy had been subject to a supranational competence ever since the Maastricht Treaty, presenting intricate problems of how to define the EU’s powers.11 Article 77(2)(a) TFEU overcame these difficulties as a result of a conscious choice to streamline the Treaty framework.12 The ‘common policy on visas and other short-stay residence permits’ covers all sorts of short-term travel, including the absence of visa requirements, procedural prescriptions, and cooperation among consular staff.13 EU institutions benefit from a certain leeway when determining what they consider to be ‘a short stay’; primary law does not set in stone the template of Schengen visas issued for periods of ninety days.14 Moreover, the legislature may have recourse to Article 79(2)(a) TFEU on ‘long-term visas and residence permit’ whenever we conclude, as in the case of humanitarian visas, that the sojourn does not qualify as a short-term stay. Primary law allows for the seamless regulation of migration statuses, if need be by means of multiple legal bases.

11.2.2 Territorial scope Table 11.1 reminds the reader that the justice and home affairs are subject to country- specific opt-outs for Ireland and Denmark (as well as the UK, before Brexit). Defining features and the motivation behind the opt-outs were presented in Chapter 2. Norway, Iceland, Switzerland, and Liechtenstein are associated with the Schengen rulebook via bilateral agreements presented in Chapter 17.2.2. Difficulties in how to define the scope of the Schengen acquis will be discussed in Chapter 12. New Member States participate in measures which are not intricately connected to the abolition of internal border controls

11 See Annalisa Meloni, Visa Policy within the European Union Structure (Springer 2005) chs 2–4; and Kay Hailbronner, Immigration and Asylum Law and Policy of the European Union (Kluwer 2000) 133–47. 12 See European Convention, ‘Final Report of Working Group X’ (CONV 426/02, 2 December 2002) http:// european-convention.europa.eu (accessed 1 March 2023). 13 See Case C-291/12 Schwarz EU:C:2013:670, paras 15–19; Daniel Thym, ‘Legal Framework for Entry and Border Controls’ in Daniel Thym and Kay Hailbronner (eds), EU Immigration and Asylum Law: Article-by-Article Commentary (3rd edn, CH Beck/Hart Publishing/Nomos 2022) MN 9–15; and Steve Peers, EU Justice and Home Affairs Law, vol I (4th edn, OUP 2016) 172–74. 14 A proposal to extend the time limit to six months for certain travellers was rejected by the Council; cf Proposal for a Directive relating to the conditions in which third-country nationals shall have the freedom to travel COM(2001) 388 final.

Visa Requirements 283 Table 11.1 Participation in the most relevant visa policy instruments Instrument

Ireland / UK (before Brexit)

Denmark

Schengen? Accession (NO, IS, CH, LI)

Visa Format Regulation yesa (EC) No 1683/95

yes

yes

yes

Carrier Sanctions Directive 2001/51/EC

no/yes

no (int. Law)

yes

yes

Visa Code Regulation (EC) No 810/2009

no

no (int. Law)

yes

no

Visa List Regulation (EU) 2018/1806

No

yesb

yes

yes

aBoth countries take part in general provisions, such as the introduction of photographs, but not, however,

in measures related to the Schengen acquis; this solution reiterates the historic pedigree of pan-European harmonisation under the Treaty of Maastricht. bDenmark participates, since visa requirements had already been supranationalised— unlike visa procedure—under the Treaty of Maastricht.

immediately.15 They have to introduce visa requirements after accession, whereas participation in the Visa Code Regulation presupposes the abolition of internal border controls.

11.3 Visa Requirements Visas are not an invention of the European Union, and, yet, the EU has played a critical role in the proliferation of visa requirements for migration control purposes. ‘Black’ and ‘white’ lists of countries whose nationals do (not) require a visa when crossing an external border embrace migration control as a major reason for introducing or lifting visa requirements (11.3.1). EU institutions use the promise of visa liberalisation as an incentive to persuade third states to support migration management by means of domestic reforms (11.3.2). A sensitive political issue is reciprocity when third states introduce asymmetric visa requirements for selected Member States (11.3.3). Chapter 10.4 on equal treatment explained that the practice of the Court of Justice does not currently support the criticism that distinctions based on country of origin amounts to illegal discrimination.

11.3.1 Rationale behind ‘black’ and ‘white’ lists Today’s Visa List Regulation (EU) 2018/1806 builds upon a complex process of harmonisation along two parallel tracks. Member States initially agreed upon a minimum list of 101 15 See eg Act concerning the conditions of accession of the Republic of Croatia [2012] OJ L112/21, art 4(1) and Annex II.

284 Visa Policy third states under the Treaty of Maastricht,16 which was readopted without modification after the instrument had been annulled by the CJEU on procedural grounds.17 Member States remained free to determine the status of nationals from third states not mentioned in the minimum black list; there was also a ‘grey’ category of visa-free foreigners.18 Coordination within the Executive Committee of the intergovernmental Schengen area proved more effective.19 All ten Schengen members had agreed upon comprehensive ‘black’ and ‘white’ lists covering all third states, except for Colombia, when the Schengen acquis was integrated into the EU framework under the Treaty of Amsterdam.20 This full harmonisation provided the template for today’s Visa List Regulation, which was adopted in 2001 with a black list comprising 134 third states.21 Criteria for the introduction or revocation of visa requirements remain abstract, comprising ‘a variety of criteria relating, inter alia, to illegal immigration, public policy and security, economic benefit . . . and the Union’s external relations’.22 Notwithstanding this plethora of relevant factors, the black list has been surprisingly stable over the years. Ecuador and Bolivia are rare examples of visa requirements being reintroduced on grounds of illegal migration risk.23 All other instances of delisting essentially concerned enlargement or visa liberalisation.24 The Commission has been criticised for the abstract character of country assessments,25 although decisions are more transparent nowadays than under the intergovernmental Schengen framework, which had combined existing visa requirements in an exercise of negative accumulation: third states subject to visa requirements for some countries were put on the black list for all others.26 Visa requirements are a comparatively crude control instrument, categorising countries as a whole on the basis of abstract risk profiles. One step further, individual characteristics of third country nationals become relevant when states examine individual visa applications. Visa requirements do not, in other words, forbid entry categorically; they relegate decisions to consular staff working in embassies abroad. To be in possession of a valid visa is a mandatory entry condition, even though visa holders can still be refused entry, in particular on the basis of new information indicating non-compliance with the entry conditions.27 In exceptional circumstances, Schengen visas can be handed out at border crossing points, and Member States may authorise admission on humanitarian grounds.28 Individuals applying for asylum must be admitted as well, in accordance with the Asylum Procedures Directive. Having said this, it is an—often undeclared—motivation behind visa requirements to prevent people from applying for asylum. Even a cursory survey shows that prominent 16 See first Visa List Regulation (EC) 2317/95. 17 See former Visa List Regulation (EC) No 574/1999; and Case C-392/95 Parliament v Council EU:C:1997:289. 18 Former Visa List Regulation (EC) No 574/1999, arts 2(1), 4(1); and Hailbronner (n 11) 133–41. 19 See generally ch 1.2.1; and Convention Implementing the Schengen Agreement (CISA), arts 9–17. 20 See Annex I to Appendix I Decision SCH/Com-ex (99)13 on the definitive versions of the Common Manual and the Common Consular Instructions [2000] OJ L239/317. 21 See former Visa List Regulation (EC) No 539/2001 [2001] OJ L81/1. 22 Visa List Regulation (EU) 2018/1806, art 1. 23 See Regulation (EC) No 453/2003 [2003] OJ L69/10, recital 1; and Regulation (EC) No 1932/2006 [2007] OJ L405/23, recital 1. 24 See the survey by Peers (n 13) 186–91; and Meloni (n 11) 100–107. 25 See Violeta Moreno-Lax, Accessing Asylum in Europe (OUP 2017) 90–92. 26 See Annalisa Meloni, ‘The Development of a Common Visa Policy under the Treaty of Amsterdam’ (2005) 42 CML Rev 1357, 1365–67. 27 See Schengen Borders Code Regulation (EU) 2016/399, arts 6(1)(b), 14(1). 28 See ch 12.5.1; and Visa Code Regulation (EC) No 810/2009, art 35.

Visa Requirements 285 countries of origin are blacklisted. Such strategic use of visa policy as a migration control instrument was spearheaded by the destination countries north of the Alps. A telling example is Turkey, which was blacklisted in several countries when political tensions before the military coup of 1980 resulted in an increase in asylum applications.29 Europeanisation was instrumental in the proliferation of visa requirements for potential asylum applicants to ever more Member States. EU institutions even promote, as we shall see, an alignment of the visa policies of neighbouring states. On the whole, migration control has become the most important motivation for lifting or introducing visa requirements, although other aspects remain relevant. Think of foreign policy or the promotion of tourism.30 ‘Golden passport’ schemes are a special scenario, which will be discussed in Chapter 15.7.3 on nationality law. The Pacific island nation of Vanuatu effectively sold passports to rich foreigners from 2015 onwards, after having been whitelisted by the EU. Visa-free access to the Schengen area was one of the benefits which rendered the golden passport scheme attractive. In 2022, the EU suspended the visa waiver agreement for all passports issued after 2015, after several years of diplomatic tensions.31 ‘Airport transit visas’ are a distinct category for transit passengers at international airports. They have traditionally been left out of the Visa List Regulation, in line with earlier limitations of competence, since they do not involve entry into the Schengen area in the administrative sense of checks at border crossing points. However, today’s Article 79(2)(a) TFEU covers airport transit visas, thus allowing for the replacement of intergovernmental coordination by supranational legislation.32 The Visa Code Regulation (EC) No 810/2009 established a minimum black list, which states may supplement by means of domestic catalogues.33 Afghans and Syrians, for example, require an airport transit visa to board a flight from Istanbul to London via Frankfurt, in line with the supranational list (for Afghans) and German rules (for Syrians).

11.3.2 Visa liberalisation as a bargaining chip Recent years have seen a cautious expansion of visa-free travel. The revised Visa List Regulation (EU) 2018/1806 contained a ‘black list’ with 104 third states, noticeably fewer than the 134 countries which had been blacklisted seventeen years earlier. What might appear as straightforward liberalisation at first sight is the product of complex dynamics upon closer inspection. To start with, enlargement was a double-faced undertaking for countries like Poland or Romania. On the one hand, their citizens gained visa-free travel in the Schengen area. On the other hand, Poland and Romania had to introduce visa requirements for neighbouring countries, such as Ukraine or Serbia, which meant that borders that had previously been permeable became difficult to cross. Local border traffic, discussed 29 See Case C-221/11 Demirkan EU:C:2013:583, paras 17–19; and Jens Vedsted-Hansen, ‘The Legal Condition of Refugees in Denmark’ (1994) 7 J Refugee Stud 249, 255. 30 See Commission, ‘Communication: Implementation and development of the common visa policy to spur growth in the EU’ COM(2012) 649 final, 3–4; and Koslowski (n 10) 120–21. 31 See Decision (EU) 2022/366 on the partial suspension of the application of the Agreement between the European Union and the Republic of Vanuatu on the short-stay visa waiver [2022] OJ 2022 L69/105. 32 See Common Consular Instructions (n 20) No 2.1.1; and Joint Action 96/197/JHA on airport transit arrangements [1996] OJ L63/8. 33 See Visa Code Regulation (EC) No 810/2009, art 3, Annex IV, with later amendments.

286 Visa Policy in Chapter 12.5.4, and visa liberalisation mitigated these negative repercussions of enlargement. Neighbouring states were promised to be put on the white list if they fulfilled a number of conditions supporting EU justice and home affairs policies.34 Negotiations proceeded comparatively quickly with the Western Balkans countries, with the exception of Kosovo, and have been completed with Eastern neighbours, such as Ukraine, Moldova, and Georgia. Negotiations with Turkey have dragged on for years, while not a single country in the Southern Mediterranean has been offered the prospect of visa liberalisation. Negotiations were modelled on the experience of enlargement conditionality: third states are expected to fulfil legislative and practical criteria contained in ‘roadmaps’ and ‘action plans’, and the Commission regularly reports about the progress made. Conditionality embraces diverse themes ranging from document security to criminal matters and the treatment of minorities.35 Moreover, EU institutions will often insist upon the parallel conclusion of readmission agreements, discussed in Chapter 16.6.4 on return, and the alignment of the visa policy of neighbouring states with the EU standard in line with comments towards the end of this chapter. Once the conditions have been fulfilled, ‘visa facilitation agreements’ will be signed as a first step. They involve lower application fees, streamlined documentary evidence, and the introduction of multi-entry visas.36 Finally, visa-free travel will be introduced on the basis of supranational legislation. While that step has been promised politically to some countries in roadmaps, it remains the unilateral decision of the EU institutions legally. Fully fledged visa waiver agreements have only been signed with small island nations, such as Mauritius and the Seychelles.37 The conditionality approach works comparatively well in relations with countries for whose citizenry visa-free entry into Schengen area has an economic and symbolic value. The Western Balkans is a case in point: cooperation on return proceeds comparatively smoothly precisely because governments know that it is connected to visa liberalisation. A generic mechanism allows for the suspension of visa-free travel when return proves problematic and irregular entry or stay increases.38 The suspension mechanism was introduced when visa liberalisation towards the Western Balkans had coincided with a rise in the number of asylum applications during the early 2010s, which were mostly rejected as unfounded.39 Effective cooperation of the countries of origin meant that the EU institutions chose not to activate it. Suspension of the visa waiver was put on the table, together with pre-accession support, as a negative incentive to persuade Serbia to reintroduce visa requirements for third states when an increasing number of their citizens used visa-free flights to Belgrade as a gateway for illegal entry into the EU during the autumn 2012.40 34 See Mathias Czaika and Florian Trauner, ‘EU Visa Policy’ in Ariadna Ripoll Servent and Florian Trauner (eds), The Routledge Handbook of Justice and Home Affairs Research (Routledge 2018) 110, 113–15. 35 By way of example see Commission, ‘Second Report on progress by Turkey in fulfilling the requirements of its visa liberalisation roadmap’ COM(2016) 140 final; and Florian Trauner and Imke Kruse, ‘EC Visa Facilitation and Readmission Agreements’ (2008) 10 EJML 411. 36 By way of example see Agreement between the European Community and the Republic of Serbia on the facilitation of the issuance of visas (adopted 18 September 2007, entered into force 1 January 2008) [2007] OJ L334/ 137; and Agreement between the European Union and the Republic of Armenia on the facilitation of the issuance of visas (adopted 17 December 2014, entered into force 1 January 2014) [2013] OJ L289/2. 37 They can be found in [2009] OJ L169. 38 See Visa List Regulation (EU) 2018/1806; introduced first by means of Regulation (EU) No 1289/2013 [2013] OJ L347/74, arts 8–11; and reformed thereafter by Regulation (EU) 2017/371 [2017] OJ L61/1. 39 See Florian Trauner and Emanuele Manigrassia, ‘When Visa-Free Travel Becomes Difficult to Achieve and Easy to Lose’ (2014) 16 EJML 125, 132–44. 40 See ‘Press Conference’ (Justice and Home Affairs Council, 14 October 2022) https://europa.eu/!rTF6xG (accessed 1 March 2023).

Visa Requirements 287 On the basis of the conditionality mechanism, the Commission continuously pursues structured dialogues with third states, thus perpetuating the strategic use of visa-free travel to monitor neighbouring countries constantly.41 Similar instruments have been put in place for third states which do not benefit from visa-free travel. The latest version of the Visa Code Regulation introduced additional ‘carrots’ and ‘sticks’ regarding visa fees or processing times to improve cooperation on return and other aspects of migration control. They will be discussed in Chapter 18.3.2. on the external dimension. Visa policy has become an essential leverage in the foreign policy arsenal of the EU institutions in support of better migration management.

11.3.3 Reciprocity in relations with third states Reciprocity is a general feature of international law and defines visa policy. Visa-free travel was an important stepping stone for the free movement of Union citizens in the single market,42 and it characterised the Nordic Passport Union and numerous other bilateral agreements. Having said this, reciprocity can also be relevant in the absence of contractual relations when states unilaterally lift visa requirements for nationals of selected countries. In the EU, reciprocity became a hot political topic after enlargement to Central and Eastern Europe. As we have seen, all new Member States applied the Visa List Regulation from day one after accession. This raised the problem how to react to asymmetric visa waiver practices of the United States, which required citizens from most new Member States to apply for entry visas. Similar disputes arose when Canada revoked the visa waiver for the Czech Republic after several hundred asylum applications, in particular by ethnic Roma.43 Some requested an activation of the comparatively strict reciprocity clause in the former Visa List Regulation to retaliate by means of blacklisting the US and Canada. Doing so might have triggered the reintroduction of visa requirements for all travel between the Schengen area and North America. EU institutions finally agreed on a revised reciprocity mechanism which gives priority to diplomatic solutions over the immediate threat of sanctions.44 Diplomatic efforts yielded some success: Canada finally whitelisted all Member States, while the US was retaining visa requirements for Cyprus, Bulgaria, Romania, and Croatia at the time of writing. The European Parliament has started proceedings for failure to act criticising the Commission’s inaction in terms of activating the mechanism towards the US; judges will probably confirm that the Commission has a wide discretion when deciding which steps it deems appropriate.45

41 See Commission, ‘Fourth report under the visa suspension mechanism’ COM(2021) 602 final; and Roman Lehner, ‘Gemeinsame Visapolitik’ in Ferdinand Wollenschläger (ed), Enzyklopädie Europarecht, Band X (2nd edn Nomos 2021) § 28 MN 14–19. 42 Today’s Free Movement Directive 2004/38/EC, art 5(1) reiterates earlier guarantees. 43 See Didier Bigo, Sergio Carrera, and Elspeth Guild (eds), Foreigners, Refugees or Minorities? (Ashgate 2013). 44 See Visa List Regulation (EU) 2018/1806, arts 7, 10–11; introduced first by Regulation (EU) No 1289/2013 amending the former Visa List Regulation [2013] OJ L347/74, 45 See AG Jean Richard de la Tour, Opinion in Case C-137/21 Parliament v Commission EU:C:2022:989; and Commission, ‘Communication: State of play as regards the situation of non-reciprocity in the area of visa policy’ COM(2020) 119 final.

288 Visa Policy In the future, reciprocity might hypothetically gain relevance if the UK decided to introduce visa requirements for selected Member States. Chapter 17.3.2 explains that the Trade and Cooperation Agreement, adopted on the eve of Brexit, does not require indefinite visa- free travel, while establishing a procedure for prior notification of the intention to introduce visa requirements. We could expect bitter diplomatic disputes if the UK reintroduced asymmetric entry rules. The Regulation adding the UK to the white list for visa-free travel after Brexit urged the institutions to apply the reciprocity mechanism strictly in such a case.46

11.4 Visa Procedure Member States waiting to join the Schengen area apply the Visa List Regulation 2018/1806 but are not bound by the procedural standards and admission criteria in the Visa Code Regulation (EC) No 810/2009. The latter apply to full Schengen members only. Readers should be careful to work with a consolidated version of the Visa Code, as described in Chapter 10.9 on practical tips. EU institutions agreed on several amendments, which must be read into the original text. A recast codifying earlier changes has not been proposed or adopted so far. Particularly relevant in practice are admission criteria (11.4.1), application procedures (11.4.2), legal remedies (11.4.3), the visa format (11.4.4), and inter-state cooperation (11.4.5).

11.4.1 Admission criteria Schengen visas will be delivered to anyone fulfilling the entry conditions under Article 6(1) Schengen Borders Code Regulation (EU) 2016/399. Ex ante verification in the visa procedure allows for a thorough examination border guards cannot usually perform at crossing points. Visa procedures provide a platform for the collection and assessment of documents, biometric identifiers, and other information in accordance with the revised version of the Visa Code, which also foresees the option of an interview. Exceptions exist for beneficiaries of free movement, in particular family members of Union citizens; they are covered by the Free Movement Directive.47 Visa requirements for refugees with a protection status from a third state are determined on the basis of residence (not nationality). By way of example, a Syrian refugee holding a travel document from Mauritius may enter the Schengen are without a visa, in the same way as Mauritian nationals.48 Humanitarian visas for those wishing for protection in the EU are discussed elsewhere. Article 10(1) Convention Implementing the Schengen Agreement had envisaged the introduction of a ‘uniform visa’ for the entire Schengen territory—an objective the Visa Code Regulation finally delivered. Its rules build upon fragmented earlier instruments adopted by the supranational institutions and the intergovernmental Schengen Executive Committee, including the Common Consular Instructions.49 The designation ‘code’ hints 46 See Regulation (EU) 2019/592 as regards the withdrawal of the United Kingdom from the Union [2019] OJ L103I/1, art 2. 47 See Visa Code Regulation (EC) No 810/2009, art 1(2); and Free Movement Directive 2004/38/EC, art 5(2). 48 See Visa List Regulation (EU) 2018/1806, art 3(2). 49 See Common Consular Instructions (n 20); and Annalisa Meloni, ‘The Community Code on Visas: Harmonisation at Last?’ (2009) 34 EL Rev 671, 673–80.

Visa Procedure 289 at a high degree of harmonisation, notwithstanding differences on the ground when consulates abroad apply the common rulebook. Visa statistics show staggering discrepancies in the rejection rates, which can hardly be explained fully by divergent characteristics of the applicants.50 Four admission criteria are particularly relevant in practice. First, the identity of the applicant and the veracity of travel documents are an important prerequisite consular staff is meant to assess with the help, amongst others, of biometric identifiers and the data stored in the VIS and the Schengen Information System (SIS).51 Interoperability of the migration databases will enhance the control options of consular staff. At the same time, information submitted during the visa procedure will be available to border guards and administrative authorities across for diverse purposes, as discussed in Chapter 9.3 on databases. Secondly, consular staff ascertains whether applicants pose a public policy or public security risk. Such an assessment may be based on entry bans, prior consultation of other states, or any other source of information.52 The bearing of ‘public policy’ and ‘public security’ was elaborated upon in Chapter 10.4 on general features, whereas rules on alerts for entry and stay will feature in Chapter 16.4 on return. We shall see that Member States benefit from a certain leeway when considering these criteria, even though blanket bans for specific countries would not usually be possible. Stricter practices could possibly be justified by invoking an ‘international relations threat’, which the Visa Code and the Schengen Borders Code recognise as an additional exception;53 judges will often exercise restraint when assessing foreign policy considerations, for instance with regard to travel restrictions for Russians after the armed attack on Ukraine.54 Thirdly, an application will be refused unless applicants demonstrate sufficient means of subsistence, both ‘for the duration of the intended stay and for the return to his country of origin’; they also need an adequate travel medical insurance.55 Chapter 10.3 addressed doctrinal uncertainties as to whether the concept of ‘means of subsistence’ authorises states to rely on abstract lump sums in the context of visa applications and entry clearance. Fourthly, applications will be rejected if there is a ‘risk of illegal immigration’. Consular staff is entrusted with verifying ‘whether the applicant intends to leave the territory of the Member States before the expiry of the visa’.56 Visa procedures are supposed to counter the phenomenon of ‘overstaying’, with third country nationals not respecting the obligation to return home before the expiry of the visa.57 To determine whether intends to return, consul staff will consider the purpose of the stay, previous visits, or indirect evidence such as job security, income levels, not family connections, amongst others.58 It is sufficient if national authorities have ‘reasonable doubts’; certainty is not required59 (here, para 68). 50 For regularly updated statistics see https://ec.europa.eu/home-affairs/policies/schengen-borders-and-visa/ visa-policy_en (accessed 1 March 2023). 51 See Visa Code Regulation (EC) No 810/2009, arts 12, 13, 21(1)–(3), with later amendments; and Schengen Borders Code Regulation (EU) 2016/399, art 6(1)(a). 52 Visa Code, arts 21(3)(c), (d), 22; and Schengen Borders Code, art 6(1)(d), (e). 53 Visa Code, arts 32(1)(a)(vi); and Schengen Borders Code, art 6(1)(e). 54 See Daniel Thym, ‘Border Closure and Visa Ban for Russians’ EU Immigration and Asylum Law and Policy (17 October 2022); and, on the example of Russia, ch 2.3.4. 55 Visa Code Regulation (EC) No 810/2009, arts 15, 21(3)(b), (e), (5), with later amendments. 56 ibid art 21(1). 57 See Case C-241/05 Bot EU:C:2006:634, para 27. 58 See Visa Code Regulation (EC) No 810/2009, arts 14(1)(d), 21(3)(b), (4), with later amendments. 59 Case C-84/12 Koushkaki EU:C:2013:862, para 68.

290 Visa Policy The abstract character of many admission criteria entails that the consular staff has leeway when deciding individual cases. In cases of doubt, it will rely on the ‘veracity and reliability’ of the statements made by the applicant, mirroring credibility assessment during asylum procedures.60 Such appraisal necessarily involves a certain degree of factual discretion. Judges accept this explicitly, while highlighting that the grounds for refusal are exhaustive.61 Administrative authorities have a certain leeway when applying the law, but they cannot invent additional conditions. Third country nationals benefit from an individual right, which can be enforced via domestic courts, to be issued a visa in case the consular staff applies supranational rules incorrectly.

11.4.2 Application procedure Applications can be lodged no more than six months and up to fifteen days before the intended stay in regular circumstances by the applicant or through intermediaries.62 Private intermediaries play an important role where applications can be difficult to submit, either as a result of long distances or difficulties in obtaining an appointment. Member States often struggle to comply with the objective of offering an appointment within two weeks. Long queues before consulates are a regular occurrence (and nuisance). It remains to be seen whether the option of electronic submission and extended cooperation with private service providers, introduced in 2019, will shorten waiting times. One step further, the Commission wants to makes a leap forward in terms of digitalising the lengthy and paper-based process by the end of 2025: electronic applications and payment of the visa fee shall become mandatory; applicants will no longer be required to go to consulates in person when they have submitted biometric identifies on a previous occasion.63 Application fees currently stand at €80 for most people, although that rate can be lowered, as we have seen, on the basis of visa liberalisation agreements and cooperation on return (failure of cooperation may result in higher fees). Private service providers may charge half of the amount on top.64 That is a lot of money, even if we recognise that fee levels are lower than in the United States. Multiple-entry visas can be provided with a validity of up to five years, in particular for those who have received consecutive Schengen visas in the past, or ‘prove their integrity and reliability’ otherwise.65 Doing so reinforces the prevalent trend towards stratification, also considering that trustworthy applicants can have other benefits as well.66 A valid passport or travel document, biometric identifiers (photographs and fingerprints), and various supporting documents shall be delivered to the consulate or to private service providers, as applicable.67 Examination involves three steps. First, the competence must be verified, since applicants cannot freely choose the consulate where to apply. By way of example, an application with 60 Visa Code Regulation (EC) No 810/2009, arts 21(7), 32(1)(a); and ch 13.5.2. 61 See ch 10.1.2; and Koushkaki (n 59) para 55. 62 See Visa Code Regulation (EC) No 810/2009, arts 9–11, Annex I, with later amendments. 63 See Proposal for a Regulation as regards the digitalisation of the visa procedure, COM(2022) 658 final, art 2(3). 64 See Visa Code Regulation (EC) No 810/2009, arts 16–17, with later amendments. 65 ibid art 24(2)–(2d). 66 ibid arts 9(3) and 14(6) by way of example. 67 ibid arts 12–14, 43(6), (c), (f).

Visa Procedure 291 the Danish consulate will be returned (together with the fee) when the Swedish consulate should have been consulted by someone who is planning to fly to Copenhagen and then take a local train to the Swedish city of Malmö to visit her brother.68 While Schengen visas have transnational effects in the sense of authorising entry and stay in the Schengen area as a whole, it has advantages to designate jurisdiction to a specific embassy. Consular staff of the main destination country will have a better knowledge of contextual factors there, and applicants cannot engage in ‘visa shopping’ by applying with the consulate of a Schengen member that is considered to be more lenient. Secondly, formalities such as the completion of the standard form are subject to an admissibility check. In case of inadmissibility, the fee will be returned, and another application can be submitted at any time.69 Thirdly, the authorities will engage in a substantive evaluation; they may ask for additional documents or request a personal interview.70 This last stage of the examination remains the prerogative of consular staff; it cannot be delegated to external service providers.71 A decision will usually be taken within fifteen calendar days after having lodged the application (and after having obtained an appointment to do so).72 Chapter 7.3.2 explained that non-compliance with the time limit does not have immediate consequences.

11.4.3 Decision and legal remedies Visas have traditionally been conceived to be a foreign policy issue closely related to core facets of state sovereignty. As a consequence, procedural safeguards were scarce or non- existent. Most Member States did not foresee legal remedies against visa rejection, nor were consulates obliged to give reasons.73 The matter had been relegated to domestic laws under the original Schengen cooperation,74 until Article 32(2) and (3) Visa Code Regulation introduced procedural safeguards, after intense political debates during the legislative procedure.75 Grounds for refusal must be communicated through a standard form. That form resembles a multiple choice test: consular staff has to tick the relevant boxes, which list the grounds for refusal in an abstract manner; additional remarks can be entered, as well as information on legal remedies.76 Space for additional remarks was extended on the occasion of the 2019 revision. Judges found the standard form to be compatible with fundamental rights, provided that states enter additional information, ‘where appropriate, [on] the essence of the reasons for th[e]‌objection’.77 To do so may be mandatory under the Charter, as demonstrated in Chapter 7.2.4 on the administrative dimension.

68 ibid arts 5(1)(b), 18. 69 ibid arts 19 and 20. 70 ibid art 21(8). 71 ibid art 43(4). 72 ibid art 24(1). 73 See Catherine Schmitter, François Julien- Laferriere, and Jean- Yves Carlier, ‘Vers quelle politique communautaire des visas de court séjour?’ in Philippe De Bruycker (ed), The Emergence of a European Immigration Policy (Bruylant 2003) 97, 152–59. 74 See Common Consular Instructions (n 20) No 2.4. 75 See Meloni (n 49) 690–92. 76 See Visa Code Regulation (EC) No 810/2009, Annex VI, with later amendments. 77 Joined Cases C-225/19 and C-226/19 Minister van Buitenlandse Zaken EU:C:2020:951, para 46.

292 Visa Policy A legal remedy against refusal exists, in accordance with national laws,78 whereas it remains unclear whether appeals can be directed against inadmissibility decisions and decisions refusing jurisdiction.79 In practice, this question has little significance, since applicants can respond to an inadmissibility decision by submitting a new application with the missing documents and challenge the refusal of this second application. Remember that the fee will be reimbursed as well. In any case, domestic court procedure has to comply with the right to an effective remedy in Article 47 Charter of Fundamental Rights (CFR) and the general principle of effectiveness, in line with the comments in Chapter 7.2 examining Court judgments on procedural matters. Inter-state cooperation can raise intricate problems of transnational legal oversight. Chapter 7.3.1 explained how judges dealt with scenarios involving more than one consulate. Notwithstanding the high number of visa applications, statistical surveys confirm that only a small majority of all refusals is challenged. This low percentage can be rationalised by practical difficulties in seizing domestic courts from abroad and the absence of immediate benefits, such as admission on the territory.80

11.4.4 Visa format One of the first legislative instruments adopted under the Treaty of Maastricht remains in force: the Visa Format Regulation (EC) No 1683/95. Several amendments and implementing decisions introduced new and improved security features against forgery and counterfeiting (some of them are secret); digital photographs were one such technical upgrade.81 A proposal to integrate fingerprints into the visa sticker was abandoned; biometric identifiers will be stored in the VIS instead.82 Physical visa stickers might disappear in the foreseeable future. The Commission has proposed legislation to move towards a ‘digital visa’ by the end of 2025, which would be stored in the VIS.83

11.4.5 Inter-state cooperation Indirect implementation entails that consulates of one country in third states deliver visas, which are valid for entry into the Schengen area as a whole. These enhanced transnational effects render inter-state cooperation particularly important. Discrepancies among national practices undermine the aspiration of uniformity and may even run counter to the objective to prevent ‘visa shopping’.84 ‘Local Schengen cooperation’ was developed as a remedy early on, between the consulates situated in a specific third state. They are expected to coordinate on formal issues, such as translation of the application form in the official language of the

78 See Visa Code Regulation (EC) No 810/2009, art 32(3), with later amendments; and Case C-403/16 El Hassani EU:C:2017:960, paras 39–41. 79 See Annalisa Meloni, ‘Visa Code Regulation (EC) No 810/2009’ in Thym and Hailbronner (n 13) Article 32 MN 7. 80 See ch 3.3.4.1; and ch 7.2.3.3. 81 Visa Code Regulation (EC) No 810/2009, arts 2, 3, 6, and Annex. 82 See Moreno-Lax (n 25) 84–85. 83 See Digitalisation Proposal (n 63), art 1(20). 84 Koushkaki (n 59) para 53, referring to Visa Code Regulation (EC) No 810/2009, recital 18.

Visa Procedure 293 host state or conversion of the application fee.85 Exchange shall extend to the factual situation in the host country to foster a common approach towards the application of abstract admission criteria, such as public policy threat or illegal migration risk (the call for official ‘risk profiles’ has been abandoned86). Effective cooperation on the ground ultimately depends on the willingness of local actors to trust each other and to search for joint solutions to common practical questions. In the early days of Europeanisation, the vision of ‘common visa offices’87 (possibly within the European External Action Service) or a ‘common European issuing mechanism’88 were promoted at the highest political level as the administrative infrastructure of a truly uniform visa policy. Such calls had always been carefully phrased to cover various forms of transnational cooperation, instead of outright federalisation.89 Whereas primary law covers closer cooperation, Treaty change would be required for the creation of veritable ‘EU consulates’ where supranational staff would deliver visas, subject to judicial oversight by the General Court in Luxembourg.90 ‘Co-location’ of national staff from different countries or the creation of ‘common application centres’ in new premises were promoted instead, and a first centre was opened in the Moldovan capital of Chișinău in 2007.91 The project was not a big success, encountering myriad practical difficulties ranging from divergent job regulations to questions of immunity.92 It proved easier (and cheaper) to pursue other means of cooperation. Representation arrangements allow countries to authorise another Schengen members to take decisions on their behalf, especially in countries where they do not have consulates. By way of example, the Netherlands entrusted Switzerland to deal with applications in Sri Lanka for which the Netherlands would normally be responsible.93 The CJEU concluded that Swiss courts must decide legal remedies in such cases, as explained in Chapter 7.3.1 on transnational legal oversight. Moreover, private service providers have gained prominence. The latest reform of the Visa Code enhanced their role in the collection of visa applications, supporting documents, and even fingerprints.94 This involvement of private service providers receives little public attention. It can have tangible benefits in terms of swift and unbureaucratic visa delivery (subject to an extra cost), but it similarly raises questions of political, financial, and administrative accountability.95

85 Visa Code Regulation (EC) No 810/2009, art 48; and Commission Implementing Decision establishing the Handbook for the administrative management of visa processing and local Schengen cooperation (Visa Code Handbook II), C(2020) 1764 final, Annex Pt II. 86 See Meloni (n 79) Article 48 MN 1. 87 The Hague Programme [2005] OJ C53/1 Pt III.1.7.3. 88 The Stockholm Programme [2010] OJ C115/1 Pt 5.2. 89 See Matthias Laas, Die Entstehung eines europäischen Migrationsverwaltungsraumes (Nomos 2008) 139–44. 90 See ch 8.2.1 on agencies; ch 3.1.3 on the court architecture; and Thym (n 13) MN 8, 11. 91 See Common Consular Instructions (n 20) Pt VII, Point 1.2, as amended by Regulation (EC) No 390/2009 [2009] OJ L131/1; and Commission, ‘Opening of a “Common Visa Application Centre” in Moldova’ (MEMO/07/ 153, 25 April 2007). 92 See Annalisa Meloni, ‘EU Visa Policy. What Kind of Solidarity?’ (2017) 24 MJECL 646, 663–64. 93 See Visa Code Regulation (EC) No 810/2009, art 8, with later amendments. 94 ibid art 43; and Visa Code Handbook II (n 85) Pt III. 95 See María-Luisa Sánchez-Barrueco, ‘Business as Usual? Mapping Outsourcing Practices in Schengen Visa Processing’ (2018) 44 JEMS 382.

294 Visa Policy

11.5 Special Case: Humanitarian Visas Visas for the entire territory of the Schengen area are known as ‘uniform visas’ in the language of the Visa Code Regulation. They must be distinguished from ‘visas with limited territorial validity’. Such national visas may be issued ‘exceptionally’ when a Member State considers it necessary to derogate from the admission criteria for Schengen visas ‘on humanitarian grounds, for reasons of national interest or because of international obligations’.96 This option has gained prominence during legal and political debates about humanitarian visas. Proponents present them as antidotes to prevalent externalisation practices in the form of safe and legal pathways for refugees. The Visa Code Regulation does not address the question directly, besides the option of national visas in Article 25(1) and an abstract reference to fundamental rights in Article 1(4). In the X and X judgment, delivered in March 2017, the Court rejected an application of the Visa Code Regulation to national humanitarian visa schemes. The case concerned a Syrian family of four, who had travelled from the war-torn city of Aleppo to Lebanon to apply for a humanitarian visa in the Belgian consulate. A Belgian family was willing to sponsor their stay under a purely discretionary humanitarian admission scheme under domestic law.97 The preliminary reference to Luxembourg pursued the objective of determining whether the consulate was obliged to deliver the visa on human rights grounds. The Grand Chamber answered in the negative: the Visa Code Regulation was not applicable, since humanitarian visas were about longer stay. To find otherwise would contradict the focus on short stays and the regulatory emphasis of the Common European Asylum System on applications made on the territory or at the border.98 The judgment has been criticised for its formalistic reasoning. Doing so allowed the Grand Chamber to evade an authoritative statement about the extraterritorial application of the Charter and obligations, under Article 4 CFR, actively to protect refugees under an imminent threat of severe ill-treatment99—in contrast to the opinion of Advocate General Mengozzi, which had attracted critical observers. Mengozzi had concluded his reasoning with an energetic plea that the Court ‘has the opportunity not only to restate, I hope vigorously, respect for the humanitarian values and human rights . . . but also, and above all, to offer the applicants in the main proceedings the hope of being spared further suffering and inhuman treatment’.100 The case put the spotlight on externalisation practices, which prevent refugees from reaching the Schengen area safely. In that respect, the political and normative repercussions were significant. When it comes to the doctrinal small print of the visa procedure, however, the impact of the X and X judgment was less drastic. Judges found domestic humanitarian admission schemes not to be covered by the Visa Code Regulation, since that instrument did not harmonise entry and stay for longer periods. In the absence of supranational harmonisation, the Charter did not apply either. This outcome effectively relegated

96 Visa Code Regulation (EC) No 810/2009, art 25(1)(a). 97 See Tristan Wibault, ‘Making the Case X & X for the Humanitarian Visa’ in Marie-Claire Foblets and Luc Leboeuf (eds), Humanitarian Admission to Europe (Nomos/Hart Publishing 2020) 271. 98 See Case C-638/16 PPU X and X EU:C:2017:173, paras 40–49. 99 See Sylvie Sarolea, ‘Is Access to Asylum the Same as Access to Justice?’ in Foblets and Leboeuf (n 97) 115, 128–35; and Meloni (n 92) 649–52. 100 AG Paolo Mengozzi, Opinion in Case C-638/16 PPU X and X EU:C:2017:93, point 168.

Other Pre-arrival Measures 295 the sensitive matter to the domestic arena, thus reinforcing that impression of an ‘administrative mindset’ on the part of the Court, as mentioned in Chapter 3.2.3. The Syrian family went to the ECtHR instead, which interpreted Article 3 ECHR not to mandate humanitarian visas, as described in Chapter 12. Reasoning before and within the Court took place against the backdrop of political discussions about the introduction of a European humanitarian visa scheme. The European Parliament had promoted the idea during the legislative debate about the reform of the Visa Code, only to meet fierce resistance among many members of the Council.101 To overcome the stalemate, the Commission withdrew the proposal and tabled new legislation, which eventually resulted in the adoption of a resolution of the Parliament calling for the introduction of a humanitarian visa mechanism. Crucially, this new scheme was to be realised in a separate instrument, thus preparing the ground for the latest reform of the Visa Code Regulation without provisions on humanitarian visas.102 They would have to be based on Article 79(2)(a) TFEU, as indicated by the Court,103 but the Commission has shied away from proposing additional legislation so far. National practices on protected entry procedures will be discussed in Chapter 13.10 on asylum.

11.6 Other Pre-arrival Measures Visas are the most visible embodiment of ‘remote control’, trying the influence the volume and composition of migratory movements towards Europe. Several measures increase the practical leverage of visa requirements. Carrier sanctions play a critical role in thwarting spontaneous arrivals (11.6.1). Less discussed and more difficult to track is the influence the European Union exerts over the visa policies of transit states, such as Turkey (11.6.2). A traditional instrument are liaison officers, even though they rarely assume operational functions (11.6.3). The upgrade and forthcoming interoperability of databases featured in Chapter 9.

11.6.1 Carrier sanctions A one-way ticket for a flight from Western Africa to the Schengen area can be bought online for less than €300, and flights from the Maghreb are available for much less. Why do Nigerians or Moroccans pay significantly more to risk their lives on old ships hardly fit for navigation on the high seas? The answer is simple. Airline staff at the point of departure will deny check-in or boarding to anyone who is not in possession of the documents required for entry into the Schengen area. Carrier sanctions featured in the original Convention Implementing the Schengen Agreement and were reinforced by one of the first pieces of supranational legislation adopted under the Treaty of Amsterdam.104 101 See LIBE Committee, ‘Working Document on humanitarian visas’ (PE619.272v01-00, 19 March 2018) 2–3; and Eugenia Relaño Pastor, ‘EU Initiatives on a European Humanitarian Visa’ in Foblets and Leboeuf (n 97) 341, 353–63. 102 See European Parliament, ‘Resolution with recommendations to the Commission on humanitarian visas’ (P8_TA(2018)0494, 11 December 2018). 103 See X and X (n 98) para 44. 104 See CISA (n 19), art 26, which remains in force; and Carrier Sanctions Directive 2001/51/EC.

296 Visa Policy Carrier sanctions are not a new invention. They have been practised for more than a century, not least by the US.105 The main asylum destination countries in Europe took up the idea four decades ago. While Denmark is usually portrayed as the pioneer from 1983 onwards, Germany had already introduced similar rules by the end of 1981.106 The UK, France, Italy, Belgium, the Netherlands, and others followed suit.107 Once again, Europeanisation served as a framework for mutual learning and the exchange of ‘best practices’ to states which had not used carrier sanctions previously by means of informal policy transfer and, later, supranational legislation. Doctrinally, carrier sanctions consist of a double obligation. First, airlines and other transport companies have to organise transportation back home at their own cost, whenever third country nationals are refused entry at the border.108 That obligation will have few practical implications if the passenger applies for asylum, since asylum applicants must be admitted on the territory, or to transit zones, in accordance with the Asylum Procedures Directive. Secondly, penalties of €3,000–5,000 for each person can be issued whenever companies do not diligently inspect ‘travel documents required for entry’, especially passports and visas.109 Passengers can bring lawsuits against transportation companies before domestic courts for denial of boarding and to claim compensation under consumer protection legislation.110 In practical terms, however, this may only help those denied boarding mistakenly, for instance when airline staff wrongly assume a visa was mandatory.111 Nigerians and Moroccans without an entry visa will not benefit from the complaint option. Airlines must not consent to boarding without visas. Carrier sanctions have been ‘subject to the obligations resulting from [the Refugee Convention]’ from the beginning.112 Nevertheless, it remains an uphill struggle to argue that they violate international refugee law or human rights. As in the case of humanitarian visas, the jurisdictional hurdle might prove insurmountable.113 Other guarantees have not traditionally been considered to hinder carrier sanctions either, in particular the prohibition of penalties for refugees on account of illegal entry under Article 31 Refugee Convention.114 The German Constitutional Court declined to engage in a substantive scrutiny of carrier sanctions,115 and political proposals to introduce an exemption for asylum seekers were rejected.116 105 See Bernard Ryan, ‘What Role for Legal Guarantees?’ in Bernard Ryan and Valsamis Mitsilegas (eds), Extraterritorial Immigration Control (Martinus Nijhoff 2010) 3, 19. 106 See Foreigners Law (Ausländergesetz), s 18(1), as amended by Law on Combating Illegal Employment (Gesetz zur Bekämpfung der illegalen Beschäftigung) [1981] Federal Law Gazette (Bundesgesetzblatt) I/1390, art 6(1). 107 See António Cruz, Shifting Responsibility. Carrier Liability in the Member States of the European Union and North America (Trentham Books 1995) 27–66; and Amnesty International, ‘No Flights to Safety. Carrier Sanctions, Airline Employees and the Rights of Refugees’ (ACT 34/021/1997, 1 November 1997) 3–6. 108 See CISA (n 19), art 26(1)(a); and Carrier Sanctions Directive 2001/51/EC, arts 2, 3. 109 CISA (n 19), art 26(1)(b), (2), (3); and Carrier Sanctions Directive 2001/51/EC, art 4. 110 See Regulation (EC) No 261/2004 establishing common rules on compensation and assistance to passengers in the event of denied boarding and of cancellation or long delay of flights [2004] OJ L46/1, arts 2(j), 7, 8; and Case C-584/18 Blue Air—Airline Management Solutions EU:C:2020:324, paras 87–98. 111 For a complex case involving a flight between two Member States waiting to join the Schengen area see Blue Air—Airline Management Solutions (n 110). 112 See CISA (n 19), art 26(1). 113 See ch 12.2.2; moreover, boarding is denied by private parties, not by state officials. 114 See Gregor Noll, ‘Article 31’ in Andreas Zimmermann (ed), The 1951 Convention Relating to the Status of Refugees and its 1967 Protocol: A Commentary (OUP 2011) MN 35; and compare Moreno-Lax (n 25) 351–52. 115 See Federal Constitutional Court (Bundesverfassungsgericht), No 2 BvL 55/92 (decision of 2 December 1997) https://www.servat.unibe.ch/dfr/bv097049.html (accessed 1 March 2023). 116 On the early 2000s see Moreno-Lax (n 25) 142–47.

Other Pre-arrival Measures 297 Carrier sanctions are intended to have preventive effects. States want transport companies not to carry those without travel documents to the external borders in the first place ‘to curb migratory flows and [to] combat illegal immigration’.117 There is little doubt that they contribute to this objective. To be sure, many will come to Europe nonetheless, risking their lives, paying piles of money, and suffering exploitation on the journey. Yet, the persistence of irregular migration does not mean that carrier sanctions are ineffective. Our inspection of the drivers of migratory movements in Chapter 4.1 on interdisciplinary perspectives showed that rising transaction costs are likely to reduce the capacity and willingness to embark on the journey. The relevant question is not whether irregular movements persist (or even increase) despite carrier sanctions but what numbers would come hypothetically if carrier sanctions were abandoned.

11.6.2 Export of visa policy to transit countries The effectiveness of pre-arrival measures is compromised if they can be circumvented by travelling to a nearby destination from where the Schengen area can be reached easily. For that reason, the EU seeks to cooperate with neighbours to buttress visa requirements and carrier sanctions. An example from the early days of externalisation were ‘deals’ West Germany and, later, Sweden and Denmark struck with the Communist regime in East Germany to require potential asylum seekers to show entry visas for Western states before boarding flights bound for East Berlin.118 At present, new Member States applying to join the EU are expected to comply with the Visa List Regulation and the Carrier Sanctions Directive from day one after accession, as mentioned previously. Enlargement to countries in Central and Eastern Europe entailed the geographic expansion of the EU’s visa regime, thus effectively creating a cordon sanitaire or ‘buffer zone’ rendering it more difficult to reach the main countries of destination. The exportation of visa policies continues. An instructive example are ongoing negotiations with Kosovo and Turkey about visa liberalisation, building upon the experience of other Western Balkans countries or Ukraine. Previous comments on visa liberalisation mentioned that the EU institutions use the incentive of visa-free travel to extract concessions. Inspection of the ‘roadmaps’ shows that Kosovo and Turkey are expected to emulate the EU’s visa regime, including carrier sanctions, as a precondition for visa liberalisation.119 Similar negotiations are taking place with neighbours to the East and South, although third states will be less likely to satisfy the EU’s demands without the incentive of visa liberalisation.120 Indeed, we should be careful not to portray neighbouring countries as passive recipients of the EU’s instructions. Turkey and Morocco are countries of destination and transit at the same time. While visa requirements may respond to the desire of the EU side, 117 Carrier Sanctions Directive 2001/51/EC, recital 2. 118 See ch 18.3.1. 119 For the original demand see Commission, ‘[First] Report on progress by Turkey in fulfilling the requirements of its visa liberalisation roadmap’ COM(2014) 646 final, 12–15; and Commission, ‘[First] Report on progress by Kosovo in fulfilling the requirements of its visa liberalisation roadmap’ COM(2013) 66 final, 8–9. 120 See Laure Delcour, ‘The EU. Shaping Migration Patterns in its Neighbourhood and Beyond’ in Dimitry Kochenov and Fabian Amtenbrink (eds), The European Union’s Shaping of the International Legal Order (CUP 2014) 261, 266–68; and Raül Hernández i Sagrera, ‘Reframing the EU Visa Cooperation with Third Countries’ in Bigo and others (n 435) 189.

298 Visa Policy they may be motivated by domestic considerations as well. Public opinion in Turkey, for instance, has become critical of further arrivals of Syrians and Afghans, and the government has constructed walls to ‘protect’ borders with Syria and Iran. Having said this, EU institutions have shown that they have additional instruments of coercion when third states are unwilling to cooperate. Chapter 10 on general features described how the EU reacted swiftly and decisively when the Belarusian dictator Lukashenko ‘instrumentalised’ migration to respond to EU sanctions against his authoritarian regime. Travel companies sold ‘tourist packages’, including entry visas, for Belarus to several thousand Iraqis, Syrians, and other nationals. In return for substantial payments, they were flown to Minsk for onward movement towards Poland or Lithuania. EU institutions and the Member States convinced commercial airlines not to cooperate any longer. The subtle threat of limiting access to the lucrative European market was sufficient to persuade Turkish Airlines and FlyDubai, amongst others, to stop flying to Minsk.121 A proposal to formalise that regime was still being considered at the time of writing.122 Moreover, EU institutions had similarly used the threat of suspending visa liberalisation, as we have seen, to press Serbia to reintroduce visa requirements for some nationalities during 2022.

11.6.3 Liaison officers abroad A conventional form of externalisation is the deployment of police officers and border guards to third states. Between 2004 and 2018, the number of liaison officers of the Member States almost quadrupled from 128 to 492 in 105 countries altogether.123 Deployment became a standard practice under the intergovernmental cooperation framework during the 1990s, both between the Member States and to third states. The main purpose was to enhance operational cooperation by means of personal contacts and better information flows. Liaison officers served as facilitators, rather than decision-takers, to render the flanking measures for the abolition of internal border controls effective.124 In the context of enlargement, they acted as trainers and role models to prepare future members to fulfil obligations under the Schengen acquis. This indirect role is the main reason why we know comparatively little about what liaison officers do.125 Liaison officers are occasionally tasked with participating in decision-making of third states or private carriers. Early instruments mentioned advice and assistance in individual cases prominently, in particular in the form of document advisers assisting airlines in checking travel documents.126 Similar tasks have not received much attention recently and no longer take centre stage in policy documents, possibly as a result of digitalisation and enhanced security features making it easier for companies to check entry requirements by 121 See ‘Turkish Airlines Risks EU Ban over Poland Border Crisis’ EUobserver (9 November 2021). 122 See Proposal for a Regulation on measures against transport operators that facilitate or engage in trafficking in persons or smuggling of migrants COM(2021) 753 final. 123 See the Explanatory Memorandum accompanying the Proposal for a Regulation on the creation of a European network of immigration liaison officers (Recast) COM(2018) 303 final, 1–2. 124 See CISA (n 19), arts 7, 47, 125. 125 See also Nicole Ostrand and Paul Statham, ‘”Street-level” Agents Operating Beyond “Remote Control” ’ (2021) 47 JEMS 25. 126 See Joint Action 96/622/JHA on pre-frontier assistance and training assignments [1996] OJ L281/1; and Decision SCH/Com-ex (98) 37 on the adoption of measures to fight illegal immigration [2000] OJ L293/203; see also IATA, ‘A Code of Conduct for Immigration Liaison Officers’ (October 2002).

Summary 299 themselves. Liaison officers seep into the background; the concentrate on advisory and support functions.127 Frontex staff will complement, and possibly replace, many national liaison officers in the years to come, once the border agency has hired all the staff to fulfil the revised mandate.128 Some officers perform operational tasks. Prominent examples include British and French border guards in railway stations and ports of the other party for transportation across the Channel. Bilateral agreements authorise extraterritorial pre-departure controls by British officials on French soil and French officials in Britain, for instance of trains crossing the Eurotunnel. These extraterritorial controls are subject to host state legislation.129 They provide the background of more than twenty years of irregular settlements in the French municipality of Sangatte near the ferry terminal of Calais, which became widely known as the ‘jungle’.130 It has caused political debates in France, and critical observers highlight the negative repercussions of externalisation. Migrants stay in the ‘jungle’ when searching for an opportunity to cross the English Channel clandestinely. Franco-British cooperation continued after Brexit and was even gathering momentum at the time of writing, in the same vein as the British opt-out from Schengen had not prevented collaboration from developing dynamically beforehand. Similar extraterritorial controls have traditionally been performed by the United States in countries worldwide.131 Chapter 8.3.3 explained that Frontex staff may be entrusted with (limited) operational assistance in the future. On the whole, however, active involvement of liaison officers in executive decisions remains the exception. States prefer to delegate migration control to third states and private carriers nowadays. Doing so evades human rights scrutiny. Limiting the role of liaison officers to advise, information exchange, and confidence building will not usually cross the jurisdictional yardstick under human rights law.

11.7 Summary Visa policy has important practical, legal, and conceptual repercussions. Member States issue more than 15 million Schengen visas for short stays of up to ninety days every year. In addition, consulates abroad deliver residence permits for longer stays, such as family reunification or work purposes, which are not covered by the legislation on visas. A defining feature of policy developments has been stratification. Smooth travel and easy access for privileged travellers coexists with hurdles and restrictions for others. Visa requirements

127 See Regulation (EU) 2019/1240 on the creation of a European network of immigration liaison officers [2019] OJ L198/88; and Moreno-Lax (n 25) 133–48. 128 See Frontex Regulation (EU) 2019/1896, art 77. 129 See [Sangatte Protocol] concerning Frontier Controls and Policing, Co-operation (adopted 25 November 1991, entered into force 2 August 1993); the Additional Protocol on the Establishment of Bureaux responsible for Controls on Persons Travelling by Train (adopted 29 May 2000, entered into force 25 May 2001); [Le Touquet] Agreement concerning the Carrying of Service Weapons (adopted 4 February 2003, entered into force 10 March 2005); [Sandhurst] Treaty concerning the Reinforcement of Cooperation for the Coordinated Management of their Shared Border (adopted 18 January 2018, entered into force 1 February 2018); and, most recently, ‘UK- France Joint Leaders’ Declaration’ (10 March 2023); see further Ryan (n 105) 16–18. 130 See Imran Rahman-Jones, ‘The History of the “Calais Jungle” Camp and How it’s Changed since 1999’ BBC. com (24 October 2016); and Henri Labayle, ‘Brexit: The Question of Borders between the European Union and the United Kingdom’ EU Immigration and Asylum Law and Policy (8 July 2016). 131 See Mau and others (n 5) 92–111.

300 Visa Policy are such an instrument of distinction between bona fide travellers and the ‘undesirable’. Irregular migration is an essential reason the EU institutions will consider when conducting an abstract risk assessment about whether to introduce, maintain, or lift visa requirements for specific third states. About 100 countries feature on the latest version of ‘black list’ in the Visa List Regulation (EU) 2018/1806. Applications for Schengen visas are made with the consulate of the competent Schengen member in accordance with the revised version of the Visa Code Regulation (EC) No 810/ 2009. Decisions on approval or rejection essentially revolve around the admission criteria in the Schengen Borders Code Regulation, in particular accuracy of travel documents, public policy threats, sufficient resources, and illegal migration risk. The latter is a decisive feature, both for the blacklisting of third states and decisions on individual applications. States aim to prevent irregular overstaying and to reduce the likelihood of an asylum application. Consular staff decide on the application on the basis of documents submitted, databases discussed previously, and, possibly, a personal interview. While decisions are taken by public servants, private service providers are increasingly being relied on to collect the material. Refusal is subject to legal remedies, and reasons must be given, albeit at a high level of abstraction in a standard form. Member States may exceptionally issue national visas with limited territorial validity in case of non-compliance with the admission criteria. In a judgment of principle, the Grand Chamber held, controversially so, that this option does not cover humanitarian visas for longer stays and that the Charter was not applicable as a result. This judgment effectively relegated the matter to the political arena and to the ECtHR. Debates among the EU institutions on humanitarian visas have yielded no results, and the Court in Strasbourg rejected, as we shall see, an obligation to deliver humanitarian visas. By contrast, the external dimension of visa policy is alive and kicking. EU institutions use the attraction of visa-free travel as a leverage to improve cooperation with neighbouring states on various aspects of migration management, including return and an alignment of visa policies. One aspect that is often overlooked is the exportation of visa requirements to neighbouring states as a precondition for visa liberalisation. Critics emphasise that visa policy serves as an instrument of externalisation, which relocates the border abroad to exercise remote control. Comments throughout this chapter have illustrated that migration management concerns do indeed dominate the EU’s visa policy agenda. Carrier sanctions are a crucial instrument, which became standard practice across the Schengen area as a result of Europeanisation. New Member States apply them from day one after accession, even before the abolition of internal border controls. Carrier sanctions require airlines and other transport companies to pay penalties when transporting third country nationals without appropriate travel documents and entry visas. While the practical effects of carrier sanctions are substantial, liaison officers seem to play a limited role; they support information exchange and confidence building.

12

Border Controls Article 3(2) Treaty on the European Union (TEU) proclaims that an area ‘without internal frontiers . . . in conjunction with appropriate measures with respect to external border controls’ shall be a core objective of the European Union. This statement hints at a basic tension. Abolishing border controls within the Schengen area is one of the Union’s proudest achievements. Realising that venture raised the question, however, what the ‘appropriate measures’ with regard to external borders and third country nationals should be. The relative degree of openness or closure remains a controversial topic. Border security is praised as the hallmark of a ‘Europe that protects’ by some, while being criticised as a move towards ‘fortress Europe’ by others.1 Corresponding debates raise formidable questions of legislative design, human rights law, and legal interpretation. When it comes to border controls, Schengen was much more than the simple projection of previous internal practices upon the external borders: control methods were being transformed. Conventional checks on persons at crossing points were complemented by pre- arrival measures, such as carrier sanctions or inter-state cooperation. These instruments turned border crossing points in one among several theatres of a multi-stage control system whose tentacles stretch beyond the geographical border area (12.1). EU institutions benefit from a broad range of supranational competences, whose realisation must comply with human rights and international obligations, including on search and rescue at sea (12.2). Anyone studying border controls must have at least a basic understanding of the idiosyncrasies of Schengen governance (12.3). Recent years have witnessed the resurgence of internal border controls and police checks within the territory, thus demonstrating that the EU institutions struggle to fulfil the original promise of free mobility within the Schengen area (12.4). External border controls comprise checks on persons at crossing points and surveillance of land and sea borders. Checks on persons are subject to detailed rules in the Schengen Borders Code Regulation, comprising legal remedies against refusal of entry (12.5). Surveillance measures at ‘green’ and ‘blue’ land and sea borders are at the centre of contemporary debates, especially with regard to allegations of ‘pushbacks’ of people entering irregularly (12.6). Document security receives little attention, despite being important in practice (12.7). Several policy initiatives in connection with border controls are discussed elsewhere in this volume: operational cooperation within Frontex and databases were elaborated upon in Chapters 8 and 9; criminalisation of illegal entry and stay will feature in Chapter 16.3; and cooperation with third states will take centre stage in Chapter 18.3.

1

See ch 1.2.6.

European Migration Law. Daniel Thym, Oxford University Press. © Daniel Thym 2023. DOI: 10.1093/oso/9780192894274.003.0013

302 Border Controls

12.1 Theory and Policy Design Border are powerful symbols. The abolition of border controls within the Schengen area is a significant achievement of European integration, which is occasionally presented as a remake of the historic model of state formation fostering a sense of common identity, in line with the political ambition of ‘ever closer union’ (12.1.1). At closer inspection, however, the territorialisation of the Schengen area remains incomplete (12.1.2). A defining feature of Europeanisation has been the transformation of control practices. Interior ministries utilise inter-state cooperation to boost leverage through the proliferation of new control instruments (12.1.3).

12.1.1 Symbolic weight of borders Schengen serves as a symbol of European integration. EU institutions started considering the idea of a passport union and the related project of Union citizenship in the 1970s, as an integral part of the broader ambition to move towards some sort of political union.2 Other actors similarly exploit the symbolic potential of borders—albeit in the opposition direction. Governments across the world rely on borders to convey a message of political power. Recent years have seen an astounding proliferation of border fences and walls in numerous countries, which, as a sort of theatrical spectacle, counter the impression of loss of control in a globalised world.3 A conspicuous example was the rallying call of former US President Donald Trump to build an ‘impenetrable, physical, tall, powerful, beautiful, southern border wall’.4 Such slogans build upon the symbolic weight of borders deeply enshrined in our cultural memory. From a very young age, we have familiarised ourselves with inter-state borders as an organising principle of international relations in the form of fine black lines drawn to separate countries on maps. EU institutions make use of this symbolic potential when they present the Schengen area as a signpost to European unity. At the same time, national governments tap into the same reservoir when ‘temporarily’ reintroducing controls within the Schengen area, often, as we shall see, in violation of EU law. Borders have become a contested terrain. Against this background, the creation of the Schengen area invites a comparison with classic episodes of state formation. Schengen marks, so it is argued, the birth of the territorial dimension of the European project.5 Judges in Luxembourg moved in this direction when they spoke, in some judgments on Union citizenship, of the ‘territory of the Union as a whole’6 or, simply, of the ‘territory of the European Union’7. There is undoubtedly a kernel of truth in such statements. Frontex and external border controls nurture a common 2 See ch 1.1.7 and 1.2.1. 3 See ch 4.2.3; and Wendy Brown, Walled States, Waning Sovereignty (Zone Books 2010)chs 1, 4. 4 ‘Donald Trump’s Mexico Wall: Who Is Going to Pay for It?’ BBC.com (6 February 2017) https://www.bbc.com/ news/world-us-canada-37243269 (accessed 1 March 2023). 5 See Lydia Lebon, La territorialité et l’Union européenne (Bruylant 2015); and Jürgen Bast, ‘Völker-und unionsrechtliche Anstöße zur Entterritorialisierung des Rechts’ (2017) 76 Veröffentlichungen der Vereinigung Deutscher Staatsrechtslehrer 277. 6 Case C-434/09 McCarthy EU:C:2011:277, para 66. 7 Case C-256/11 Dereci and others EU:C:2011:734, para 50.

Theory and Policy Design 303 identity (through joint criticism, in the eyes of some). The external travel ban during the Covid-19 pandemic reinforced, after initial setbacks, a sense of commonality by presenting the EU as a unitary actor.8

12.1.2 Incomplete ‘territorialisation’ of the EU At the same time, the territorialisation remains incomplete. The Court did not develop further the abstract references to Union territory.9 The identificatory potential of the Schengen area is compromised by the asymmetric non-participation of some countries. Frontex has received substantial powers and capacities, yet Chapter 8 argued that the agency is a far cry from a truly federal border agency. Similarly, the creation of an ‘area’ of freedom, security, and justice, analysed in light of different language versions, need not embody far-reaching territorial aspirations.10 In a symbolic move, Article 77(4) of the Treaty on the Functioning of the European Union (TFEU) reaffirms national responsibilities for the demarcation of state borders, thus excluding, in particular, definite statements, on the part of the supranational institutions, about the international legal status of Gibraltar.11 Most importantly, episodes of state building differed in important respects: the supranationalisation of border controls did not coincide with an extensive identificatory, bureaucratic, and military internal territorialisation.12 Core policies which are commonly associated with territorial statehood remain the responsibility of the Member States.13 A telling example are equal treatment provisions in secondary legislation, which will be discussed in Chapter 13.3 on migrant integration. They prescribe various degrees of equal treatment with nationals or other persons without guaranteeing specific levels of benefits. Member States retain the authority to determine protection levels, which EU migration law extends to third country nationals. Some observers have argued that the EU mirrors an empire more than a neo-federal ‘super state’ in the Westphalian tradition. That comparison was common at the turn of the millennium when enlargement and the prospect of a neighbourhood policy indicated the emergence of ‘imperial’ authority beyond the external borders.14 Such interpretation can even be traced back to darker episodes of European history, in particular Nazi academics— some continued their careers as experts in EU law—reflecting on a continental ‘Großraum’ (greater area).15 Arguably, Schengen reflects such broader ideas of organising the European space in a post-national and non-totalitarian manner better than simple comparisons with 8 See Stephen Coutts, ‘Citizenship, Territory and COVID-19’ in Dora Kostakopoulou and Daniel Thym (eds), Research Handbook on European Union Citizenship Law and Policy (Edward Elgar Publishing 2022) 116. 9 See Niamh Nic Shuibhne, ‘The “Territory of the Union” in EU Citizenship Law’ (2019) 38 Yearb Eur L 267. 10 See ch 7.1.1. 11 See Henri Labayle, ‘L’espace de liberté, sécurité et justice dans la Constitution pour l’Europe’ (2005) 41 Revue trimestrielle de droit européen 437, 453. 12 cf Stein Rokkan, State Formation, Nation-Building, and Mass Politics in Europe (OUP 1999) 97–107; and Christer Jönsson, Sven Tägil, and Gunnar Törnqvist, Organizing European Space (Sage 2000). 13 See William Walters, ‘Mapping Schengenland. Denaturalizing the Border’ (2002) 20 Envt Plan D: Soc & Space 561; and Massimo Fichera, ‘Sketches of a Theory of Europe as an Area of Freedom, Security and Justice’ in Maria Fletcher and others (eds), The European Union as an Area of Freedom, Security and Justice (Routledge 2018) 34, 40–46. 14 See Jan Zielonka, Europe as Empire (OUP 2006). 15 See Christian Joerges, ‘Europe a Großraum? Shifting Legal Conceptualisations of the Integration Project’ in Christian Joerges and Navraj Singh Ghaleigh (eds), Darker Legacies of Law in Europe (Hart Publishing 2003) 167.

304 Border Controls classic state formation.16 To decide on the degree of openness or closure towards the outside world is one of the challenges such an organisation is confronted with—as the example of Schengen illustrates.

12.1.3 Towards a new ‘culture’ of border controls Millions benefit from the freedom-enhancing appeal of the Schengen area, especially those living near an internal border (as does the author of this book). Abolishing internal controls does not, however, infer a specific vision of external controls. Frontex serves as a powerful reminder that the initial expectation, among some observers, proved incorrect that the EU might replicate internal free movement externally. Chapter 2.1.1 explained that interior ministries jumped on the Schengen bandwagon when they realised that doing so allowed them to promote their agenda through ‘flanking measures’ compensating states for the loss of control at the internal borders. Recent years have seen further strengthening of migration control. In this overall context, the ‘flanking measures’ were no simple continuation of previous internal practices. Schengen served as a laboratory for the design of new instruments, thus gradually establishing a new ‘culture of border controls’.17 Chapter 9 demonstrated that the Schengen Information System (SIS) served as the prototype of justice and home affairs databases, which have expanded significantly. Operational cooperation between national border guards lay the foundations for the creation of Frontex, as analysed in Chapter 8. Countries like Germany and, later, Spain and Italy spearheaded international cooperation, thus effectively externalising border controls to third states, as described in Chapter 18.3. EU institutions promoted visas requirements and carrier sanctions, which have been addressed in the previous chapter. Schengen does not mark the first transformation of border controls, nor is the EU alone in designing new instruments. Border officers checking identity documents at crossing points superseded the earlier focus on post-arrival expulsion. The ‘invention of the passport’ during the long nineteenth century replaced earlier forms of private or public travel documents, which were often delivered in a decentralised manner (think of the town gate as an epitome of local control).18 This transformation is ongoing. The ‘control’ of individuals by border guards is increasingly being replaced by generalised forms of ‘management’ based on abstract risk assessments.19 A decisive development has been the move beyond the external borders: pre-arrival measures and cooperation with third states, discussed in Chapters 11 and 19, effectively established a ‘border abroad’.20

16 See Hans Lindahl, ‘Finding a Place for Freedom, Security and Justice’ (2004) 29 EL Rev 461. 17 Ruben Zaiotti, Culture of Border Controls (University of Chicago Press 2011)chs 1, 2, 5. 18 See ch 1.1.2. 19 See Thomas Spijkerboer, ‘Changing Paradigms in Migration Law Research’ in Carolus Grütter and others (eds), Migration on the Move (Brill 2017) 13, 15–18. 20 Elspeth Guild, ‘The Border Abroad: Visas and Border Controls’ in Kees Groenendijk and others (eds), In Search of Europe’s Borders (Kluwer 2003) 87.

Constitutional Foundations 305

12.2 Constitutional Foundations There are few—albeit politically sensitive—limitations of competence to pursue common action in accordance with Article 77 TFEU (12.2.1). Human rights raise a series of complex questions. A first controversy concerns the extraterritorial reach of the Charter, the ECHR, and multilateral human rights conventions (12.2.2). Search and rescue obligations in the Mediterranean and elsewhere are primarily governed by the law of the sea (12.2.3). Of great practical significance and difficult to disentangle are interpretatory uncertainties over the procedural dimension of the double prohibition of refoulement and collective expulsion (12.2.4). Finally, readers are reminded of the special status of Ireland and Denmark (12.2.5).

12.2.1 Supranational competences The notion of ‘border controls’ in Article 77 TFEU does not lend itself to being easily defined. In any case, it is not limited to checks on persons at crossing points. The historic template of Schengen cooperation illustrates that the drafters of the Treaty competence conceived flanking measures to be covered, with checks on persons being mentioned as an example amongst others. Moreover, the Treaty of Lisbon discontinued the earlier connection to free movement within the Schengen area.21 In light of Irish opt-out, a flight between Belgium and Ireland crosses an external border for the purposes of Article 77 TFEU. EU institutions benefit from a certain flexibility when determining the exact geographic location of control measures. Interpretation in light of the objective of ‘efficient22 management’ of migration flows ‘at all stages’, enshrined in Article 79(1) TFEU, indicates that border controls need not be confined to the immediate proximity of the demarcation line. This conclusion is reinforced by the reference to the management concept, which embodied the reorientation of the international debate towards inter-state cooperation during the early 2000s.23 Thus, the External Sea Borders Regulation (EU) No 656/2014 could be adopted on the basis of Article 77(2)(b) TFEU, even though it applies to coastal waters and the high seas. In cases of doubt, international measures, such as support to third states by the border agency Frontex, can be based on other or additional legal bases.24 Career sanctions are covered by Article 77(2)(a) TFEU, since they complement visa requirements for transportation crossing an external border. Flexibility in the interpretation of the term ‘border controls’ entails that control activities can contribute to search and rescue de facto, even though primary law does not provide for an autonomous competence for search and rescue. In line with case law on the delimitation of competences, Article 77(2)(b) TFEU covers corollary rules on safety at sea, search and rescue, or disembarkation provided the centre of gravity of the measure concerns border controls.25 Telling examples are the Sea Borders Regulation and Frontex missions.

21 cf EC Treaty, art 61(a), as amended by the Treaty of Amsterdam. 22 While the English language version employs the process-oriented term ‘efficient’, others use the outcome- oriented ‘effective’, or wording that can mean both; see the German wirksam, the French efficace, or the Spanish eficaz. 23 See ch 18.1.1. 24 See ch 8.3.3. 25 See Case C-411/06 Commission v Council EU:C:2009:518, paras 45–46.

306 Border Controls

12.2.2 Extraterritorial reach of human rights A prominent criticism of EU policies concerns the ‘externalisation’ of border controls by means of visa requirements and international cooperation. This criticism renders the reach of human rights obligations a pertinent question, which, unfortunately, remains subject to some interpretative ambiguity. What is certain is that Article 1 ECHR obliges state parties to secure human rights ‘to everyone within their jurisdiction’—a condition not limited to the territory. Settled case law, which oftentimes concerned military activities, holds that a state’s ‘jurisdictional competence is primarily territorial’, although extraterritorial action may clear that hurdle in ‘exceptional circumstances’.26 Jurisdiction in this sense covers effective control of an area, as well as factual authority over individuals abroad, for instance after arrest.27 On that basis, the European Court of Human Rights (ECtHR) concluded, in the Hirsi judgment, that Italy held jurisdiction over migrants on board of a coastguard vessel.28 An open question is whether the Court’s statement about migrants on board of ships ‘registered in, or flying the flag of, that State’29 can be extended to other forms of de facto control, in particular when border guards are not in physical contact with migrants.30 This question is far from academic considering the relevance of private vessels and neighbouring states, including Libya, for search and rescue in the Mediterranean. In a case about a delayed rescue operation, the Human Rights Committee found a jurisdictional link even without physical contact ‘in the particular circumstances of the case’.31 It remains to be seen whether the ECtHR will follow that example and whether the same conclusion might even be reached in scenarios involving support given to the Libyan coast guard.32 With regard to humanitarian visas, the Grand Chamber concluded that the rejection of an application in a consulate abroad does not qualify as an exercise of state jurisdiction. To be sure, the Court highlighted that it might find otherwise in ‘exceptional circumstances’, for instance for family reunification or nationals seeking diplomatic protection.33 Involvement of domestic authorities in the visa procedure or the risk of inhuman or degrading treatment were not, however, considered sufficient.34 In a statement of principle, the ECtHR emphasised that ‘to find otherwise would amount to enshrining a near‑universal application of the Convention . . . and therefore to create an unlimited obligation on the Contracting States to allow entry to an individual who might be at risk of ill-treatment’.35 This conclusion came as a disappointment for proponents of functional or joint jurisdiction as an alternative jurisdictional benchmark.36

26 Reaffirmed by MN and others v Belgium App no 3599/18 (ECtHR [GC], 5 May 2020) §§ 98, 102. 27 See the summary in Hirsi Jamaa and others v Italy App no 27765/09 (ECtHR [GC], 23 February 2012) §§ 73–75. 28 ibid §§ 76–82. 29 ibid § 75. 30 See also Solon Solomon, ‘Migrant Boats on the High Seas and their Interception through Psychologically Coercive Measures’ (2019) 37 Netherlands Q Hum Rts 36. 31 See AS and others v Italy Communication no 3042/2017 (HRC, 27 January 2021) Nos 7.4–8. 32 See SS and others v Italy App no 21660/18 (pending). 33 See MN and others v Belgium (n 26) §§ 101–06. 34 ibid §§ 110–23. 35 ibid § 123–24 (references omitted). 36 For different views see Thomas Gammeltoft-Hansen and James C Hathaway, ‘Non-Refoulement in a World of Cooperative Deterrence’ (2015) 53 Colum J Transnatl L 235; and Violeta Moreno-Lax, ‘The Architecture of Functional Jurisdiction’ (2020) 21 German LJ 385.

Constitutional Foundations 307 With the wisdom of hindsight, we may conclude that observers might have misjudged the strategic value of the Hirsi judgment, which had combined a spectacular outcome with conventional legal arguments.37 In particular, the reasoning remained grounded in a ‘primarily territorial’ understanding of jurisdiction, which judges projected upon the high seas. States cannot evade their obligations by relocating border controls outwards. Along similar lines, the ECtHR had rebutted earlier attempts at carving out segments of the territory from legal oversight, for instance border procedures at airports or in transit zones.38 To control the fringes of territorial jurisdiction is a different matter than to recognise functional or joint jurisdiction, which would involve judges in complex foreign policy matters on diverse subject matters.39 We cannot, therefore, expect outsourcing migration control to neighbouring states to cross the jurisdictional yardstick.40 There might be room for dynamism concerning search and rescue, but the Grand Chamber ruling on humanitarian visas arguably reined in the potential of extraterritorial jurisdiction. The Refugee Convention does not define its territorial scope. There is growing support to extend the prohibition of refoulement to rejection at the border and—notwithstanding the position of the US Supreme Court41—extraterritorial de facto control over persons or territory, even though the Convention had originally applied to persons within the state territory only.42 One step further, proposals have been put forward to use the doctrinal concept of positive obligations to limit state discretion,43 especially for multilateral human rights instruments not subject to the jurisdictional benchmark.44 When it comes to legal practice, however, the ECtHR’s position on the ‘primarily territorial’ notion of jurisdiction remains essential. It entails that visa requirements and carrier sanctions cannot be held liable under the Convention.45 Uncertainties persist regarding the scope of the Charter. On the one hand, it can be argued that the Charter emulates the jurisdictional threshold of Article 1 ECHR, at least for guarantees, such as the prohibition of refoulement or the right to asylum, which may have to be interpreted in light of international obligations.46 On the other hand, the open formulation, in Article 51(1) Charter of Fundamental Rights (CFR), about the Charter being ‘addressed to the institutions’ and applying to Member States ‘when implementing Union law’, could be read as an implicit statement about extraterritorial scope, also considering that the Court has found secondary legislation to reach beyond the external borders.47 Arguably, 37 See Moritz Baumgärtel, Demanding Rights (CUP 2019) 87–98. 38 Reaffirmed by ND and NT v Spain App nos 8675/15 and 8697/15 (ECtHR [GC], 13 February 2020) §§ 102–11. 39 See also Daniel Thym, ‘The End of Human Rights Dynamism?’ (2020) 32 IJRL 569, 584–88. 40 See MB and RA v Spain App no 20351/17 (ECtHR, 5 July 2022) §§ 16. 41 See Sale v Haitian Centers Council 509 US 155 [1993] (judgment of 21 June 1993). 42 See Guy S Goodwin-Gill and Jane McAdam, The Refugee in International Law (4th edn, OUP 2021) 307–28; Thomas Gammeltoft-Hansen, Access to Asylum (CUP 2011) ch 3. 43 See Thomas Spijkerboer, ‘Stretching the Limits’ in Marleen Maes and others (eds), External Dimensions of EU Migration and Asylum Law and Policy (Bruylant 2011) 387, 399–408. 44 For the ICESCR see Annick Pijnenburg, At the Frontiers of State Responsibility: Socio-Economic Rights and Cooperation on Migration (Intersentia 2021). 45 cf Frank McNamara, ‘Member State Responsibility for Migration Control within Third States’ (2013) 15 EJML 319. 46 See CFR, recital 4, art 52(3), (4); Case C-239/14 Tall EU:C:2015:824, paras 53–54; and Stephanie Law, ‘Humanitarian Admission and the Charter of Fundamental Rights’ in Marie-Claire Foblets and Luc Leboeuf (eds), Humanitarian Admission to Europe (Nomos/Hart Publishing 2020) 77, 97–109. 47 See Cathryn Costello and Violeta Moreno-Lax, ‘The Extraterritorial Application of the EU Charter of Fundamental Rights’ in Steve Peers and others (eds), The EU Charter of Fundamental Rights. A Commentary (1st edn, CH Beck/Hart Publishing/Nomos 2014) 1657.

308 Border Controls one reason for the convoluted reasoning in the X and X judgment on humanitarian visas, mentioned in Chapter 11.5, may have been the desire to evade a definite statements on these matters. For the time being, the scope of the Charter remains indistinct.

12.2.3 Search and rescue at sea The obligation to safe life at sea is firmly embedded in international law, including in Article 98 United Nations Convention on the Law of the Sea (UNCLOS), which the EU has ratified. Regulatory standards of the Maritime Safety Committee of the International Maritime Organisation (IMO) specify and supplement the general obligation to rescue, on the basis of the International Convention on Maritime Search and Rescue (SAR Convention).48 These standards establish interstate coordination requirements and define the meaning of ‘distress’ and ‘place of safety’.49 A place of safety, where disembarkation may take place, is generally understood to require physical security and the fulfilment of basic human needs, including compliance with the prohibition of refoulement.50 These obligations are not directly binding on the EU but specify the meaning of Article 98 UNCLOS, which the EU must respect.51 To apply these legal standards remains a practical and legal challenge. First, the definition of ‘safety’ defies easy definition at the interplay between the law of the sea and human rights. Similar problems persist regarding the meaning of ‘distress’ and its application to the factual situation on the ground, for instance concerning rubber boats.52 Secondly, the obligation to search and rescue does not imply an authorisation for disembarkation in a specific country, for example on grounds of geographic proximity or rescue zones. States have to cooperate in finding a place of safety—there is an obligation of conduct, not of result.53 States are not required to open ports to private vessels, unless in situations of humanitarian urgency.54 Migrants rescued at sea have repeatedly been passed around like the proverbial hot potato, with no one assuming responsibility. We shall see that the Sea Borders Regulation does little to overcome these shortcomings. Rescue at sea cannot usually be enforced, as an international obligation, through human rights courts or treaty bodies. Nevertheless, national courts have decided cases involving the law of the sea, especially with regard to the criminalisation of humanitarian support. Moreover, advocates have filed a complaint with the International Criminal Court.55 Finally, customary rules on state responsibility may be relied upon to extend responsibility. It could

48 Adopted 27 April 1979, entered into force 22 June 1985, 1405 UNTS 118, which is binding on most Member States. 49 See ‘Guidelines on the Treatment of Persons Rescued at Sea’ (Resolution MSC.167(78), 20 May 2004) No 6.12–18. 50 See Revised Annex to SAR Convention (Resolution MSC.70(69), 20 May 1998; Resolution MSC.155(78), 20 May 2004) https://www.imo.org/en/KnowledgeCentre/IndexofIMOResolutions/Pages/MSC.aspx (accessed 1 March 2023). 51 See also Joined Cases C-14/21 and C-15/21 Sea Watch EU:C:2022:604, paras 89–94. 52 See Violeta Moreno-Lax, ‘Seeking Asylum in the Mediterranean’ (2011) 23 IJRL 174, 186–99. 53 See Jasper van Berckel Smit, ‘Taking Onboard the Issue of Disembarkation’ (2020) 22 EJML 492. 54 See Jasmine Coppens, ‘Interception of Migrant Boats at Sea’ in Violeta Moreno-Lax and Efthymios Papastavridis (eds), ‘Boat Refugees’ and Migrants at Sea (Brill Nijhoff 2016) 199. 55 See Ioannis Kalpouzos, ‘International Criminal Law and the Violence against Migrants’ (2020) 21 German LJ 571.

Constitutional Foundations 309 be argued, for instance, that aid or assistance to the Libyan coastguard brings about international legal responsibility or that states have due diligence obligations to safe life at sea.56 That being said, such international legal arguments based upon customary law must be distinguished from the notion of jurisdiction under human rights law. Moreover, the Court of Justice of the European Union (CJEU) confined the legal effects of Article 98 UNCLOS to the rescue operation sensu stricto, instead of deriving due diligence obligations from it with regard to the treatment of rescue ships in-between seaborne missions.57

12.2.4 Rejection at the border Chapter 5.1 explained that the ECtHR maintains that states enjoy the right ‘as a matter of well-established international law . . . to control the entry, residence, and expulsion of aliens’. This assertion of state sovereignty has been limited considerably, without bringing about a generic human right to be admitted. Only nationals and, possibly, settled migrants must not be arbitrarily deprived of the human right to enter their ‘own country’.58 Attempts at vitalising the right to leave any country have not been fruitful so far;59 the ECtHR rejected an activation of the right to leave against interception at sea.60 This leaves the debate about rejection at the border with one essential question: the scope of procedural human rights guarantees.

12.2.4.1 Right to asylum and prohibition of refoulement While the prohibition of refoulement obliges states not to return anyone to unsafe territories, the right of asylum is generally understood to embrace access to status determination and, in case of recognition, a set of socioeconomic rights.61 For entry purposes, access to status determination is crucial but legally unproblematic under Union law as a matter of principle. We shall see in the following chapter that the Asylum Procedures Directive establishes an individual guarantee to access the asylum procedure ‘at the border’ or ‘in the territorial water’. Compliance if far from guaranteed in practice, but these shortfalls do not affect the contents of the legislation. Nevertheless, some legal queries remain. Which procedural guarantees apply when individuals do not apply for asylum or if they are pushed back on the high seas? Such scenarios do not come within the geographic ambit of the Asylum Procedures Directive, thus turning the prohibition of refoulement into the decisive yardstick for the legal analysis. Inspection of the procedural guarantees underlying the principle of non-refoulement in Chapter 13.2.2 will demonstrate that any invocation presupposes a ‘real risk’ of a violation. 56 See Achilles Skordas, ‘A “Blind Spot” in the Migration Debate?’ EU Immigration and Asylum Law and Policy (30 January 2018); and Guy S Goodwin-Gill, ‘Drowning in the Mediterranean. Time to Think and Act Regionally’ EJIL:Talk! (12 April 2021). 57 See Sea Watch (n 51) paras 105–108, 115–19; and ch 16.3.2. 58 ICCPR, art 12(4); Additional Protocol No 4 to the ECHR, art 3(2); and HRC, ‘General Comment No 27’ (CCPR/C/21/Rev.1/Add.9, 2 November 1999) No 20; in Union law, this guarantee may exist as an unwritten general principle of Union law in line with TEU, art 6(3). 59 See Vladislava Stoyanova, ‘The Right to Leave Any Country and the Interplay between Jurisdiction and Proportionality in Human Rights Law’ (2020) 32 IJRL 403; and Nora Markard, ‘The Right to Leave by Sea’ (2016) 27 EJIL 591. 60 See Xhavara and Others v Italy and Albania App no 39473/98 (ECtHR, 11 January 2001) § 3. 61 See ch 18.2.3.

310 Border Controls A major doctrinal conundrum are the procedural consequences of this threshold where individuals are returned to countries, such as Morocco, which are—unlike Libya at present— not generally unsafe. Are states allowed to send migrants to such places without performing a basic screening, if the available information does not indicate a real risk? To answer that question to the negative, as settled case law can be read to suggest, would be a logical dead from a practical perspective. Without basic procedural safeguards, it is notoriously difficult to know whether someone is confronted with a real risk of refoulement or not. The ECtHR recognised this problem in the ND and NT ruling on the Spanish policy of hot returns in the exclave of Melilla: generous interpretation of the prohibition of collective expulsion filled the procedural gap by requiring the rudimentary screening of anyone rejected at the border, irrespective of whether there was a danger of refoulement or not.62 One step further, the Grand Chamber insisted, in a ruling about Hungary pushing asylum seekers back to Serbia, that states should usually embark on a ‘thorough examination’ and stipulated that doing so will usually require a ‘legal procedure resulting in a legal decision’;63 these procedures will usually take place on the territory to which the individual must be admitted.64 At the same time, however, the Court introduced a back door. Article 3 ECHR can be complied with ‘[i]‌n the absence of such [individual procedure]’ on the basis of information about the general level of safety, which may emanate from country reports;65 irregular entry by large numbers using force may be another element justifying summary rejection.66 It can be frustrating that these essential questions about the outer limits of the prohibition of refoulement remain unresolved at the time of writing.

12.2.4.2 Prohibition of collective expulsion Article 19(1) CFR replicates the prohibition of collective expulsion enshrined in Article 4 of the Additional Protocol No 4 to the ECHR, in the light of which the Charter must be interpreted.67 In a series of judgments, the ECtHR postulated a wide interpretation of a provision which had originally been meant to prohibit mass expulsion on ethnic, religious, or political grounds. It is interpreted nowadays to require states to provide for procedural safeguards in cases of non-admission, irrespective of the number of people involved.68 The Hirsi judgment extended these safeguards to rejection at the border, including on the high seas,69 thereby rendering obsolete the limitation of the separate procedural guarantee against expulsion to those ‘lawfully resident’.70 Judges effectively established freestanding procedural safeguards against rejection at the border, even for those without a prima facie claim to a real risk of refoulement. Crucially, the Grand Chamber introduced an exception when it found the Spanish policy of hot returns not to violate the prohibition of collective expulsion in the ND and NT judgment. Conditions under which the exception applies remain somewhat unclear. The Court 62 See ND and NT v Spain (n 38) §§ 171–81. 63 See Ilias and Ahmed v Hungary App no 47287/15 (ECtHR [GC], 21 November 2019) §§ 134–38. 64 See MK and others v Poland App nos 40503/17, 42902/17 and 43643/17 (ECtHR, 23 July 2020) §§ 178–79. 65 ibid §§ 137–41. 66 See MB and RA v Spain (n 40) §§ 19–20. 67 See CFR, art 52(3); note that the Charter has a broader geographic scope, since Additional Protocol No 4 has not been ratified by Greece (as well as Turkey, amongst others). 68 See ND and NT v Spain (n 38) §§ 166–88; reiterating the narrow 4:3 decision in Čonka v Belgium App no 51564/99 (ECtHR, 5 February 2002) §§ 59–63. 69 See Hirsi Jamaa and others v Italy (n 27) §§ 166–82. 70 See ECHR, Additional Protocol No 7, art 3, which is binding on most state parties.

Schengen Governance 311 had relied upon the ‘own conduct’ of the individuals and gave contradictory explanations of why the ‘own conduct’ justified an exemption from basic screening: the availability of legal pathways, large numbers of individuals involved, and the use of force.71 Later judgments can be read to conceive of these criteria as alternatives; moreover, the availability of legal pathways was specified primarily to concern the option to apply for asylum at border crossing points.72 If these criteria are met, states can reject migrants at the border without basic screening. Whenever the prohibition of collective expulsion applies, procedural standards can be lower compared to cases involving a real risk under Article 3 ECHR. Judges accepted an unspecific opportunity to bring forward arguments on an individual basis to be sufficient; they also found brief and identical decisions to comply with human rights and did not require remedies with suspensive effect.73 These basic safeguards stay short of the guarantees in the Asylum Procedures Directive about personal interviews with trained personnel. They apply at the point of entry when border guards determine whether to authorise entry into the territory. A higher level of protection will be available under Article 3 ECHR whenever the basic screening indicates a real risk of refoulement. We shall see that the screening obligations in the Sea Borders Regulation mirror the rudimentary procedural safeguards of the prohibition of collective expulsion.

12.2.5 Territorial scope Supranational legislation is subject to country-specific opt-outs for Ireland and Denmark, as summarized in Table 12.1. Chapter 2.4 presented the distinct legal patters of these opt- outs and discussed the motivation behind them. Denmark, for instance, is usually bound by measures building upon the Schengen acquis as a matter of public international law. Norway, Iceland, Switzerland, and Liechtenstein are associated members on the basis of association agreements, mentioned in Chapter 17.2.2. Difficulties in defining the scope of the Schengen acquis will be elaborated upon in the context of Schengen governance. Some instruments apply to all members from day one after accession in accordance with the Act of Accession. Whenever this contribution employs the term ‘Schengen area’, it should be read to refer to those countries applying the acquis fully.

12.3 Schengen Governance The abolition of internal border controls is often praised as a hallmark of European integration. Nevertheless, recent years have exposed structural deficits. The frequent reintroduction of internal border controls signalled broader governance deficits. Geography underlies

71 See the abstract findings in ND and NT v Spain (n 38) §§ 200, 231 and the case-specific arguments in §§ 201, 208, 167–68; see also Thym (n 39) 576–79. 72 See AA and others v North Macedonia App no 55798/16 (ECtHR, 5 April 2022) §§ 113–23; Shahzad v Hungary App no 12625/17 (ECtHR, 8 July 2021) §§ 59–62; MH and Others v Croatia App nos 15670/18 and 43115/18 (ECtHR, 18 November 2021) §§ 294–304; MK and others v Poland (n 64) §§ 207–09. 73 See Khlaifia and others v Italy App no 16483/12 (ECtHR [GC], 12 December 2016) §§ 237–42, 276–77; and Asady and others v Slovakia App no 24917/15 (ECtHR, 24 March 2020) §§ 57–71.

312 Border Controls Table 12.1 Participation in the most relevant entry and border control instruments Denmark

Schengen? Accessiona (NO, IS, CH, LI)

Residence Permit no/yes Format Regulation (EC) No 1030/2002

no (int. law)

yes

yes

Passports and Travel Documents Regulation (EC) No 2252/2004

no

no (int. law)

yes

yes

Local Border Traffic Regulation (EC) No 1931/2006

no

no (int. law)

yes

yes

Sea Borders Regulation (EU) No 656/2014

no

no (int. law)

yes

no

Schengen Borders Code Regulation (EU) 2016/399

no

no (int. law)

yes

no

Instrument

Ireland / UK (before Brexit)

aSome

instruments apply immediately in accordance with, for instance, Annex II Act concerning the conditions of accession of the Republic of Bulgaria and Romania [2005] OJ L157/203, art 4(1).

many debates, as the Schengen area stretches from the Aegean Sea, Sicily, and Ceuta to the North Cape and the external borders with Russia, Ukraine, and Belarus nowadays (12.3.1). Crises and reform efforts have been triggered by recurring debates about irregular secondary movements from 2011 onwards (12.3.2). A technical question of great practical importance for associated countries and Member States with an opt-out is the definition of the ‘Schengen acquis’ (12.3.3).

12.3.1 Membership and phased implementation ‘Schengen’ started among five Member States, three of which had abolished border controls amongst themselves already. Chapter 1.2.1 illustrated that relations between France, Germany, and the Benelux on the one hand and the supranational institutions on the other hand were defined by rivalry throughout the 1980s and 1990s, until the Treaty of Amsterdam integrated the Schengen acquis into the EU framework. The period until 2008 saw geographic expansion and optimism, with the discontinuation of internal border controls towards more and more Member States. While the media had hardly taken notice of the signature of the first Schengen agreement, barriers were removed along the former iron curtain with much fanfare.74 Nowadays, the Schengen area comprises 26 countries, including the associated countries.75 New Member States accede after enlargement in

74 75

See Angela Siebold, ZwischenGrenzen (Ferdinand Schöningh 2013) 43–49, 90–93. For the status of micro-states see ch 17.2.3.

Schengen Governance 313 line with the Act of Accession. Croatia joined the inner circle in 2023, whereas Romania, Bulgaria, and Cyprus had been in the antechamber for years. Ireland has an opt-out. We must distinguish between the obligation to join the Schengen area and the actual abolition of internal border controls. The latter requires a decision of the Council concluding that ‘the preconditions for its implementation have been fulfilled’.76 Such decision used to be a formality, once an evaluation mechanism found that the legal, administrative, and technical adjustments had been completed. On that basis, the application of the Schengen acquis is usually phased in, starting with databases, continuing with airports, and concluding with land borders.77 Bulgaria and Romania, for instance, participate in the SIS but have been waiting for full membership since 2011. Adoption of the final decision had originally been frozen when Finland and the Netherlands voiced their opposition at the last minute, whereas Austria vetoed an attempt to authorise admission in December 2022.78 Irregular migration was a major reason,79 which was perpetuated when the asylum policy crisis of 2015/16 reinforced political concerns about a ‘land bridge’ between Greece and the European heartland. From a legal perspective, one might argue that blockage of the Council Decision, which requires unanimity, violates the obligation of loyal cooperation under Article 4(3) TEU. The Act of Accession presupposes approval ‘after verification in accordance with the applicable Schengen evaluation procedures that the necessary conditions for the application . . . have been met’.80 Nevertheless, success of an action for failure to act under Article 265 TFEU is not a foregone conclusion considering the element of discretion the unanimity requirement arguably involves.81 Having said this, parts of the Schengen acquis apply immediately after accession, as explained in the previous section on the territorial scope of the legislation on border controls.

12.3.2 Crises and reform efforts The creation of the Schengen area had seen occasional tensions, for example French criticism of the liberal Dutch drug policy.82 On the whole, however, collaboration initially took place in an atmosphere of mutual trust that paved the way for the gradual expansion of border-free travel. By way of example, the terrorist attacks of 11 September 2001 did not trigger enhanced border controls.83 Things changed in 2011 when tens of thousands of 76 Joint Declaration on Article 139(2) and Final Act of the Convention Implementing the Schengen Agreement; and, nowadays, the Act of Accession. 77 For Central and Eastern Europe see Council Decision 2007/801/EC on the full application of the provisions of the Schengen acquis [2007] OJ L323/34, art 1. 78 See ‘Draft Decision’ (Council doc 14302/3/11, 7 November 2011); and European Parliament, ‘Resolution on the full application of the provisions of the Schengen acquis in Bulgaria and Romania’ (P8_TA(2018)0497, 11 December 2018). 79 See ‘Schengen Zone: Delay for Bulgaria and Romania to Join’ BBC.com (9 June 2011); and ‘Austria blocks Schengen accession of Romania and Bulgaria’ euronews.com (9 December 2022). 80 Act concerning the conditions of accession of the Republic of Bulgaria and Romania [2005] OJ L157/203, art 4(2). 81 See Jonas Bornemann, ‘Joining Schengen: Cutting a Gordian Knot for Romania and Bulgaria?’ EU Immigration and Asylum Law and Policy (26 April 2018). 82 See Ruben Zaiotti, ‘The Italo-French Row over Schengen, Critical Junctures, and the Future of Europe’s Border Regime’ (2013) 28 J Borderlands Stud 337, 341–44. 83 See Kees Groenendijk, ‘Reinstatement of Controls at the Internal Borders of Europe’ (2004) 10 ELJ 150, 166.

314 Border Controls Tunisians arrived on Italian shores, spurring hectic responses, including cooperation with Tunisia for swift return. Italy finally granted humanitarian residence to many Tunisians, thus effectively sponsoring onward movement, which met the fierce resistance of France. Border controls and ‘pushbacks’ at the land border and in trains between Ventimiglia and Menton made headline news.84 In the same year, the Danish government announced customs controls towards Germany.85 With the wisdom of hindsight, these events during 2011 appeared as the harbinger of things to come. EU institutions reacted with operational and financial support for Italy, including a Frontex mission called ‘Hermes’ on the island of Lampedusa.86 Moreover, they agreed, after bitter disputes between the Parliament and the Council about alleged procedural tricks, on a Schengen governance package: new rules on temporary internal border controls, to be discussed hereafter, and a substantial revision of the Schengen evaluation mechanism.87 A defining feature was proceduralisation. EU institutions, notably the Commission, were involved in decisions about reintroducing border controls and partake in the evaluation, which had been purely intergovernmental hitherto.88 The outcome was a complex and lengthy evaluation procedure: each Schengen member is assessed once every five years, and the Council may adopt formal recommendations.89 Such recommendations have covered diverse issues ranging from external border controls to visa policy. The Commission concedes that the evaluation mechanism has generally proven to lack teeth.90 Terrorist attacks and the tensions about secondary movements provoked the proliferation of ‘temporary’ internal border controls. They were continued for ever longer periods and did not always comply, as we shall see, with the wording and spirit of the law. These developments created a sense of uncertainty about the future of ‘Schengen’. Political responses concentrated on asylum reform, Frontex, and international cooperation91—together with an attempt to revise the rules on internal border controls. This reform did not proceed as a result of inter-institutional disputes and political insistence, among Northern governments, that the future of Schengen was closely connected to the prevention of secondary movements.92 Whereas the original legal connection between asylum jurisdiction vanished with supranationalisation,93 the political linkage between Dublin and Schengen is real. Dutch

84 Zaiotti (n 82) 343–46; and Elspeth Guild and others, ‘Internal Border Controls in the Schengen Area: Is Schengen Crisis-Proof?’ (Study for the European Parliament, PE 571.356 June 2016) 23–25. 85 See Carsten Willemoes Jørgensen and Karsten Engsig Sørensen, ‘Internal Border Controls in the European Union’ (2012) 37 EL Rev 249. 86 See Bruno Nascimbene and Alessia Di Pascale, ‘The “Arab Spring” and the Extraordinary Influx of People who Arrived in Italy from North Africa’ (2011) 13 EJML 341, 343–46. 87 See Guild and others (n 84) 25–37; and Henri Labayle, ‘La suppression des contrôles aux frontières intérieures de l’Union’ in Claude Blumann (ed), Les frontières de l’Union européenne (Bruylant 2013) 19, 35–45. 88 See Executive Committee Decision (98)26 setting up a Standing Committee on the evaluation and implementation of Schengen [2000] OJ L239/138. 89 See Regulation (EU) No 1053/2013 establishing an evaluation and monitoring mechanism to verify the application of the Schengen acquis [2013] OJ L295/27. 90 See Commission, ‘Report on the functioning of the Schengen evaluation and monitoring mechanism’ COM(2020) 779 final, 14–17. 91 For an overview see Ruben Zaiotti, ‘Border Management. The Schengen Regime in Times of Turmoil’ in Ariadna Ripoll Servent and Florian Trauner (eds), The Routledge Handbook of Justice and Home Affairs Research (Routledge 2018) 99; and Yves Pascouau, ‘Quel avenir pour Schengen?’ in Josiane Auvret-Finck and Anne-Sophie Millet-Devalle (eds), Crise des Réfugiés, Crise de l’Union Européenne? (Pedone 2017) 261. 92 See Marie De Somer, ‘Schengen: Quo Vadis?’ (2020) 22 EJML 178, 184–86. 93 Non-compliance with take back requests under Articles 28–39 CISA might possibly have justified the reciprocal reintroduction of internal border controls under the Vienna Convention on the Law of Treaties, art 60(2).

Schengen Governance 315 politicians have been particularly vocal in questioning the status quo, going as far as calling for a ‘Mini-Schengen’.94 The Schengen area went through a near-death experience during the first months of the Covid-19 pandemic in the spring of 2020, when Member States reinstated border controls and adopted unprecedented travel restrictions. Pictures of Romanian and Latvian citizens being prevented, temporarily at least, from transit through Hungary and Poland by car would have been deemed unthinkable only a few weeks earlier.95 The symbolic weight of borders, presented at the outset, threatened to turn itself against the European project when national governments conveyed control signals. The situation improved after a few months and the later waves of the pandemic did not see a repeat of the drastic restrictions. Nevertheless, the pandemic reinforced the impression that the governance structure was insufficient and that the supranational institutions were ill-placed to sustain the Schengen area without the continued support of the Member States.96 As a response to these governance deficits and failures, the Commission presented another reform package in June 2021. A regular ‘Schengen Forum’ shall build political momentum in support of border-free travel, and the evaluation mechanism is to be rendered more effective. Frequent and flexible evaluation shall be completed in shorter time frames, thus ideally creating peer pressure to remedy shortcomings at short notice97— to be reinforced by the vulnerability assessments of Frontex and the Asylum Agency.98 There was broad political support for reforming the governance structure. The idea of a Schengen Forum was upgraded to ministerial meetings in a ‘Schengen Council’ as an informal gathering, mirroring the ‘eurogroup’ for monetary union, to discuss the state of the Schengen area.

12.3.3 Scope of the ‘Schengen acquis’ The French noun ‘acquis’ is widely used in EU jargon to describe a set of rules on a specific theme, notably in the context of accession. It became a legally relevant concept when the Treaty of Amsterdam turned the intergovernmental Schengen rulebook into supranational legislation, which can be amended by the EU institutions.99 Doing so required the definition of the Schengen acquis, which had been partly confidential and dispersed across multiple intergovernmental decisions.100 Yet, defining the acquis was no singular undertaking. 94 See ‘Mini-Schengen not an Option, for Now’ EURACTIV (25 January 2016) https://www.euractiv.com/sect ion/central-europe/news/mini-schengen-not-an-option-for-now (accessed 1 March 2023); and De Somer (n 92) 190–96. 95 On national practices see Sergio Carrera and Ngo Chun Luk, In the Name of COVID: An Assessment of the Schengen Internal Border Controls and Travel Restrictions in the EU (Study for the European Parliament, PE 659.506, September 2020) 18–23, Annex I. 96 See Daniel Thym and Jonas Bornemann, ‘Schengen and Free Movement Law During the First Phase of the Covid-19 Pandemic’ (2020) 5 European Papers 1143, 1146–55. 97 See Commission, ‘Communication: A strategy towards a fully functioning and resilient Schengen area’ COM(2021) 277 final, 16; and Proposal for a Regulation on the establishment and operation of an evaluation and monitoring mechanism to verify the application of the Schengen acquis, COM(2021) 278 final. 98 See ch 8.3.2 and 8.4.3. 99 See ch 1.2.3; today’s Protocol [No 19] integrating the Schengen acquis into the framework of the European Union [2008] OJ C115/290; and its predecessor [1997] OJ C340/93. 100 See Council Decision 1999/435/EC concerning the definition of the Schengen acquis [1999] OJ L176/1; and Council Decision 1999/436/EC determining the legal basis for each of the provisions or decisions which constitute the Schengen acquis [1999] OJ L176/17.

316 Border Controls The question comes up regularly whenever new measures are adopted. The scope of the Schengen acquis determines the legal position of Member States with an opt-out and of the associated countries. Unfortunately, preferences diverge. The associated countries will usually support a generous reading to be able to participate, and Denmark will take the same view to evade exclusion. By contrast, Ireland and the United Kingdom (an influential player before Brexit) habitually call for a narrow understanding, since their flexible opt-outs cover non-Schengen measures only, as explained in Chapter 2.4.2. Finally, countries waiting for full membership may similarly prefer narrow interpretation, since they are bound by non-Schengen instruments immediately. Two scenarios are particularly relevant. First, when new measures have no direct precursor. Frontex and access to the VIS for law enforcement purposes are a case in point. The linkage to Schengen might have been politically obvious for Frontex, but the agency was a new instrument, nonetheless; VIS access could similarly be presented as an autonomous public security matter. The UK unsuccessfully challenged exclusion before the Court. Judges confirmed that the ‘Schengen acquis’ is not confined to legislation amending or replacing pre-existing instruments; in cases of doubt, the classification follows the aim and content.101 That interpretation effectively gives the institutions some leeway to define the scope of the Schengen acquis. Secondly, some measures may qualify as ‘hybrid’ instruments, whose content and aim concerns both Schengen and other matters. Asylum procedures at the external borders and feedback loops between internal border controls and asylum jurisdiction are prominent examples. Can Ireland participate in and vote on these matters? And would the coherence of Union law, emphasised by the Court, be undermined if an instrument was adopted as a hybrid act? Similar questions may arise with regard to the ‘Dublin acquis’.102 These questions became relevant during negotiations on asylum reform. In a memorandum, the Council Legal Service maintained that two proposals amounted to hybrid acts, which threatened to undermine the coherence of Union law as a result of asymmetric participation; it also suggested solutions how the alleged difficulties could be avoided, for instance by means of splitting the legislative proposals.103

12.4 Internal Movements within the Schengen Area Abolishing internal border controls was the main motivation behind Europeanisation— an objective the proliferation of internal border controls has put the back on the agenda. Irregular secondary movements of asylum applicants and other third country nationals are a matter of concern for some Member States, thus turning the spotlight on legal rules for their movement within the Schengen area (12.4.1). Complex provisions govern the temporary reintroduction of internal controls, which many governments have repeatedly 101 See Case C-482/08 United Kingdom v Council EU:C:2010:631, paras 54–58, 77–78; AG Verica Trstenjak, Opinion in ibid, EU:C:2007:419, points 102–15; Case C-482/08 United Kingdom v Council EU:C:2010:631, paras 45–49; and Galina Cornelisse, ‘What’s Wrong with Schengen?’ (2014) 51 CML Rev 741, 752–56. 102 See Council Legal Service, ‘The Proposed new Pact on Migration and Asylum: “Variable geometry” ’ (Council doc 6357/21, 19 February 2021) 10–14. 103 ibid 15–39.

Internal Movements within the Schengen Area 317 handled leniently, in violation of the Schengen Borders Code Regulation (12.4.2). Police checks and other measures within the territory may help rendering temporary internal border controls superfluous (12.4.3).

12.4.1 Circulation of third country nationals The individual right to cross internal borders ‘at any point without a border check on persons, irrespective of their nationality, being carried out’104 does not render internal borders irrelevant. Third country nationals benefit from the absence of checks on persons ‘irrespective of nationality’ and need not use border crossing points when travelling from Slovenia to Austria. However, the absence of physical controls is not the same as the legal authorisation to enter another country. Factual freedom of travel and the legal guarantee not to be checked do not bring about a generic right of entry and stay valid throughout the Union. The ‘conditions governing the movement of aliens’105 are among the few provisions of the original Convention Implementing the Schengen Agreement (CISA) which are still in force.106 Some articles have been amended or repealed by secondary legislation, thus rendering the acquis somewhat non-transparent.

12.4.1.1 Free travel for short stays Several groups of people have to be distinguished when it comes to temporary stays of up to ninety days. Longer periods are governed by the rules on legal migration under national or supranational legislation.107 When it comes to short stays of up to ninety days in another Member State, an obvious category are Schengen visa holders. They may travel freely within the Schengen area in accordance with Article 19(1) CISA during the period of validity of the visa, provided they still fulfil the entry conditions. Third country nationals not subject to visa requirements may similarly circulate freely for up to 90 days in any 180-day period; an amendment reversed the effects of an earlier Court ruling about the calculation of the relevant period.108 A peculiarity is pre-existing agreements, some Member States have signed with third states (notably Portugal with Brazil), concerning an extension of visa-free travel beyond ninety days; an amendment clarified that the effects of such an extension are limited to the territory of the country concerned.109 Of great practical importance is the revised version of Article 21 CISA, which authorises anyone holding a valid residence document of another Schengen country to go elsewhere for short stays. Free circulation covers any national residence permit or long-term visa issued by full members of the Schengen area.110 Third country nationals residing in Ireland have to apply for Schengen visas when visiting the continent—and vice versa for a Nigerian 104 Schengen Borders Code Regulation (EU) 2016/399, art 22; in line with CISA, art 2(1). 105 Title of Chapter 4 CISA; see also the former CISA, art 2(3). 106 The Proposal for a Directive on the right of third country nationals to travel in the Community, COM(95) 346 final was not adopted; see Kay Hailbronner, Immigration and Asylum Law and Policy of the European Union (Kluwer 2000) 289–99. 107 See ch 14.4.5. 108 See CISA, art 20(1), as amended by Regulation (EU) No 610/2013 [2013] OJ L182/1; in response to Case C- 241/05 Bot EU:C:2006:634, which had highlighted in para 42 that the legislature might change the rules. 109 See CISA, art 20(2)–(2d), as amended by EES Regulation (EU) 2017/2226, art 60. 110 See CISA, art 21(1), (2a), as amended by Regulation (EU) No 265/2010 as regards movement of persons with a long-stay visa [2010] OJ L85/1.

318 Border Controls national with a French residence permit who wants to fly to Dublin for a weekend. This geographic limitation is inherent in the variable geometry of the Schengen area, including for countries waiting to become full members.111 A special scenario concerns transit through Bulgaria, Croatia, Cyprus, or Romania, as well as travel in-between these countries.112 Moreover, school pupils may travel freely among all Member States, including Ireland, for purposes of school excursions.113

12.4.1.2 Secondary movements of asylum applicants Secondary movements are politically controversial. While they are illegal on many occasions, there are exceptions. Free travel for short stays under Article 21 CISA covers anyone holding a valid residence permit, including beneficiaries of international protection. National statuses on humanitarian grounds and the regularisation of illegal stay equally bring about a right to move freely within the Schengen area for up to ninety days. States are obliged to notify the Council, as the successor to the Executive Committee, a list of documents which they issue as residence permits.114 Secondary movements are legal whenever these conditions are being complied with. By way of example, a Syrian national with Greek refugee status may take a flight to Frankfurt, and a Tunisian with an Italian humanitarian permit can cross the border into France. Nevertheless, it is precisely these movements which have caused heated controversies ever since the public clash between the governments of France and Italy over Tunisians during 2011. Upon closer inspection, the legal picture becomes complex. Article 21(1) CISA makes free travel on the basis of national residence permits conditional upon the continued fulfilment of entry conditions under today’s Article 6(1), (c), (e) Schengen Borders Code Regulation with regard to means of subsistence and public policy threats; moreover, admission may be refused if the purpose and duration of the intended stay does not comply with the ninety-day rule. These caveats apply indiscriminately to all third country nationals, but they are particularly relevant for ‘secondary movements’ of beneficiaries of international protection and third country nationals with humanitarian residence permits, since their movements will not always comply with all entry conditions. France has repeatedly relied on these legislative caveats when refusing admission or proceeding with swift returns at the internal borders towards Spain and Italy.115 Asylum seekers whose application is still being processed do not benefit from free movement at all. Their right to remain under Article 9(1) Asylum Procedures Directive qualifies as neither legal residence nor as a provisional residence permit pending an ongoing application.116 As a result, an explicit obligation to refrain from secondary movements, which the Commission proposes introducing in the context of the ongoing legislative reform debate, would be declaratory in character; it would confirm the status quo under the revised version 111 CISA, art 21(1) is not listed in Act of Accession (n 80) art 4(1), Annex II. 112 See Decision No 565/2014/EU introducing a simplified regime for the control of persons at the external borders [2014] OJ L157/23. 113 See Decision 94/795/JHA on a joint action concerning travel facilities for school pupils from third countries resident in a Member State [1994] OJ L327/1; see also Executive Committee Decision (98) 19 on Monegasque residence permits [2000] OJ L239/199. 114 See CISA, art 21(3). 115 See Sara Casella Colombeau, ‘Crisis of Schengen?’ (2020) 46 JEMS 2258, 2264–66. 116 See CISA, art 21(1), (2), read in conjunction with Schengen Borders Code Regulation (EU) 2016/399, art 2(16)(b)(1), which distinguishes asylum applications from both residence permits and temporary permits; they are mentioned before the ‘or’, thus indicating a distinction.

Internal Movements within the Schengen Area 319 of Articles 19–24 CISA.117 Consequences of irregular secondary movements will be discussed in Chapter 16.6.3 on return and in the comments on the Dublin system in the next chapter.

12.4.2 Internal border controls The option of ‘temporarily’ reintroducing border controls is an integral part of the Schengen area, in accordance with the historic template of the Schengen agreements.118 Corresponding rules can be found in Articles 25–35 Schengen Borders Code Regulation (EU) 2016/399, as amended in 2013 in response to bitter disputes about secondary movements. Another reform was presented by the Commission after the events of 2015/16 but failed to muster political support as a result of ongoing political tensions mentioned previously. At the time of writing, it looked as if the latest reform package, presented at the end of 2021, could result in a compromise about how to revise the rules on the temporary reintroduction of internal controls.

12.4.2.1 Requirements for temporary reintroduction Existing rules distinguish between two sets of justifications: a serious threat to public policy or internal security (Articles 25–28) and the malfunctioning of the Schengen area as a whole (Articles 29–30); additional reporting obligations cover both scenarios (Articles 31, 33–35). Public policy or internal security are classic criteria to justify a derogation from an obligation under Union law. As a result, we may have recourse in settled case law on the definition of the public policy and internal security. Chapter 10.3 demonstrated that uniform interpretation leaves room for sector-specific outcomes, which, in the case of internal border controls, arguably give Member States some leeway considering the abstract character of the risk assessment under Article 25(1) and their primary responsibility for the maintenance of law and order, as confirmed by Article 72 TFEU.119 ‘Public health’ is not mentioned, although the social and economic consequences of a serious pandemic arguably cross the threshold of public policy risk.120 Having said this, a certain leeway on the part of states should not be misconstrued as a carte blanche to do as they please. Articles 25 to 28 limit discretion substantively and procedurally. Border controls are to be instituted as a measure of ‘last resort’, whose activation requires a proportionality assessment.121 Controls shall not exceed thirty days, although they may be prolonged for renewable periods up until a maximum of six months (or two years ‘in exceptional circumstances’).122 In cases requiring immediate action, Member States may proceed 117 See also Case C-606/10 ANAFE EU:C:2012:348, paras 29, 39; Case C-444/17 Arib and others EU:C:2019:220, para 38; and the CISA, former art 23(1), which has been supplanted by the Return Directive 2008/115/EC; Joined Cases C-322/19 and C-385/19 The International Protection Appeals Tribunal and others EU:C:2021:11, para 82 concerned the timing of asylum applications, rather than the legality of secondary movements. 118 See CISA, art 2(2); and Executive Committee Decision (95) 20 on the procedure for applying Article 2(2) [2000] OJ L239/133; and, on primary law, Daniel Thym, ‘Legal Framework for Entry and Border Control Policies’ in Daniel Thym and Kay Hailbronner (eds), EU Immigration and Asylum Law. Article-by-Article Commentary (3rd edn, CH Beck/Hart Publishing/Nomos 2022) MN 22–23. 119 See ch 10.7; and Joined Cases C- 368/ 20 and C- 369/ 20 Landespolizeidirektion Steiermark and Bezirkshauptmannschaft Leibnitz EU:C:2022:298, para 86. 120 See Thym and Bornemann (n 96) 1147–49. 121 See Schengen Borders Code Regulation (EU) 2016/399, art 26. 122 ibid art 25(1), (2), (4).

320 Border Controls unilaterally for up to ten days, up to a maximum of two months.123 EU institutions must be consulted four weeks in advance.124 At the end of the day, however, national governments decide what to do, since consultation does not give EU institutions a veto. The Commission could initiate infringement proceedings anytime, but it has performed its oversight function negligently.125 A novelty, introduced in 2013, was exceptional circumstances where ‘the overall functioning’ of the Schengen area is put at risk as a result of ‘persistent serious deficiencies relating to external border control’—a rather direct reference to the policy crisis of 2015/16.126 While admission of asylum seekers in accordance with the Asylum Procedures Directive does not constitute a serious deficiency in itself, the context of administrative deficits and prolonged irregular crossings via the ‘blue’ or ‘green’ border may meet the threshold. Non- compliance with a Council decision instructing Member States to accept the support of Frontex may similarly serve as a ground for internal border controls.127 To activate Article 29(1) remains the prerogative of the Council, acting by qualified majority.128 Deficiencies which were observed in Greece served as a justification for the first Council decision on this basis in the summer of 2016, which was prolonged three times, thus almost exhausting the maximum period of two years.129 Each decision specified the countries and border sections where controls could be reintroduced.

12.4.2.2 Extensive and illegal state practice Substantive and procedural requirements for the temporary reintroduction were unsuccessful in reigning in the spread of internal border controls. Member States have repeatedly introduced controls as if it was their sovereign prerogative. Justifications varied between terrorism, structural deficiencies of the Schengen area, secondary movements, and the Covid-19 pandemic—contrary to the initial practice before 2011 when controls were primarily instigated during summit meetings or large sport events.130 Governments have repeatedly relied on the idea, supported by a Commission Recommendation, that new threats may justify prolonged border controls on the basis of an accumulation of legal bases131—a position confirmed by the CJEU.132 It is a different matter what states do in practice. By way of example, Austria had been controlling its border with Slovenia for more than seven consecutive years at the time of 123 ibid art 28. 124 ibid art 27. 125 See ch 3.3.2; and AG Henrik Saugmandsgaard Øe Opinion in Joined Cases C-368/20 and C-369/20, Landespolizeidirektion Steiermark and Bezirkshauptmannschaft Leibnitz EU:C:2021:821, point 73. 126 Schengen Borders Code Regulation (EU) 2016/399, art 29(1), introduced first by Regulation (EU) No 1051/ 2013 [2013] OJ L295/1. 127 ibid art 29(1), as amended by former Frontex Regulation (EU) 2016/1624, art 80. 128 ibid arts 29–30; and TEU, art 16(3); see also Pierre Berthelet, ‘Les conséquences des crises migratoires de 2011 et de 2015’ (2018) 54 Cahiers de droit européen 395, 418–32. 129 See Implementing Decision (EU) 2016/894 setting out a recommendation for temporary internal border control [2016] OJ L151/8; and the follow-up by Implementing Decision (EU) 2016/1989 [2016] OJ L306/13; Implementing Decision (EU) 2017/246 [2017] OJ L36/59; and Implementing Decision (EU) 2017/818 [2017] OJ L122/73. 130 See Commission, ‘Report on the application of Title III (Internal Borders)’ COM(2010) 554 final, Annex I; and Groenendijk (n 83) 158–63. 131 See Commission Recommendation (EU) 2017/1804 on the implementation of the provisions of the Schengen Borders Code on temporary reintroduction of border control at internal borders in the Schengen area [2017] OJ L259/125, recital 2. 132 See Landespolizeidirektion Steiermark and Bezirkshauptmannschaft Leibnitz (n 119) paras 71, 79–81.

Internal Movements within the Schengen Area 321 writing: on the basis of a public policy threat in accordance with Article 25 at the height of the asylum policy crisis from September 2015 onwards; following the above-mentioned Council Decisions on systemic deficits in Greece between May 2016 and November 2017; thereafter, the Austrian government reverted to Article 25 once again, relying on more or less the same public policy risks that had defined all previous prolongations; the Covid-19 pandemic provided a new reason in March 2020. This succession of alternating justifications reached the Court on the basis of domestic proceedings two individuals, including one expert in EU law, had started against administrative fines of less than €50 they had to pay after refusing to show identity papers at the Austrian-Slovenian border. This act of civil disobedience gave the CJEU the opportunity to highlight the mandatory character of the maximum period of six months for internal border controls based upon public policy and internal security threats.133 Judges rejected the proposal of Advocate General Øe to set aside the maximum time limit in light of Article 72 TFEU.134 The Grand Chamber highlighted, amongst others, that free movement was a fundamental objective and that the EU institutions had introduced a ‘comprehensive framework’ when adopting the reform package of 2013.135 Public policy and internal security threats may persist for longer than six months but such a situation has to be addressed by other means than by border controls.136 These findings support the view that internal border controls were illegal on many occasions—for Austria and other Member States.137 At the time of writing, there were no indications that Austria and other Member States would abandon their earlier practices. While the armed attacked against Ukraine provided a new justification initially, the autumn of 2022 saw a return to the old habit of blanket justifications. If this continues, respect for Union law would have to be enforced via individuals seizing domestic courts, or the Commission initiating infringement proceedings. That does not appear to be happening so far. Reintroducing internal border controls on the basis of the same threat after a waiting period of a few weeks will regularly fall foul of the effet utile of Union law, in line with comments in Chapter 6.4.5 on the doctrinal foundations. In the medium run, the judgment may feed the appetite for legislative reform to give states more flexibility. Indeed, the Commission had proposed in December 2021, four months before the ruling, a fundamental overhaul, which would effectively legalise previous national practices de lege ferenda. Unilateral border controls may be initiated for up to two years and may even persist beyond that date indefinitely.138 The Council appeared to stick with the revision, notwithstanding some cosmetic changes to strengthen conditions for longer border controls in response to the Court judgment, but the Parliament might insist on further limitations. To introduce strict procedural notification and consultation requirements as a counterbalance cannot be expected to change much in practice. After all, similar provisions, adopted in 2013, had not prevented the proliferation of internal border controls. The fate of the latest legislative proposal was uncertain at the time of writing, also 133 ibid paras 57–70. 134 ibid paras 83–90; contra AG Saugmandsgaard Øe (n 125) points 53–68. 135 Landespolizeidirektion Steiermark and Bezirkshauptmannschaft Leibnitz (n 119) paras 59, 63–66, 72, 74, 87. 136 ibid para 77. 137 See Guild and others (n 84) 38–58; and Stefan Salomon and Jorrit J Rijpma, ‘A Europe Without Internal Frontiers’ (2023) 24 German LJ section B 1. 138 See Proposal for a Regulation amending the Schengen Borders Code Regulation (EU) 2016/399, COM(2021) 891 final, art 25a(4)–(5), 27a(5).

322 Border Controls considering that an earlier reform had failed.139 The Commission’s latest initiative might be more successful for the simple reason that it gives Member States what they want: flexibility. In the case of temporary internal controls, the provisions on external border controls ‘shall apply mutatis mutandis’.140 At first sight, this looks like a clear-cut prescription. Using crossing points becomes mandatory, sanctions for non-compliance are permissible, and everyone must show travel documents (provided states have enough personnel to equip crossing points). At the same time, the Court found that the parallelism was limited to Articles 5–21 Schengen Borders Code Regulation. Member States cannot set aside the Return Directive at the internal borders when refusing entry during border surveillance.141 A judgment banning carrier sanctions at internal borders in regular circumstances should be extended to the temporary reintroduction of internal controls,142 although the Commission proposes revisiting this requirement.143 Transit zones and asylum border procedures are similarly permissible at the external borders only.

12.4.3 Police checks in border areas Flanking measures compensating states for the loss of control options at internal borders have defined the Schengen acquis from the beginning. They support the realisation of public policy objectives, such as the prevention of crime, and may serve as alternatives to internal border controls.144 It is this second component the Commission has emphasised as a response to the proliferation of internal border controls. The Commission is eager to promote transnational cooperation, the use of technology, and strengthened internal security measures as an incentive for governments not to reinstate internal border controls.145 Notwithstanding this broader political function, flanking measures may raise legal questions as well, especially with regard to police checks in border areas. By contrast, technological border surveillance, such as the computer-based identification of number plates of cars crossing an internal Schengen border, cannot easily be qualified as having equivalent effects as physical border controls, even though they may raise legal concerns from the perspective of data protection standards.146 Police checks in border areas have triggered a series of rulings on today’s Article 23(1)(a) Schengen Borders Code Regulation (EU) 2016/399, which the Commission proposes reformulating slightly.147 The security apparatus is particularly keen on random checks without reasonable suspicion. Several countries used the abolition of internal border controls to 139 See Proposal for a Regulation amending Regulation (EU) 2016/399 as regards the rules applicable to the temporary reintroduction of border control at internal borders, COM(2017) 571 final; and De Somer (n 92) 184–86. 140 Schengen Borders Code Regulation (EU) 2016/399, art 32. 141 See ch 16.5.2. 142 See AG Yves Bot, Opinion in Joined Cases C-412/17 and C-474/17 Touring Tours und Travel and Sociedad de Transportes EU:C:2018:671, points 117–64 passim. 143 See Temporary Border Controls Proposal (n 139) art 23(b). 144 For the first two decades of Schengen see Commission (n 130) 3–6; and Kees Groenendijk, ‘New Borders behind Old Ones’ in Groenendijk and others (n 20) 131–46. 145 See Commission Recommendation (EU) 2017/820 on proportionate police checks and police cooperation in the Schengen area [2017] OJ C122/79; and Commission (n 97) 10–16. 146 See Temporary Border Controls Proposal (n 139) recital 21; and Julien Jeandesboz, ‘Ceci n’est pas un contrôle: PNR Data Processing and the Reshaping of Borderless Travel in the Schengen Area’ (2021) 23 EJML 431, 443–49. 147 Temporary Border Controls Proposal (n 139) art 23(a).

Internal Movements within the Schengen Area 323 introduce legislation to this effect, with a particular focus on the border area. Preliminary references from France, the Netherlands, and Germany brought the matter to the Court to identify legal limits. As a starting point, judges in Luxembourg accepted the very idea of more stringent police checks in border areas to comply with the Treaty objective to ensure the absence of ‘any controls’ for border crossings, also in light of Article 72 TFEU.148 This finding is based on a narrow understanding of the term ‘border control’, following the definition in the Schengen Borders Code, although one might possibly have defended a different reading of the Treaty objective.149 Article 23(1)(a) limits state discretion by forbidding an exercise of police powers having ‘an effect equivalent to border checks’. Domestic rules and practices on random police checks in the ‘hinterland’ were not considered to have such equivalent effect per se, in light of the wording and the rejection of the Commission’s original proposal to require the same rules for border areas as for the territory as a whole.150 States may even carry out random checks with a view to preventing irregular migration. To do so may be a crucial objective of border controls, yet Article 79 TFEU shows that such an objective is not intrinsically confined to the border area. The pursuit of other, notably crime-related, objectives was sufficient to distinguish police checks within the territory from border controls at crossing points.151 Such formalistic approach reiterates our earlier finding about an ‘administrative mindset’ characterising many rulings. The assessment whether random police checks have ‘an effect equivalent to border checks’ has to be performed by domestic courts in light of the criteria put forward in Article 23(1)(a)(i)–(iv). By way of example, controls at the railway station in the German city of Kehl, just 500 meters from the border with France, were considered to be ‘within the territory’ and to comply with the prohibition of equivalent effects whenever their frequency and intensity marked a difference to border controls.152 Such open-ended formulae give national governments much leeway in practice, which judges reined in somewhat by insisting on proceduralisation: the prevention of equivalent effects requires sufficiently precise domestic rules, which are legally binding and amenable to guide individual police units.153 Critical scholars highlight that the Court’s insistence on the random character of police checks, which are less systematic than border controls, has side effects, since it heightens the risk of racial profiling as a result of selective control practices.154

148 See Case C-278/12 Adil EU:C:2012:508, paras 51–52 passim; and Joined Cases C-188/10 and C-189/10 Melki and Abdeli EU:C:2010:363, paras 63–65. 149 Adil (n 148) paras 60–61; Schengen Borders Code Regulation (EU) 2016/399, art 2(2); and Michael Wilderspin, ‘Article 77 TFEU’ in Manuel Kellerbauer and others (eds), The EU Treaties and the Charter of Fundamental Rights. A Commentary (OUP 2019) MN 18–30. 150 Adil (n 148) para 71; referring to the Proposal for a Regulation establishing a Community Code, COM(2004) 391 final, art 19(a). 151 Adil (n 148) paras 64–67; and Case C-9/16 A EU:C:2017:483, paras 44–59. 152 See A (n 151) paras 63, 75; Adil (n 148) paras 76–87; and Melki and Abdeli (n 148) paras 71–74. 153 See Adil (n 148) paras 64–67; A (n 151) paras 59–62; and the follow-up ruling on the rather technical assessment of the German situation Case C-554/19 Staatsanwaltschaft Offenburg EU:C:2020:439, paras 42–56. 154 See also Maartje van der Woude and Joanne van der Leun, ‘Crimmigration Checks in the Internal Border Areas of the EU’ (2017) 14 Eur J Criminol 27.

324 Border Controls

12.5 External Controls at Crossing Points Border guards verifying identity papers are the image most readers will associate intuitively with the term ‘border control’. Indeed, provisions on checks at crossing points take centre stage in the latest version of the Schengen Borders Code Regulation (EU) 2016/399, which has seen minor amendments and will be supplemented further by the Entry/Exit System (EES) in the near future. Readers should be careful, therefore, to consult the consolidated version of the Borders Code which includes the recent changes; this can be found online, as described in Chapter 10.9 on practical tips. When it comes to checks on persons, entry conditions are the decisive legal hurdle (12.5.1), which are assessed by border guards at crossing points (12.5.2). In case of non-compliance, the border police will issue a refusal of entry, subject to legal remedies without suspensive effect (12.5.3). Special rules for local border traffic mitigate negative repercussions of the strict control regime for the regions on both sides of the external borders (12.5.4).

12.5.1 Entry conditions The designation ‘code’ signals a high degree of harmonisation, based on the understanding that systematic border control is ‘in the interest . . . of all Member States which have abolished internal border control’.155 Judges consider ‘strict compliance with the conditions of entry’ to be essential, since ‘[e]‌ach Member State . . . must have confidence that the controls carried out by every other State in the Schengen area are effective and stringent’.156 Member States cannot complement entry conditions with additional requirements; the list in Article 6(1) Schengen Borders Code Regulation is exhaustive.157 Three sets of entry conditions command our attention, mirroring previous comments on the admission criteria for Schengen visas. First, border guards must ascertain the veracity of travel documents and visas. Data stored in the SIS, VIS, an Interpol database, and national document registers are essential in this respect, as a legislative amendment requiring ‘thorough checks’ stipulated.158 Once the EES starts operating, border guards shall compare fingerprints and the live facial image with the biometrics stored in the databases (or make a new entry, for visa-exempt travellers).159 Secondly, border guards verify whether someone poses a public policy or security risk. As stated previously, alerts or other information in the databases plays a critical role, together with the passenger name record communicated by airlines before arrival.160 The bearing of the ‘public policy’ and ‘public security’ exception was elaborated upon in Chapter 10.3 on the general features of secondary legislation, including case law circumscribing the degree 155 Schengen Borders Code Regulation (EU) 2016/399, recital 6. 156 ANAFE (n 117) para 26. 157 See Case C-575/12 Air Baltic Corporation EU:C:2014:2155, paras 57–70. 158 See Schengen Borders Code Regulation (EU) 2016/399, arts 6(1)(a), (b), 8(3)(a)(i)–(ii), (b)–(f), as amended by Regulation (EU) 2017/458 as regards the reinforcement of checks against relevant databases at external borders [2017] OJ L74/1. 159 ibid art 6(1)(f), as amended by Regulation (EU) 2017/2225 as regards the use of the Entry/Exit System [2017] OJ L327/28, read in conjunction with EES Regulation (EU) 2017/2226, arts 16–17, 23(2), (4). 160 See Council Directive 2004/82/EC on the obligation of carriers to communicate passenger data [2004] OJ L261/24; and Practical Handbook for Border Guards, Annex to Commission Recommendation, C(2019) 7131 final, points 2.5, 3.4, which replaced earlier versions.

External Controls at Crossing Points 325 of overlap with the free movement of Union citizens. The unprecedented travel ban for third country nationals during the Covid-19 pandemic, which was based on the ‘public health’ exception,161 was presented in Chapter 2.3.4 on the ancillary role of soft law. It is inherent in the case law that refusal will usually require an individualised assessment, although Member States arguably benefit from some additional leeway under the additional benchmark of ‘international relations threat’.162 Thirdly, entry will be refused in the absence of sufficient means of subsistence for the duration of the stay and return home.163 Border guards are also expected to verify the purpose of the intended stay to prevent overstaying or irregular work.164 In practice, however, border guards will rarely have the time and information to assess these matters extensively at crossing points. That is why third states whose nationals are considered to pose an abstract ‘illegal migration risk’ are subject to visa requirements, in line with the comments in the previous chapter. Doing so allows consular staff to check the admission criteria in an administrative procedure beforehand. Arrival at the border without a visa will result in a refusal of entry.165 Entry conditions are mandatory.166 Border guards cannot derogate from them unless the Schengen Borders Code authorises them to do so. Three exceptions are explicitly foreseen: third country nationals holding a residence permit or long-term visa for another Member State must be permitted to go there; Schengen visas can be delivered at the external border in exceptional circumstance; and states are free to authorise entry ‘on humanitarian grounds’ on a purely discretionary basis.167 Note that doing so does not exempt states at the external borders from assuming responsibility for asylum applications submitted by migrants admitted on humanitarian grounds; the Courts interprets the ‘irregular entry’ criterion in Article 13 Dublin III Regulation (EU) No 604/2013 to cover anyone not fulfilling the entry conditions.168 Those applying for asylum must be authorised to enter in accordance with the comments on border surveillance that follow.

12.5.2 Checks on persons Anyone entering or leaving the Schengen area shall be checked by border guards, irrespective of nationality. Having said this, the Schengen Borders Code continues the entrenched distinction between lenient rules for beneficiaries of free movement, especially Union citizens, and third country nationals who shall be subject to ‘thorough checks’.169 This differential approach towards the density of entry clearance lies beneath the separate 161 See Schengen Borders Code Regulation (EU) 2016/399, art 6(1)(e), with later amendments. 162 ibid and ch 11.4.1. 163 Schengen Borders Code, art 6(1)(c), (4); and further ch 10.2. 164 Schengen Borders Code399, arts 6(1)(c), 8(3)(a)(iv), Part B Annex V; and Practical Handbook for Border Guards (n 160) point 3.7 s I Pt II. 165 Schengen Borders Code, arts 6(1)(b), 14(1); note that visa holders can still be refused entry, especially on the basis of new information. 166 ibid art 6(1): ‘shall be’; see also Air Baltic Corporation (n 157) paras 50, 67. 167 Schengen Borders Code, art 6(5)(a)–(c). 168 See Case C-646/16 Jafari EU:C:2017:586, paras 75–89; and the alternative opinion of AG Eleanor Sharpston, Opinion in ibid EU:C:2017:443, points 155–90. 169 The distinction goes back to CISA, art 6(2), whose abstract rules were complemented by the former Common [Schengen] Manual [2022] OJ C313/97.

326 Border Controls lanes for ‘EU/EEA/CH Citizens’ on the one hand and for ‘all passports’ on the other hand, which can be found at most international airports (and in many third states pursuing similar practices).170 It would violate Article 18 TFEU to introduce stricter standards for the citizens of Member States not (yet) fully participating in the Schengen area, such as Ireland and Bulgaria171—whereas British nationals count as third country nationals after Brexit. Separate lanes for beneficiaries of free movement are a sign of stratification among (un) privileged travellers, observed in Chapter 11.1, which will be reinforced by self-service terminals and e-gates for registered travellers under the Entry/Exit System.172 Privileges of Union citizens are not limited to these separate lanes. Border guards must not stamp passports, nor can they ask questions about the purpose of the trip, with the exception of potential public policy threats.173 Of great practical importance are judgments on family members having the nationality of a third state. They may even have to be admitted without valid travel documents, and visas may be granted at the border; entry alerts in the SIS will not result in automatic refusal.174 However, EU institutions have established an obligation for border guards to check the veracity of travel documents systematically and the absence of public security risks for beneficiaries of free movement upon entry and exit.175 Doing so was the response to Union citizens joining terrorist groups, such as the ‘Islamic State’. Convoluted legislative provisions give permission temporarily to relax these stringent document checks to prevent congestion.176 Once the Entry/Exit System starts operating, third country nationals will be subject to additional checks, as described in Chapter 9.2 on databases. The underlying idea is that authorities across the Schengen area are informed about the time and place of entry and exit, thus allowing them to identify overstayers quickly and to use the information in the EES for other migration control purposes. The new mechanism will replace the systematic stamping of travel documents.177 Stamping requirements have caused practical problems, in particular with regard to a rebuttable presumption that those without entry stamps do not fulfil the conditions governing the length of permissible stay.178 Two preliminary references illustrate the practical relevance of the matter. They dealt with special problems emanating from the Spanish language version, which did not include the option of rebuttal,

170 See Annex III Schengen Borders Code Regulation (EU) 2016/399, art 10; and previously Decision 2004/581/ EC determining the minimum indications to be used on signs at external border crossing points [2004] OJ L261/ 119. 171 See Daniel Thym, Ungleichzeitigkeit und Europäisches Verfassungsrecht (Nomos 2004) 254–58 http:// www.ungleichzeitigkeit.de (accessed 1 March 2023); and Case C-754/18 Ryanair Designated Activity Company EU:C:2020:478, paras 41–47. 172 See Schengen Borders Code Regulation (EU) 2016/399, arts 8a–8d, as amended Regulation (EU) 2017/2225 (n 159). 173 See Case 157/ 79 Pieck EU:C:1980:179, paras 3– 10; and Case C- 68/ 89 Commission v Netherlands EU:C:1991:226, paras 11–15. 174 See Case C-459/99 MRAX EU:C:2002:461, paras 53–62; Case C-754/18 Ryanair Designated Activity Company EU:C:2020:478, paras 25–47; and Case C-503/03 Commission v Spain EU:C:2006:74. 175 See Schengen Borders Code Regulation (EU) 2016/399, art 8(2), (2e), as amended by Regulation (EU) 2017/ 458 (n 158). 176 Schengen Borders Code, art 8(2a)–(2d). 177 ibid art 11, as amended by Regulation (EU) 2017/2225 (n 159). 178 Schengen Borders Code, art 12 builds upon the provisions first introduced by Regulation (EC) No 2133/2004 on the requirement to stamp systematically the travel documents [2004] OJ L369/5; see also Steve Peers, EU Justice and Home Affairs Law, vol I (4th edn, OUP 2016) 221, 125–27.

External Controls at Crossing Points 327 and stamping requirements for seamen working on ships not usually leaving the international section of Europe’s biggest harbour in Rotterdam.179 Another peculiarity are passengers located in the transit area of airports. They are staying on the territory during transit and are subject to domestic laws and human rights as a result, even though they do not pass a border crossing point. We must distinguish, as a result, between the ‘administrative’ admission to the Schengen area at border crossing points and the ‘constitutional’ presence under human rights law. When it comes to secondary legislation, EU institutions determine the scope ratione loci independently.180 Similarly, the legislature may lay down special rules for transit areas, for instance with regard to asylum border procedures. This distinction between administrative admission and constitutional presence informs the ongoing debate about the ‘fiction of non-entry’, which will be mentioned in Chapter 13.8.4 on asylum.

12.5.3 Refusal of entry and legal remedies Third country nationals not complying with entry conditions ‘shall be refused entry’.181 Reasons must be given in the standard form in Annex V of the Schengen Borders Code, which was introduced at the bequest of the European Parliament and lists abstract criteria for justifying refusal. Additional remarks can be entered whenever appropriate, and Chapter 7.2 illustrated that it can be mandatory to do so in light of judgments interpreting the fundamental right to good administration. Legal remedies before domestic courts are explicitly foreseen not to have suspensive effect.182 Chapter 7.2 demonstrated that doing so can be compatible with human rights provided that there is no danger of refoulement. It would have to be assessed in light of the comments in Chapter 6.5 on the doctrinal foundations whether the EU institutions could hypothetically abandon the need for legal remedies against refusal of entry altogether. Legal remedies under the Schengen Borders Code are confined to refusal of entry. EU legislation does not govern the behaviour of border guards concerning, by way of example, inappropriate statements, overtly lengthy checks, or instances of ‘racial profiling’. In a judgment on such accusations, the Court confirmed the applicability of domestic laws, while not declaring definitely—in the absence of sufficient information in the order for reference— whether the Charter applies in accordance with Article 51(1) CFR.183 Chapter 5.4 explained that it can be challenging to define the parameters for the application of the Charter to operational activities of national border guards which are not covered by the Schengen Borders Code. These uncertainties concern checks at border crossing points as well as border surveillance. In case of ‘second line checks’ at crossing points in a separate location, third country nationals must be given information on the identity of the officer.184 Having said 179 See Joined Cases C-261/08 and C-348/08 Zurita García and Choque Cabrera EU:C:2009:648, paras 52–57; and Case C-341/18 Staatssecretaris van Justitie en Veiligheid EU:C:2020:76. 180 Case C-17/16 El Dakkak and Intercontinental EU:C:2017:341 found that it follows Article 52 TEU and Article 355 TFEU in cases of doubt. 181 Schengen Borders Code Regulation (EU) 2016/399, art 14(1). 182 ibid art 14(2); the former Article 5(2) CISA and the Common [Schengen] Manual (n 169) had relegated these matters to domestic law. 183 See Case C-23/12 Zakaria EU:C:2013:24, paras 33–41. 184 See Schengen Borders Code, arts 2(13), 8(5).

328 Border Controls this, legal remedies against misbehaviour will subject to domestic legislation, in the absence of rules in the Schengen Borders Code.

12.5.4 Local border traffic External border controls complicate the life of persons and companies residing in the border area. Think of the close connections between Romania and Moldova, or between the Eastern parts of Slovakia and Poland and Western Ukraine. To mitigate negative consequences of the enlargement of the Schengen area, EU institutions adopted the Local Border Traffic Regulation (EC) No 1931/2006, which realised an idea that had featured in the Schengen acquis from the beginning.185 Benefits of local border traffic are restricted to third country nationals who have resided within 30km of the border for at least one year; they may receive a special permit for cross-border visits valid between one and five years. Judges have confirmed that the ‘letter and spirit’ of the Local Border Traffic Regulation entail that the sequence of several short term visits may exceed the limit of 90 days during any 180-day period for Schengen visas.186 Details are subject to bilateral arrangements between Member States and third states, which are a prerequisite for the application of the local border traffic regime. There are several such agreements between countries in Central and Eastern Europe, as well as between Norway and Russia, which appear to function reasonably smoothly after some initial arguments.187 The Spanish exclaves of Ceuta and Melilla are not governed by the Regulation.188 Member States remain free, moreover, to conclude treaties with neighbours on other border-related matters, such as the location of crossing points. Protocol [No 23] on External Relations of the Member States with regard to the Crossing of External Borders prevents the emergence of an exclusive external treaty-making competence.189 Protocol [No 23] obliges bilateral agreements to ‘respect Union law’; derogations from existing legislation, such as visa requirements, thus require an express statutory authorisation mirroring the example of the local border traffic permit.

12.6 Surveillance of the External Borders External borders can be compared to the wall of a house: doors in the form of crossing points serve as official gateways to the Schengen area, while the remainder of the ‘green’ land and ‘blue’ sea border is closed: ‘borders may be crossed only at border crossing points and during the fixed opening hours’.190 Control activities in-between crossing points are called ‘border surveillance’ and serve the primary purpose of countering irregular crossings and other illegal activities.191 In contrast to checks at crossing points, the Schengen rulebook on 185 See CISA, art 3(1), which had not been implemented. 186 See Case C-254/11 Shomodi EU:C:2013:182, paras 22–26. 187 See Commission, ‘Second report on the implementation and functioning of the local border traffic regime’ COM(2011) 47 final; and, with regard to Kaliningrad, COM(2014) 74 final. 188 See Local Border Traffic Regulation (EC) No 1931/2006, art 16. 189 See Geert De Baere, ‘The Basics of EU External Relations Law’ in Maes and others (n 43) 121, 170–71; and Thym (n 118) MN 21. 190 Schengen Borders Code Regulation (EU) 2016/399, art 5(1). 191 ibid arts 2(12), 13.

Surveillance of the External Borders 329 border surveillance remains loosely knit, thus leaving Member States much leeway (12.6.1). Interaction with asylum legislation is crucial for the legal assessment of ‘pushback’ allegations against the border guards of several Schengen countries (12.6.2). Special rules exist for external sea borders, which, unfortunately, refrain from defining core terms with sufficient precision (12.6.3).

12.6.1 Land borders Article 13 Schengen Borders Code Regulation (EU) 2016/399 remains abstract in its description of border surveillance, nor does the non-binding Practical Handbook contain further detail.192 Additional information may be found in the operational plans for Frontex missions, in line with comments in Chapter 8.3.1, which apply to national units operating within their framework. In other circumstances, states retain the prerogative to define border surveillance practices. Such leeway has the side-effect of rendering it difficult to conclude that the Charter applies and that the Court holds jurisdiction to answer preliminary references on border surveillance, as described in the section on refusal of entry on the previous pages. The Commission proposes amending Article 13, including an authorisation to adopt delegated acts concerning additional measures governing border surveillance.193 If adopted, this amendment would give the Commission additional leverage to influence controversial national surveillance practices. In accordance with Article 13(1), anyone who enters irregularly via the ‘green’ external land border and does not apply for asylum shall be apprehended and made subject to procedures under the Return Directive, or the basic rump guarantees described in Chapter 16.5.2 on return. These minimum standards encompass basic protection during presence on the territory, not, however, procedural safeguards against rejection at the border.194 ‘Pushback’ allegations against Croatia, Poland, Hungary, and Spain, in particular, are not procedurally curtailed under the Schengen Borders Code and the Return Directive, provided governments decide not to apply ‘regular’ return procedures. These limitations of the legislative provisions on border controls explain why interaction with asylum legislation is crucial for the assessment of pushback allegations. One of the reasons why asylum legislation takes centre stage are the procedural uncertainties about the outer limits of the double prohibition of refoulement and collective expulsion, discussed in the constitutional foundations. Moreover, one might try to activate procedural guarantees in Articles 43 and 47 CFR, although it remains a challenge to maintain that the abstract reference to ‘procedures respecting Directive 2008/115/EC’, in Article 13(1) Schengen Borders Code, is sufficient to bring domestic practices within the scope of the Charter. Even if judges followed that argument, the obligation to give reasons can be restricted and the right to an effective remedy presupposes the violation of an individual right and does not necessarily require suspensive effect, as explained in Chapter 7.2 on the administrative dimension.

192

Practical Handbook for Border Guards (n 160) Pt III lists different control modalities on two short pages. See ch 2.3.3; and Temporary Border Controls Proposal (n 139) art 13(7). 194 See Return Directive 2008/115/EC, arts 2(2)(a), 4(4). 193

330 Border Controls None of the above becomes practically relevant if physical barriers, such as border fences or walls, prevent foreigners from crossing an external border. It may sound cynical but physical carriers can be employed by states to avoid the politically sensitive and legally complex surveillance operations by border guards, which require human resources and can give rise to legal challenges. The Schengen Borders Code contains no provisions on physical barriers, apart from the obligation to use the official crossing points. Recent years have witnessed the proliferation of border fences and walls, starting with Spain towards Morocco in the 1990s and Hungary towards Serbia in 2015. Greece, Lithuania, Poland, Estonia, and Bulgaria followed the ‘example’, as did Hungary and Slovenia towards Croatia before full Schengen membership. After several years of debates, the European Council decided in 2023 that the EU budget can be used for financing ‘infrastructure, means of surveillance . . . and equipment’.195 That was a face-saving exercise for the opponents of direct payments for fences. National governments finance the fences as such, whereas the EU gives money for the control environment.

12.6.2 Interaction with asylum law ‘Pushback’ does not have a precise meaning, although the term is usually employed for state practices which prevent entry onto the territory or force foreigners, including potential asylum applicants, to leave without any procedural safeguards. EU legislation does not currently outlaw such practices during border surveillance explicitly, and uncertainties persist, as we have seen, about the outer limits of human rights law. The Schengen Borders Code contains an abstract reference to refugee law, not, however, any further instruction as to what border guards are supposed to do.196 That is why asylum legislation takes centre stage when dealing with pushback scenarios. Articles 3 and 6(1) Asylum Procedures Directive 2013/32/EU lay down an individual right to apply for asylum. Comments in the next chapter will mention that his option may be mandatory under Article 18 CFR, notwithstanding ambiguous wording. For practical purposes, these uncertainties about human rights are irrelevant for the simple reason that secondary legislation is precise: asylum applications can be lodged ‘at the border’ or ‘in the territorial waters’. They can be submitted with any state authority, including border guards, in line with comments in Chapter 13.4.1 on access to the procedure. Asylum seekers must be allowed to enter the territory and receive a procedure at the external border or elsewhere. Yet, the devil is in the detail. Articles 6(1)(3), Article 8(1), and Recital 28 Asylum Procedures Directive require states to actively provide information on the possibility to make an application and to ensure basic communication via unspecified ‘interpretation arrangements’.197 All this is to be done by border guards, mirroring basic procedural safeguards under the prohibition of collective expulsion in line with previous comments. Specifically trained personnel and the presence 195 See https://en.wikipedia.org/wiki/Border_barrier (accessed 1 March 2023); and European Council, ‘Conclusions’ (EUCO 1/23, 9 February 2023) No 23. 196 See Schengen Borders Code Regulation (EU) 2016/399, arts 3(b), 4, 15(1); and Violeta Moreno-Lax, Accessing Asylum in Europe (OUP 2017) 70–75. 197 See also Case C-808/18 Commission v Hungary EU:C:2020:1029, paras 104–05; and MK and others v Poland (n 64) §§ 207–09.

Surveillance of the External Borders 331 of translators are foreseen at a later stage within the territory: during the personal interview in accordance with Article 15. The wording confines these positive obligations to crossing points. During border surveillance, Recital 26 demands less: neither proactive provision of information nor interpretation, just training for border guards so that they can inform individuals about what to do next once they have made an application. Apparently, the onus to apply for asylum rests with applicants. There is not much case law elucidating the bearing of these provisions, let alone their interaction with Article 18 CFR. That is unfortunate, as their interpretation is of utmost practical significance. Under which conditions should border guards assume that third country nationals wish to apply for asylum. Do they have to say so more or less explicitly (and if so, in which language)? Judges recognised that the use of specific terminology, such as the term ‘asylum’, is not warranted, although it remains unclear whether circumstantial evidence alone, without individuals having expressed an active desire for asylum of some sort, may be sufficient to conclude that an application has been made.198 Migrants and refugees sitting on boats in the Mediterranean cannot apparently be categorised as asylum seekers automatically. States exploit corresponding factual uncertainties to justify summary rejection during border surveillance. We shall see in Chapter 13.4.1 that Commission proposal to postpone the ‘lodging’ and ‘registration’ of asylum applications whenever neighbouring states ‘instrumentalise’ migration would not invalidate the option to apply for asylum. Member States would be authorised to defer formal processing only. Having said this, governments may possibly try to argue that border guards could be authorised prevent entry via the ‘green’ land border, provided that those wishing to apply for asylum could realistically do so at a border crossing point nearby, mirroring case law on the prohibition of collective expulsion mentioned previously. Lithuania seems to have implemented such ‘diversion’ to crossing points during 2021.199 Even if one accepted that option in the abstract, such requirement would regularly fall foul of the obligation to guarantee effective access to asylum procedures if states closed crossing points or if crossing points could not be reached easily in practice. That is precisely what Greece and Poland did during clashes about the ‘instrumentalisation’ of migration with Turkey in 2020 and Belarus in 2021.200

12.6.3 Sea borders Control of the external sea borders reiterates the basic distinction between crossing points and border surveillance. International ports serve as the official gates of entry.201 Minor exceptions exist for cruise ships, pleasure boating, and coastal fishing; they may, in certain circumstances, enter via the ‘blue’ sea border.202 For others, it is illegal to do so. Corresponding 198 Case C-36/20 PPU Ministerio Fiscal EU:C:2020:495, paras 56–68 passim; and also MA v Lithuania App no 59793/17 (ECtHR, 11 December 2018) § 109; and Khlaifia and others v Italy (n 73) §§ 238, 247–48. 199 See ECRE, ‘Legislative Changes in Lithuania’ (Legal Note 11/2021, 3 September 2021); Case C-72/22 PPU Valstybės sienos apsaugos tarnyba EU:C:2022:505 concerned a different aspect of the reform, namely the denial of the option to apply for asylum for anyone residing illegally on the territory. 200 See Grażyna Baranowska, ‘The Deadly Woods’ Verfassungsblog (29 October 2021). 201 See Schengen Borders Code Regulation (EU) 2016/399, Annex VI, arts 2(2), 5(1), (2), 19, Pt III, with later amendments. 202 ibid Nos 3.2.3(c), 3.2.4, 3.2.7 Part III Annex VI; and also CISA, art 3(1).

332 Border Controls debates about boat arrivals in Spain, Italy, Malta, Greece, and Cyprus have persisted for decades. Chapter 8.3.1 discussed Frontex missions in response to the surge in such arrivals, and Chapter 18.3.1 on the external dimension will highlight controversial national practices. The best-known example may be cooperation between Italy and Libya, which gave rise to the Hirsi judgment.

12.6.3.1 Sea Borders Regulation After laborious negotiations, the Council and the European Parliament agreed on the Sea Borders Regulation (EU) No 656/2014, after the predecessor instrument had been annulled by the Court on procedural grounds.203 The short title used in this collection should not be misread as indicating that the Regulation contains generic prescriptions for sea border surveillance: the scope is limited ratione materiae to national activities in the context of Frontex missions.204 Other national practices are—as in the case of land borders—governed by the abstract provisions on border surveillance in the Schengen Borders Code and need not be subject to the Charter as a result,. The Sea Borders Regulation revolves around two sets of rules: one on ‘interception’ and one on ‘search and rescue’, which responded to the Hirsi judgment. The reference to both ‘interception’ and ‘search and rescue’ hints at a conceptual tension between deterrence and protection which defines the Regulation.205 While the element of protection was modelled on human rights and the law of the sea, the component of deterrence builds upon the ‘Palermo Protocol’ against the smuggling of migrants, which will be mentioned in Chapter 16.3.1 on return. The ensuing tension between control and protection cannot be resolved easily. Articles 9 and 10 on search and rescue do little in terms of defining the notion of ‘distress’ or introducing inter-state cooperation procedures to overcome the shortcomings of the law of the sea. This silence on controversial issues was a deliberate choice at a time when Southern Member States were involved in controversies over disembarkation.206 An important novelty was Article 10(1), which requires Frontex to include modalities for disembarkation into operational plans by specifying the state responsible; disembarkation in third states is mentioned as an option, provided a country qualifies as a ‘place of safety’ in accordance with international law. In the absence of alternative solutions, the host state of the Frontex mission shall assume responsibility for disembarkation, which led Malta to refuse to host Frontex missions henceforth.207 Given that the Regulation covers Frontex missions only, it did not prevent heated inter-state disputes over disembarkation in other scenarios. As stated previously, no generic obligation to permit disembarkation exists under international law, except for situations of urgency.

203 See ch 2.4.3; and Decision 2010/252/EU supplementing the Schengen Borders Code as regards the surveillance of the sea external borders [2010] OJ L111/20; which was annulled by Case C-355/10 European Parliament v Council EU:C:2012:516; for the drafting history see Bernard Ryan, ‘Sea Borders Regulation (EU) No 656/2014’ in Thym and Hailbronner (n 118) Article 1 MN 5–8a. 204 See Sea Borders Regulation (EU) No 656/2014, art 1. 205 See Maarten den Heijer, ‘Frontex and the Shifting Approaches to Boat Migration in the European Union’ in Ruben Zaiotti (ed), Externalizing Migration Management (Routledge 2016) 53, 54–58. 206 See ‘Position on Articles 9 and 10’ (Council doc 14612/13, 10 October 2013); and Enkelejda Koka and Denard Veshi, ‘Irregular Migration by Sea’ (2019) 21 EJML 26, 43–46. 207 See den Heijer (n 205) 63–66.

Surveillance of the External Borders 333 The migration control component takes centre stage in the provisions on interception in the territorial sea, in the contiguous zone, or on the high seas. Interception modalities were modelled on the facts of the Hirsi judgment and the contents of the Palermo Protocol, which both concerned operational powers regarding the boarding, searching, and seizure of vessels engaged in smuggling.208 Border guards may order vessels ‘not to enter the territorial sea’ or to leave it, and they may even escort the ship to a third state.209 Governments often rely on this authorisation to justify pushback practices to prevent, in the language of the supranational legislature, ‘smuggling of migrants by sea’ ‘for the purposes of detecting, preventing and combating illegal immigration’.210

12.6.3.2 Basic procedural safeguards Article 4(3) is the most important provision in the Sea Borders Regulation. It establishes basic procedural safeguards for ‘intercepted or rescued persons’. The wording of Article 4(3) refers to identification, assessment of personal circumstances, information about the destination, and the option to bring forward arguments about a danger of refoulement—‘in a way that those persons understand or may reasonably be presumed to understand’. These procedural safeguards mirror the ECtHR’s case law on the prohibition of collective expulsion and are, therefore, less time-consuming than asylum status determination. Once a third country national applies for asylum, she has to be brought onshore. Nothing in the Regulation requires Member States to foresee legal remedies against interception or disembarkation before domestic courts.211 Notwithstanding these caveats, Article 4(3) establishes a basic safety net against pushback practices. Unfortunately, the Sea Borders Regulation refrains from defining the notion of ‘interception’, which triggers the procedural safeguards. There are at least two possibilities as to what ‘interception’ means. On the one hand, one could interpret the term broadly to cover any prevention of onward travel; the heading of Articles 6–8 supports such a reading.212 On the other hand, one might require a certain level of control over the vessel. Doing so would coincide with the general scheme of the Sea Borders Regulation, which is primarily concerned with boarding and searching; moreover, procedural safeguards in Article 4(3) presuppose the presence of migrants on border guard ships (like in the case of ‘rescue’, to which this provision refers as well). One might draw a parallel between ‘interception’ and the notion of ‘jurisdiction’ under human rights law, for instance in the Hirsi ruling, which provided the backdrop for the legislative debate. At the time of writing, there had been no widely reported domestic judgments, let alone preliminary references, exploring the notion of interception. This silence reiterates our earlier finding about the practical limitations of judicial oversight by the CJEU. Whenever the Sea Borders Regulation does not apply ratione materiae, individuals may rely on the double prohibition of refoulement and collective expulsion, discussed at the outset of this chapter. Practical difficulties in reaching the official border crossing points in international 208 See Sea Borders Regulation (EU) No 656/2014, arts 6–8; and ch 16.3.1. 209 Ibid arts 6(2)(b), 7(2)(b), (c). 210 ibid recital 6, arts 6 and 7. 211 See also den Heijer (n 205) 59–62. 212 See also UNHCR ExCom, ‘Protection Safeguards in Interception Measures’ (Conclusion No 97 (LIV), October 2003); and remember that terminology of EU legislation often has an autonomous meaning independent of international law.

334 Border Controls ports can be used as an argument why the ‘own conduct’ exception, which permits summary pushbacks under human rights law, does not apply to vessels carrying migrants in the Mediterranean.

12.7 Document Security Passports replaced other identity documents when migration law emerged throughout the long nineteenth century. Anti-forgery and counterfeiting had always been a priority to ensure authenticity and to prevent fraud and abuse for diverse illicit purposes.213 In the aftermath of the terrorist attacks of 2001, the Passports and Travel Documents Regulation (EC) No 2252/2004 set out minimum security standards for passports of Union citizens. Security features comprise a storage medium with a facial image; two fingerprints from anyone above the age of twelve years were added later.214 These biometric identifies must also be included in ID cards for Union citizens and residence permits for their family members coming from third states.215 Judges confirmed that the collection and storage of fingerprints complies with fundamental rights.216 Political preference, on the part of the Commission, to introduce a pan-European register for travel documents of Union citizens was not pursued further, thus effectively relegating this matter to the domestic level.217 Many Member States have such databases, which transcend the scope of application of the Passports and Travel Documents Regulation and are not, as a result, subject to CJEU jurisdiction.218 Nevertheless, these national databases must be consulted during entry and exit checks at border crossing points.219 Questions regarding the entry of names and other information into travel documents can be subject to Court oversight, as two cases concerning names at birth and the transcription of the Greek alphabet illustrate.220 Today, Article 77(2)(a) TFEU serves as a legal basis for the Passports and Travel Documents Regulation. A previous common accord on a uniform passport pattern was adopted on an intergovernmental basis; it introduced the burgundy red cover page bearing the words ‘European Union’ (formerly European Community).221 Article 21(2) TFEU may be activated as an alternative or additional legal basis, when the link to border controls is not evident, for instance for identify cards, whereas Article 79(3) TFEU, requiring unanimity, appears superfluous.222 Biometrics and security features are similarly prescribed for visas and residence permits, including for purposes not harmonised in sectoral EU legislation.223 213 See generally Mark B Salter, Rights of Passage (Lynne Rienner 2003). 214 See Passports and Travel Documents Regulation (EC) No 2252/2004, art 1, as amended by Regulation (EC) No 444/2009 [2009] OJ L142/1. 215 See Regulation (EU) 2019/1157 on strengthening the security of identity cards of Union citizens [2019] OJ L188/67, which replaced an Intergovernmental Resolution (Council doc 15356/06, 15 November 2006). 216 See Case C-291/12 Schwarz EU:C:2013:670, paras 24–62. 217 See Peers (n 178) 146–49. 218 See Joined Cases C-446/12–C-449/12 Willems and others EU:C:2015:238, paras 47–50. 219 See Schengen Borders Code Regulation (EU) 2016/399, art 8(2)(a)(3), with later amendments. 220 See Case C-101/13 U EU:C:2014:2249, paras 48–50; and Case C-168/91 Konstantinidis EU:C:1993:115, paras 11–17, decided on the basis of single market provisions. 221 See the Resolution of the Representatives of the Governments of the Member States, meeting within the Council [1981] OJ C241/1, with later amendments. 222 See Thym (n 118) MN 24. 223 See Residence Permit Format Regulation (EC) No 1030/2002, as amended by Regulation (EC) No 380/2008 [2008] OJ L114/88; and ch 11.3.4.

Summary 335 Uniform security standards facilitate the recognition of residence permits by border guards and other authorities across the Union. A new stand-alone template for residence permits started operating in 2019.224 Unlike in the case of Union citizens, many documents of third country nationals will be stored centrally in the Common Identity Repository, once the databases described in Chapter 9 will become interoperable.

12.8 Summary External border controls are focal points of contemporary debates about asylum, in contrast to the earlier emphasis on the abolition of internal border controls as the main motivation behind ‘Schengen’. Common rules on checks on persons entering the Schengen area from a third state were one of the flanking measures compensating states for the loss of internal control options. At the same time, Europeanisation was much more than the simple projection of earlier control practices upon the external border. The Schengen framework served as a laboratory for the design and proliferation of new control instruments which have defined European cooperation ever since, especially databases, visa requirements, and inter- state cooperation. Moreover, borders serve as symbolic markers in the policy debate. In the EU, the absence of internal controls embodies European unity, and the fortification of the external borders coveys a message of protection towards the population. Human rights are vital for the legal analysis. The notion of ‘jurisdiction’ defines the application of the ECHR whenever states exercise effective control over persons or territory. Member States must respect human rights in transit areas and towards persons on coast guard vessels on the high seas, not however when assessing an application for a humanitarian visa in a consulate abroad. The territorial scope of the Charter defies easy definition and has not been resolved authoritatively. A crucial question of great practical importance concerns procedural guarantees against rejection at the border under human rights law: doctrinal uncertainties persist as to the overlap and distinction of the double prohibition of collective expulsion and non-refoulement. It remains unclear whether and if so under which circumstances border guards may ‘push back’ migrants entering irregularly without basic screening. By contrast, the obligation to safe life at sea is firmly embedded in the international legal material, even though it can be difficult to apply these standards in practice and to convince states to accept disembarkation. Not all Member States participate in the Schengen area. While Ireland has an opt-out, others are waiting to join the inner circle after accession; a Council decision on the full membership of Bulgaria and Romania has been postponed for more than a decade. Political disagreement about enlargement reiterates the governance deficits that have shaken the Schengen area on several occasions, ever since a public dispute between France and Italy over secondary movements in 2011. Attempts to render the Schengen evaluation mechanism more effective have yielded mixed results so far. The asylum policy crisis of 2015/16 and the Covid-19 pandemic triggered the multiplication of internal border controls, which several governments retained over prolonged periods by invoking different legal bases. The

224 See Residence Permit Format Regulation (EC) No 1030/2002, as amended by Regulation (EU) 2017/1954 [2017] OJ L286/9.

336 Border Controls Court confirmed that many of these practices were illegal. A legislative revision extending time limits was ongoing at the time of writing. Along similar lines, police checks in border areas are subject to legislative caveats, even though judges give Member States considerable leeway to organise random police checks in the ‘hinterland’. Rules governing movements within the Schengen area for short stays of up to 90 days are straightforward for Schengen visa holders and third country nationals with domestic residence permits. By contrast, secondary movements of (former) asylum applicants are politically controversial. While asylum seekers are not usually authorised to travel freely for short stays, beneficiaries of international or humanitarian protection can do so if they fulfil entry conditions. At the external borders, EU law distinguishes checks on persons at crossing points from border surveillance. The archetypical image of border guards checking identity papers at crossing points is regulated in detail, revolving around compliance with the entry conditions set forth in Article 6(1) Schengen Borders Code Regulation. In case of non- compliance, a written refusal of entry will use a standard form for underlying reasons, even though human rights can oblige border guards to handle procedural guarantees generously. Refusal of entry is subject to judicial review, albeit without suspensive effect. While third country nationals are checked thoroughly in separate lanes, Union citizens and their family members are subject to basic checks upon entry and exit. Entry into operation of the Entry/Exit System will supplant previous rules on the stamping of passports with digitalised checks via databases. In contrast to administrative decisions about refusal of entry, the behaviour of border guards is loosely regulated at the supranational level; it depends on the circumstances of the individual case whether the CJEU holds jurisdiction. Surveillance of the ‘green’ and ‘blue’ external land and sea borders in between crossing points is governed by domestic legislation for the most part. Deference to the Return Directive in the Schengen Borders Code Regulation has little effects in practice, and the procedural guarantees against refusal of entry apply at crossing points only. That is why the legal analysis of ‘pushback’ allegations concentrates on asylum legislation: third country nationals must be admitted to the territory if we conclude that they have applied for asylum during border surveillance. Notwithstanding this express obligation, practical and legal ambiguities concern the factual conditions under which an application has been made. The Sea Borders Regulation (EU) No 656/2014 applies to national units operating in the context of a Frontex mission. It goes further than the provisions on land borders by embracing basic screening obligations, to be performed at sea, once a person has been ‘intercepted’ or ‘rescued’. Unfortunately, the interpretation of these crucial triggers for basic procedural safeguards is subject to interpretative uncertainty.

13

Common European Asylum System Asylum law and policy preoccupy many practitioners and academics. Questions of legal interpretation, the search for an appropriate policy design, and the assessment of factual circumstances have significant implications for the lives of refugees and cause heated controversies. In this overall context, this chapter will concentrate on the supranational legislation. Directives and regulations on asylum jurisdiction, procedures, reception conditions, and the criteria for refugee status and subsidiary or temporary protection are the hallmark of the Common European Asylum System (CEAS), which the EU has set up in accordance with Article 78(2) TFEU. Speaking about the CEAS habitually involves an analytical focus on the legislative dimension,1 in contrast to the wider notion of ‘asylum policy’. The latter embraces the external dimension and interaction with other segments of migration law. Visa policy and border controls, for instance, have a profound impact on the number of asylum applications but they are not commonly discussed under the heading of ‘Common European Asylum System’. As a result, anyone aspiring to have a holistic understanding of EU asylum policy ought to look beyond the legislation discussed in this chapter. International cooperation will take centre stage in Chapter 18; the Asylum Agency, Frontex, and hotspot practices were examined in Chapter 8; Eurodac and other databases can be found in Chapter 9; visas and border controls featured in Chapters 11 and 12; and return after a negative decision will be elaborated upon in Chapter 16. Legislation on asylum has to be analysed in this broader context, especially when it comes to access to the territory. Once an asylum applicant has reached European soil, she will be treated in accordance with the instruments presented below, provided domestic authorities are able and willing to respect them in practice. Controversies have accompanied the Europeanisation of asylum policy from the beginning. The legislation discussed in this chapter presents us with the path-dependent continuation of decisions taken early on (13.1). Constitutional guarantees on human rights and inter-state solidarity are of crucial significance (13.2). ‘Dublin’ has caused political debates about the first entry criterion and secondary movements, as well as legal battles about the limits of mutual trust (13.3). Procedures before national authorities and domestic courts are essential for the examination of individual cases (13.4), in light of the intricate criteria for refugee status and subsidiary protection (13.5 and 13.6). By contrast, temporary protection is based on abstract criteria, without long procedures (13.7). Standards on the treatment of asylum applicants and beneficiaries of international protection differ before and after recognition, as well as between Member States (13.8 and 13.9). Resettlement and other legal pathways mitigate the predominance of non-arrival policies (13.10).

1 See TFEU, art 78(2)(a)–(f); and European Council, ‘Presidency Conclusions of the Meeting in Tampere’ (15/ 16 October 1999) Nos 13–17. European Migration Law. Daniel Thym, Oxford University Press. © Daniel Thym 2023. DOI: 10.1093/oso/9780192894274.003.0014

338 Common European Asylum System

13.1 Theory and Policy Design Observers of contemporary asylum policy are often frustrated with what they perceive to be a lack of commitment to uphold the rights of refugees. While the intensity and salience of the debate have increased, ambivalence is not a novelty. Western countries developed ‘non-arrival’, ‘protection elsewhere’, and ‘non-admission’ strategies when refugees from the Global South started moving northwards in greater numbers (13.1.1). Europeanisation was a by-product of these restrictive endeavours during the early years of Schengen, Dublin, and intergovernmental cooperation (13.1.2). Legislative harmonisation under the Treaty of Amsterdam resulted in the adoption of complex rules with enhanced guarantees for applicants (13.1.3). Repeated policy crises have shaken the CEAS, with the events of 2015/ 16 as the climax of political attention (13.1.4). Another round of legislative negotiations was ongoing at the time of writing; it partly replaced an earlier, futile attempt at policy reform (13.1.5). Recent years have witnessed the return of geopolitics, with neighbouring states ‘instrumentalising’ migration for political purposes and the first ever activation of the Temporary Protection Directive (13.1.6). Structural deficits hamper policy reform and complicate compliance with EU asylum legislation on the ground (13.1.7).

13.1.1 A history of half-hearted commitment Notwithstanding the widespread designation of the Refugee Convention as the ‘magna carta’ for refugee protection, historical studies emphasise the element of ambivalence that has defined the emergence of international refugee law. It is widely known that the Refugee Convention had originally been based on a geographic and temporal limitation: it concerned only refugees from Europe who were residing in another country already; millions of displaced persons on the Indian subcontinent and victims of late colonial suppression were not included (with the exception of Palestinian refugees under the auspices of the United Nations Relief and Works Agency for Palestine Refugees in the Near East (UNRWA)).2 This changed in 1967 when the Convention was universalised in a rather haphazard manner; the core motivation was to secure the predominance of Western-style multilateralism in the age of decolonisation.3 States did not change the narrow refugee definition, although it was manifestly insufficient to deal with mass displacement in the Global South, for instance during the Biafran War. At the time, refugees from the Global South remained in neighbouring countries predominantly. Western states welcomed refugees from the Communist East with open arms and sponsored generous resettlement of Vietnamese boatpeople escaping communism, thus accentuating their normative credentials.4 For refugees elsewhere, however, the West assumed responsibility indirectly, via financial support for humanitarian actors. The mandate of UNHCR was extended on several occasions to take care of refugees in regions of 2 See Nevzat Soguk, States and Stranger (Minnesota UP 1999) ch 4; and Emma Haddad, The Refugee in International Society (CUP 2008) ch 6. 3 See Sara E Davies, ‘Redundant or Essential? How Politics Shaped the Outcome of the 1967 Protocol’ (2007) 19 IJRL 703, 715–26. 4 See James C Hathaway, ‘A Reconsideration of the Underlying Premise of Refugee Law’ (1990) 31 Harv Intl LJ 129, 144–56; and Guy S Goodwin-Gill, ‘The Politics of Refugee Protection’ (2008) 27 Refugee Surv Q 8.

Theory and Policy Design 339 origin.5 When migrants and refugees from the Global South started moving northwards in greater numbers from the 1980s onwards, destination countries reacted by means of domestic reform and inter-state cooperation to ‘stem’ the inflow.6 The demise of communism and the emergence of ‘global governance’ reinforced the trend towards restrictions: it became normatively attractive to focus on the eradication of root causes.7 Externalisation will be traced in Chapter 18.3. Restrictive practices are often described in terms of ‘non-arrival’, ‘protection elsewhere’, and ‘non-admission’. These labels designate inter-related approaches to asylum policy. ‘Non-arrival’ (French non-entrée) serves as an umbrella term for policy initiatives preventing arrivals at the border.8 Visa requirements, carrier sanctions, and financial or operational support for third states, discussed in Chapters 11 and 18, are classic examples of non-arrival. ‘Protection elsewhere’ is used for asylum legislation that allows states to reject asylum applications on the basis of protection applicants have received, or could have received, in transit countries. Provisions on safe third countries and first countries of asylum will be elaborated upon in the section on asylum procedures. The notion of ‘non-admission’ is less common and can be employed to describe procedural restrictions in the form of border procedures, administrative detention, or accelerated processing.9 They play a central role in the ongoing legislative debate in Brussels. Europeanisation reinforced the proliferation of restrictions, with the focus of attention shifting in response to changing circumstances. One factual change was the reorientation of migratory patterns, which have occasionally resembled a ‘cat-and-mouse game’: migrants and smugglers adapt to policy reform in line with the interdisciplinary comments in Chapter 4.2.5. By way of example, visa requirements and carrier sanctions forced migrants to reach Europe irregularly, in the absence of legal entry options. Having said this, movements are not simply ‘diverted’ as a result of restrictions. To use irregular entry channels is costly and dangerous, thus causing a rise in ‘transaction costs’ in the language of economic theory. In light of the interdisciplinary debate about the drivers of migration, such an increase in financial expenditure and personal risks will reduce the number of migrants who venture the journey, compared to those who would have tried to reach Europe in the absence of policy restrictions.

13.1.2 Early years of intergovernmental cooperation The historic background to Europeanisation was what might be called the EU’s first asylum policy crisis during the late 1980s and early 1990s when several countries witnessed an increase in asylum applications governments deemed to be problematic.10 Profound 5 See Alexander Betts and others, UNHCR (2nd edn, Routledge 2012) ch 2. 6 See BS Chimni, ‘The Geopolitics of Refugee Studies’ (1998) 11 J Refugee Stud 350. 7 See GJL Coles, ‘Approaching the Refugee Problem Today’ in Gil Loescher and L Monahan (eds), Refugees and International Relations (OUP 1989) 373, 383–410; and Alexander T Aleinikoff, ‘State-Centered Refugee Law’ (1992) 14 Mich J Intl L 120. 8 The term was coined by James C Hathaway, ‘The Emerging Politics of Non-entrée’ (1992) 9 Refugees 40, in a publication of the UNHCR Public Information Section, which is not widely available in university libraries. 9 See Jens Vedsted-Hansen, ‘Non-Admission Policies and the Right to Protection’ in Frances Nicholson and Patrick Twomey (eds), Refugee Rights and Realities (CUP 1999) 269. 10 See Adam Luedtke, ‘Migration Governance in Europe’ in Agnieszka Weinar and others (eds), The Routledge Handbook of the Politics of Migration in Europe (Routledge 2019) 15, 16–19.

340 Common European Asylum System asymmetries characterised the asylum statistics at the time: during 1985–2000, no fewer than two-thirds of all applications were made in a single country, Germany, followed by France, the UK, the Netherlands, and Switzerland. In relative terms, Sweden, Germany, Denmark, Austria, and the Benelux countries were disproportionately affected, receiving ten times as many applications per capita as Italy or Spain.11 Countries in the south and east had ratified the Refugee Convention but did not possess complex procedures or administrative expertise in dealing with asylum matters.12 Important policy choices were made at the time and have been continued in a path- dependent manner until today. Europeanisation served as a platform for the promotion of ‘non-arrival’, ‘protection elsewhere’, and ‘non-admission’ policies. Classic examples are provisions on safe third countries and accelerated procedures for manifestly unfounded applications, which had been spearheaded by countries like Denmark, Germany, Belgium, or the Netherlands and via intergovernmental consultations.13 The ‘London Resolutions’ on these matters were adopted among the Member States in a purely intergovernmental setting in November 1992.14 Other issues were covered by non-binding resolutions under the intergovernmental ‘third pillar’ of the Maastricht Treaty, mentioned in introductory Chapter 1.2.2, which delivered few tangible results but prepared the ground for legislative harmonisation later on. Additional, non-binding activities took place in the framework of the Council of Europe, which have occasionally been taken up by the case law of the European Court of Human Rights (ECtHR).15 The proliferation of ‘best practices’, designed by asylum destination countries, mostly occurred without a legal obligation to do so. Horizontal policy transfer from north to south and east by means of ‘copying and pasting’16 continued in the run-up to enlargement and under the first generation of the CEAS.17 Germany, Denmark, and the Netherlands, amongst others, dominated the policy debate on the basis of their practical expertise, thus leading by way of example. In this overall context, Europeanisation was an end in itself to establish functioning asylum systems across the continent, but it also pursued the—not so—hidden agenda to create a ring of safe neighbours. The German Constitutional Court accepted a constitutional amendment on safe third countries and authorising the ratification of Schengen and Dublin, noting that the legislature had ‘abandoned the previous idea

11 See the dataset of Eiko R Thielemann, ‘Why Asylum Policy Harmonisation Undermines Refugee Burden- Sharing’ (2004) 6 EJML 47, 48–54. 12 See Andrew Geddes and Peter Scholten, The Politics of Migration and Immigration in Europe (2nd edn, Sage 2016) chs 8–10; and Russel King, Gabriella Lazaridis, and Charalambos Tsardanidis (eds), Eldorado or Fortress? Migration in Southern Europe (Palgrave Macmillan 2010). 13 See Maryellen Fullerton, ‘Restricting the Flow of Asylum-Seekers in Belgium, Denmark, the Federal Republic of Germany, and the Netherlands’ (1988/89) 29 Va J Intl L 29, 33–114; Hélène Lambert, Seeking Asylum: Comparative Law and Practice in Selected European Countries (Martinus Nijhoff 1995); and Christina Oelgemöller, The Evolution of Migration Management in the Global North (Routledge 2017). 14 See Kay Hailbronner, Immigration and Asylum Law and Policy of the European Union (Kluwer 2000) 443– 63; and Daniele Joly, ‘The Porous Dam. European Harmonization on Asylum in the Nineties’ (1994) 6 IJRL 159, 166–74. 15 See Kay Hailbronner, Möglichkeiten und Grenzen einer europäischen Koordinierung des Einreise-und Asylrechts (Nomos 1989) 27–34. 16 Ségolène Barbou des Places, ‘EU Asylum Policy and Regulatory Competition’ (2004) 24 J Public Policy 75, 93–94. 17 See Cathryn Costello, ‘Administrative Governance and the Europeanisation of Asylum and Immigration Policy’ in Herwig Hofmann and Alexander Türk (eds), EU Administrative Governance (Edward Elgar Publishing 2006) 287, 313–14.

Theory and Policy Design 341 of solving the problems resulting from worldwide movements of refugees and migrants solely by means of domestic legislation’.18 Schengen and Dublin were the only legally binding output of the early years of intergovernmental cooperation. The Convention Implementing the Schengen Agreement set the tone by establishing rules on asylum jurisdiction.19 They were replicated by the Dublin Convention among all twelve Member States, which was negotiated at short notice and whose substance followed the Schengen model.20 Rules on asylum jurisdiction were put into effect in March 1995 (for Schengen) and in September 1997 (for Dublin) after a drawn-out ratification process at a moment when the EU’s first asylum policy crisis was subsiding.21 They established a pattern that haunts European asylum policy to this day: the infamous residual asylum jurisdiction of the country of first arrival whose origin will be traced in the Dublin context.

13.1.3 First and second phase of harmonisation Asylum jurisdiction was coordinated under the Schengen and Dublin Conventions without any harmonisation of the national legislation on procedures, reception conditions, or recognition criteria besides an abstract reference to the Refugee Convention. That changed under the Treaty of Amsterdam when the EU institutions made a leap forward on the basis of the political momentum initiated by the Tampere Programme. Heads of state or government also introduced the notion of a ‘Common European Asylum System’, which is mentioned in Article 78(2) TFEU nowadays. It served as an umbrella term for the legislation on asylum and signalled an intermediate level of ambition, as described in Chapter 7.1.1. EU Treaties originally provided for ‘minimum standards’ only, raising legal arguments about the permissibility of national deviation under the first generation of asylum legislation.22 This debate is no longer relevant. We should be careful not to read history backwards. Many of the provisions and policy concepts we may take for granted these days were the outcome of long and often controversial debates about the way forward. Think of subsidiary protection, which was a novelty, or detailed prescriptions for the abstract notion of persecution. National practices on these matters varied greatly.23 While the Commission kick-started the debate with communications and proposals,24 the destination countries in Western and Northern Europe dominated the negotiations for the simple reason that they had the practical expertise in how to set up functioning asylum systems. A first generation of asylum legislation was adopted 18 Federal Constitutional Court (Bundesverfassungsgericht), Joined Cases 2 BvR 1938/93 and 2 BvR 2315/93 (judgment of 14 May 1996) para 160 (own translation) https://www.servat.unibe.ch/dfr/bv094049.html (accessed 1 March 2023). 19 See CISA, arts 28–38. 20 See Convention Determining the State Responsible for Examining Applications for Asylum Lodged in one of the Member States of the European Communities (adopted 15 June 1990, entered into force 1 September 1997) [1997] OJ C254/1 (Dublin Convention). 21 See Kay Hailbronner and Claus Thiery, ‘Schengen II and Dublin’ (1997) 34 CML Rev 957. 22 See ch 10.6.1. 23 See Lambert (n 13); and Miriam Wolter, Auf dem Weg zu einem gemeinschaftlichen Asylrecht in der Europäischen Union (Nomos 1998). 24 See eg Commission, ‘Communication: Towards a common asylum procedure and a uniform status, valid throughout the Union’ COM(2000) 755 final.

342 Common European Asylum System unanimously by the Council, after consultation of the Parliament. It comprised the former Asylum Reception Conditions Directive 2003/9/EC, the former Asylum Qualification Directive 2004/83/EC, the former Asylum Procedure Directive 2005/85/EC, and the former Dublin II Regulation (EC) No 343/2003, together with the Eurodac database. Critical voices showed disappointment about the spread of restrictions.25 Notwithstanding the criticism, harmonisation exceeded the lowest common denominator and modernised national asylum laws. By way of example, Germany and France embraced a generous reading of the refugee definition, including gender-specific considerations and persecution by non-state actors. Political scientists attribute this outcome to the depoliticised character of decision-making and the efforts of the main destination countries to project their established practices on the EU level.26 Widespread fear of a race to the bottom gave way to moderately liberal supranational legislation with sophisticated procedural safeguards mirroring the domestic practices of ‘experienced’ asylum destination countries like Germany or the Netherlands. Procedural safeguards, amongst others for vulnerable persons, were enhanced during the second phase of harmonisation. Negotiations took place between 2008 and 2013; they required a majority vote in the Council and the consent of the European Parliament.27 The second phase established today’s Qualification Directive 2011/95/EU, Asylum Procedures Directive 2013/ 32/ EU, Reception Conditions Directive 2013/ 33/ EU, and Dublin III Regulation (EU) No 604/2013. The legislature put an emphasis on replacing ‘minimum standards’ by a higher degree of uniformity, while recognising the need for effective implementation.28 The former Asylum Office, presented in Chapter 9.3 on agencies, was set up to improve administrative practices on the ground. Nevertheless, implementation deficits have haunted EU asylum policy from its inception, in particular in states with little previous experience in setting up asylum systems.

13.1.4 Policy crisis of 2015/16 Images of the dead toddler Alan Kurdi, German citizens welcoming refugees at Munich main station, and thousands marching over the fields of the Western Balkans spring to the mind of anyone who witnessed what is often called the ‘migration and asylum (policy) crisis’.29 To be sure, asylum policy had seen periods of difficulties, real or perceived, before and continues to be defined by recurring instability. Think of the early 1990s, seaborne arrivals in Italy and Spain from the mid-2000s onwards, and contemporary debates about search and rescue or pushback practices. Yet, the events of 2015/16 arguably constituted a turning point for EU asylum policy in terms of administrative challenges, conceptual 25 See Geoff Gilbert, ‘Is Europe Living Up to its Obligations to Refugees?’ (2004) 15 EJIL 963–87; and Catherine Teitgen-Colly, ‘The European Union and Asylum. An Illusion of Protection’ (2006) 43 CML Rev 43 1503. 26 See ch 2.2.3; and Natascha Zaun, ‘Why EU Asylum Standards Exceed the Lowest Common Denominator’ (2016) 23 J Eur Pub Pol 136. 27 See Patricia Van de Peer, ‘Negotiating the Second Generation of the Common European Asylum System Instruments’ in Vincent Chetail and others (eds), Reforming the Common European Asylum System (Brill 2016) 55. 28 See generally Commission, ‘Green Paper on the future of the Common European Asylum System’ COM(2007) 301 final; and Commission, ‘Communication: Policy plan on asylum’ COM(2008) 360 final. 29 On the use of language see ch 4.3.3; on the course of events see Agustín Menéndez, ‘The Refugee Crisis. Between Human Tragedy and Symptom of the Structural Crisis of European Integration’ (2016) 22 ELJ 388, 395– 403 https://en.wikipedia.org/wiki/European_migrant_crisis (accessed 1 March 2023).

Theory and Policy Design 343 repercussions, and political fallout. Migration became a highly contested topic, both domestically and supranationally, thus complicating, as we have seen, interinstitutional policy debates to the point where reform appears beyond reach. There are at least three components of the policy response30 that deserve our attention, besides cooperation with third states discussed elsewhere. First, numerous deaths at sea put the Mediterranean in the spotlight. Member States and EU institutions first reacted to devastating shipwrecks with a mixture of state-sponsored search and rescue, including the Italian operation ‘Mare Nostrum’ during 2013/14 and the follow-up missions ‘Sophia’ and ‘Irini’, which were coordinated by Frontex and retreated from search and rescue gradually.31 International and supranational legal standards for search and rescue were presented in Chapter 12 on border controls, and Chapter 16.3 on return will discuss the criminalisation of humanitarian assistance. Secondly, the situation at the external borders received much attention. ‘Hotspot’ became shorthand for the elusive desire for fast and fair asylum procedures and adequate reception conditions on the Greek islands and elsewhere. They have been discussed in Chapter 8.4.2 on the Asylum Office, which, together with Frontex, assumed enhanced operational responsibilities. Comments throughout this chapter will deal with reception conditions, (closed) accommodation centres, and the move towards border procedures. A widely known element of the policy response was the EU–Turkey Statement of 16 March 2016, which brought together domestic and international measures to ‘manage’ movements in the Aegean Sea.32 The international components of EU–Turkey cooperation will be presented in Chapter 18 on the external dimension, whereas comments on asylum border procedures and resettlement can be found below. Thirdly, the autumn of 2015/16 saw what was dubbed the ‘wave-through approach’33 along the ‘Western Balkans Route’; countries along that route tolerated or organised onward movements. Political criticism of the ‘wave-through approach’ epitomises an aspect of the policy debate that is often less visible: secondary movements. They will be touched upon in the section on Dublin that follows. Secondary movements subsided when borders were gradually closed during the first months of 2016;34 the camp in Idomeni near the Greek Northern border signalled the end of the ‘Western Balkans Route’.35 Border closures were the harbinger of things to come in the following years: allegations of pushbacks, discussed in Chapter 12. The policy debate saw—and continues to see—the confrontation of two competing narratives which render the search for compromise difficult if not impossible. While countries at the external borders complain about having to shoulder the ‘burden’ without meaningful solidarity, politicians further North often decry the alleged incapacity of peers to run 30 For the basic policy outline see Commission, ‘Communication: A European Agenda on Migration’ COM(2015) 240 final; and ‘Presidency Conclusion’ (Council doc 12002/15, 14 September 2015); they were developed further in numerous later communications and (European) Council conclusions. 31 See ch 18.3.4; and https://en.wikipedia.org/wiki/Operation_Mare_Nostrum (accessed 1 March 2023). 32 See ‘EU-Turkey Statement’ (Press Release 144/16, 18 March 2016); ch 3.3.1.1 on the absence of direct legal effects; and Andrea Ott, ‘EU–Turkey Cooperation in Migration Matters’ in Francesca Ippolito and others (eds), Bilateral Relations in the Mediterranean (Edward Elgar Publishing 2020) 184. 33 European Council, ‘Conclusions’ EUCO 1/16 (19 February 2016) No 8(d). 34 See ‘Declaration of the Conference “Managing Migration Together” ’ (Council doc 6481/ 16, 24 February 2016). 35 See ‘In Pictures: Life for Migrants in Idomeni Refugee Camp in Greece’ BBC.com (7 March 2016) https:// www.bbc.com/news/world-europe-35742796 (accessed 1 March 2023).

344 Common European Asylum System functioning asylum systems and to prevent secondary movements. The first narrative is fed by the well-known pictures of arrivals at the external borders and the structural asymmetry of the Dublin system, while the second narrative is sustained by long procedures, inacceptable reception conditions, and the lived experience of onward movements. EU institutions try to mitigate these tensions under the compromise formula of ‘solidarity and responsibility’, which takes up basic principles of primary law for mutual assistance (Article 80 TFEU) and loyal application of Union law (Article 4(3) TEU). In practical terms, policy reform will have to accommodate both aspects.

13.1.5 Elusive reform efforts EU institutions have embarked on a flurry activities to respond to the conceptual and practical difficulties of the CEAS. They are defined by a basic dichotomy: while cooperation with third states and financial or operational support yielded tangible results, discussed elsewhere, the reform of asylum legislation has caused protracted controversies, starting with the adoption of two Relocation Decisions in the summer of 2015. A third round of legislative harmonisation began in 2016 when the Commission presented proposals for the reform of the instruments discussed in this chapter.36 They will be mentioned in the thematic sections that follow. With the exception of Dublin, the proposals essentially perpetuated the structure of existing rules in a path-dependent manner, with more mandatory harmonisation and restrictions to streamline procedures.37 Debates proceeded reasonably well at a technical level within the Council, and the European Parliament adopted positions on all the legislative files.38 Nevertheless, no overall consensus was found, not least as a result of political stalemate over how to reform the Dublin system. The prevalent ‘package approach’ meant that nothing was agreed until everything was agreed. In September 2020, the Commission tried to reinvigorate reform efforts under the ‘New Pact on Migration and Asylum’.39 The Commission presented a legislative patchwork that can be difficult to keep an oversight of: some proposals of the 2016 package remained on the table (reception conditions, qualification criteria, resettlement); others required the combined reading of the original proposal and additional suggestions (asylum procedures, Eurodac); some were new or replace earlier proposals (screening, crisis and force majeure, Dublin was rebranded as ‘asylum and migration management’). They added up to hundreds of pages of legislative text and accompanying documents raising manifold questions of policy design and legal interpretation.40 Principled opposition of some countries to mandatory relocation was one factor complicating the search for an overall compromise, but by far

36 See generally Commission, ‘Communication: Towards a Reform of the Common European Asylum System’ COM(2016) 197 final; and Hugo Brady, ‘Openness versus Helplessness. Europe’s 2015–2017 Border Crisis’ in Tómas JoensenIan Taylor (ed), Small States and the European Migrant Crisis (Springer 2021) 67, 73–76. 37 For an overview see European Parliamentary Research Service, ‘Addressing migration in the European Union. Compendium of briefings’ (February 2017). 38 See the European Parliament’s ‘Legislative Train Schedule’ https://www.europarl.europa.eu/legislative-train/ theme-towards-a-new-policy-on-migration (accessed 1 March 2023). 39 See Commission, ‘Communication on a New Pact on Migration and Asylum’ COM(2020) 609 final. 40 See the open access publication Daniel Thym and Odysseus Academic Network (eds), Reforming the Common European Asylum System (Nomos 2022).

Theory and Policy Design 345 not the only one; tensions between northern and southern states were another factor why reform efforts did not get off the ground.41 At the time of writing, political debates in Brussels were approaching a sort of ‘end game’. European elections in the spring of 2024 entail that the negotiations will have to be concluded in February that year, at the latest, to leave time for technical redrafting and formal adoption in accordance with a ‘roadmap’ the Parliament and the Council have agreed upon.42 To move towards this objective, the institutions pursue a ‘gradual approach’; it is a classic negotiating strategy to break deadlock to identify elements on which a compromise might be reached. At the end of 2022, negotiations were provisionally concluded on resettlement, reception conditions, and the Qualification Directive, although formal adoption was delayed in light of the ‘package approach’. Moreover, Parliament and Council were making an effort to find a common internal position on other files to be able to start interinstitutional negotiations.43 In March 2023, a likely scenario was that an agreement could be reached on Eurodac, return, and screening as well (together with the Schengen Borders Code, which does not technically belong to the ‘asylum package’). All these instruments could then be adopted as a ‘mid-sized package’, thus allowing the relevant actors to show that the EU can deliver. By contrast, national governments struggled, to say the least, to agree on a common line, let alone an inter-institutional compromise, on asylum jurisdiction, asylum procedures (in particular border procedures), crisis and force majeure, as well as instrumentalisation. The odds were against a substantial reform of these core pieces of legislation. An important element of the reform package of 2020, which does not receive much attention, are better governance structures. The ‘Asylum and Migration Management Regulation’, the successor to the Dublin III Regulation, would establish a framework for early warning, reporting obligations, supranational oversight, financial and operational support, and emergency response.44 A ‘Crisis and Force Majeure Regulation’ would introduce the option of implementing acts triggering, amongst others, stricter time limits for asylum procedures and returns.45 A Recommendation is intended to render some of these governance structures immediately effective.46 Experience with similar efforts to improve Schengen governance, portrayed in Chapter 12.3.2 on border controls, show that practical success is far from guaranteed.

13.1.6 Geopolitics enter the political equation There had always been a contrast between the humanitarian vocation of refugee law and the political character of state response. A trigger for the Europeanisation of asylum policy was, as we have seen, the desire of the destination countries in Western and Northern Europe to reduce the number of arrivals after the fall of the iron curtain. State interests have always 41 See Daniel Thym, ‘Never-Ending Story?’ in Thym and Odysseus Academic Network (n 40) 11, 13–20. 42 See ‘Migration and Asylum: Roadmap on way forward’ (European Parliament Press Release, 7 September 2022). 43 See ‘Gradual approach on migration and asylum’ (Council doc 5393/22, 20 January 2022); and ‘Overview of the current legislative proposals’ (Council doc. 6838/23, 3 March 2023). 44 See Proposal for an Asylum and Migration Management Regulation COM(2020) 610 final, arts 3–6. 45 See Proposal for a Crisis and Force Majeure Regulation COM(2020) 613 final. 46 See Commission Recommendation (EU) 2020/1366 on a Migration Preparedness and Crisis Blueprint [2020] OJ C317/26.

346 Common European Asylum System influenced asylum policies. Nevertheless, recent years have seen an additional twist which gives greater weight to the political dimension. Destination countries in the Global North do not dominate the domestic and international policy agenda any longer. The externalisation of migration management, traced in Chapter 18.3, means that countries of origin and transit can use their gatekeeper function to their advantage. State interests and geopolitical considerations gain visibility as a result, often to the detriment of humanitarian objectives. Policy debates about asylum are not immune to the reorientation of world politics towards interest-based realism in recent years.47 Third states ask for a return on their ‘services’, be it financial support or political concessions in the domain of migration or beyond, such as silence on human rights abuses.48 Susceptibility to extortion had been a recurrent criticism of the EU–Turkey Statement, in particular, ever since the authoritarian turn of President Erdoğan after the attempted military coup of July 2016. The entry of geopolitics into asylum policy was raised to a new level when the Turkish President encouraged migrants to leave the country en masse in February 2020, in what appeared like a potential remake of the events during 2015/16. Greek policemen closed the border, and the Commission President praised them as the ‘shield’ of Europe.49 Morocco used similar tactics during May 2021 to put pressure on Spain not to insist on the independence of the Western Sahara any longer. The Spanish government gave in a few months later, mentioning ‘management of migratory flows’ as a core area of future cooperation.50 Later that year, the ‘instrumentalisation’ of migration by the Belarusian dictator resulted in months of violent clashes at the border with Lithuania and Poland. Chapter 10.7 on general features explained that Member States relied on Article 72 TFEU on all these occasions to justify restrictive practices. One step further, the Commission proposed an ‘Instrumentalisation Regulation’, whose adoption would permit lower standards on reception conditions and asylum procedures.51 In the summer of 2022, the Council was moving towards a common position, thus setting the scene for heated interinstitutional negotiations with the Parliament. There were only a few weeks between frontage news on the violent ‘pushbacks’ at the Polish-Belarusian border and the outbreak of the Russian war of aggression against Ukraine (possibly, the action of the Belarusian dictator was even conceived of as a prelude, in terms of causing European disunity, like during the mass influx of 2015/16). The war triggered the biggest and fastest refugee emergency Europe had experienced ever since the Second World War. More than five million people left Ukraine within weeks, several million more were internally displaced.52 The contrast to the standoff with Belarus was palpable: a wave of solidarity swept across the continent, and the Council activated the Temporary Protection Directive 2001/55/EC for the first time ever. After years of bitter disputes over reform efforts, Member States stood united in solidarity with Ukrainians. This time around, refugees

47 See Achilles Skordas, ‘The Rise of the Neo-Hobbesian Age’ (2019) 79 Heidelberg J Intl L 469. 48 See ch 18.1.4. 49 See ch 1.2.6. 50 See ‘Spain changes tune on Western Sahara’ AfricaNews.com (19 March 2022) https://www.africanews.com/ 2022/03/19/spain-changes-tune-on-western-sahara (accessed 1 March 2023). 51 See Proposal for an Instrumentalisation Regulation, COM(2021) 890 final. 52 See https://data.unhcr.org/en/situations/ukraine (accessed 1 March 2023).

Theory and Policy Design 347 reaped the benefit of geopolitics. Like during the cold war, the victims of the enemy were welcomed with open arms.

13.1.7 Structural deficits and their fallout EU asylum legislation has never worked particularly well in practice, and attempts at improving compliance have yielded mixed results. These protracted deficiencies indicate that the reform of asylum policy should think outside the box. Basic policy choices which have been continued in a path-dependent manner for more than two decades should not be sacrosanct. More of the same, or a focus on ‘better implementation’ of the existing rules, might not be enough in the long-run.53 The degree of practical difficulties indicates that something more fundamental needs to be done. We can identify at least seven overarching structural deficits. First, there are flagrant implementation deficits in core areas of asylum legislation, especially regarding reception conditions, return, and Dublin procedures. It is not enough to agree on a directive or regulation in Brussels and to assume that domestic authorities will ensure implementation in practice. That is why agencies, financial assistance, databases, and governance structures are so important. They will not bring about a brave new world of compliance single-handedly but may influence developments on the ground nonetheless. Doing so will also facilitate political negotiations, which, at present, are complicated by two-level games if governments consider both the contents of the proposals and their putative real-life functioning.54 Secondly, the profound solidarity deficit underlying the Dublin system is the best-known example of a legislative design weakness. Rules on asylum jurisdiction are structurally unfair and often disrespected in practice. Continued failure to reform the Dublin system means that an essential piece of legislation connecting national asylum systems remains practically and conceptually flawed. Some critics argue that the very idea of combining internal free movement with a stable allocation of asylum jurisdiction, underlying the original Schengen Convention, presents the ‘original sin’ at the heart of EU asylum policy.55 Other sectoral legislation has design deficits as well, even though their impact is less drastic than in the case of Dublin. Thirdly, statistics and comparative surveys demonstrate that reception conditions and recognition quotas differ markedly across the Union. The EU’s claim to a level playing field is undermined if harmonisation does not result in comparable outcomes. These entrenched discrepancies are one motivation, amongst others, for irregular secondary movements and undermine the claim to establish a ‘common’ area of protection. We shall see that the predominance of equal treatment with nationals in the asylum legislation perpetuates differences between the Member States, instead of reducing discrepancies with regard to living conditions. 53 See generally Daniel Thym, ‘The “Refugee Crisis” as a Challenge of Legal Design and Institutional Legitimacy’ (2016) 53 CML Rev 1545, 1554–58. 54 See Thym (n 41) 13–32. 55 See Evangelia (Lilian) Tsourdi and Cathryn Costello, ‘The Evolution of EU Law on Refugees and Asylum’ in Paul Craig and Gráinne de Búrca (eds), The Evolution of EU Law (3rd edn, OUP 2021) 793, 805–807; and Vincent Chetail, ‘The Common European Asylum System: Bric-à-brac of System?’ in Chetail and others (n 27) 3, 11.

348 Common European Asylum System Fourthly, EU asylum legislation resembles a ‘fair weather construction’:56 it looks reasonably good on paper but fails to deliver in challenging circumstances. Reliance on complex procedures and court oversight in literally thousands of cases has proven unrealistic, in particular in the geographical periphery or when the number of arrivals increases. One explanation for procedural complexity is the influence of the classic destination countries. They dominated the early policy debate not as political hegemons but as a result of profound expertise, thus effectively uploading their domestic models to Brussels.57 States in southern and, later, Eastern Europe took over sophisticated procedural patterns, which worked reasonably well in the West and North but proved not fit for purpose elsewhere.58 The first ever activation of the Temporary Protection Directive during 2022 may be perceived an act of liberation from procedural rigidity. Fifthly, observers have noticed a mismatch between the internal and external dimension. The increase in protection standards and procedural safeguards domestically coincided with the proliferation of non-arrival policies and the externalisation of migration control.59 Events during 2015/16 can be presented as a temporary collapse of the externalisation architecture,60 which was rebuilt even stronger in the years thereafter. Disagreement over how to reform the Dublin system arguably increased the appetite for externalisation, which reduces the need for politically toxic disputes about solidarity. Refugee protection resembles the proverbial hot potato: governments pass it around, until it falls to the floor. More legal pathways can be an instrument to balance the preference for externalisation, provided that the volumes involved are more than a humanitarian fig leave (something they have not usually been so far). Sixthly, the influence of human rights and international refugee law remains ambiguous. On the one hand, EU legislation embraced a broad definition of refugee status and subsidiary protection, thus enhancing the rights of refugees. On the other hand, Member States ‘outsourced’ migration management to countries of origin and transit. Critical academics castigate these attempts at evasion as ‘hyper-legalism’61 or ‘legal black holes’.62 Governments exploit the jurisdictional yardstick enshrined in human rights law to exercise control indirectly, by means of externalisation. Cooperation with third states can be presented as an effort to comply with the letter of the law, while contrasting with the alleged spirit. Finally, structural deficits may be a slippery slope towards partial disintegration, with Member States taking supranational legislation less and less seriously. Willingness to apply common standards may decline, as protracted allegations of ‘pushback’ practices and the continuation of internal border controls within the Schengen area illustrate. Some national measures were illegal, but they continued nonetheless.63 Informal arrangements of 56 Kay Hailbronner, ‘Asyl in Europa: wenn, wie, wann, wo?’ Frankfurter Allgemeine Zeitung (12 October 2015) 6 (own translation); see also Brunessen Bertrand, ‘De quelques paradoxes de la directive procédures’ [2016] Revue trimestrielle de droit européen 21. 57 See Natascha Zaun, EU Asylum Policies: The Power of Strong Regulating States (Palgrave 2017). 58 See Rosemary Byrne, Gregor Noll, and Jens Vedsted-Hansen, ‘Understanding the Crisis of Refugee Law’ (2020) 33 Leiden J Intl L 871, 881–88. 59 See Sandra Lavenex, ‘“Failing Forward” Towards Which Europe?’ (2018) 56 JCMS 1195, 1199–201. 60 See Maarten den Heijer, Jorrit Rijpma, and Thomas Spijkerboer, ‘Coercion, Prohibition, and Great Expectations’ (2016) 53 CML Rev 607, 618–23. 61 Daniel Ghezelbash, ‘Hyper-Legalism and Obfuscation’ (2020) 68 Am J Comp L 479. 62 Itamar Mann, ‘Maritime Legal Black Holes’ (2018) 29 EJIL 347. 63 See generally Menéndez (n 29) 408–13; and Iris Goldner Lang, ‘No Solidarity without Loyalty’ (2020) 22 EJML 39.

Constitutional Foundations 349 ‘coalitions of the willing’ complementing the Dublin system are, like any soft law instrument, double-edged: they might prepare the ground for policy reform or undermine the doctrinal force of existing laws.64

13.2 Constitutional Foundations Failure to reform asylum policy is the result of institutional blockage, not lack of competences for law-making (13.2.1). Important constitutional limitations emanate from human rights law, especially with regard to the prohibition of refoulement (13.2.2) and the right to asylum (13.2.3). These asylum-specific comments complement the procedural human rights aspects, which were discussed in Chapter 7.2 on the administrative dimension, and build upon the general features of human rights law, presented in Chapter 5. Article 80 TFEU calls for intra-European solidarity, although the provision will rarely result in a judiciable obligation of result (13.2.4). Finally, readers are reminded of the special status of Ireland and Denmark (13.2.5).

13.2.1 Supranational competences Article 78(2) TFEU was revised substantially by the Treaty of Lisbon, which established a broad range of competences for a ‘uniform status of asylum’, subsidiary protection, procedures, asylum jurisdiction, reception conditions, and cooperation with third states. These provisions allow, as we have seen, for the introduction of powerful databases and for a substantial upgrade of the Asylum Agency short of fully federalised administrative decision-making. When it comes to legislative harmonisation, primary law contains few caveats limiting the room for manoeuvre of the EU institutions. Failure to agree on policy reform results from political stalemate over the design of new instruments, not from lack of competences. The drafting history and the general scheme of the Treaties, including the reference to the Refugee Convention, indicate that a ‘status of asylum’ under Article 78(2)(a) TFEU matches the criteria of refugee status; the formulation does not hint at a protection status independent of international law.65 The Court of Justice of the European Union (CJEU) insists that asylum legislation must be interpreted in light of the Refugee Convention, even though the absence of authoritative interpretation of the Convention by an international court or UNHCR means that EU institutions have factual leeway when illuminating its meaning.66 Different options of how a ‘uniform status . . . valid throughout the Union’ might look like will be discussed in the section on the status after recognition. Article 78(2)(b) TFEU goes beyond the Refugee Convention by enshrining the notion of subsidiary protection in primarily law.67 The provision was modelled on the Qualification 64 See ch 2.3.4; and Luisa Marin, ‘Waiting (and Paying) for Godot’ (2020) 22 EJML 60. 65 See Daniel Thym, ‘Legal Framework for Asylum Policy’ in Daniel Thym and Kay Hailbronner (eds), EU Immigration and Asylum Law: Article-by-Article Commentary (3rd edn, CH Beck/Hart Publishing/Nomos 2022) MN 15. 66 See ch 5.6. 67 The previous EC Treaty, art 63(2)(a), as amended by the Treaty of Amsterdam, had referred, somewhat ambiguously, to persons who ‘otherwise’ are in need of protection; see Hailbronner (n 14) 81.

350 Common European Asylum System Directive, and yet the contours of subsidiary protection could be altered or amended in response to practical demands or political priorities.68 Some outer limits of the legislative discretion can be deduced from the notion of ‘international protection’, which responds to risks of a general or individual nature in countries of origin or transit.69 By way of example, Article 78(2)(b) TFEU could be used to extend subsidiary protection to ‘complementary’ humanitarian needs,70 whereas the regularisation of illegal stay independent of the situation in countries of origin or transit is covered by Article 79(2)(a) TFEU. Judgments defining the ‘rationale of international protection’ to exclude residence ‘on compassionate or humanitarian grounds’ in reaction to economic or social shortcomings in countries of origin concerned the Qualification Directive; they cannot, as a result, be projected onto primary law without qualification.71 Article 78(2)(c) TFEU with regard to ‘temporary protection’ for ‘displaced persons’ in ‘the event of a massive inflow’ provides additional leeway beyond the contents of the Temporary Protection Directive 2001/55/EC. Along similar lines, Article 78(2)(e) TFEU serves as a legal basis of the Dublin system, even though the abstract reference to substantive ‘criteria’ and procedural ‘mechanisms’ indicates that the legislature can modify existing rules or introduce a quota system. Primary law also permits parallel mechanisms for distinct scenarios. Careful wording of ‘common’ (not ‘uniform’ or ‘single’) asylum procedures indicates that the level of harmonisation will usually be lower than for qualification criteria. We shall see that primary law is silent on the geographic location of asylum procedures, thus authorising external processing. For the same reason, Article 78(2)(d) TFEU covers resettlement or humanitarian visas.72 None of these options has resulted in binding legislation so far. Civil wars or other crises can give rise to in ‘one or more Member States being confronted by an emergency situation characterised by a sudden inflow’.73 In such scenarios, Article 78(3) TFEU authorises the Council to adopt ‘provisional measures’ by qualified majority,74 as it did with the Relocation Decisions in September 2015. Judges emphasised that the Council has leeway when determining for how long they shall apply; relocation during a 24- month period was found to be ‘provisional’.75 Institutions also have flexibility when deciding whether an ‘emergency situation’ exists and regarding the causal nexus with the measures taken.76 Crucially, the Court confirmed that ‘provisional measures’ may amend secondary legislation outside the ordinary legislative procedure.77 Whereas Article 78(3) TFEU concerns actions by the supranational institutions, Member States may possibly rely on Article 72 TFEU to justify non-compliance, as discussed in Chapter 10.7 on general features. An

68 See Thym (n 65) MN 20; and Gerhard Muzak, ‘Artikel 78 AEUV’ in Thomas Jaeger and Karl Stöger (eds), Kommentar zu EUV und AEUV (141th update, Manz 2012) MN 23–24. 69 See Hemme Battjes, European Asylum Law and International Law (Martinus Nijhoff 2006) 541–42. 70 ibid 544–47; and Julia Schieber, Komplementärer Schutz (Nomos 2013) 303–10. 71 See Case C-542/13 M’Bodj EU:C:2014:2452, paras 35–37, 44. 72 See ch 11.5; and Catharina Ziebritzki, ‘The Objective of Resettlement in an EU Constitutional Perspective’ in Marie-Claire Foblets and Luc Leboeuf (eds), Humanitarian Admission to Europe (Nomos/Hart Publishing 2020) 285, 319–23. 73 Similarly, the earlier EC Treaty, art 64(2), as amended by the Treaty of Amsterdam. 74 cf TEU, art 16(3). 75 See Joined Cases C-643/15 and C-647/15 Slovakia and Hungary v Council EU:C:2017:631, paras 78, 89–103. 76 ibid paras 113–28. 77 ibid paras 70–77.

Constitutional Foundations 351 emergency measure to respond the ‘instrumentalisation’ of migration by Belarus was not adopted, as Poland considered it to be insufficient.78

13.2.2 Prohibition of refoulement Human rights have replaced the Refugee Convention as the focal point of refugee protection, largely as the result of institutional muscle of international courts and treaty bodies. Chapter 5.3 and 5.6 illustrated that the ECtHR has traditionally assumed a lead function, not only in Europe. Judgments on Article 3 European Convention on Human Rights (ECHR) exceed the level of protection under the Refugee Convention and guide the interpretation of Articles 4 and 19(2) Charter of Fundamental Rights (CFR). Comments that follow will concentrate on guarantees for third country nationals staying on the territory of the Member States already, since the extraterritorial reach of human rights and procedural guarantees against rejection at the border were discussed in the previous chapter. Standards and burden of proof for status determination will be mentioned in the context of the refugee definition. Article 3 ECHR applies a different benchmark than the notion of ‘persecution’ under the Refugee Convention. Over the years, the ECtHR developed ever finer standards, mirroring the criteria for status determination under the Qualification Directive.79 To start with, judges construe Article 3 ECHR as an absolute guarantee covering anyone who, despite being persecuted, is excluded from refugee status.80 Interpretation of the Charter in light of the European Convention involves that the general limitation clause in Article 52(1) CFR does not apply to Articles 4 and 19(2) CFR. Return may be justified whenever the country of destination gives diplomatic assurances against inhuman treatment, including procedures certifying practical compliance.81 The Court recognised that Article 3 ECHR protects against ill-treatment emanating from non-state actors if state authorities are unable or unwilling to provide protection,82 and it requires individuals to avail themselves of internal protection alternatives.83 Crucially, the ECtHR also embraced situations of indiscriminate violence, mirroring subsidiary protection under Article 15(c) Qualification Directive 2011/95/EU. After some far-reaching judgments, which had met the opposition of governments and domestic courts, the ECtHR stressed that ‘a general situation of violence will not normally in itself entail a violation of Article 3’, since such a conclusion is warranted ‘only in the most extreme cases’.84 During the 2010s, it carefully applied these standards to different countries: return to Afghanistan required a case- by- case assessment;85 parts of Somalia and Iraq were considered 78 See Proposal for a Decision on provisional emergency measures for the benefit of Latvia, Lithuania and Poland COM(2021) 752 final. 79 See Marie- Bénédicte Dembour, When Humans Become Migrants (OUP 2015) ch 7; and Andreas Zimmermann and Björn Elberling, ‘Ausweisungsschutz’ in Oliver Dörr and others (eds), Konkordanzkommentar EMRK/GG (2nd edn, Mohr Siebeck 2013) ch 27. 80 Contrast Refugee Convention, arts 1F, 33(2) with Saadi v Italy App no 37201/06 (ECtHR [GC], 28 February 2008) §§ 124–27. 81 See Othman (Abu Qatada) v United Kingdom App no 8139/09 (ECtHR, 17 January 2012) §§ 187–89. 82 This has been settled case law since HLR v France App no 24573/94 (ECtHR [GC], 29 April 1997) §§ 32, 40. 83 See Salah Sheekh v Netherlands App no 1948/04 (ECtHR, 11 January 2007) § 141. 84 NA v United Kingdom App no 25904/07 (ECtHR, 17 July 2008) §§ 114–15. 85 See AM v Netherlands App no 29094/09 (ECtHR, 5 July 2016).

352 Common European Asylum System safe;86 Syria, by contrast, was generally unsafe.87 Future scenarios will similarly have to distinguish between the absence of widespread indiscriminate violence, the opposite finding of general unsafety, and case-by-case assessments of the country as a whole, or parts thereof. Individual factors may be combined with general considerations,88 mirroring the CJEU’s ‘sliding scale’ approach to subsidiary protection discussed below. Notwithstanding this overlap, there may be ancillary discrepancies between human rights and the Qualification Directive, for instance with regard to internal protection alternatives or cessation. In the absence of general violence, a real risk of individual mistreatment may still prohibit return under Article 3 ECHR. In cases concerning the Dublin system, the ECtHR found that a situation of ‘extreme material poverty’89 can amount to a violation of Article 3 ECHR, thus extending non- refoulement to socio-economic living conditions. Unfortunately, the bearing of the poverty test remains uncertain. So far, it has only been applied to transfers within the Dublin system and the reception of asylum seekers in the Member States, reflecting the consensus expressed in Directive 2013/33/EU.90 By contrast, return to third states remained subject to stricter standards. The CJEU followed the ECtHR in judgments on the Dublin system and reception conditions. Articles 4 CFR prohibits intra-European transfers in situations of extreme material poverty where an individual cannot ‘meet his most basic needs, such as, inter alia, food, personal hygiene and a place to live’.91 The additional reference, on the part of the CJEU, to ‘a state of degradation incompatible with human dignity’ might possibly be used to broaden the protective reach to third states. When it comes to third states, the ECtHR assumes that socio-economic living conditions will prevent deportation only ‘in very exceptional cases’, when the situation would deteriorate massively, including the danger of imminent death or a serious, rapid, and irreversible decline in the state of health resulting in intense suffering upon return—a standard the CJEU has followed explicitly, thus refusing the idea of a higher level of protection under Union law.92 Individuals cannot ‘in principle claim any entitlement to remain in the territory of a Contracting State in order to continue to benefit from medical, social, or other forms of assistance and services’, even if the standard of living in the country of origin is much lower.93 It remains an open question whether ‘non-removable’ third country nationals staying on Union territory can rely, in disputes with Member States about living conditions within Europe, on the higher benchmark for asylum seekers, as a particularly

86 See MYH and others v Sweden App no 50859/10 (ECtHR, 27 June 2013) §§ 62–67; and Sufi and Elmi v United Kingdom App nos 8319/07 and 11449/07 (ECtHR, 28 June 2011) §§ 212ff. 87 See SK v Russia App no 52722/15 (ECtHR, 14 February 2017). 88 See NA v United Kingdom (n 84) §§ 130, 143–44, 146. 89 MSS v Belgium and Greece App no 30696/09 (ECtHR [GC], 21 January 2011) § 252. 90 ibid §§ 250–53; Tarakhel v Switzerland App no 29217/12 (ECtHR [GC], 4 November 2014) §§ 93–99; and NH and others v France App nos 28820/13, 75547/13 and 13114/15 (ECtHR, 2 July 2020) §§ 158–64; and ch 5.3.3. 91 Case C-163/17 Jawo EU:C:2019:218, para 92, referring to the ECtHR’s MSS judgment; and also Case C-233/ 18 Haqbin EU:C:2019:956, paras 46–47. 92 See Paposhvili v Belgium App no 41738/10 (ECtHR [GC], 13 December 2016) §§ 172–93; on procedural obligations, Savran v Denmark App no 57467/15 (ECtHR [GC], December 2021) § 130; and Case C-69/21 Staatssecretaris van Justitie en Veiligheid EU:C:2022:913, paras 60–72. 93 Paposhvili v Belgium (n 92) §§ 176, 188; Staatssecretaris van Justitie en Veiligheid (n 92) paras 90–99 for Article 8 ECHR; and Vladislava Stoyanova, ‘How Exceptional Must “Very Exceptional” Be?’ (2017) 29 IJRL 580.

Constitutional Foundations 353 vulnerable group, or on the lower level of protection guiding the assessment whether return to a third state is legal.94 Other human rights may also bring about a prohibition of refoulement. However, unlike in the case of Article 3 ECHR, only the ‘flagrant denial’ of essential guarantees will prevent return. By way of example, extreme unfairness of judicial proceedings or dreadful detention conditions may give rise to an issue under Articles 5 and 6 ECHR, especially in extradition cases.95 A comparable, though doctrinally distinct, test applies under the Charter. The CJEU has recourse to the ‘essence of rights’ to determine whether treatment abroad hinders extradition.96 For asylum law, Article 3 ECHR remains the central yardstick.

13.2.3 Right to asylum A doctrinal distinction, which is not always acknowledged, concerns overlap and distinction of the prohibition of refoulement and the right to asylum. As a matter of principle, neither the Refugee Convention nor the ECHR ‘protect, as such, the right to asylum’,97 in the habitual reading at least.98 References to a right to ‘seek’ asylum, including in the practice of UNHCR, retain a deliberate vagueness from a legal perspective. They stay short of a legally enforceable guarantee to ‘receive’ an asylum procedure and to ‘be granted’ a protection status,99 which also the ‘right to seek and to enjoy’ asylum in the non-binding Universal Declaration of Human Rights does not embrace.100 Article 18 CFR may potentially go further.

13.2.3.1 Article 18 CFR While the prohibition of refoulement obliges states not to return anyone to unsafe territories, the meaning of the right of asylum is rarely illustrated. Core components are generally understood to embrace access to status determination procedures, a provisional right to stay, and, in case of recognition, a set of socio-economic guarantees.101 Status after recognition features prominently in Articles 12–24 Refugee Convention, not, however, in the ECHR, which is mainly concerned with non-refoulement. In this overall context, the meaning of the Article 18 CFR defies easy classification as a result of arguments, during the drafting process, about the possible creation of an individual guarantee transcending

94 See ch 16.5.5; Fabian Lutz, ‘Return Directive 2008/115/EC’ in Thym and Hailbronner (n 65) Article 14 MN 3–10; and Lieneke Slingenberg, ‘The Right Not to be Dominated. The Case Law of the European Court of Human Rights on Migrants’ Destitution’ (2019) 19 Hum Rts L Rev 291. 95 See Othman (Abu Qatada) v United Kingdom (n 81) §§ 231–35, 258–62. 96 Case C-216/18 PPU Minister for Justice and Equality EU:C:2018:589, paras 59–60 did not follow the position of the Advocate General. 97 ND and NT v Spain App nos 8675/15 and 8697/15 (ECtHR [GC], 13 February 2020) § 188; asylum was excluded from the scope of the original ECHR and later Protocols; see ‘Report on the Work of the Council of Europe in the Field of Human Rights’ (1962) 5 Yearb ECHR 42. 98 See Atle Grahl-Madsen, The Status of Refugees in International Law, vol 2 (AW Sijthoff 1972) §§ 161–82; and the dynamic proposal by William Thomas Worster, ‘The Contemporary International Law Status of the Right to Receive Asylum’ (2014) 26 IJRL 477. 99 See Alice Edwards, ‘Human Rights, Refugees, and the Right “to Enjoy” Asylum’ (2005) 17 IJRL 17 293, 302–304. 100 The formulation ‘and to enjoy’ was a deliberate choice to highlight state discretion; see Felice Morgenstern, ‘The Right of Asylum’ (1949) 26 Brit Yearb Intl L 327, 336–37. 101 See UNHCR ExCom, ‘Safeguarding Asylum’ (Conclusion No 82 (XLVIII), October 1997) Pt B.

354 Common European Asylum System the Refugee Convention.102 Several doctrinal ambiguities define the interpretation of the open-ended invocation of a ‘right to asylum . . . with due regard for the rules of the Refugee Convention . . . and in accordance with the [TFEU]’. Some authors conclude that the provision contains an individual guarantee, since the wording speaks of a ‘right’ and since the absence of the right-bearer may be rationalised by disputes about Union citizens.103 Others maintain that the reference to the Refugee Convention refutes an individual guarantee, which the latter does not contain. Moreover, the additional submission ‘with due regard’ (French dans le respect, German nach Maßgabe) to the EU Treaties can be relied upon to conceive of Article 18 as a principle, which, unlike a fundamental right, cannot be applied independent of secondary legislation in accordance with Article 52(5) CFR.104 Even if the provision was found to contain an individual right, the scope would have to be ascertained: either access to status determination for anyone irrespective of a real risk of refoulement and/or the promise of status rights after a positive decision. In practical terms, none of this has been relevant so far for the simple reason that Article 3 Asylum Procedures Directive 2013/32/EU and Articles 20–35 Qualification Directive 2011/95/EU embrace both guarantees. In a sign of pragmatism, the CJEU has focused on secondary legislation. Judges mentioned Article 18 CFR in several rulings but evaded principled statements about its meaning. They found that the provision does not prevent transfers under the Dublin system, even if procedural guarantees are violated.105 Similarly, asylum seekers cannot choose the country where to lodge a claim, thus indicating that the EU could abandon the option to apply for asylum several times.106 At the same time, the Court emphasised that secondary legislation rendered more effective the right to asylum,107 thus reinforcing the contents of the Refugee Convention.108 A judgment censuring Lithuania for denying access to status determination to anyone staying irregularly came close to recognising an individual human right to receive an asylum procedure irrespective of the likelihood of recognition, although the indiscriminate amalgamation of legislative and human rights considerations could also be rationalised as an interpretation of the Asylum Procedures Directive in light of a constitutional principle in accordance with Article 52(5) CFR.109

13.2.3.2 Protocol on Union citizens A peculiarity is the Protocol [No 24] on Asylum for Nationals of Member States of the European Union, which was introduced first by the Treaty of Amsterdam to accommodate 102 See Norbert Bernsdorff, ‘Artikel 18’ in Jürgen Meyer and Sven Hölscheidt (eds), Die Charta der Grundrechte der Europäischen Union (5th edn, Nomos 2019). 103 See María-Teresa Gil-Bazo, ‘The Charter of Fundamental Rights of the European Union and the Right to be Granted Asylum in the Union’s Law’ (2008) 27 Refugee Surv Q 33, 37–45; and Violeta Moreno-Lax, Accessing Asylum in Europe (OUP 2017) 363–75. 104 See Martijn den Heijer, ‘Article 18’ in Steve Peers and others (eds), The EU Charter of Fundamental Rights: A Commentary (2nd edn, CH Beck/Hart Publishing/Nomos 2021) MN 25–50; and Salvatore Fabio Nicolosi, ‘Going Unnoticed? Diagnosing the Right to Asylum in the Charter of Fundamental Rights’ (2017) 23 ELJ 94, 98–105. 105 See Joined Cases C-297/17, C-318/17, C-319/17 and C-438/17 Ibrahim and others EU:C:2019:219, para 99. 106 See Slovakia and Hungary v Council (n 75) para 343. 107 See Case C-36/20 PPU Ministerio Fiscal EU:C:2020:495, para 82; Case C-821/19 Commission v Hungary EU:C:2021:930, para 99; and Joined Cases C- 924/ 19 PPU and C- 925/ 19 PPU Országos Idegenrendészeti Főigazgatóság Dél-alföldi Regionális Igazgatóság EU:C:2020:367, paras 192, 196–97, which could be read to hint at the interpretation of secondary legislation in light of the Charter in line with CFR, art 52(5). 108 See Joined Cases C-391/16, C-77/17 and C-78/17 M and others EU:C:2019:403, paras 73–75, 100, 111–12. 109 See Case C-72/22 PPU Valstybės sienos apsaugos tarnyba EU:C:2022:505, paras 58–64.

Constitutional Foundations 355 Spanish concerns about Belgian courts refusing to extradite Basque terrorist suspects.110 Domestic authorities apply national legislation in such scenarios, since both Article 78(2) TFEU and the Asylum Procedures Directive concern third country nationals only. The sole operative article states that Member States ‘shall be regarded as constituting safe countries of origin in respect of each other for all legal and practical purposes in relation to asylum matters’.111 This designation has the rank of primary law, as confirmed by Article 51 TEU, and benefits from supremacy over national legislation in cases of conflict, including in Ireland and Denmark. Having said this, the Protocol does not outlaw asylum applications. It only requires domestic authorities to examine them ‘on the basis of the presumption that [they are] manifestly unfounded without affecting in any way . . . the decision-making power of the Member State’.112 This wording is open enough to comply with the prohibition of refoulement under human rights law and the Charter. Domestic courts may consult the Court of Justice on how to interpret primary law in these scenarios. In practice, the Protocol has gained limited relevance after the accession of countries like Bulgaria and Romania, whose nationals had often applied for asylum beforehand.113 Debates about the extradition of the leaders of the Catalan independence movement after the—unsuccessful—secession of 2017 concentrated on the immunity of MEPs and extradition for criminal charges under the European Arrest Warrant or domestic laws.

13.2.4 Solidarity (Article 80 TFEU) Solidarity has defined debates about EU asylum policy for years. The European Convention drafting the erstwhile Constitutional Treaty responded to these concerns by introducing Article 80 TFEU, which builds on a weaker earlier provision.114 Doing so buttressed pleas for solidarity in The Hague Programme and the Stockholm Programme and echoes the call for inter-state cooperation in Recital 4 Refugee Convention.115 Similar provisions exist for other policy areas.116 Attempts to translate these calls for solidarity into specific policy designs failed repeatedly, not only in migratory matters.117 Among the reasons explaining that outcome are doctrinal ambiguities and the theoretical open-endedness of the concept of inter-state solidarity.118 Comparable to ‘justice’ or ‘democracy’, solidarity is an essentially contested concept. States agree on the desirability in the abstract but disagree on what it means in practice.

110 See Violeta Moreno- Lax and Madeline Garlick, ‘Qualification’ in Steve Peers and others (eds), EU Immigration and Asylum Law (Text and Commentary), vol 3 (2nd edn, Brill Nijhoff 2015) 65, 80–81. 111 [2008] OJ C115/305. 112 See further Daniel Fröhlich, Das Asylrecht im Rahmen des Unionsrechts (Mohr Siebeck 2011) 280–98. 113 See Karin Zwaan, ‘The Aznar Protocol’ in Paul Minderhoud and others (eds), Caught in Between Borders (Wolf 2019) 303, 308–10. 114 EC Treaty, art 63(2)(b), as amended by the Treaty of Amsterdam, authorised EU institutions to ‘promot[e]‌a balance of effort between Member States’. 115 TFEU, art 80 was proposed first by the former Italian Foreign Minister Gianfranco Fini; see Presidium of the Convention, ‘Area of freedom, security and justice’ (CONV 614/03, 14 March 2003) 25. 116 See TFEU, arts 122(1), 194(1), 222. 117 See Daniel Thym and Evangelina (Lilian) Tsourdi, ‘Searching for Solidarity in the EU Asylum and Border Policies’ (2017) 24 MJECL 605, 606–609; and Esin Küçük, ‘Solidarity in EU Law’ (2016) 23 MJECL 965. 118 See Andrea Sangiovanni, ‘Solidarity in the European Union’ (2013) 33 OJLS 213.

356 Common European Asylum System Article 80 TFEU establishes, unlike the predecessor provision, no freestanding legal basis for the adoption of measures promoting solidarity.119 Rather, the wording refers to ‘Union acts adopted pursuant to this Chapter’, thereby indicating that Article 80 TFEU establishes a horizontal proviso guiding the interpretation of Articles 77–79 TFEU. In doing so, Article 80 TFEU creates a legal obligations of the EU institutions, which are mentioned explicitly, and for the Member States, when read in conjunction with Article 4(3) TEU.120 The Court confirmed that the Treaty provisions on solidarity are legally binding.121 Having said this, the institutions have, like in the case of Treaty objectives, a broad margin of appreciation when deciding which measures are ‘appropriate’; the open-ended wording, the practical diversity of potential solidarity measures, and the theoretical open-endedness support this conclusion.122 While Article 80 TFEU concerns solidarity among the Member States, wider commitments towards third states or individuals can emanate from the human rights law or foreign policy objectives. In practice, the institutions may opt for and have established various forms of financial, logistical, operational, or legislative support, in line with the classic typology of ‘sharing money’, ‘sharing people’, and ‘sharing norms’ (we might add ‘sharing resources’, for instance via the agencies).123 While some instruments are of a general character (eg training tools for civil servants developed by the Asylum Agency), others relate to specific scenarios (eg asylum support teams). Note that the relocation of asylum seekers or the free movement of beneficiaries of international protection are an option amongst others. Nothing in the Treaties indicates preference for ‘sharing people’, which, unlike financial assistance, is not mentioned. The abstract character entails that there are, in regular circumstances at least, no judiciable standards that the EU institutions or national governments must opt for a specific solution, although they are legally obliged to respect Article 80 TFEU as a matter of constitutional law. Seminal judgments on the events during 2015/16 reaffirmed that finding. Judges generally reminded the institutions and the Member States to act in a ‘spirit of solidarity’.124 However, it was political actors the Court called upon to develop the ‘spirit’ of solidarity—in striking contrast to earlier invocations of the ‘spirit’ in cases such as Van Gend en Loos or Les Verts, which had paved the way for constitutional innovations on the part of judges.125 They did not follow the proposal of an Advocate General, or academic authors, that Article 80 TFEU should be operationalised to develop judiciable standards, which might, for instance, result in an annulment of the Dublin III Regulation or a legal obligation to activate the discretionary clause.126 119 See Steve Peers, EU Justice and Home Affairs Law, vol I (4th edn, OUP 2016) 244; and Jürgen Bast, ‘Deepening Supranational Integration’ (2016) 22 EPL 289, 291–94; contra Antonio Caiola, ‘Une base juridique pour la solidarité’ (2018) 54 Cahiers de droit européen 437, 458–94. 120 See Case C-848/19 P Germany v Poland EU:C:2021:598, para 41. 121 ibid paras 42–45. 122 See Thym and Tsourdi (n 117) 609–12. 123 See Gregor Noll, Negotiating Asylum (Martinus Nijhoff 2000) 267–74. 124 Case C-646/16 Jafari EU:C:2017:586, para 100; similarly Slovakia and Hungary v Council (n 75) paras 251– 53; and Joined Cases C-715/17, C-718/17 and C-719/17 Commission v Poland and others EU:C:2020:257, paras 97, 180–81. 125 See Daniel Thym, ‘Between “Administrative Mindset” and “Constitutional Imagination”’ (2019) 44 EL Rev 138, 153–57. 126 cf AG Yves Bot, Opinion in Joined Cases C-643/15 and C-647/15 Slovakia and Hungary v Council EU:C:2017:618, points 16–24, 242; and also Esin Küçük, ‘The Principle of Solidarity and Fairness’ (2016) 22 ELJ 448, 454–63.

Asylum Jurisdiction 357 Table 13.1 Participation in core CEAS legislation Instrument

Ireland/UK

Denmark

Schengen acquis? (NO, IS, CH. LI)

Temporary Protection Directive 2001/55/EC

yesa/yes

no

no

Asylum Qualification Directive 2011/95/EU

no (yes)b

no

no

Asylum Procedures Directive 2013/32/EU

no (yes)c

no

no

Asylum Reception Conditions Directive 2013/32/EU

no (yes, for the UKd)

no

no

Dublin III Regulation (EU) No 604/2013

yes

yes (int. law)

yes

Relocation Decisions (EU) 2015/ 1523 & (EU) 2015/1601

no

no

no

Temporary Protection Implementing yes Decision (EU) 2022/382

no

no

Proposal for a Resettlement Framework Regulation

no

no

tbd

aIreland joined on the basis of Commission Decision 2003/690/EC [2003] OJ L251/23. bIreland (and the UK, before Brexit) continues to apply the former Asylum Qualification Directive 2004/83/EC. cIreland (and the UK, before Brexit) continues to apply the former Asylum Procedures Directive 2005/85/EC. dThe UK—unlike Ireland—had participated in the former Asylum Reception Conditions Directive 2003/9/EC.

13.2.5 Territorial scope In accordance with the opt-outs presented in Chapter 2.4, Ireland and Denmark are not obliged to participate in asylum legislation. Nevertheless, Table 13.1 illustrates that Ireland (and the UK, before Brexit) partook in the first phase of legislative harmonisation, while abstaining from follow-up instruments. As a result, the directives agreed upon during the period from 2003 to 2005 continue to apply on the Emerald Isle. In a similar twist of unnecessary complexity, Denmark is associated with the Dublin system and the Eurodac Regulation on the basis of public international law, mirroring the status of the associated countries of Norway, Iceland, Switzerland, and Liechtenstein. Since none of the asylum instruments qualifies as belonging to the Schengen acquis, new Member States participate unconditionally from day one after accession. A separate notion of Dublin acquis, mentioned in Chapter 12.3.3 on border controls, determines the scope of the association agreements; they cover any future amendment of the rules on asylum jurisdiction and Eurodac.

13.3 Asylum Jurisdiction Rules on asylum jurisdiction are essential, since they connect national asylum systems horizontally. They were the historic starting point of Europeanisation under the

358 Common European Asylum System Schengen and Dublin Conventions, whose standards had emerged in a different geopolitical context than the one we live in today (13.3.1). Policy debates ever since have focused, to little avail, on inter-state solidarity (13.3.2). Widely known are the substantive criteria for state jurisdiction in the Dublin III Regulation (EU) No 604/2013 (13.3.3). Less visible but practically as relevant are the limited scope ratione materiae, which does not cover all forcibly displaced migrants at present (13.3.4), and legal remedies against transfer decisions (13.3.5). Judgments limiting Dublin transfers on human rights grounds have received much attention (13.3.6), whereas the driving forces and legal effects of secondary movements are rarely discussed (13.3.7). Comments on Eurodac can be found in Chapter 9 on databases. The Commission has tabled a proposal for a future ‘Asylum and Migration Management Regulation’ in September 2020 whose adoption was highly unlikely at the time of writing; its contents will be integrated into our analysis whenever appropriate.

13.3.1 Origin of the first entry rule Jurisdictional rules in today’s Dublin III Regulation reiterate the model set by the original Schengen cooperation, which were replicated by the Dublin Convention among the twelve Member States of the European Economic Community.127 Similar rules also featured in the draft text of a pan-European agreement on asylum jurisdiction, examined in the framework of the Council of Europe during 1988/89,128 which failed to must widespread political support as a result of protracted differences of interests among national governments.129 Debates among a smaller group of like-minded countries within the original Schengen area proved to be productive in terms of establishing a regulatory standard which suited the interests of the main destination countries. The exclusion of Italy, in particular, meant that the government in Rome was effectively forced to take over the template set among the inner Schengen circle when it joined the Schengen area.130 Negotiations on the Dublin Convention were swift and eventually reiterated a compromise which defines EU asylum law to date: the residual jurisdiction of the state of first entry was combined with the transfer of jurisdiction in scenarios of secondary movements.131 I am unaware of any archival study on the origin of the first entry rule during negotiations on the Schengen Conventions among France, Germany, and the Benelux countries in the late 1980s, where this compromise formula was developed first. One plausible explanation concerns a path-dependent variation on the regulatory regime for border-free

127 See CISA, arts 28–38; and Dublin Convention (n 20) arts 4–9; and, on Dublin replicating Schengen, Agnes Hurwitz, The Collective Responsibility of States to Protect Refugees (OUP 2009) 35. 128 A ‘Draft European Agreement on Responsibility for Examining Asylum Applications’ was discussed by the Ad Hoc Committee of Experts on the Legal Aspects of Territorial Asylum, Refugees and Stateless Persons (CAHAR); it is reproduced in Hailbronner (n 15) 28–29, Annex II. 129 See Gaia Lott, ‘The Dublin Convention and the Introduction of the “First Entry Rule”’ (2022/23) 31/32 Contemp Eur Hist 1. 130 See Simone Paoli, ‘The Schengen Agreements and their Impact on Euro-Mediterranean Relations’ (2015) 21 J Eur Integr Hist 125. 131 See CISA, arts 30(1)(e), 31(3); Dublin Convention (n 20) arts 7(1), 11(1); and Lott (n 129).

Asylum Jurisdiction 359 travel among the Benelux countries.132 Implementing decisions of the Benelux rulebook specified that countries that had issued a visa or residence permit were obliged to take back ‘unwanted’ foreigners, as were countries across whose external borders someone had entered irregularly; there were inter-state consultation requirements with time-limits and documentary evidence, which resemble the basic contours of the contemporary Dublin III Regulation (EU) No 604/2013.133 Similar rules existed for the Nordic Passport Union.134 To be sure, these provisions concerned take back after expulsion, but they might have served as a template for asylum seekers nonetheless. Be that as it may, the factual context is as relevant as the lost story of the Dublin criteria: they were developed for a different world from the one we live in today. Negotiations on Schengen were completed during the summer of 1989 when few people would have predicted that the iron curtain would fall a few months later. The Eastern border was hermetically sealed, and there were few irregular arrivals from the South. Many asylum applicants entered regularly at the time, in the absence of visa requirements or on other legal bases.135 As a result, the first entry rule, according to which countries at the external border have to take back those entering irregularly, was—unlike today—not the most relevant criterion when negotiations took place. Signature was postponed once German reunification became within reach and took place a few months later on the basis of what had been agreed upon previously.136 European states sleepwalked into structurally biased rules on asylum jurisdiction. Ratification proved cumbersome but was not blocked indefinitely. One factor facilitating ratification might have been that Germany was the central gateway to both the original Schengen area and the twelve state parties of the Dublin Convention in the early 1990s, when most asylum seekers came from or via South East Europe.137 Throughout the 1990s, the German government made repeated attempts at convincing European partners to sign up to ‘burden-sharing’, echoing the example of its domestic distribution key based on GDP and population size. These initiatives were discussed under the heading of ‘temporary protection’ at the time, in response to persons fleeing the wars in the former Yugoslavia. Discussions were cumbersome and dragged on for years.138 The end result did not mention quotas and went as far as recognising other ‘solidarity contributions’, such as military crisis prevention (an early predecessor of contemporary ‘flexible solidarity’).139 Follow-up proposals by the Commission were not adopted. The Temporary

132 See Daniel C Turack, ‘Freedom of Movement and the Travel Document in Benelux’ (1968) 17 ICLQ 191, 199–206. 133 See Decision of the Working Party on the Movement of Persons regarding the expulsion and transfer of persons, arts 1–2, 4–7, 11, 13–16 (M/P (67) 1, 28 June 1967), published in the Official Journal of the Kingdom of the Netherlands (Tractatenblad van het Koninkrijk der Nederlanden) [1978] No 171/6 https://zoek.officielebekendm akingen.nl/trb-1978-171 (accessed 1 March 2023). 134 See the first entry rule in Agreement for the Readmittance of Aliens who Have Illegally Entered the Territory of Another Contracting Party (adopted 14 July 1952, entered into force 12 July 1951 retroactively) 198 UNTS 47, art 2; and Daniel C Turack, ‘The Scandinavian Passport Union’ (1968) 38 Nordic J Intl L 171. 135 See den Heijer, Rijpma, and Spijkerboer (n 60) 618–23. 136 See Angela Siebold, ZwischenGrenzen (Ferdinand Schöningh 2013) 50–56. 137 See (German) Federal Government (Bundesregierung), ‘Migrationsbericht’ (2001) 36–47 https://www.bund esregierung.de/breg-de/service/archiv/alt-inhalte/migrationsberichte-460442 (accessed 1 March 2023). 138 See Noll (n 123) 285–311; and Hailbronner (n 14) 413–37. 139 See Council Resolution on burden-sharing with regard to the admission and residence of displaced persons [1995] OJ C262/1, No 4; and Decision 96/198/JHA on an alert and emergency procedure for burden-sharing [1996] OJ L63/10.

360 Common European Asylum System Protection Directive 2001/55/EC took up the idea on the basis of a compromise formula, which effectively adjourned the design of a solidarity mechanism to the implementation stage.140 Absence of pre-defined solidarity was one reason why the Directive was not activated during the first twenty years of its existence.

13.3.2 Futile quest for solidarity Supranationalisation of justice and home affairs put the Dublin criteria back on the political agenda. Greece and Italy opposed their path-dependent continuation, without, however, exercising a veto when the former Dublin II Regulation (EC) No 343/2003 was adopted unanimously in the written procedure.141 Proposals for a temporary suspension mechanism resurfaced during negotiations of the Dublin III Regulation but did not find their way into the final text.142 Tolerated non-compliance arguably explains the longevity of the Dublin system.143 From the start, there was a massive gap between the policy design and the situation on the ground. During 1998/99, actual transfers concerned no more than 1.7 per cent of all asylum applications;144 things improved slightly with Eurodac, even though an enormous implementation deficit has persisted ever since.145 Dublin procedures take time and transfers are performed in a minority of cases, indicating that the whole system suffers from bureaucratic exuberance.146 Somewhat paradoxically, it is this discrepancy between the lop-sided criteria for asylum jurisdiction and the practical reality of secondary movements which might explain the survival of the Dublin system, on paper at least. During 2015/16, the sheer number of arrivals prompted the German Federal Asylum Office to declare the factual suspension of Dublin procedures for Syrians in a tweet.147 The event was the visible result of the pragmatic recognition that administrative resources could be better invested in rendering asylum procedures more effective. Along the same line, German Chancellor Angela Merkel declared that ‘[t]‌he Dublin process, in its current form, is obsolete’.148 For a few weeks, it seemed as if a distribution key, essentially based on GDP and population size, might show a way out of the structural and practical dysfunctionality of the Dublin system. ‘Relocation’ became the new buzzword, which had been proposed first by the Commission in May 2015 after a series of devastating shipwrecks in the Central Mediterranean. 140 See Temporary Protection Directive 2001/55/EC, arts 5(3)(c), 25(1), (3), 26. 141 See Jonathan P Aus, ‘The Mechanisms of Consensus’ in Daniel Naurin and Helen Wallace (eds), Games Governments Play in Brussels (Palgrave Macmillan 2008) 99. 142 Contrast Dublin III Regulation (EU) No 604/2013, art 33(4) with Proposal for a Dublin III Regulation, COM(2008) 820 final, art 31. 143 See Tsourdi and Costello (n 55) 797–98; and Domenica Dreyer-Plum, ‘Dublin Revisited’ (2020) 22 EJML 386, 389–98. 144 See Commission, ‘Evaluation of the Dublin Convention’ SEC(2001) 756 final; and Hailbronner (n 14) 397–401. 145 See Francesco Maiani and Vigdis Vevstad, Setting up a Common European Asylum System: Study for the European Parliament (PE 425.622, August 2010) 157–62; and the constantly updated Eurostat datasets ‘MIGR_ DUBRI’, ‘MIGR_DUBRO’, ‘MIGR_DUBDI’, ‘MIGR_DUBDO’, ‘MIGR_DUBTI’, ‘MIGR_DUBTO’ https://ec.eur opa.eu/eurostat (accessed 1 March 2023). 146 See Francesco Maiani, ‘The Reform of the Dublin System and the Dystopia of “Sharing People”’ (2017) 24 MJECL 622. 147 See https://twitter.com/bamf_dialog/status/636138495468285952 (accessed 1 March 2023). 148 ‘François Hollande and Angela Merkel Face MEPs’ (Press Release 94921, 7 October 2015) https://www.europ arl.europa.eu/news/en/press-room/20150929IPR94921 (accessed 1 March 2023).

Asylum Jurisdiction 361 Governments agreed on a non-binding intergovernmental resolution on the relocation of 40,000 persons from Italy and Greece.149 When the Eastern Mediterranean became the focus of attention, two Relocation Decisions were adopted by qualified majority against the express will of several countries, raising relocation numbers to 160,000 people with a high statistical likelihood of receiving a positive decision; relocation of other applicants was not even considered opportune in September 2015.150 The idea had been spearheaded by the ‘Pilot Project for intra-EU Relocation from Malta’ (EUREMA), which had relocated slightly more than 600 people rescued at sea from the small island nation on a purely voluntary basis in two phases from 2009 to 2013.151 An action for annulment against the Relocation Decisions of 2015, initiated by Hungary and Slovakia, failed.152 Infringement proceedings against reticent Member States for non- compliance were successful, albeit without tangible effects since the two-year period for relocation had elapsed already.153 Implementation proved cumbersome for other Member States as well, with fewer than 35,000 actual relocations taking place in practice.154 Practical difficulties arguably indicate that the very idea of relocating tens of thousands of people to destinations throughout the continent might not be feasible in practice and could even be described as an exercise in ‘technocratic overreach’.155 Quota-based relocation was a central feature of the Commission’s Proposal for a Dublin IV Regulation, which comprised a ‘corrective allocation mechanism’ in situations of disproportionate pressure.156 Tolerated non-compliance and the transfer of jurisdiction after onward movements were supposed to give way to structured relocation. Negotiations went far at a technical level, with advanced debates in the Council and a report of the European Parliament’s LIBE Committee suggesting to take into account the preferences of asylum seekers (in addition to sanctions for secondary movements).157 Nevertheless, the search for compromise failed, not least since the Visegrád countries vehemently opposed relocation. Adoption by means of majority voting would have entailed structural risks, as explained in Chapter 2.2.3 on the institutional practice. Moreover, a qualified majority might not have been within reach anyway; several Member States which do not receive many asylum applications quietly opposed extensive relocation as well. The New Pact on Migration and Asylum, presented in September 2020, made another attempt at overcoming the political stalemate. On paper, it looked like a ‘fresh start’.158 The Proposal for a Dublin IV Regulation was replaced by the blueprint for an Asylum and 149 See Commission (n 30) 3; and ‘Resolution of the Representatives of the Governments of the Member States’ (Council doc 11097/15, 20 July 2015), adopted unanimously after the quota for Austria and Hungary had been set to zero. 150 See Relocation Decision (EU) 2015/1523, arts 3(2), 4; and Relocation Decision (EU) 2015/1601, arts 3(2), 4, with later amendments. 151 See Commission, ‘Staff Working Document accompanying the 5th Annual Report on Immigration and Asylum’ SWD(2014) 165 final, 25. 152 See Slovakia and Hungary v Council (n 75); and Bruno de Witte and Evangelia (Lilian) Tsourdi, ‘The Court of Justice Endorses the Emergency Scheme for Compulsory Relocation of Asylum Seekers within the European Union’ (2018) 55 CML Rev 1457. 153 See Commission v Poland and Others (n 124). 154 See Commission, ‘Progress Report on the Implementation of the European Agenda on Migration’ COM(2018) 250 final, 16, Annex 4. 155 Luuk van Middelaar, Alarums and Excursions (Agenda 2019) 100; see also Maiani (n 146) 640–44. 156 See Proposal for a Dublin IV Regulation COM(2016) 270 final, arts 5(3), 20, 34–44. 157 See Report on the Proposal for a Dublin IV Regulation A8-0345/2017 (6 November 2017); Presidency, ‘Compromise proposal’ (Council doc 8895/18, 17 May 2018); and Maiani (n 146) 628–40. 158 See Commission (n 39) 1.

362 Common European Asylum System Migration Management Regulation (AMMR). It presents a new label for more of the same upon closer inspection. The first entry criterion persists, and the transfer of jurisdiction in case of secondary movements shall be abolished for beneficiaries of international protection only.159 Instead of mandatory relocation, the Commission backed flexible solidarity, including administrative support and ‘return sponsorships’, which—somewhat wittingly— might result in mandatory relocation if third states do not take back nationals (as they rarely do).160 Quasi-permanent inter-state negotiations on the form and volume of solidarity contributions would replace fixed quotas.161 At the time of writing, it looked highly unlikely that the reform would be adopted in the foreseeable future, although the institutions might possibly agree on a ‘mini deal’, mirroring the recast of 2013, by abandoning most of the controversial elements. Experiences with the voluntary relocation of people disembarked in the Central Mediterranean among a ‘coalition of the willing’ during 2019/20 showed how cumbersome such negotiations can be.162 The mechanism was quietly abandoned when others refused to join and the number of arrivals in Italy fell below the number of applications further North, often after secondary movements. In the summer of 2022, the Council breathed new life into voluntary ad hoc solidarity of persons with protection needs arriving in the Mediterranean to break the deadlock over policy reform by means of ‘mini deals’, in line with previous comments. It remains to be seen whether the new mechanism would prove more successful than the predecessor.

13.3.3 Hierarchy of the substantive criteria Dublin has never functioned well in practice, and yet it has given rise to a significant number of preliminary references: 38 judgments had been delivered by the CJEU until the end of 2022. That dynamism largely results from the well-functioning court architecture in Central and Northern Europe, which triggers most preliminary references.163 These judgments demonstrate that Dublin is about much more than the eye-catching first entry rule. The complex procedural provisions are as relevant as the jurisdictional criteria which dominate the public discourse. The basic principles are straightforward: Articles 7–15 Dublin III Regulation set out a hierarchy of criteria to be checked in the ascending order.164 Whenever one of the leges speciales applies, the first entry criterion is supplanted. Statistics show that this happens frequently in relations between countries in Central and Northern Europe, which regularly transfer individuals between themselves. The Dublin system works reasonably well among

159 See AMMR Proposal (n 44) arts 21(1), 27(1), 35(2). 160 ibid art 55(2); and ch 16.6.1. 161 AMMR Proposal (n 44) arts 45–61; and Francesco Maiani, ‘Into the Loop’ in Thym and Odysseus Academic Network (n 40) 43, 51–56. 162 On the ‘Malta declaration’ of five Member States see the ‘Joint Declaration of Intent’ (23 September 2019); and Eleonora Frasca and Francesco Luigi Gatta, ‘The Malta Declaration on Search & Rescue, Disembarkation and Relocation’ EU Immigration and Asylum Law and Policy (3 March 2020). 163 See ch 3.3.3; more specifically, twenty-four out of the thirty-eight judgments responded to references from Germany (8), the Netherlands (7), Austria (5), and Sweden (4). 164 See Constantin Hruschka and Francesco Maiani, ‘Dublin III Regulation (EU) No 604/2013’ in Thym and Hailbronner (n 65) Articles 7–15.

Asylum Jurisdiction 363 these countries, in contrast to the protracted dysfunctionality in relations with the states of first arrival. When it comes to the substance of the criteria, unaccompanied minors shall be reunited with parents, siblings, or other relatives as a matter of principle. In the absence of family members, responsibility rests with the country where the minor resides, in accordance with a Court ruling interpreting vague wording in light of the Charter.165 If, however, the unaccompanied minor moves to another country after having received a positive asylum decision, he cannot rely on the residence rule—an outcome one might possibly question in light of a judgment limiting the return of minors to third states.166 Moreover, the country of residence must consider the asylum application of a new-born child even if the parent had received international protection elsewhere; the asylum application cannot be rejected as inadmissible.167 The Commission has proposed facilitating the transfer of minors to the country of first entry in the future, although the abstract reference to the best interests of the child may result in the continuation of the status quo, in light of the increasing number of judgments relying on the best interests formula.168 Ethnic and family networks are a core factor determining where people want to go, as explained in the interdisciplinary Chapter 4.1.3. Against this backdrop, the definition of ‘family member’169 comprising spouses and minor children constitutes a crucial bottleneck for the practical relevance of Articles 9–11. Confining family life to the core family reflects social practices in contemporary Europe and corresponds to human rights law,170 but it frustrates the preference of asylum seekers to join siblings and other relatives nonetheless. These other relatives can rely on family bonds only in situations of dependency, in accordance with Article 16(1).171 Not surprisingly, Mediterranean states and non-governmental organisations (NGOs) support a broader definition of family member embracing other relatives, in particular siblings, thus obliging countries further North actively to take charge of applicants arriving in Greece or Italy whose relatives are residing elsewhere. Such an amendment was proposed by the Commission in 2016 and 2020; it was rejected in the Council on the first occasion, while discussions on the latest proposal were ongoing at the time of writing.172 Applicants in possession of a visa or residence permit will be transferred to the Member State which had issued the document. Judges interpret the terminology strictly. State support for onwards movement along the ‘Western Balkans Route’ during 2015/16 was no equivalent of entry with a visa; the first entry rule applied instead, since the notion of ‘irregular’ border crossings covers anyone who does not fulfil the entry conditions in Article 6 Schengen Borders Code Regulation (EU) 2016/399.173 Note that the jurisdiction based on 165 See Case C-648/11 MA and others EU:C:2013:367, paras 53–60 on today’s Dublin III Regulation, art 8(4). 166 MA and Oothers (n 165) para 64 on the Dublin system; differs from Case C-441/19 Staatssecretaris van Justitie en Veiligheid EU:C:2021:9, paras 53–56 on return to third states; see further ch 5.4.2. 167 See Case C-720/20 Bundesrepublik Deutschland EU:C:2022:603. 168 See AMMR Proposal (n 44) art 15(5). 169 Dublin III Regulation (EU) No 604/2013, art 2(g); in line with the original Dublin Convention (n 20), art 4(2); and Case C-194/19 État belge EU:C:2021:270, para 30. 170 See ch 14.2.3. 171 See Case C-245/11 K EU:C:2012:685, paras 46–53. 172 See AMMR Proposal (n 44), art 2(g); Proposal for a Dublin IV Regulation (n 142), art 2(g); and Presidency (n 157) 26–27. 173 See Jafari (n 124) paras 59–92; contra AG Eleanor Sharpston, Opinion in ibid, EU:C:2017:443, points 155– 210; see also ch 6.2.1.

364 Common European Asylum System irregular entry is subject to time limits: it ceases after twelve months, or after a continuous stay in another country for at least five months. The second situation authorised the UK to return applicants crossing the Channel directly to France, instead of having to transfer them to Italy or Spain (which rarely works in practice).174 Moreover, failure of the transfer will usually result, as we shall see, in the transfer of jurisdiction. Additional criteria for visa- free entry and transit zones at airports are not practically relevant. Chapter 11.3 explained that an increase in the number of asylum applications can be a reason to introduce visa requirements.

13.3.4 Scope and procedure Article 77(2)(d) TFEU gives the European Union broad powers to define criteria and mechanisms for determining asylum jurisdiction. At present, the Dublin III Regulation uses this competence for third country nationals whose application is being considered or has been withdrawn.175 Beneficiaries of international protection and unsuccessful applicants cannot be transferred, on the basis of the Dublin III Regulation, to the country that had decided the application.176 Bilateral arrangements apply instead; they will be discussed in Chapter 16.6.3 on return. This limitation ratione materiae has gained increasing relevance in the early 2020s as a result of secondary movements after a positive asylum decision in countries at the external borders. Absence of internal border controls facilitates these movements, even though statutory limitations, described in Chapter 12.4.1, limit mobility rights to those with a residence permit and sufficient resources. As a response, the Commission wants to extend the scope of the Dublin system to people with a positive or negative decision,177 thus closing a loophole of the current legislation. Dublin procedures precede the regular asylum procedure.178 They assess jurisdiction on the basis of distinct procedural guarantees, with a personal interview, protection for minors, and the provision of information.179 Domestic authorities performing Dublin procedures need not be identical with the asylum authorities.180 Border guards can be assigned this task, provided they comply with procedural safeguards. Having said this, Member States remain free to skip the Dublin assessment, at least under the current legal framework.181 Individuals cannot insist on Dublin procedures being performed, even if the effects of this omission on the asylum procedure are negative.182 Similarly, the transfer of a person does not pre-empt the outcome of the asylum procedure thereafter.183 Dublin is about the verification of jurisdiction, no more.

174 See Peers (n 119) 299. 175 See Dublin III Regulation (EU) No 604/2013, art 18(1); and Agnes Hurwitz, ‘The 1990 Dublin Convention. A Comprehensive Assessment’ (1999) 11 IJRL 646, 650–52 on further limitations of earlier legislation. 176 See Case C-36/17 Ahmed EU:C:2017:273. 177 See AMMR Proposal (n 44) art 26(1). 178 See Dublin III Regulation (EU) No 604/2013, art 20(1); and Asylum Procedures Directive 2013/32/EU, art 33(1). 179 Dublin III Regulation (EU) No 604/2013, arts 4–6. 180 See Asylum Procedures Directive 2013/32/EU, art 4(2)(a). 181 See, however, Amended Proposal for an Asylum Procedures Regulation COM(2020) 611 final, art 37(1). 182 See Case C-56/17 Fathi EU:C:2018:803; and Ibrahim and others (n 105) paras 75–80. 183 See Case C-695/15 PPU Mirza EU:C:2016:188.

Asylum Jurisdiction 365 Member States may voluntarily assume responsibility on the basis of the ‘discretionary clause’ in today’s Article 17(1) Dublin III Regulation. Judges described the discretion to be ‘extensive’,184 or even ‘absolute’.185 National authorities define relevant criteria autonomously, in particular political, humanitarian, or practical considerations.186 By way of example, the provision may be activated for relatives beyond the nuclear family, for vulnerable people, or for practical reasons. Suspension of Dublin procedures for Syrians by the German authorities during 2015/16 may be interpreted as an en bloc activation of the discretionary clause (possibly in violation of the general scheme and the effet utile of the Regulation, which foresees individual decisions187). Member States determine the competent authority and need not provide for legal remedies.188 Applicants do not benefit from an individual right that Article 17(1) should be activated;189 nor does the Charter seem to apply.190 In short, Member States decide on the basis of domestic laws.

13.3.5 Take charge/back requests Rules on asylum jurisdiction had always pursued two objectives which often compete with one another: effective access to asylum procedures and the prevention of ‘forum shopping’ and secondary movements. In a staggering elasticity, described in Chapter 6.4.3 on the doctrinal foundations, the Court fluctuates between these teloi depending on the outcome. Both objectives similarly pervade secondary legislation. On the one hand, the prevention of ‘forum shopping’ entails that individuals cannot chose the country where to apply for asylum; they must comply with the jurisdictional rules.191 On the other hand, Member States have to examine each application.192 This obligation responded to the previous phenomenon of ‘refugees in orbit’, with applicants being ‘referred successively from one Member State to another without any of these states acknowledging itself to be competent’.193 In the only infringement proceedings that have reached the Court so far, the Commission complained against Greece not examining the claims of returnees, on grounds of secondary movements having discontinued earlier applications.194 In this overall context, provisions on inter-state procedures serve as an intermediate device. They support the transfer to the country responsible, while, at the same time, being a ‘weapon’ in the armoury of lawyers to thwart this outcome. Two inter-state procedures have to be distinguished: ‘take charge’ concerns transfers to states where applicants have not been so far, for instance Spain asking Sweden to assume responsibility for the parent of an

184 See K (n 171) para 27. 185 Case C-661/17 MA and others EU:C:2019:53, para 58. 186 See Case C-528/11 Halaf EU:C:2013:342, para 37. 187 See Mattias Wendel, ‘The Refugee Crisis and the Executive’ (2016) 17 German LJ 1005. 188 See MA and others (n 185) paras 60–69, 70–79. 189 ibid para 60; and ch 6.5. 190 MA and others (n 185) paras 70–72; Halaf (n 186) paras 40–42; see also chs 5.4.4 and 7.2. 191 See ch 12.4.1.2; and AG Jean Richard de la Tour, Opinion in Joined Cases C-323/21, C-324/21 and C-325/21 Staatssecretaris van Justitie en Veiligheid EU:C:2022:651, points 76–79. 192 See Dublin III Regulation (EU) No 604/2013, art 3(1); in line with the original Dublin Convention (n 20) art 3(1). 193 Dublin Convention (n 20) recital 4. 194 See the order for Case C-130/08 Commission v Greece EU:C:2008:854, which struck the case from the register after it had been resolved; and the Commission’s pleas in fact and law [2008] OJ C128/25.

366 Common European Asylum System unaccompanied minor who lives in Malmö; ‘take back’ responds to secondary movements. Procedural guarantees are particularly relevant for take back requests, where the interests of state authorities and applicants usually collide. Chapter 6.5.4 explained that the exclusion of legal remedies against the refusal of a take charge request, under Article 27(1) Dublin III Regulation, was found to be incompatible with Article 47(1) CFR in situations involving unaccompanied minors. Experience of the author shows that the general public, as well as politicians and legal experts without an expertise in migration, find it hard to recognise that the widely-known first entry rule can be supplanted by seemingly obscure procedural provisions. They often misread the visibility of the hierarchy of criteria as an authorisation summarily to reject asylum applications in case of secondary movements. Such arguments dominated debates in Germany and Austria during and after 2015/16.195 Court judgments insisting on the wording and general scheme of the Regulation rejected such arguments. The Dublin III Regulation may be structurally unfair and practically dysfunctional, but it remains the law of the land in the absence of legislative change. Procedures were streamlined during the first and second phase of harmonisation. By way of example, failure to respond to a take back request within a time limit of two weeks or one month, depending on the scenario, shall be tantamount to accepting the transfer.196 The Commission proposes to simplify procedures further through a simple notification, together with a one-week deadline for objections, which can be extended—or even suspended—in vaguely defined situations of force majeure.197 Realisation of the transfer is a different matter: procedural simplification need not result in more returns. Moreover, Dublin can be a perpetuum mobile. German authorities are obliged to start a second tack back procedure if someone returns within a matter of days or weeks after having been transferred to Italy for a first time, while the original transfer decision remains valid when the applicant leaves the country temporarily before the actual transfer or the time limit for the transfer of jurisdiction.198 At the same time, the Dublin III Regulation obliged Member States to introduce an effective remedy against the transfer decision—a matter that had been left to domestic law under the Dublin II Regulation, subject to the condition that suspensive effect shall not usually be automatic.199 Few people would have predicted that this amendment ensued a spectacular about-turn of the case law: seemingly obscure procedural guarantees contain individual rights asylum seekers may rely upon to challenge the transfer decision. Chapter 6.5 discussed the doctrinal foundations of these rulings, uncertainties about the scope of Article 47 CFR, and how the judgments interact with the Commission’s proposal to limit legal remedies in the future.

195 Among those questioning Dublin were Alexander Peukert, Christian Hillgruber, Ulrich Foerste, and Holm Putzke, ‘Einreisen lassen oder zurückweisen?’ [2016] Zeitschrift für Ausländerrecht 131; for the reverse position see Daniel Thym, ‘Der Rechtsbruch-Mythos und wie man ihn widerlegt‘ Verfassungsblog (2 May 2018); and also Stephan Detjen and Maximilian Steinbeis, Die Zauberlehrlinge (Klett-Cotta 2019). 196 See Dublin III Regulation (EU) No 604/2013, art 25(2). 197 See AMMR Proposal (n 44) art 31; and Proposal for a Crisis and Force Majeure Regulation (n 45) art 8(1), (3). 198 See Case C-360/16 Hasan EU:C:2018:35, paras 21–45, 71–88; and Joined Cases C-323/21, C-324/21, and C- 325/21 Staatssecretaris van Justitie en Veiligheid EU:C:2023:4, paras 57–61. 199 See Dublin III Regulation (EU) No 604/2013, art 27; former Dublin II Regulation (EC) No 343/2003, art 19(2); and Dublin Convention (n 20) art 11(5).

Asylum Jurisdiction 367 For the time being, judges emphasise the mandatory nature of legal remedies on grounds of procedural errors. First, take back requests must be submitted within two months after a Eurodac hit; delay result in the transfer of jurisdiction to the state asking for the transfer.200 This becomes practically relevant if an increase in the number of asylum applications means that several months pass between the initial registration of the asylum request (usually with border guards or other authorities) and the formal application (with the asylum authorities).201 An alternative recourse to bilateral arrangements to perform the transfer was found to be illegal, since take back procedures under the Dublin system shall not be circumvented.202 In cases of doubt or unusual circumstances, the scope of judicial oversight has to be determined by means of statutory interpretation.203 Secondly, transfers are subject to a six-month time limit after which jurisdiction shall be transferred onto the requesting state.204 The period will start running once legal remedies against the decision have been concluded.205 Member States cannot suspend the Dublin procedure for reasons which are not related to doubts about the legality of the transfer decision, such as factual difficulties to realise the transfer during the Covid-19 pandemic or onward movement to another Member State.206 Once the six-month deadline for the execution of the transfer has expired, jurisdiction changes automatically, even if the failure of the transfer was caused by the lack of cooperation of the authorities of the receiving state (they cannot voluntarily accept the take back after the expiry of the time limit either).207 Individuals can invoke the transfer of jurisdiction as an individual right, thus receiving a full asylum procedure.208 Obstruction of the transfer on the part of the applicant does not result in an extension of the take back period, except when he absconds. Note, however, that the notion of ‘absconds’ does not coincide with the ‘risk of absconding’ under the Return Directive. The Court construes the meaning of ‘absconds’ autonomously and insists on a strict interpretation, thus limiting the capacity of states to extend the time limit.209 Placement in closed psychiatric facilities does not justify an extension of the six-month deadline either; the notion of ‘imprisonment’ concerns the detention of asylum applicants for the purpose of criminal proceedings.210 Chapter 16.3.4 on return will explain that the six-month time persists when the applicant puts forward plausible arguments that he is victim of human trafficking, thus triggering a reflection period under the Victims of Trafficking Directive 2004/81/EC during which Dublin transfers cannot be executed. Mobility within the Schengen area need not be a one-way street from the country of first arrival to the preferred destination. Some applicants make several applications and may 200 See Dublin III Regulation (EU) No 604/2013, art 21(1)(3). 201 See Case C-670/16 Mengesteab EU:C:2017:587, paras 41–62, 74–103. 202 See Joined Cases C-582/17 and C-583/17 H and R EU:C:2019:280, paras 51–53; Joined Cases C-47/17 and C-48/17 X and X EU:C:2018:900, paras 56ff; and Case C-647/16 Hassan EU:C:2018:368, paras 57–62. 203 H and R (n 202) paras 38–50; État belge (n 169) paras 33–36; and Staatssecretaris van Justitie en Veiligheid (n 198) paras 93–94. 204 See Dublin III Regulation (EU) No 604/2013, art 29(2). 205 ibid art 29(1); and Case C-490/16 AS EU:C:2017:585, paras 56–57; see also Case C-19/08 Petrosian EU:C:2009:41, paras 42–45 on the predecessor provision. 206 See Joined Cases C-245/21 and C-248/21 Bundesrepublik Deutschland EU:C:2022:709, paras 51–62; and Staatssecretaris van Justitie en Veiligheid (n 198) paras 66–72. 207 See Case C-201/16 Shiri EU:C:2017:805, paras 26–34. 208 ibid paras 35–46. 209 See Jawo (n 91) paras 52–65. 210 See Case C-231/21 Bundesamt für Fremdenwesen und Asyl EU:C:2022:237, paras 44–61.

368 Common European Asylum System even move back and forth between, by way of example, Germany and the Netherlands several times after having entered Union territory via Italy. This raises the question about how to deal with a plurality of tack back requests. Judges confirmed that only one state shall be responsible, although that does not entail, against the Advocate General, that the first take back request takes precedence over those submitted later. Rather, each request shall be analysed independently, including the potential transfer of jurisdiction in accordance with the regular time limits.211 Whenever a second country assumes jurisdiction as a result of the late submission of the take back request or the failure of the transfer, any third or fourth Member State cannot ask the first state to assume jurisdiction any longer in order to prevent conflicting outcomes. Instead, they may ask the second country to take back the applicant; time limits shall start running anew (paras 74–85).

13.3.6 Mutual trust and its limits Human rights limits to Dublin transfers are the single most widely discussed question of EU migration law. Dozens of articles have been written about a series of ECtHR and CJEU judgments. Widespread attention was warranted given the practical and conceptual relevance of a debate which saw the confrontation of two powerful courts, before a common ground emerged after years of judicial, academic, and political discussion. The starting shot was fired by the ECtHR in Strasbourg. Its MSS judgment of January 2011 famously declared transfers to Greece to be illegal due to miserable living conditions.212 Doing so broadened human rights-based non-refoulement to socio-economic factors, as described in the constitutional foundations. National courts had discussed similar questions under the original Dublin Convention at a time when asylum procedures, recognition criteria, and reception conditions had not yet been harmonised.213 The British House of Lords prohibited returns to France and Germany, since both countries did not recognise persecution by non-state actors (the judgment was later overturned by an Act of Parliament).214 The ECtHR was less critical at the time, declaring inadmissible an application in a similar case.215 Things changed with the MSS ruling, which heralded a period of tensions between the judges in Luxembourg and Strasbourg. The CJEU blocked the EU’s accession to the ECHR, emphasising, amongst others, that human rights should not undermine inter-state cooperation on asylum, since mutual trust relates to the ‘raison d’être’ of European integration.216 Nevertheless, judges shied away from open confrontation when interpreting the former Dublin II Regulation (EC) No 343/2003 in light of human rights to prohibit transfers to Greece.217 Rhetoric emphasis on the exceptional character of the outcome left room for discussion. The CJEU initially held that ‘only’218 systemic deficiencies may stand in the way of Dublin 211 See Staatssecretaris van Justitie en Veiligheid (n 198) paras 50–54, 62–65. 212 See MSS v Belgium and Greece (n 89) §§ 335–61. 213 See Gregor Noll, ‘Empiricism vs Formalism’ (2001) 70 Nordic J Intl L 161. 214 See R v Secretary of State for the Home Department, ex parte Adan and Aitseguer [2000] UKHL 67; and Andrew Nicol, ‘Between Immigration and Policing’ (2004) 10 ELJ 171. 215 See TI v United Kingdom App no 43844/98 (ECtHR, 7 March 2000). 216 See ch 5.3.3. 217 See Joined Cases C-411/10 and C-493/10 NS and others EU:C:2011:865, paras 79–106; and, later, Jawo (n 91) paras 80–85. 218 Case C-394/12 Abdullahi EU:C:2013:813, para 60.

Asylum Jurisdiction 369 transfers.219 We may conclude, with the wisdom of hindsight, that judges gradually agreed on common standards.220 The CJEU accepted—contra the Commission and the Advocate General—that Article 4 CFR covered individual risks short of systemic deficiencies.221 Follow-up rulings of the ECtHR carefully assessed individual cases, allowing, for instance, the transfer of asylum applicants to Italy under difficult but not outright unacceptable conditions.222 In Tarakhel, the Grand Chamber stopped the return of a family with five small children to Italy.223 These diverse outcomes emphasise that we have to distinguish, in light of the factual situation on the ground, between different categories: families with young children may have to be treated in a different way than healthy adults; general information on a country may have to be combined with individual factors increasing, or lowering, the risk of ill-treatment.224 Such differentiation is typical of human rights-based non-refoulement assessments in line with previous comments. In any case, Dublin transfers will only be prohibited when ill-treatment attains a level of severity that brings the situation under of Article 4 CFR. Disrespect for the Reception Conditions Directive does not prevent transfers;225 nor do procedural deficiencies and divergences in the application of the Qualification Directive.226 Guarantees by the receiving state may provide a foundation for a transfer, with judges granting Member States more flexibility as to the form and contents of such guarantees, in comparison with diplomatic assurances by third states.227 All the rest depends on the factual situation. Domestic courts have to apply the—not so crystal clear—human rights standards diligently. The outcome depends on human rights law, not a balance with mutual trust, which does not constitute a constitutional meta-principle of Union law.228 EU institutions codified the case law in Article 3(2)(2) Dublin III Regulation, whose reference to both ‘systemic flaws’ and Article 4 CFR must be read in light of the above- mentioned judgments. Whenever the provision applies, domestic authorities where the applicant is residing must verify whether the Dublin III Regulation authorises the transfer to a different country from the one where the applicant cannot be sent; they are not obliged to activate the discretionary clause.229 When no other country is responsible, jurisdiction rests with the country where an individual is staying in accordance with Article 3(2) (1). Even beneficiaries of international protection will receive a second asylum procedure

219 See Iris Canor, ‘My Brother’s Keeper? Horizontal Solange’ (2013) 50 CML Rev 383, 405–13. 220 See Koen Lenaerts, ‘La vie après l’avis: Exploring the Principle of Mutual (yet not Blind) Trust’ (2017) 54 CML Rev 805; and Henri Labayle, ‘Droit d’asile et confiance mutuelle’ (2014) 49 Cahiers de droit européen 501. 221 See Case C-578/16 PPU CK and others EU:C:2017:127, paras 59–95; and also Case C-646/16 Jafari EU:C:2017:586, para 101. 222 Among a series of inadmissibility decisions see Mohammed Hussein and others v Netherlands and Italy App no 27725/10 (ECtHR, 2 April 2013); and Daytbegova and Magomedova v Austria App no 6198/12 (ECtHR, 4 June 2013). 223 See Tarakhel v Switzerland (n 90) §§ 93–122; and MT v Netherlands App no 46595/19 (ECtHR, 23 March 2021) §§ 53–58. 224 See Anna Lübbe, ‘“Systemic Flaws” and Dublin Transfers’ (2015) 27 IRJL 135. 225 See NS and others (n 217) para 84; and AG Vera Trstenjak, Opinion in ibid, EU:C:2011:611, points 124– 26; the reference to secondary legislation was ambivalent regarding the interpretation of human rights in MSS v Belgium and Greece (n 89) § 250. 226 See Ibrahim and others (n 105) paras 95–100. 227 See Tarakhel v Switzerland (n 90) §§ 119–22; and Jens Vedsted-Hansen, ‘Current Protection Dilemmas in the European Union’ in Carolus Grütter and others (eds), Migration on the Move (Brill 2017) 95, 105–109. 228 See Francesco Maiani and Sara Mogliorini, ‘One Principle to Rule Them All!’ (2020) 57 CML Rev 7; and Frank Meyer, ‘Der Grundsatz des gegenseitigen Vertrauens’ [2017] Europarecht 163. 229 See NS and others (n 217) paras 96–97, 107–108; and CK and others (n 221) para 97.

370 Common European Asylum System if they cannot be returned on human rights grounds. Article 33(2)(a) Asylum Procedures Directive 2013/32/EU does not apply.230 A special scenario concerns applicants who maintain that living conditions would be insufficient once they have received international protection, as opposed to reception conditions during the asylum procedure. We shall see that Articles 26–34 Qualification Directive 2011/95/EU essential give beneficiaries of international protection the same treatment as nationals, which may result in less support than during the procedure. Judges accepted that Article 4 CFR covers treatment after recognition.231 Uncertainties persist, as we have seen, what exactly human rights law requires states to guarantee in addition to equal treatment with nationals.

13.3.7 Secondary movements and the transfer of jurisdiction Secondary movements are, together with ‘solidarity’, at the heart of political disputes between ‘frontline’ Member States and destination countries further North. Statistics are notoriously unreliable, but indicators of recurring secondary movements include the high number of take back requests under the Dublin system and discrepancies between the numbers of asylum applications and administrative first instance decisions in countries of first arrival; asylum applicants move on before the Italian or Greek authorities deliver a decision. Furthermore, recent years have witnessed increasing numbers of beneficiaries of international protection going elsewhere after the completion of the asylum procedure. The (il)legality of these movements under the Schengen acquis was discussed in Chapter 12.4.1 on border controls. An agreement on the reform of the Eurodac Regulation would improve our knowledge base. As mentioned previously, procedural safeguards in the Dublin III Regulation (EU) No 604/2013 mitigate the first entry criterion. Article 29(2) mandates that ‘responsibility shall . . . be transferred’ to the country that had initiated the take back procedure ‘[w]‌here the transfer does not take place within . . . six months’ (or eighteen months in case of absconding). Similarly, jurisdiction will shift mandatorily when the take back request is delayed for more than two or three months.232 Third country nationals may submit several asylum applications, and they will receive a regular procedure in the country of residence, once jurisdiction has been transferred. In its former Proposal for a Dublin IV Regulation, the Commission had proposed eliminating the transfer of jurisdiction. Asylum seekers were expected to return to the state responsible. While the Parliament supported this change (together with mandatory relocation), the Council discussed the idea of ‘stable responsibility’ during a period of five or eight years.233 Somewhat surprisingly, the new ‘Pact on Migration and Asylum’ of September 2020 basically returned to the status quo ante of the Dublin III Regulation, at least for the failure

230 See Joined Cases C-540/17 and C-541/17 Hamed and Omar EU:C:2019:964, para 42, which ignored the complex argument put forward in the order for reference. 231 See Ibrahim and others (n 106) paras 93–94; and Maarten den Heijer, ‘Transferring a Refugee to Homelessness in Another Member State’ (2020) 57 CML Rev 539, 549–51. 232 See Dublin III Regulation (EU) No 604/2013, art 21(1)(3); and n 201and accompanying text. 233 See LIBE Committee, ‘Report on the Proposal for a Dublin IV Regulation’ (PE 599.751v03-00, 6 November 2017); and Presidency (n 157) art 9a.

Asylum Jurisdiction 371 of the take back procedure. It continued the transfer of jurisdiction, while discontinuing that outcome in case of late submission of the take back request, absconding, and for beneficiaries of international protection.234 Additional guarantees may possibly be available for the latter under the Council of Europe Agreement on Transfer of Responsibility, mentioned in the context of the Qualification Directive. Politically speaking, the Commission might have had little choice than to continue the status quo, given that it similarly refrained from proposing meaningful solidarity. Other legislative changes are meant to prevent secondary movements by means of positive incentives or negative sanctions: the proverbial ‘carrots and sticks’.235 Positive incentives include a broad definition of family life, embracing siblings as well as the option of legal onward movement, which will be mentioned in Chapter 13.9.3 below. In addition, the Commission invites Member States to consider ‘meaningful links’ when determining whom to relocate under the future system for flexibly solidarity.236 Doing so takes up the European Parliament’s position on the former Proposal for a Dublin IV Regulation regarding relocation on the basis ‘in particular [of] family, cultural or social ties, language skills or other meaningful links which would facilitate his or her integration into that other Member State’.237 Note that the ‘meaningful links’ criterion would be one element amongst others, thus staying substantially short of a guarantee of free choice. Sanctions would include the restriction of legal remedies discussed previously. Furthermore, social benefits are to be reduced after secondary movements.238 The reduction would remain subject to a threefold caveat. First, it shall apply once a transfer decision has been notified, not automatically whenever someone files a second application. Secondly, the general scheme of the draft legislation indicates that the reduction will come to an end with the transfer of jurisdiction. Thirdly, the Council’s compromise text introduces a caveat for ‘the need to ensure a standard of living in accordance with [the Charter]’, whose exact requirements remain uncertain. Interdisciplinary studies show that benefit levels have a certain impact on migratory patterns, even though other elements, such as economic prospect or family ties, are more significant.239 Since these other factors are beyond the direct reach of asylum legislation, the phenomenon would persist, albeit possibly at a lower scale, if the reform was adopted in the years to come. In December 2021, the Commission proposed a novel mechanism for the direct transfer to neighbouring countries ‘to counter irregular movements’.240 By way of example, Austria could cooperate with Slovenia to return people apprehended ‘as part of cross-border police operational cooperation, in particular, during joint police patrols’.241 This wording excludes unilateral returns without the consent of neighbouring states. Individuals would be returned on the basis of a standard form within a short period of no more than one day; 234 See AMMR Proposal (n 44) arts 27(1), 35(1), (2). 235 See Maiani (n 146) 640–44; and Daniel Thym, ‘Secondary Movements’ in Thym and Odysseus Academic Network (n 40) 129, 131–35. 236 See AMMR Proposal (n 44) arts 57(3)(1), 49(2). 237 LIBE Committee (n 233) art 24b, read in conjunction with art 19(2). 238 See AMMR Proposal (n 44) art 10(1); Proposal for a Recast of the Reception Conditions Directive, COM(2016) 465 final, art 17a; and Presidency, ‘Conditional confirmation of the final compromise text’ (Council doc 10009/18, 18 June 2018). 239 See ch 4.1; and Thym (n 235) 137–44. 240 Proposal for an Amendment of the Schengen Borders Code Regulation (EU) 2016/399 COM(2021) 891 final, recital 25. 241 ibid art 23a(1)(c).

372 Common European Asylum System legal remedies should be available, albeit without suspensive effect.242 Even though the instrument has apparently been designed to apply to beneficiaries of international protection and asylum applicants, the proposal remains silent on how it relates to the asylum acquis. It remains unclear, in particular, whether the transfer could take place if the person applied for asylum in the country sponsoring the return.

13.4 Asylum Procedures Anyone familiar with administrative practices understands how important procedures are for the final outcome. Common standards can be found in the Asylum Procedures Directive, which has given rise to an increasing number of preliminary references. Having said this, it was no foregone conclusion that the EU would introduce far-reaching procedural prescriptions. The Refugee Convention remains silent, and national practices had diverged greatly.243 Negotiations of the former Asylum Procedures Directive 2005/85/EC proved complex, despite the pedigree of a Council Resolution.244 EU institutions finally agreed on a neatly balanced combination of state interests and individual rights.245 Today’s Asylum Procedures Directive 2013/32/EU follows this compromise, while reducing the leeway of the Member States.246 Seven themes define the legal practice. The Directive applies once an application has been made (13.4.1) and revolves around the personal interview, in line with the European tradition of complex administrative procedures (13.4.2). A defining feature of the Asylum Procedures Directive are distinct rules for the accelerated examination of some claims (13.4.3), with border procedures being at the centre of contemporary debates (13.4.4). EU legislation proved essential in the proliferation of national rules on safe countries of origin and transit (13.4.5). Judicial oversight is an essential safeguard, which features in many Court cases (13.4.6). External processing of asylum applications in third states regularly surfaces in the political debate, although it is not permitted under the Asylum Procedures Directive de lege lata (13.4.7). The Commission’s proposal to replace the Asylum Procedures Directive by a Regulation would increase harmonisation further, even though Chapter 7.2.1 explained the abstract distinction between directives and regulations gives way to gradual variation in the institutional practice. Reform proposals will be mentioned on the pages that follow whenever the Commission puts forward important changes. They can be found in a first Proposal, tabled in 2016, and an Amended Proposal, presented in 2020, which have to be read jointly.247 An additional instrument on screening shall be introduced as well, on which an interinstitutional negotiations might be reached before the European elections in 2024.248 242 ibid art 23(a), read in conjunction with Annex XII. 243 See Guy S Goodwin-Gill and Jane McAdam, The Refugee in International Law (4th edn, OUP 2021) 600–607; and Lambert (n 13) chs 2–4. 244 See Resolution on minimum guarantees for asylum procedures [1996] OJ C274/13. 245 See Doede Ackers, ‘The Negotiations on the Asylum Procedures Directive’ (2005) 7 EJML 1; and Jens Vedsted-Hansen, ‘Common EU Standards on Asylum’ (2005) 7 EJML 369. 246 See Cathryn Costello and Emily Hancox, ‘The Recast Asylum Procedures Directive 2013/32/EU’ in Chetail and others (n 27) 377. 247 See Proposal for an Asylum Procedures Regulation, COM(2016) 467 final; and Amended APR Proposal (n 181). 248 See Proposal for a Screening Regulation COM(2020) 612 final.

Asylum Procedures 373 By contrast, an adoption of the Asylum Procedures Regulation seemed unlikely, since stricter border procedures are politically linked to a reform of the Dublin III Regulation, which appeared beyond reach. Irrespective of these reform initiatives, comments hereafter must be read against the backdrop of the horizontal description of the administrative dimension in Chapter 7.2. Asylum procedures reiterate the inherent tension between supranational guidance and distinct national traditions.

13.4.1 Access to the procedure Articles 3 and 6(1) Asylum Procedures Directive lay down an individual guarantee to apply for asylum in the territory or at the border. This may be mandatory under Article 18 CFR, in line with previous comments. Judges conceive effective access to the procedure to be one of the objectives of the Directive,249 which does not depend on the prospects of success of such a claim.250 Asylum applications may be submitted with any state authority, including border guards, the police, or judges reviewing the legality of detention.251 Member States are obliged to train the personnel and to provide individuals with information how and where to lodge the official application.252 Chapter 12.6.2 described that it remains a formidable challenge to determine what this abstract obligation entails in scenarios of external border surveillance. We have to distinguish between ‘making’ and ‘lodging’ an application, as well as the official ‘registration’. The act of ‘making’ refers to the abstract expression of the wish to apply for asylum with any state authority, while ‘lodging’ entails formal submission with the competent personnel, usually at a designated place.253 ‘Registration’ shall take place no later than three working days after the application has been made in regular circumstances, or ten working days in times of crises.254 Non-compliance with the time limits does not have immediate benefits for applicants, while preclusion deadlines for lodging applications, laid down in national laws, can fall foul of the principle of effectiveness, as described in Chapter 7.2.2.2. In practical terms, applications are often ‘made’ with different state authorities than the ones responsible for ‘lodging’ and ‘registering’, especially border guards or the police. Crucially, the Court interprets an individual to be an ‘applicant’ from the moment the wish to apply for asylum has been ‘made’.255 Guarantees in the Asylum Procedures Directive, the Reception Conditions Directive, and the Dublin III Regulation can be relied upon in future. The underlying distinction between making, lodging, and registering is particularly relevant in the context of external border controls. Member States cannot deny reception

249 See Ministerio Fiscal (n 107) para 63. 250 See Commission v Hungary (n 107) para 136. 251 See Ministerio Fiscal (n 107) paras 52–68, 79–82. 252 ibid paras 73–78; and Asylum Procedures Directive 2013/32/EU, art 6(1)(3). 253 Compare Asylum Procedures Directive 2013/32/EU, arts 2(c), 6(1)(2)–(3), (2) with art 6(3)–(4); and see AG Maciej Szpunar, Opinion in Case C-36/20 PPU Ministerio Fiscal EU:C:2020:331, points 78–89; see also APR Proposal (n 247) arts 25–29. 254 Asylum Procedures Directive 2013/32/EU, art 6(1)(1), (5). 255 See Ministerio Fiscal (n 107) paras 86–94.

374 Common European Asylum System conditions, refuse entry at the border, or revert to blanket detention contending that the formal requirements for lodging have not been complied with.256 Whenever neighbouring states instrumentalise migration for political purposes, the Commission wants to authorise Member States to defer ‘registration’ for up to four weeks and to require applications to ‘lodge’ claims at official border crossing points.257 Doing so gives state authorities practical leeway to manage movements and limits procedural guarantees for applicants, but it does not absolve Member States from complying the remainder of the asylum acquis. The Instrumentalisation Regulation would not sanction pushback practices whenever we conclude that an application has been ‘made’ with the border police. Quick adoption of the measures failed partly because the Polish government, in particular, had hoped that the proposal would go further.258 In the absence of EU measures on instrumentalisation, Poland and Lithuania tried—unsuccessfully—to invoke Article 72 TFEU as a justification, as described in Chapter 10.3 on overarching features. Another novelty is the Proposal for a Screening Regulation. What may sound like an important innovation at first would effectively formalise existing obligations to verify the identity, consult the Eurodac database, find out whether someone poses a public security threat, and wishes to apply for asylum; an additional health screening would formalise something most countries had introduced in response to the Covid-19 pandemic. Screening during five to ten working days is to prepare formalised decision-making under existing instruments, based on a ‘debriefing form’.259 Legal remedies shall be available against the final decision.260 Several elements have been criticised for undermining the rights of asylum seekers de facto,261 notably the ‘fiction of non-entry’, which will be discussed in the context of the Reception Conditions Directive. The Commission Proposal remained silent as to whether reception conditions would be available during the screening.

13.4.2 Personal interview and individualised assessment A core element is the personal interview, which must be performed by the competent authorities in accordance with Articles 14–17 Asylum Procedures Directive 2013/32/EU. It can be omitted in exceptional circumstances only.262 The personal interview covers both the admissibility and the merits, while the assessment of jurisdiction under the Dublin III Regulation is subject to distinct standards and may be performed by another authority.263 Such individual examinations are the hallmark of European tradition of asylum, which

256 See Case C-808/18 Commission v Hungary EU:C:2020:1029, paras 95–106; and Valstybės sienos apsaugos tarnyba (n 109) paras 57–62. 257 See Proposal for an Instrumentalisation Regulation (n 51) art 2(1); and also Proposal for a Crisis and Force Majeure Regulation (n 45) art 7. 258 See UNHCR, ‘Observations on the Draft Law’ UD265 (16 September 2021) https://www.refworld.org/ docid/61434b484.html (accessed 1 March 2023). 259 See Proposal for a Screening Regulation (n 248) arts 11–14. 260 As found to be legal in Case C-69/10 Samba Diouf EU:C:2011:524, paras 40–44, 54, 57ff. 261 See Lyra Jakuleviciene, ‘Pre-Screening at the Border in Asylum and Migration Pact’ in Thym and Odysseus Academic Network (n 40) 81; and Maarten den Heijer, ‘The Pitfalls of Border Procedures’ (2022) 59 CML Rev 641, 655–57. 262 Tens of thousands of (mostly positive) German asylum decisions based on a written questionnaire during the second half of 2015 tested the outer limits of Asylum Procedures Directive 2013/32/EU, art 14(2). 263 ibid art 34; and Case C-517/17 Addis EU:C:2020:579, paras 46–55.

Asylum Procedures 375 contrasts with the prevalence of abstract assessments elsewhere.264 The sheer complexity of asylum procedures presents itself as a structural deficit of the asylum acquis in line with introductory comments. State practices at the external borders, in particular, have proven to be deficient on many occasions. Member States have to ensure the impartiality of the personnel performing the interview, which shall receive appropriate training and must consult interpreters to facilitate communication.265 A second interview may be warranted, when the initial hearing does not foreclose all relevant information.266 Standards for how to assess credibility are discussed in the context of the criteria for refugee status. Chapter 7.2.4 on the administrative dimension explained that the absence of a meaningful interview will usually result in the illegality of the administrative decision, while states retain leeway as to whether domestic courts may substitute the decision with their own, instead of referring the file back to the authorities. The European Parliament insisted on detailed provisions on procedural support. Information free of charge shall be available during the administrative proceedings, which, in the terminology of the Asylum Procedures Directive, are referred to as the ‘first instance’.267 One step further, legal assistance and representation shall be made available during ‘appeals procedures’ (the first instance of judicial oversight); it shall be free of charge in regular circumstances, thus going beyond minimum requirements of human rights law.268 In 2016, the Commission proposed extending free legal assistance to the administrative procedure, reflecting the positive experience of Switzerland in terms of improving quality and accelerating the overall process.269 Finally, applicants may consult their own lawyers and seek the assistance of non-governmental organisations, or the UNHCR, which shall have access to reception or detention facilities.270 While these legislative guarantees for information, assistance, and representation are far- reaching on paper, the reality on the ground often looks different. In many cases, trained personnel may be a scarce resource, especially in hotspots and other remote locations. There might simply not be enough supply in the form of civil society organisations or lawyers to respond adequately to the demand for legal advice and support. Moreover, effective access to support services is occasionally obstructed by state authorities. Hungary flagrantly violated Union law by criminalising legal assistance;271 the government’s claim that procedural guarantees were ‘abused’ did not satisfy the high judicial threshold for such a finding.272 Additional safeguards exist for unaccompanied minors and other vulnerable groups.273 Distinct ‘vulnerability assessments’ must be performed, which can be combined with 264 See Costello and Hancox, ‘Recast Asylum Procedures Directive’ (n 246) 379–81; and Bruce Burson, ‘Refugee Status Determination’ in Cathryn Costello and others (eds), The Oxford Handbook of International Refugee Law (OUP 2021) 569, 577–80. 265 See Asylum Procedures Directive 2013/ 32/ EU, recitals 16– 17, arts 14– 16; and Pieter van Reenen, ‘Impartiality in the EU Asylum Procedure’ (2018) 20 EJML 338. 266 See Case C-560/14 M EU:C:2017:101, paras 47–53. 267 See Asylum Procedures Directive 2013/32/EU, arts 8(1), 19, 20(2), 21. 268 ibid arts 20–21; and ch 7.2.3.3; see also Reception Conditions Directive 2013/33/EU, art 26(2)–(6). 269 See APR Proposal (n 247) art 15; and Laurent Bernhard and David Kaufmann, ‘Coping with the Asylum Challenge’ (2018) 44 JEMS 2506. 270 See Asylum Procedures Directive 2013/32/EU, arts 8(2), 22–23, 29; and Reception Conditions Directive 2013/33/EU, art 10(4). 271 See Commission v Hungary (n 107) paras 73–108. 272 ibid paras 111–33; and ch 10.8. 273 See Asylum Procedures Directive 2013/32/EU, arts 24–25; and Reception Conditions Directive 2013/33/EU, arts 21–25.

376 Common European Asylum System the corresponding processes under the Qualification Directive. They are one of the tasks today’s Asylum Agency performed in the hotspots on the Greek islands, as described in Chapter 8.4.2. The Agency will be authorised to assume additional procedural functions under the new mandate, which entered into force in 2022. Nothing in the Asylum Procedures Directive requires that the person who conducts the personal interview or prepares a decision shall be identical with the officer who formally adopts the decision concluding the asylum procedure. Directive 2013/32/EU requires applications for refugee status and subsidiary protection to be assessed jointly, unlike the predecessor instrument. This modification remains relevant insofar as Ireland continues to be bound by the former Asylum Procedures Directive 2005/ 85/EC, resulting in Court judgments on how the two procedures relate to each other.274 EU legislation employs the term ‘application for international protection’ as the overarching category nowadays.275 Nevertheless, this volume takes up everyday semantic practices and speaks of an ‘asylum application’—a terminology open enough to cover refugee status, subsidiary protection, and complementary protection statuses under national laws. For the latter, Member States may either apply the Asylum Procedures Directive or introduce separate procedural avenues.276

13.4.3 Sixfold procedural differentiation Debates about asylum procedures are defined by an inbuilt-in tension between the protection of asylum seekers, as a vulnerable group, and concerns about the unwarranted use of the asylum system by people without protection needs. Legal developments can be compared to a pendulum swinging back and forth in response to changing circumstances. Intergovernmental restrictions during the 1990s gave way to intermediate solutions in the former Asylum Procedures Directive 2005/85/EC, before today’s Asylum Procedures Directive 2013/32/EU reinforced protective elements moderately.277 For the future, the Commission Proposals of 2016 and 2020 focus on accelerated decision-making. A defining feature of supranational harmonisation has been procedural differentiation, introducing distinct avenues for specific scenarios and subject to stricter rules. Such differentiation has been criticised for ‘normalising’ second-rate standards.278

13.4.3.1 Asylum jurisdiction and admissibility First, authorities are expected to verify jurisdiction in accordance with the Dublin III Regulation, which may be omitted in line with previous comments. Secondly, an admissibility check is supposed to filter out applications where an inspection of the merits is not warranted. Admissibility checks must be distinguished from Dublin procedures for the simple that distinct standards apply: procedural guarantees in the Asylum Procedures 274 See M (n 266); and Case C-604/12 N EU:C:2014:302. 275 See Asylum Procedures Directive 2013/32/EU, art 2(b), (i). 276 ibid art 3(3). 277 See Costello and Hancox (n 246) 377; and Peers (n 119) 282–94. 278 See Cathryn Costello, ‘The Asylum Procedures Directive in Legal Context’ in Helen Toner and others (eds), Whose Freedom, Security and Justice? (Hart Publishing 2007) 151–94 passim; and Minos Mouzourakis, ‘More Laws, Less Law’ (2020) 26 ELJ 171, 173–79.

Asylum Procedures 377 Directive differ from the contents of the Dublin III Regulation.279 Article 33(2) Asylum Procedures Directive fully harmonises the conditions under which applications can be rejected as inadmissible; their meaning is determined by the Court.280 Member States cannot, for instance, introduce an additional inadmissibility standard for ‘safe transit countries’, which are not mentioned in Article 33(2).281 In practice, admissibility checks are mainly relevant for safe countries, as defined by Union law, and persons who have been granted international protection in another Member State. This inadmissibility criterion covers refugee status and subsidiary protection, not, however, complementary protection under national laws; it applies to Ireland by analogy.282 Exceptionally, Member States may be obliged to examine the merits of an application by a beneficiary of international protection if return to the country that had granted the status would amount to unhuman treatment incompatible with Article 4 CFR.283 By contrast, the Court considered family unity and the best interests of the child not to mandate such an exception284—in contrast to the opposite finding with regard to Dublin procedures.285 Judges apparently put more emphasis on the stability of state jurisdiction, once the procedure has been concluded in one country.

13.4.3.2 Regular and accelerated examination Thirdly, Member States will examine the merits of each application for which they have jurisdiction and which is not inadmissible. This shall usually be done within six months in regular circumstances.286 When deciding individual cases, refugee status takes priority over subsidiary protection.287 Applicants can challenge a positive outcome granting subsidiary protection if refugee status was refused erroneously and if the switch of status has advantages; limitations of the right to appeal in Article 46(2) must be interpreted narrowly.288 Along similar lines, subsidiary protection takes priority over complementary national statuses. Domestic authorities are obliged to the give reasons in fact and in law when rejecting an application. Such enhanced reasoning goes beyond the standard forms used in other segments of EU migration law and confirms that asylum procedures are generally subject to a comparatively high degree of sophistication.289 Member States may combine a negative asylum decision with a return decision, thus effectively aggregating the procedural requirements under the Asylum Procedures Directive and the Return Directive in a single administrative process.290 Doing so may facilitate return. The Commission wants to make the combination mandatory.291 279 See Ahmed (n 176) paras 38–39. 280 See ch 6.4.4; and Case C-564/18 Bevándorlási és Menekültügyi Hivatal (Tompa) EU:C:2020:218, paras 29–30. 281 See Országos Idegenrendészeti Főigazgatóság Dél-alföldi Regionális Igazgatóság (n 107) paras 148–65. 282 See Case C-616/19 Minister for Justice and Equality EU:C:2020:1010. 283 See Hamed and Omar (n 230) para 35. 284 See Case C-483/20 Commissaire général aux réfugiés et aux apatrides EU:C:2022:103, paras 35–37; against the Opinion of AG Priit Pikamäe. 285 See n 167 and accompanying text. 286 See Asylum Procedures Directive 2013/32/EU, art 31(3)–(5); and also APR Proposal (n 247) art 34. 287 See N (n 274) paras 29–35. 288 See Case C-662/17 EG EU:C:2018:847, paras 49, 63–69. 289 See ch 7.2.4.1; and Asylum Procedures Directive 2013/32/EU, art 11(2); in contrast to former Asylum Procedures Directive 2005/85/EC, art 9(2). 290 See Case C-181/16 Gnandi EU:C:2018:465, paras 35–59; and Case C-166/13 Mukarubega EU:C:2014:2336, paras 60–61. 291 See Amended APR Proposal (n 181) art 35a.

378 Common European Asylum System Fourthly, some applications are examined in an accelerated procedure, which the Commission similarly suggests rendering mandatory.292 Legislation has to specify to what extent accelerated procedures are subject to special rules; the abstract reference to an examination ‘in accordance with the basic principles and guarantees of Chapter II’293 does not generally authorise lower standards, since the formulation reiterates the title of Chapter II verbatim. The most important restriction concerns time limits, which must respect general principles of Union law and human rights, as explained in Chapter 7.2.2 on the administrative dimension. EU legislation fully harmonises the grounds for accelerated procedures, unlike the predecessor instrument.294 They primarily cover people from safe countries of origin, various instances of alleged bad faith, security threats, and incomplete or contradictory submissions.295 Such scenarios are prime example of what is often called ‘abuse’ in the policy discourse, to convince the EU legislature to lay down stricter standards. Accelerated procedures will often find applications to be ‘manifestly unfounded’, thus limiting the right to remain during legal remedies.296 Vulnerable persons may be exempted from accelerated procedures, depending on the individual case.297 Note that Member States may prioritise some applications in the framework of the regular procedure in scenarios where Article 31(8) does not apply.298 The effect will often be the same, although statutory exceptions, such as shorter time limits, are usually unavailable. Such (de)prioritisation of some claims over others can be an effective means to use scarce resources effectively. By way of example, asylum authorities may decide ‘easy’ cases swiftly when high numbers do not allow them to process all applicants at once. UNHCR recommends clustering as an instrument of fast and fair procedures, which had been spearheaded by the Netherlands.299

13.4.3.3 Subsequent applications and withdrawal Fifthly, subsequent applications are telling examples of the tension inherent in the legislative design of asylum procedures. On the one hand, they are essential for the effective exercise of the right to asylum when new developments take place.300 On the other hand, subsequent applications can be used to adjourn return after a first rejection.301 To distinguish both scenarios, Article 40 Asylum Procedures Directive foresees an examination in two stages. The purpose of the first stage, which is preliminary, is to verify admissibility, whereas the second stage concerns the merits of any subsequent application found to be admissible.302 It is an open question whether Member States can qualify an application as

292 See APR Proposal (n 247) art 40. 293 Asylum Procedures Directive 2013/32/EU, art 31(8) (emphasis added). 294 cf Case C-175/11 D and A EU:C:2013:45, paras 71–70. 295 See Asylum Procedures Directive 2013/32/EU, art 31(8). 296 ibid arts 32(2), 46(6)(a); additional restrictions may exist in national laws in areas not harmonised at the EU level. 297 ibid art 24(3)(2). 298 ibid art 31(4), (7). 299 See UNHCR, ‘Fair and Fast. Accelerated and Simplified Procedures in the European Union’ Discussion Paper (25 July 2018) 8–9 https://www.refworld.org/docid/5b589eef4.html (accessed 1 March 2023). 300 See Országos Idegenrendészeti Főigazgatóság Dél-alföldi Regionális Igazgatóság (n 107) para 192. 301 See Case C-921/19, EU:C:2021:478 Staatssecretaris van Justitie en Veiligheid, paras 49–51. 302 ibid paras 29–38.

Asylum Procedures 379 being ‘subsequent’ if the applicant has received an asylum decision in another Member State bound by the Asylum Procedures Directive.303 Conditions under which Member States shall proceed to the second stage are harmonised, albeit at a high level of abstraction. ‘New elements or findings’ cover a change of circumstance in the country of origin or the emergence of additional evidence in relation to past events, which had been assessed during the preceding asylum procedure already.304 Examination of these developments follows the regular rules on the burden and standards of proof, although states may restrict, according to Article 40(4), admissibility to new findings the applicant was incapable of submitting earlier.305 Additional restrictions under domestic laws have to be disapplied in the absence of an authorisation in the Asylum Procedures Directive.306 The admissibility threshold is crossed whenever new developments ‘significantly add to the likelihood’ of receiving a positive decision.307 Examination at the second stage unfolds in an accelerated procedure.308 A judgment of the Court supporting a different interpretation of the Qualification Directive than the one on which the original asylum decision had been based will not usually present a new development justifying a second examination. Settled case law on other segments of Union law recognises that legal certainty and the principle of res iudicata do not require otherwise.309 A different situation may arise when domestic asylum laws blatantly violate core principles of EU asylum legislation (as opposed to subtle discrepancies). In such a scenario, the effet utile of Union law may require domestic courts to examine subsequent applications fully.310 Finally, international protection can be withdrawn whenever ‘new elements or findings’ arise. Mirroring the rulings on subsequent applications, withdrawal may concern a change of circumstance in the country of origin or the emergence on new evidence indicating that the original decision was incorrect. Judges found that the substantive criteria for revocation cover not only misrepresentations or omissions by the applicant but also mistakes on the part of the authorities. Articles 14(3) and 19(3) Qualification Directive 2011/95/EU should be interpreted accordingly.311 Withdrawal must be performed by qualified personnel working in the asylum authorities.312 Articles 44 and 45 Asylum Procedures Directive set out basic procedural guarantees for withdrawal, which must always consider the situation of each person separately.

13.4.4 Border procedures States have restricted access to asylum procedures by means of ‘non-admission’ strategies from the 1980s onwards. Governments carved out segments of their territories where 303 See Case C-497/21 Bundesrepublik Deutschland EU:C:2022:721, paras 43–47. 304 See Case C-18/20 Bundesamt für Fremdenwesen und Asyl EU:C:2021:710, paras 35–43. 305 See Staatssecretaris van Justitie en Veiligheid (n 301) paras 40–44, 57–62. 306 See Bundesamt für Fremdenwesen und Asyl (n 304) paras 54–67. 307 See Asylum Procedures Directive 2013/32/EU, art 40(3). 308 ibid art 31(8)(f). 309 See ch 7.2.2.1. 310 See Országos Idegenrendészeti Főigazgatóság Dél-alföldi Regionális Igazgatóság (n 107) paras 185– 87, 193–202. 311 See Case C-720/17 Bilali EU:C:2019:448, paras 40–52. 312 See Case C-159/21 Országos Idegenrendeszeti Főigazgatóság and others EU:C:2022:708, paras 62–69.

380 Common European Asylum System stricter rules would apply, thus effectively erecting a second administrative border behind the territorial one.313 Such de facto extra-territorialisation remains subject to human rights, in accordance with settled case law on state jurisdiction, but legislative restrictions exist nonetheless. They are reinforced by practical deficits on the ground. Border procedures at international airports or in transit zones at the external borders are at the centre of contemporary reform debates which appeared not to move towards a conclusion. Discussions within the Parliament and the Council indicate that the draft legislation would be amended significantly; it might turn out to be a complex and unworkable procedural labyrinth. Article 43 Asylum Procedures Directive 2013/32/EU establishes a time limit of four weeks for border procedures, which concern admissibility checks and accelerated procedures only. After four weeks, access to the territory shall be granted. When many people arrive, the border procedure may be performed inside the territory, but the four-week limit stays the same.314 Vulnerable groups may be exempted, depending on personal circumstances.315 This comparatively strict set of rules followed the example of airport procedures, even though these standards apply at land borders as well. Note that the procedures in the hotspots on the Greek islands did not always qualify as a border procedure within the meaning of the Directive, not least for non-compliance with the four-week rule. National legislation might use the term ‘border procedure’ for other situations; the ‘regular’ supranational procedural standards apply in these cases.316 Border procedures are a modality of examining an application more than a distinct procedural avenue. By way of example, admissibility checks for Syrians on the Greek islands may cover arrivals from Turkey as a safe third country. To do so in the context of a border procedure does not change the standards applicable to admissibility assessments and the safe third country rule. Similarly, accelerated procedures for applicants from safe countries of origin are the same within or outside border procedures. In all these cases, however, the modalities of border procedures may differ significantly from an examination elsewhere: movement will usually be restricted in accordance with the ‘fiction of non-entry’, to be discussed in the context of reception conditions; applicants are habitually accommodated in camps, detached from local communities; reception conditions will mostly be provided in kind, by either state authorities or subcontractors; remote geographic locations undermine legal advice and other support. In 2020, the Commission proposed to significantly increase the scope and length of border procedures. Swift procedures in camp-like structures are the hallmark of the reform initiative. Border procedures shall last for up to twelve weeks (including legal remedies, unlike today’s four-week rule); the scope ratione materiae shall be extended to applicants coming from countries with an average Union-wide recognition rate of 20 per cent or lower; the ‘fiction of non-entry’ will be formalised; and Member States are scheduled to

313 See Vedsted-Hansen, ‘Non-Admission Policies’ (n 9) 269–88; and Ayelet Shachar, ‘The Shifting Border’ in Ayelet Shachar (ed), Shifting Border (Manchester UP 2020) 3, 19–25, 31–34, 41–46 on North America and Australia. 314 See Asylum Procedures Directive 2013/32/EU, art 43(2), (3). 315 ibid art 24(3)(2). 316 On the diversity of national practices see Wouter van Ballegooij and Katharina Eisele, ‘Asylum Procedures at the Border: European Implementation Assessment’ (European Implementation Assessment, (PE 654.201, November 2020).

Asylum Procedures 381 loose discretion not to apply border procedures.317 A new border procedure for carrying out return should facilitate return.318 In vaguely defined crisis situations, border procedures may be extended to applicants with an average recognition rate of up to 75 per cent and may last for up to twenty weeks, followed by the same period for return.319 Additional proposals addressing the instrumentalisation of migration by third states replicated core elements of the Commission’s border procedure proposal, albeit subject to additional leeway on the part of the Member States. An ‘emergency procedure’ was meant to last for up to 16 weeks and cover all asylum applications, irrespective of average recognition quotas; in case of rejection, the Commission wants to apply neither the Return Directive 2008/115/EC nor the border return procedure, except for some basic guarantees.320 A parallel proposal regarding confrontation with Belarus suggested limiting the right to remain during judicial oversight.321 Irrespective of whether the instruments are adopted, they will be referred to as prototypes for procedural restrictions in the future, thus effectively turning the idea of procedural differentiation upside down: stricter rules for specific scenarios gradually become the new standard.

13.4.5 Safe countries Classic destination countries in Central and Northern Europe developed the concept of safe countries, which was later emulated across the continent, as stated at the outset. Corresponding provisions can be found in Articles 35–39 Asylum Procedures Directive 2013/32/EU nowadays. They have featured prominently in policy debates over the past twenty years and form part of the ongoing reform initiative. We have to distinguish four sets of rules: (1) home states whose nationals are generally conceived not to have a well- founded fear of persecution or other serious harm (safe country of origin); (2) third states where someone was recognised as a refugee before, or otherwise enjoyed sufficient protection, independent of whether the country was listed as a safe country of origin (first country of asylum); (3) third states where applicants could have received the same treatment (safe third country); (4) countries belonging to the Council of Europe with an asylum procedure prescribed by law (European ‘supersafe’ third country).

13.4.5.1 Safe countries of origin and common lists Safe countries of origin are the least controversial element, reflecting widespread experience with (very) low recognition rates for some nationalities. Member States can adopt domestic lists of safe countries of origin, provided the situation there ‘generally and consistently’ indicates, on the basis of a range of sources, the absence of factors that would justify granting international protection.322 Such listing introduces a presumption of adequate protection, which can be rebutted by the applicant where he submits overriding reasons.323 317 See Amended APR Proposal (n 181) arts 41 (2), 41c, 41b; see also art 41(1) in conjunction with art 40(1) (i); further Jens Vedsted-Hansen, ‘Border Procedure on Asylum and Return’ in Thym and Odysseus Academic Network (n 40) 99; and den Heijer, ‘Border Procedures’ (n 261) 659–72. 318 Amended APR Proposal (n 181) art 41g; and ch 16.7.2.2. 319 See Proposal for a Crisis and Force Majeure Regulation (n 45) arts 4, 5(1). 320 See Proposal for an Instrumentalisation Regulation (n 51) arts 2(1)(b), (c), 4. 321 See Proposal for provisional emergency measures (n 78) arts 2(6), 4(1), which was not adopted. 322 See Asylum Procedures Directive 2013/32/EU, art 37, Annex I. 323 ibid art 36(1); and Case C-404/17 A EU:C:2018:588, paras 25–27.

382 Common European Asylum System Designation of safe countries of origin does not, in other words, exempt authorities from assessing the situation of each applicant individually, albeit in an accelerated procedure. Procedural and other restrictions are the main focus of the safe country of origin concept from a policy perspective. Domestic laws may introduce additional restrictions, such as a prohibition of work in case of rejection. Before activating the safe countries of origin concept, Member States have to adopt implementing legislation. EU legislation does not apply directly. Authorisation for common minimum lists of safe countries of origin and safe third countries in Articles 29 and 36 former Asylum Procedures Directive 2005/85/EC were annulled by the Court on institutional grounds.324 An initiative, on the part of the Commission, to adopt legislation designating the Western Balkan countries and Turkey as safe countries of origin at the height of the policy crisis of 2015 was withdrawn later.325 The idea of common minimum lists features in the ongoing reform debate.326

13.4.5.2 Rebuttal of the presumption of safety Provisions on safe third countries and first countries of asylum reiterate that applicants cannot chose freely where to receive protection. They are the external equivalent of the EU’s domestic Dublin system, with third states providing sufficient protection on their own initiative or in the context of interpretation cooperation. The idea behind the EU–Turkey Statement of 2016 shows how the safe third country might operate in practice. Greek authorities were expected to adopt an inadmissibility decision on grounds of the protection applicants might enjoy in the neighbouring country, partly as a result of extensive EU funding.327 This was supposed to allow swift return to Turkey on the basis of a bilateral readmission agreement. However, Syrians frequently made use of the option, enshrined in EU legislation, to challenge the presumption of safety. Court proceedings delayed the procedure significantly and overturned the administrative decision frequently. Many applicants received protection, and the number of actual returns to Turkey was small, until Turkey stopped cooperating completely; the Greek Council of State has asked the CJEU whether systematic non-cooperation means that Turkey cannot be qualified as a safe third country any longer.328 Leaving aside the Turkish case, the statutory option to rebut the presumption of safety extends to those coming from European ‘supersafe’ third countries; the legislature abandoned the original idea behind supersafety as an irrefutable presumption.329 Member States cannot unilaterally introduce stricter domestic standards anymore.330

324 See ch 2.3.3; and Case C-133/06 Parliament v Council EU:C:2008:257. 325 See Proposal for a Regulation establishing an EU common list of safe countries of origin COM(2015) 452 final; and, on withdrawal, [2019] OJ L210/13. 326 See APR Proposal (n 247) arts 46, 48–49. 327 See ‘EU–Turkey Statement’ (n 32). 328 On the situation in 2016/17 see Commission, ‘Seventh Report on the Progress made in the implementation of the EU–Turkey Statement’ COM(2017) 470 final, 5–7; and the (Greek) Council of State (Συμβούλιο της Επικρατείας), App no 177/2023 (decision of 3 March 2023). 329 Contrast Asylum Procedures Directive 2013/32/EU, arts 38(2), 39(3) with the former Directive 2005/85/EC, arts 27(2), 36(3); and see ch 12.2.4.1 on human rights. 330 See Bevándorlási és Menekültügyi Hivatal (Tompa) (n 280) paras 29–41.

Asylum Procedures 383

13.4.5.3 Controversies about safe third countries While the concept of safe countries is widely accepted to comply with the Refugee Convention and human rights law,331 conditions under which a country may be qualified as safe are controversial. That is significant, since the European Council decided in June 2017 that EU legislation ‘should be aligned with the effective requirements arising from the Geneva Convention and EU primary law’, thus calling for a decrease to the constitutional minimum.332 Debates in the Council indicate that governments prefer altering the earlier Commission proposal.333 Three elements are politically and practically relevant for the status quo and the ongoing reform debate. First, the presumption of safety must follow from legal guarantees that are effective in practice.334 The example of Turkey illustrates that a safe third state may possibly not have ratified the Refugee Convention, something Turkey has not done with regard to people coming from outside Europe. Nevertheless, the Greek Council of State concluded—without referring the matter to Luxembourg—that the designation of Turkey as safe met the condition in Article 38(1)(c) and (e) Directive 2013/32/EU of protection ‘in accordance with’ the Convention.335 This outcome forms the basis of the latest reform proposal and might comply with international minimum requirements.336 It is irrelevant whether the Refugee Convention has been ratified, as long as effective protection is available in law and fact in a safe third country. Secondly, safe third countries must stand ready to readmit the person. In the absence of readmission, Member States cannot reject an application as inadmissible.337 One step further, the Asylum Procedures Directive currently requires applicants to have a ‘connection’ to the third country on the basis of which ‘it would be reasonable for [the applicant] to go to that country’338—a condition the Court interprets to require more than transit.339 The Commission wants to continue the connection benchmark, while indicating that transit was sufficient.340 It could arguably be abandoned altogether, on account of not being mandatory under international law.341 Moreover, the Council proposes that parts of a country can be qualified as safe and that the presumption of safety could even be activated without a third state having been qualified as being generally safe.342 The latter proposal mirrors the notion of first country of asylum. 331 See Goodwin-Gill and McAdam (n 243) 436–47. 332 See European Council, ‘Conclusions’ (EUCO 8/17, 23 June 2017) para 23; and, on the background, Daniel Thym, ‘Expert Opinion on the Reform of the Common European Asylum System’ (January 2017) 23–26 http:// ssrn.com/abstract=3163014 (accessed 1 March 2023) 5–40. 333 See Presidency, ‘Amended proposal’ (Council doc 7166/23, 10 March 2023); building upon APR Proposal (n 247) arts 43a–50. 334 See James Hathaway, The Rights of Refugees under International Law (2nd edn, CUP 2021) 366–75. 335 See Council of State (Συμβούλιο της Επικρατείας), App nos 2347/2017 and 2348/2017 (judgments of 22 September 2017); and Roman Lehner, ‘The EU–Turkey-“Deal”’ (2019) 57 Int Migration 176. 336 See APR Proposal, art 45(1)(a)–(c), as revised by Presidency (n 333); and UNHCR, ‘Summary Conclusions on the Concept of “Effective Protection”’ (Lisbon Expert Roundtable, February 2003) No 15(3) https://www.unhcr. org/3e5f323d7.html (accessed 1 March 2023). 337 See Asylum Procedures Directive 2013/32/EU, arts 38(4), 39(6). 338 ibid art 38(2)(a). 339 See Bevándorlási és Menekültügyi Hivatal (Tompa) (n 280) paras 44–50. 340 See APR Proposal, art 45(2b)(b), recital 37, as revised by Presidency (n 333). 341 See UNHCR, ‘Legal Considerations Regarding Access to Protection and a Connection between the Refugee and the Third Country’ (April 2018) No 6 https://www.refworld.org/docid/5acb33ad4.html (accessed 1 March 2023). 342 See APR Proposal, as revised by Presidency, (n 336) arts 45(1a), (2)(b).

384 Common European Asylum System Thirdly, EU legislation currently requires safe third countries to offer the possibility to request refugee status and ‘to receive protection in accordance with the Geneva Convention’.343 Doing so effectively requires third states to guarantee a set of socioeconomic rights, including access to the labour market and social assistance under Articles 17– 24 Refugee Convention. A core element of the ongoing reform debate is the introduction of a generic notion of ‘effective protection’ covering an adequate standard of living, emergency healthcare, and primary education, not, however, labour market access and domestic treatment with regard to social assistance.344 The idea builds upon international standards below the threshold of Articles 17–24 Refugee Convention, which—unlike the prohibition of refoulement—apply only to people who have been granted formal refugee status, not other refugees.345 An adoption of the reform would facilitate the application of the safe third country provision to neighbouring states, building on the existing notion of ‘sufficient protection’ with regard to first countries of asylum,346 which, unlike safe third countries, covers only individuals who have enjoyed such treatment before leaving for the EU.

13.4.6 Legal remedies Article 46(3) Asylum Procedures Directive 2013/32/EU obliges Member States to make available an effective legal remedy that ‘provides for a full and ex nunc examination of both facts and points of law’. Statistics show that this guarantee is essential: many asylum decisions are corrected by domestic courts. Comments on the achievements and limitations of the preliminary reference procedure in asylum cases can be found in Chapter 3.3 on the court architecture. In recent years, the Court has delivered important judgments disapproving of domestic laws restricting judicial oversight under recourse to general principles Articles 41 and 47 CFR. They were elaborated upon in Chapter 7.2 on the administrative dimension. Commission proposals aim at further harmonisation, for border procedures in particular: time limits for appeals shall be reduced to at least one week; domestic courts are supposed to take a decision within two months; and judicial oversight is scheduled to be reduced to one instance.347 While the second and third reform are constitutionally unproblematic, provided that domestic courts retain the option to prolong their deliberations in complex cases, short time limits have been found to comply with human rights in exceptional circumstances only.348 Moreover, the Commission wants to limit the right to remain during legal remedies against negative decisions taken in the border procedure. Applicants shall have a very short time-limit of five days, mitigated by free legal assistance, to ask domestic courts to stay return proceedings; they can be deported once a court has rejected that request.349 This reform 343 Asylum Procedures Directive 2013/32/EU, art 38(1)(e). 344 See APR Proposal, art 43a, as revised by Presidency (n 336). 345 See ch 13.9.4; Stephen H Legomsky, ‘Secondary Refugee Movements and the Return of Asylum Seekers to Third Countries’ (2003) 15 IJRL 567, 633–53; and Thym (n 328) 18–27. 346 See Asylum Procedures Directive 2013/32/EU, art 36(1)(b); and Case C-585/16 Alheto EU:C:2018:584, paras 137–43. 347 See Amended APR Proposal (n 181) art 53(7)(a), (9); and APR Proposal (n 247) art 55(1)(b). 348 See ch 7.2.2.2 and 7.2.3.1. 349 See Amended APR Proposal (n 181) art 54(3)(3), (4)–(5).

Asylum Procedures 385 project was closely modelled on case law, described in Chapter 7.2.3.3, circumscribing the outer limits of the right to remain in light of Article 47 CFR. At present, the right to remain endures during the entire first instance of judicial oversight, except for certain manifestly unfounded or inadmissible applications.350 Once the right to remain expires, migrants can be returned; applicants challenging a negative asylum decision are staying illegally for the purposes of the Return Directive.351 Proposals to mandatorily combine asylum and return decisions are meant to speed up the overall process.

13.4.7 Recurring ‘vision’: external processing Debates about outsourcing asylum applications to third states have accompanied the evolution of EU asylum policy, building upon the examples of the United States and Australia.352 A British initiative, put forward in 2003, spurred considerable debate, including a feasibility study commissioned by the Commission; it was taken up on several occasions in the following years.353 In 2018, the European Council officially embraced the idea in the form of ‘regional disembarkation platforms’ in third states, where migrants rescued in the Mediterranean would be brought to assess protection needs.354 The idea was not realised, however, since no neighbouring state was willing to host such a disembarkation platform for people rescued at sea. More recently, Denmark and the United Kingdom have adopted laws which would allow for the transfer of asylum seekers elsewhere.355 In April 2022, the British government signed a Memorandum of Understanding with Rwanda. External processing will be performed by Rwandan authorities in accordance with national laws and the standards in the MoU; refugees would be given the prospect of permanent settlement, whereas ‘a portion’ of Rwanda’s most vulnerable refugees would be resettled in the UK.356 This controversial initiative waited to be implemented at the time of writing, after the ECtHR had provisionally stopped the first return flight to give domestic courts time to examine the legality of the transfer arrangements; the first instance held the scheme to be legal.357 Hungarian practices requiring applicants to seek asylum in consulates abroad violate Union law, since the country does not, unlike Denmark, have an opt-out from the Asylum Procedures Directive.

350 See Directive 2013/32/EU, art 46(5)–(8); and Commission v Hungary (n 256) paras 281–314. 351 See ch 16.5.1. 352 See Daniel Ghezelbash, Refuge Lost (CUP 2018) ch 5; and Maarten den Heijer, Europe and Extraterritorial Asylum (Hart Publishing 2012) ch 7. 353 See Gregor Noll, ‘Visions of the Exceptional’ (2003) 5 EJML 303, 304–24; Violeta Moreno-Lax, ‘External Dimension’ in Peers and others (n 110) 617, 644–64; and Carl Levy, ‘Refugees, Europe, Camps/State of Exception’ (2010) 29 Refugee Surv Q 92. 354 See European Council, ‘Conclusions’ (EUCO 9/18, 28 June 2018) para 4; and the Commission’s non-paper, attached to ‘Managing migration’ (Press Release IP 18/4629, 24 July 2018) https://ec.europa.eu/commission/pres scorner/detail/en/IP_18_4629 (accessed 1 March 2023). 355 See Nikolas Feith Tan and Jens Vedsted-Hansen, ‘Denmark’s Legislation on Extraterritorial Asylum in Light of International and EU Law’ EU Immigration and Asylum Law (15 November 2021); and UNHCR, ‘Observations on the Nationality and Borders Bill’ (October 2021). 356 See Memorandum of Understanding for the provision of an asylum partnership arrangement (signed 14 April 2022) https://www.gov.uk/government/publications/memorandum-of-understanding-mou-between-the- uk-and-rwanda (accessed 1 March 2023). 357 See KN v United Kingdom App no 28774/22 (ECtHR, 14 June 2022); and High Court AAA and others v Secretary of State for the Home Department [2022] EWHC 3230.

386 Common European Asylum System As a matter of principle, the EU Treaties do not categorically prohibit external processing. Article 78(2)(d) and (g) TFEU were drafted in an open manner, in parallel to the debate about the British initiative of 2003.358 Having said this, external processing would require an amendment of the Asylum Procedures Directive.359 Moreover, third states appear mostly unwilling to consent to the transfer, since they do not want to be perceived as servants of EU interests. Rwanda arguably used the deal with the UK to present itself as a modern and safe country. Finally, the examples of the EU–Turkey Statement (based on the safe third country provision, not external processing) and of the evacuation of refugees from Libya to Niger, mentioned in the resettlement context, illustrate that variations on external processing are possible. Formal external processing will be confronted with myriad political, practical, and legal challenges. Politically, third states would have to be convinced and domestic and international criticism of burden-shifting would have to be overcome. UNHCR, for instance, supports inter-state cooperation but worries about the fallouts of lopsided initiatives.360 Practically, external processing could come in many forms. Status determination and reception conditions could be provided, jointly or cooperatively, by third states, international organisations, private service providers, EU agencies, or Member States. Different options exist as to the applicable laws (EU legislation, international standards, domestic laws of the third country), judicial oversight (in the EU or abroad), and the fate of applicants after processing (resettlement, return, protection in situ). Legally, the danger of severe human rights violations is real. Reception conditions could turn out to be sub-standard; procedural deficiencies might result in illegal refoulement; and indefinite detention would be unacceptable.361 Much would depend on whether practical arrangements amounted to an exercise of jurisdiction by the Member States or the EU institutions, thus triggering the applicability of human rights standards.362 These multifaceted challenges may be reason enough not to pursue the idea further. Yet discussions alone can have a tangible impact even in the absence of direct realisation. The British initiative of 2003, for example, paved the way for refugee protection initiatives in the Global South and the spread of control-oriented externalisation practices, which will be traced in Chapter 18.3 and 18.4. Instead of external processing, European actors invested considerable resources in reducing the number of arrivals.

13.5 Refugee Status Recital 23 Qualification Directive 2011/95/EU states paradigmatically that the Directive shall ‘guide the competent national bodies . . . in the application of the Geneva Convention’. To specify the meaning of its highly abstract provisions was an essential added value 358 See Thym (n 65) MN 26–27, 35. 359 Only persons rescued on the high seas are not covered by Asylum Procedures Directive 2013/32/EU, arts 3, 8(1). 360 See ‘UNHCR’s Three-Pronged Proposal’ (January 2003) pt B https://www.refworld.org/docid/3efc4b834. html (accessed 1 March 2023). 361 See den Heijer, Extraterritorial Asylum (n 352) 268–81; and Orla Lynskey, ‘A Legal Analysis of the Emerging Extraterritorial Elements of EU Refugee Protection Policy’ (2006) 31 EL Rev 230, 240–47. 362 See ch 12.2.2; Noll (n 353) 325–38; and Violeta Moreno-Lax and Martin Lemberg-Pedersen, ‘Border- Induced Displacement’ (2019) 56 Questions Int L 5, 12–25.

Refugee Status 387 of Europeanisation. Chapter 5.6 explained that the absence of an international body interpreting the Convention authoritatively is a structural weakness. EU migration law filled this lacuna. For all practical purposes, the Qualification Directive serves as the primary frame of reference for decision-making nowadays. Nevertheless, the introduction of common standards need not result in uniform outcomes (13.5.1). Standards and burden of proof govern the assessment of each individual case (13.5.2). A critical novelty of the Qualification Directive was legislative guidance on the meaning of ‘persecution’ (13.5.3), actors of protection (13.5.4), and reasons for persecution (13.5.5). Serious wrongdoing or a change of circumstance may justify the refusal or ending of refugee status (13.5.6). These themes are, together with the Dublin system and asylum procedure, among the topics that have featured most prominently in the supranational case law. At the time of writing, it appeared likely that a minor revision of the Qualification Directive would be adopted before the European elects in the spring of 2024.

13.5.1 Preventing disparate recognition quotas Stark discrepancies in recognition rates for important countries of origin are a constant reminder that agreement on legal criteria for refugee status is a necessary but insufficient condition for coherent outcomes. By way of example, recognition rates in Italy have traditionally been at the upper end for most nationalities, standing at 30 per cent for Nigerians during 2020 (as opposed to hardly exceeding 10 per cent elsewhere). In Austria, more Afghans received refugee status than subsidiary or complementary protection, while the situation was the reverse in neighbouring Germany.363 Some of these discrepancies may be rationalised by distinct profiles of applicants, but the degree and endurance of the asymmetries hint at a deeper problem. Asylum decision-making requires a forward-looking prognosis as to whether the applicant would be persecuted upon return to the home state. Doing so involves an assessment of the general situation and of individual circumstances in distant places, which are often subject to an enhanced degree of uncertainty. Reliable information and the need for case- specific conclusions are critical variables that explain asymmetric outcomes. Administrative convergence requires practical guidance and coordination, on top of the uniform interpretation of the legal standards. As a result, transforming the Qualification Directive into a Regulation, as the Commission proposed in 2016, would not, in itself, change much in practice.364 Directive 2011/95/EU had achieved a high degree of harmonisation already. Chapter 10.6 explained that important provisions embody fully harmonisation. The rest depends on administrative practices. That is why, in addition to legislative harmonisation, attempts at improving coherence should focus on practical measures, such as training and information gathering. Chapter 8.4.1 illustrated that the Asylum Agency increasingly performs valuable tasks in this respect, such as the collection and constant update of ‘country of origin information’ (COI). Readers may similarly benefit from the case law database, set up by the Agency, and

363 See EASO Asylum Report 2021 (EASO 2021) 205–206. 364

See Proposal for a Qualification Regulation COM(2016) 466 final.

388 Common European Asylum System the non-governmental equivalent, coordinated by ECRE, to identify relevant supranational and domestic judgments.365 National case law remains highly relevant, both for questions of legal interpretation that have not reached the Court and for assessing the factual situation. Chapter 3.3.3 illustrated that the CJEU’s output fluctuates between micromanagement and vagueness regarding the assessment of the case behind a preliminary reference. In the field of asylum, these general limitations are reinforced by the well-established division of labour according to which Luxembourg concentrates on interpreting Union law, while referring courts have to assess the facts. The ECtHR will rarely be able to step in to establish a pan-European standard for how to assess specific scenarios. Judgments are usually delivered with years of delay after the exhaustion of local remedies. Moreover, they concern human rights, not the refugee definition.366 A distinct ‘country guidance system’ with lead judgments on the situation in specific third states, following the British model,367 does not exist elsewhere, let alone at the supranational level. National courts may partly bridge the gap by actively reflecting the position of peers in other states on how to assess the situation in specific states in an exercise of transnational dialogue. All of this indicates that some degree of disparity on the ground is bound to stay. Empirical studies on asylum decision-making in the United States and Germany have unearthed similar discrepancies in the practices of administrative authorities and courts within one country, resulting in a sort of ‘refugee roulette’.368 Don’t get me wrong. To recognise the structural limitations of the search for practical coherence is not meant to vindicate the stark discrepancies in recognition rates within the EU. The Asylum Agency, national asylum authorities, the CJEU, the ECtHR, and domestic courts should do their best to reduce the incoherence.

13.5.2 Standards and burden of proof Standards and the burden of proof are of critical relevance for asylum decision-making, especially in borderline scenarios where the claim to protection hangs in the balance. Article 4 Qualification Directive 2011/95/EU lays down basic principles which must be read in light of the case law: the benchmarks for refugee status (‘well-founded fear’) and subsidiary protection (‘real risk’) establish a comparable standard of proof, mirroring settled ECtHR case law on Article 3 ECHR. Status determination must be carried out on an individual basis with vigilance and care, in the light of the applicant’s personal circumstances.369 The assessment is meant to be forward-looking considering what would happen in case of return to the country of origin. 365 See https://caselaw.euaa.europa.eu and https://asylumineurope.org (accessed 1 March 2023). 366 See Hugo Storey, ‘Briefing Note for Participants’ (2013) 25 IJRL 328, 334–46. 367 See Sir Nicholas Blake, ‘Luxembourg, Strasbourg and the National Court’ (2013) 25 IJRL 349. 368 See Jaya Ramji-Nogales, Andrew I Schoenholtz, and Philip G Schrag, Refugee Roulette (NYU Press 2009); and Gerald Schneider, Nadine Segadlo, and Miriam Leue, ‘Forty-Eight Shades of Germany’ (2020) 29 German Politics 546. 369 See Joined Cases C-175/08, C-176/08, C-178/08 and C-179/08 Abdulla EU:C:2010:105, para 90; Joined Cases C-71/11 and C-99/11 Y and Z EU:C:2012:518, para 68; and FG v Sweden App no 43611/11 (ECtHR [GC], 23 March 2016) §§ 111–12.

Refugee Status 389 According to Article 4(1) and ECtHR case law, it is in principle for the applicant to adduce evidence of ill-treatment, while state authorities are obliged to dispel any doubts; such evidence must provide a substantiated and credible account but need not encompass direct documentary proof, which can be difficult to obtain.370 Notwithstanding the obligation of individuals to substantiate their claim, publicly available information about third states, including country of origin reports and guidance by the Asylum Agency, must be considered proprio motu.371 The wording of Article 4(3)(a) mentions this crucial obligation rather briefly, unlike Article 8(1) Qualification Directive on internal protection alternatives and Article 37(3) Asylum Procedures Directive 2013/32/EU on safe countries of origin. Domestic authorities and courts can also request external expertise, without being obliged to follow the expert advice.372 It ultimately depends on the nature of the claim to protection whether the general situation or personal circumstances will determine the outcome; often, both elements will coalesce to varying degrees. In many cases, the unavailability of direct documentary or other evidence complicates the prognosis of a ‘well-founded fear’ or ‘real risk’. In such situations, domestic authorities and courts retain the responsibility to ascertain whether the information available makes it sufficiently plausible that there is a well-founded fear or real risk.373 Judgments have shed light on several factors that may be relevant: explanations for the lack of direct evidence; failure to authenticate a document or other source cannot, in itself, justify the exclusion from the examination; initial silence on sensitive matters need not be interpreted to the disadvantage of the applicant.374 Rather, credibility has to be assessed on an individual basis taking into account the criteria listed in Article 4(5) Qualification Directive, in particular whether the applicant’s general credibility has been established and whether his statements are found to be coherent and plausible.375 Despite of this, the credibility assessment inevitably retains an element of context-dependent fluidity.376 In a judgment of principle, the Court restricted the methods authorities may use when establishing the facts and assessing credibility.377 Questions based on stereotyped notions as to the behaviour of homosexuals, films of intimate acts, and pseudo-scientific ‘tests’ cannot be relied upon, in light of Article 7 CFR.378 At the same time, judges did not demand reliance on the applicant’s declarations alone; other information shall be taken into account.379 The personal interview may consider sensitive personal information, for instance regarding

370 See the summary in JK v Sweden App no 59166/12 (ECtHR [GC], 23 August 2016) §§ 91–97; Case C-349/ 20 Secretary of State for the Home Department EU:C:2022:151, paras 63–65; and also Ida Staffans, Evidence in European Asylum Procedures (Martinus Nijhoff 2012). 371 JK v Sweden (n 370) § 98; Fathi (n 182) para 98; Case C-238/19 Bundesamt für Migration und Flüchtlinge EU:C:2020:945, paras 54–56; and Joined Cases C-199/12–C-201/12 X and others EU:C:2013:720, paras 58–59. 372 See Case C-473/16 F EU:C:2018:36, paras 27–46. 373 See Case C-472/13 Shepherd EU:C:2015:117, paras 42–43. 374 See M (n 266) para 36; Staatssecretaris van Justitie en Veiligheid (n 301) paras 40–44; and Joined Cases C-148/ 13–C-150/13 A, B and C EU:C:2014:2406, paras 67–71. 375 See Fathi (n 182) paras 86–87; A, B and C (n 374) paras 48–49; and International Association of Refugee Law Judges, ‘Assessment of Credibility in Refugee and Subsidiary Protection Claims’ (March 2013) https://www.refwo rld.org/docid/557028564.html (accessed 1 March 2023). 376 See ch 4.2.4; and Carolus Grütters, Elspeth Guild, and Sebastiaan de Groot (eds), Assessment of Credibility by Judges in Asylum Cases in the EU (Wolf 2013). 377 See A, B and C (n 374) paras 53–54, 59. 378 ibid paras 60–66; and F (n 372) paras 27–46. 379 A, B and C (n 374) paras 53–59 was stricter than AG Eleanor Sharpston.

390 Common European Asylum System religious affiliation, as long as the contents and context does not amount to a violation of the right to privacy.380 Notwithstanding the forward-looking character of the prognosis of ‘well-founded fear’, past ill-treatment often plays a critical role. Article 4(4) Qualification Directive qualifies such earlier instances as a ‘serious indication’ that persecution would occur again. If past ill-treatment has been established with sufficient plausibility, state authorities need ‘good reasons’ to conclude otherwise, thus effectively introducing a sort of rebuttable presumption of well-founded fear.381

13.5.3 Notion of persecution Article 2(d) Qualification Directive 2011/95/EU reproduces the basic features of the refugee definition in Articles 1A(2) and 1F Refugee Convention. For analytical purposes, the definition can be broken down into five elements, which inform the structure of our comments: ‘well-founded fear’—‘of being persecuted’—‘unable or unwilling to avail oneself of the protection of the home country’—‘for reasons of race, religion, etc.’—‘the exclusion clause does not apply’. Most of these elements are not specified further in the Refugee Convention, and it was an important achievement of EU harmonisation to have established common standards.

13.5.3.1 Severe violation of basic human rights International practice on the meaning of ‘persecution’ has developed over time and continues to cause controversies.382 Article 9(1) Qualification Directive opted for a generous reading: acts of persecution are defined as a ‘severe violation of basic human rights’, or an accumulation of measures affecting individuals in a similar manner.383 This interpretation had been spearheaded by James Hathaway in the early 1990s, reflecting the predominance of human rights in the legal discourse at the time; it effectively prepared the ground for an extension of the refugee definition to, amongst others, persecution by non-state actors and gender-specific considerations.384 The refugee definition relates to foreigners who are ‘outside the country of nationality’ and who cannot avail themselves of the protection ‘of that country’. As a result, the forward- looking prognosis of persecution concerns return to the home state, not the country of previous residence. For example, applications by Afghan nationals having spent all their lives in Iran will be assessed in light of the situation in Afghanistan; return to Iran would require its classification as a safe third country. However, the country of former habitual residence is the relevant benchmark for de jure stateless persons; applicants with dual nationality

380 See Fathi (n 182) para 89. 381 See Harald Dörig, ‘Asylum Qualification Directive 2011/95/EU’ in Thym and Hailbronner (n 65) Article 4 MN 30–33; and also Abdulla (n 369) paras 92–98. 382 See James C Hathaway and Michelle Foster, The Law of Refugee Status (2nd edn, CUP 2014) ch 3; and Goodwin-Gill and McAdam, International Law (n 243) 67–76. 383 This definition built upon Joint Position 96/196/JHA on the harmonized application of the definition of the term ‘refugee’ [1996] OJ L63/2, No 4. 384 See James C Hathaway, The Law of Refugee Status (1st edn, Butterworths 1991) 101–105; and, for critical reflection, Romit Bhandari, Human Rights and The Revision of Refugee Law (Routledge 2020).

Refugee Status 391 are expected to return to any home state.385 A Protocol on the examination of asylum applications made by Union citizens was mentioned in the context of the constitutional foundations. While it may be of intuitive appeal to define persecution as a ‘severe violation of basic human rights’, the linkage gives rise to uncertainties. It was originally conceived of as referring primarily to civil and political rights (not economic, social, and cultural ones).386 That view is supported by the reference ‘in particular’ to non-derogability under Article 15(2) ECHR in Article 9(2)(a) Qualification Directive, as well as by the invocation of threats to ‘life and freedom’ in Article 33(1) Refugee Convention. In spite of this, the wording of the Qualification Directive remains open-ended and may possibly be read to embrace other harm. This will be exemplified with regard to climate change and poverty on the coming pages. The specimen of acts of persecution listed in a non-exhaustive manner in Article 9(2) are similarly flexible, for instance when it comes to the definition of ‘disproportionate or discriminatory’ punishment.387 The Court transcended the focus on non-derogability early on, while distinguishing categorically, as we shall see, civil and political rights from economic hardship. Freedom of religion was classified as one of the foundations of a democratic society, even though it is not listed as non-derogable right in Article 15(2) ECHR.388 Nevertheless, not any infringement will constitute persecution. Rather, judges relied on the ‘severity’ of the measure as the benchmark to distinguish human rights violations below and above the threshold of persecution.389 In casu, the Court rejected the idea that restrictions of religious activities in public will not usually amount to persecution, provided that applicants could practice their faith in private. On that basis, it deferred to the domestic court to determine whether the restrictions on the public exercise of certain religions in Pakistan were severe enough to qualify as persecution or not.390 The basic contours of the ‘severity test’ were reaffirmed with regard to homosexuality, were the Court stated explicitly: ‘not all violations of fundamental rights suffered by a gay applicant will necessarily reach that level of seriousness’.391 More specifically, judges required an examination of whether the legislation and practices of the home state sanctioning homosexuality. While criminalisation as such does not meet the severity threshold, imprisonment was found capable of constituting an act of persecution, provided that it is actually applied.392 Moreover, the Court reiterated an earlier finding, with regard to public religious activities, that applicants cannot be expected to refrain from exercising certain acts to prevent punishment.393 Judges remained silent on whether other sanctions than imprisonment would amount to persecution. Such instances will have to be assessed in light of the open-ended severity benchmark.

385 See Qualification Directive 2011/95/EU, art 2(d), (n); and Case C-91/20 Bundesrepublik Deutschland EU:C:2021:898, paras 29–33. 386 See Hathaway (n 384) 112–24; the second edition adopted a broader outlook; see Hathaway and Foster (n 382) 228–38. 387 See generally Shepherd (n 373) paras 49–55. 388 See Y and Z (n 369) paras 53–57. 389 ibid paras 58–61, 65–67. 390 ibid paras 62–64, 68–70; and also Fathi (n 182) paras 78–83. 391 X and others (n 371) paras 51–54. 392 ibid paras 55–57. 393 ibid paras 65–76; and Y and Z (n 369) paras 73–80.

392 Common European Asylum System

13.5.3.2 Refugees sur place and family members It is inherent in a forward-looking prognosis that the fear of persecution may arise after someone left the home state—a phenomenon the international debate habitually refers to as refugees sur place (French for ‘on the spot’). It may arise in two scenarios: a change of circumstance in the country of origin, such as a military takeover, or activities on the part of the applicant outside the home state. The second scenario can give rise to legal and practical controversies if the applicant is accused of having ‘constructed’ the case to prevent return; classic examples are religious conversion and political activities in exile. Article 5(2) and (3) Qualification Directive can be read to require a causal link between these activities and previous convictions, although the open formulation (‘in particular’, ‘should not normally’) and the international legal context support a narrow interpretation.394 Activities after departure can give rise to refugee status but do require a particularly careful credibility assessment.395 A unique occurrence was Case C-563/10: the domestic court withdrew the preliminary reference whether the applicant could be expected to live as a gay man in secret in Iran, since the publication of his full name on the CURIA website meant that everybody could know, so he received international protection. Applications by family members can be assessed jointly, provided that state authorities distinguish the situation of each individual.396 EU asylum legislation does not follow the example of some Member States which automatically extend the refugee status of one family member to the other members of the nuclear family; they may, however, be eligible for the same status if the actor of persecution threatens family members as well.397 If not, they will benefit from the rights mentioned in Article 23 Qualification Directive. Moreover, Member States may automatically extend the protection status under more favourable domestic legislation in accordance with Article 3 Qualification Directive, as discussed in Chapter 10.6 on overarching features. 13.5.3.3 Poverty Climate change and poverty raise interrelated questions. To start with, the example of poverty reiterates the indeterminacy of the EU’s definition of persecution as a ‘severe violation of basic human rights’. The context and background may support, as we have seen, a focus on grave violations of civil and political rights. However, this interpretation is not automatic. Arguments have been put forward that the refugee definition should be extended to socio-economic deprivation, notably failure to guarantee minimum core obligations under the International Covenant on Economic, Social, and Cultural Rights and, possibly, the European Social Charter.398 The Court has not yet dealt with the issue explicitly but has hinted at a narrow reading on at least two occasions. First, judges stated categorically, in the context of cessation, that ‘economic hardship cannot, as a general rule, be classified as “persecution” ’.399 Judges did not 394 See Hugo Storey, ‘EU Refugee Qualification Directive’ (2008) 20 IJRL 1, 27–28; and Hathaway and Foster (n 382) 75–90. 395 See Qualification Directive 2011/95/EU, art 4(3)(d). 396 See Case C-652/16 Ahmedbekova EU:C:2018:801, paras 52–65. 397 ibid paras 44–51, 68. 398 See Michelle Foster, International Refugee Law and Socio-Economic Rights (CUP 2007); Meta Oepen-Mathey, Internationaler Schutz im Falle extremer Armut (Nomos 2021) 104–265; and generally Alexander Betts, ‘Survival Migration: A New Protection Framework’ (2010) 16 Global Governance 361. 399 Case C-255/19 Secretary of State for the Home Department EU:C:2021:36, para 49.

Refugee Status 393 rely on the ‘severity test’ but stated that the concepts of persecution and protection ‘are intrinsically linked’400, thus emphasising, indirectly at least, the absence of an actor of persecution to be held responsible for economic hardship in situations of poverty. Secondly, this argument reiterates the basic premise of the case law on medical cases, discussed in the next section on subsidiary protection. Judges found the ‘rationale of international protection’ to respond to serious harm inflicted by an actor of persecution under Article 6, not, however, ‘general shortcomings in the health system of the country of origin’.401 A different conclusion would require intentional or discriminatory access to social services, for instance to the detriment of ethnic minorities.402

13.5.3.4 Climate change The legal categorisation of the consequences of climate change can build upon our comments on poverty. Why? Persons do not usually escape from climate change in the abstract. Rather, climate-induced migration responds to its effects, either direct ones like natural disasters (floods, fires, rising sea levels) or indirect consequences in the form of economic hardship or civil wars (some claim that the Syrian civil war was caused, partly at least, by draughts and disputes about water usage). While it may be possible to show a direct linkage between climate change and natural disasters, subject to uncertainties about the degree of causality, the connection between climate change and other economic, social, or political ‘push factors’ will rarely be identified with sufficient plausibility. Uncertainties about the drivers of migration, discussed in the interdisciplinary Chapter 4.1, are amplified by scientific arguments about the impact of climate change.403 This matters, since the refugee definition offers protection, for legitimate reasons of legal certainty, against discernible acts of persecution, irrespective of distant root causes. Against this background, it is irrelevant, for the purposes of examining asylum applications, whether, or to which degree, the civil war in Syria or dire living conditions in the Sahel zone were caused by climate change. What matters is whether the factual situation is covered by the doctrinal standards for refugee status or subsidiary protection enshrined in Articles 6–10 and 15 Qualification Directive. While civil wars may be covered, settled case law does not usually allow us to reach the same conclusion for poverty and other hardship caused by natural disasters or creeping environmental change, in the absence of an actor of persecution.404 Enhanced protection may be available under human rights law, albeit subject to the high threshold for socio-economic deprivation.405 The Human Rights Committee supported this view with regard to rising sea levels, while rejecting the complaint of a Pacific islander in the absence of an immediate threat.406 One may use, moreover, the analytical perspective of climate change to try to reverse the established wisdom, underlying the CJEU’s and ECtHR’s case law, that natural disasters, economic hardship, and environmental degeneration do not 400 ibid paras 54–62. 401 See M’Bodj (n 71) paras 35–37, 44. 402 ibid paras 35–36; and Case C-353/16 MP EU:C:2018:276, paras 51, 57. 403 See Michael Berlemann and Max Friedrich Steinhardt, ‘Climate Change, Natural Disasters, and Migration. A Survey of the Empirical Evidence’ (2017) 63 CESifo Econ Stud 353. 404 See Jane McAdam, Climate Change, Forced Migration, and International Law (OUP 2012) 42–48. 405 See ch 13.2.2; McAdam, Climate Change (n 404) ch 3; and Isabel M Borges, Environmental Change, Forced Displacement and International Law (Routledge 2019) ch 2. 406 See Ioane Teitiota v New Zealand Communication no 2728/2016 (HRC, 7 January 2020) Nos 9.11–14.

394 Common European Asylum System amount to persecution, given that they are the indirect, albeit multi-causal, effect of human behaviour and the inaction of states.407 For the time being, however, doing so remains an uphill struggle. The political debate about ‘climate refugees’—or ‘climate migration’, as some prefer to say408—does not primarily concern the refugee definition anyway. The term ‘climate refugee’ is often used, for instance by the Council of Europe, to gather political momentum in support of policy initiatives in diverse domains, amongst others to strengthen economic and social resilience of local communities in the areas most affected by climate change.409 Chapter 5.5.4 on international law mentioned that the non-binding Global Compacts, adopted in 2018, discussed the matter in the context of safe and orderly migration, not the parallel document on refugee protection.

13.5.4 Actors of protection Protection features prominently in the refugee definition, yet the exact meaning had often remained nebulous. Protection is commonly discussed in relation to two scenarios nowadays, covered explicitly by Articles 6–8 Qualification Directive 2011/95/EU: persecution by non-state actors and the internal protection alternative. Europeanisation played a critical role, as we have seen, in the proliferation of these concepts, thus partly extending and partly restricting the level of protection that had been available under domestic laws previously.

13.5.4.1 Home state and international organisations Refugee law is commonly described to be subsidiary insofar as applicants must avail themselves of the protection that is available by or in the home state, notably against persecution by non-state actors. To embrace persecution by non-state actors had been a major result of Europeanisation, bringing to an end the former contrast between ‘accountability’ and ‘protection’ theories in the interpretation of the Refugee Convention.410 Examples of persecution by non-state actors are female genital mutilation or gang violence, provided we conclude, with regard to the latter, that the level of harm clears the severity hurdle for an act of persecution. Extending the refugee definition to persecution by non-state actors leaves us with the question when home states, or other actors, offer sufficient protection against non-state perpetrators. Article 7(2) Qualification Directive defines the relevant level of protection through a combination of obligations of effect and conduct: home states must be willing and able to provide protection effectively and non-temporarily;411 this threshold will usually be met when reasonable steps have been taken to prevent non-state persecution ‘by

407 See Matthew Scott, Climate Change, Disasters, and the Refugee Convention (CUP 2020); and Matthew Scott, ‘Natural Disasters, Climate Change and Non-Refoulement’ (2014) 26 IJRL 404. 408 See Elisa Fornalé, ‘A l’envers. Setting the Stage for a Protective Environment to Deal with “Climate Refugees” in Europe’ (2020) 22 EJML 518, 521–25. 409 See PACE, ‘A Legal Status for “Climate Refugees” ’ (Resolution 2307 (2019), 3 October 2019); and Presidency, ‘Discussion Paper’ (Council doc 13376/21, 9 November 2021). 410 See Thomas Spijkerboer, ‘Subsidiarity in Asylum Law’ in Daphné Bouteillet-Paquet (ed), Subsidiary Protection of Refugees in the European Union (Bruylant 2002) 19, 29–36. 411 See also JK v Sweden (n 370) § 80.

Refugee Status 395 operating a functioning legal system’, provided the applicant has access to it.412 Whether that is the case or not ultimately depends on the general situation in the country concerned and personal circumstances. While the inclusion of non-state persecution was generally welcomed by critical observers, the extension of the actors of protection to ‘international organisations’ in Article 7(1) (b) Qualification Directive has been met with criticism.413 In a judgment on cessation, the Court embraced the wording and interpreted it generously to include military actors in two short sentences, without mentioning discrepancies between this outcome and other views on how to interpret the Refugee Convention.414 The case concerned a classic instance of regime change: an US-led international military presence in Iraq was found to provide sufficient protection in the mid-2000s, once the fall of the old government had terminated the fear of persecution. Domestic courts will usually have to invest more energy than the Court in assessing the ‘effectiveness’ and ‘non-temporary character’ of protection, as prescribed by Article 7(2). By contrast, support given by private actors was found to be generally irrelevant for protection purposes.415

13.5.4.2 Internal protection alternative In everyday administrative practices, the internal protection alternative (IPA) often plays an important role. Again it may become relevant in different circumstances: parts of a country may be under the control of an official or de facto regional government (think of the autonomous Kurdish region in Northern Iraq); moreover, individuals seeking refuge from persecution in rural parts of a country may possibly relocate to the relative anonymity of bigger cities. To assess such scenarios, Article 8 Qualification Directive requires, first, that the alternative location offers protection against persecution and other serious harm and, secondly, that the applicant ‘can safely and legally travel [there] and gain admittance’. Safe travel was established as a necessary condition during the last revision of the Directive, in response to an ECtHR judgment; travel through unsafe civil war territory to reach internal safety cannot be expected.416 Controversies revolve around the third criterion, in Article 8(1), whether the applicant ‘can reasonably be expected to settle there’.417 What may sound innocent at first, serves as an entry door for calls to distinguish the necessary level of protection under Article 8 for the IPA from the one under Article 7 for persecution by non-state actors. UNHCR endorses such an additional safety requirement and expects an IPA to offer socio-economic living conditions without undue hardship.418 By contrast, the ECtHR does not support such an interpretation. Several judgments throughout the 2010s found internal flight alternatives to be available in Iraq, Afghanistan, and Somalia (after the situation there had improved), recognising 412 See Hathaway and Foster (n 382) 292–331; and Constantin Hruschka, ‘Anerkennung und Aufnahme als Flüchtling’ in Ferdinand Wollenschläger (ed), Enzyklopädie Europarecht, Band X (2nd edn, Nomos 2021) § 21 MN 46–51. 413 See Julian Lehmann, ‘Protection’ in European Union Asylum Law (Brill Nijhoff 2020) 122–38. 414 See Abdulla (n 369) paras 74–75. 415 See Secretary of State for the Home Department (n 399) paras 40–47. 416 Contrast former Qualification Directive 2004/83/EC, art 8(3); and see Sufi and Elmi v United Kingdom (n 86) §§ 268–71; and Jessica Schultz, The Internal Protection Alternative in Refugee Law (Brill Nijhoff 2019) ch 5. 417 Similarly, Salah Sheekh v Netherlands (n 83) § 141. 418 See UNHCR, ‘Guidelines on International Protection No 4: Internal Flight or Relocation Alternative’ (HCR/ GIP/03/04, 23 July 2003) Nos 22–30; see also Lehmann (n 413) 152–95; and Reinhard Marx, ‘The Criteria of Applying the “Internal Flight Alternative” Test’ (2002) 14 IJRL 179, 199–211.

396 Common European Asylum System that ‘[i]‌nternal relocation inevitably involves certain hardship’.419 Notwithstanding the practical relevance, no preliminary reference has reached Luxembourg so far. Insistence on parallelism with the criteria for status determination in Articles 6, 7 and 9 and the categorical rejection of socio-economic considerations for purposes of cessation may indicate that a similar answer would be defended for Article 8.420

13.5.5 Reasons for persecution Article 9(3) Qualification Directive 2011/95/EU takes up the wording of the refugee definition by requiring a connection between the acts of persecution and the grounds of ‘race, religion, nationality, political opinion, or membership of a particular social group’. That connection is mandatory; ill-treatment for other purposes does not give rise to refugee status (subsidiary protection is a different matter). Space precludes an extensive analysis of the five criteria, especially with regard to scenarios that have not been subject to supranational case law so far.421 We will concentrate on two situations that give rise to structural challenges. Additional comments can be found in the passages on religion and homosexuality on the previous pages.

13.5.5.1 Membership of a particular social group Membership of a particular social group defies easy definition. Conceptual ambiguity means that it serves as a residual category to integrate novel protection claims into the refugee definition, as the drafting of the Qualification Directive amply illustrated. Article 10(1)(d) bears testimony to the choice, on the part of the legislature, to cover sexual orientation and gender identity. In order to assess specific situations, international practice had developed two competing methodologies of how to classify a ‘particular social group’: internal characteristics or external social perception.422 The wording of Article 10(1)(d) combined both approaches (‘and’), and judges confirmed that they should be conceived of as being cumulative,423 contrary to UNHCR’s opinion to treat them as alternatives.424 Silence, on the part of judges, about the divergence exemplifies the introverted style of reasoning, assessed in Chapter 5.6.2 on international law. The silence may be rationalised, in part at least, by the focus of the parties to the proceedings to the qualification of gay men as a distinct social group. While this outcome was uncontroversial, in light of the wording of the Directive, it was a novelty at the time and required some discussion.425 Thus, the divergence from other views might have ‘slipped through’ largely unnoticed, with the Court following the wording of secondary legislation— as it often does—without considering the non-binding views of UNHCR. Be it as it may, 419 See DNM v Sweden App no 28379/11 (ECtHR, 27 June 2013) § 59; and Schultz (n 416) 297–318. 420 cf Abdulla (n 369) paras 65–68; Secretary of State for the Home Department (n 399) paras 48–62; and see Dörig (n 381) Article 8 MN 14–18c. 421 See further Goodwin-Gill and McAdam (n 243) 95–143; and Hathaway and Foster (n 382) ch 5. 422 Goodwin-Gill and McAdam (n 243) 100–18. 423 See X and others (n 371) paras 41–49; and Ahmedbekova (n 396) para 89. 424 See UNHCR, ‘Guidelines on International Protection No 2: Membership of a Particular Social Group’ (HCR/ GIP/02/02, 7 May 2002) Nos 10–13. 425 See AG Eleanor Sharpston, Opinion in Joined Cases C-199/12 and C-201/12 X and others EU:C:2013:474, points 27–36, who had not raised the issue either, despite her track record in ‘progressive’ opinions; on domestic practices see Dörig, ‘Directive 2011/95/EU’ (n 381) Article 10 MN 12–18.

Refugee Status 397 a ‘characteristic so fundamental to a person’s identity that the persons concerned cannot be required to renounce it’426 serves as the relevant benchmark to determine which other groups are covered by Article 10(1)(d). Note that there will often be an overlap with other reasons, such as political opinion or religion. Acts of persecution on such grounds may come within the ambit of the refugee definition even if we conclude that, for instance, those filing an individual application before the ECtHR against the home state do not constitute a distinct social group in the meaning of the case law.427

13.5.5.2 Refusal to perform military service Refusal to perform military service is governed by fairly detailed guidance in Article 9(2)(e) Qualification Directive 2011/95/EU: prosecution or punishment will amount to an act of persecution if the refusal concerns ‘a conflict, where performing military service would include [war crimes or crimes against humanity]’. This benchmark came up in a high-profile case of a mechanic working for the US armed forces in Iraq. He defected and applied for asylum, after having voluntarily extended the service, arguing that his unit might be involved in war crimes. Judges held indirect participation of logistical personnel and a sufficient likelihood of future transgressions to be sufficient, while concluding, at the same time, that military missions in the UN context cannot usually be suspected of participating in war crimes; moreover, the applicant might have failed to use alternative means to prevent involvement by re-enlisting voluntarily.428 A question open for debate is whether lower ranking troops can rely on the war crime criterion when illegality concerns participation in an act of aggression, which only the political leadership can commit under the Rome Statute of the International Criminal Court.429 An example is the war of aggression the Russian Federation launched against Ukraine in 2022. Another ruling on conscientious objection addressed the situation of Syrian refugees who had left the country to evade military service. Judges concluded that they need not have recourse to a practically dysfunctional procedure to refuse military service and that, in the Syrian case, repeated and systematic war crimes meant that any service in the armed forces would fulfil this requirement.430 At the same time, disproportionate punishment for refusal to perform military service was not sufficient to support refugee status, since this conclusion required a causal connection to one of the reasons put forward in the refugee definition.431 For Syrians, there was a ‘strong presumption’ that the connection existed, not least since the regime treated defection as an act of political opposition, although the German Federal Administrative Court reminded domestic courts that this ‘strong presumption’ did not substitute a thorough analysis of each case, which could result in a negative conclusion in light of individual characteristics or a change of circumstance.432 The judgment demonstrates the significance of the causal linkage, which might be less evident in other scenarios. Similarly, the question might arise whether disproportionate punishment for refusal to 426 X and others (n 371) para 70. 427 See Ahmedbekova (n 396) paras 84–90. 428 See Shepherd (n 373) paras 33–46. 429 See Andreas Zimmermann and Claudia Mahler, ‘Article 1A, para 2’ in Andreas Zimmermann (ed), The 1951 Convention Relating to the Status of Refugees and its 1967 Protocol: A Commentary (OUP 2011) MN 536–59. 430 See Bundesamt für Migration und Flüchtlinge (n 371) paras 26–38. 431 ibid paras 39–44; and further Hathaway and Foster (n 382) 367–90. 432 Bundesamt für Migration und Flüchtlinge (n 371) paras 46– 60; and Federal Administrative Court (Bundesverwaltungsgericht), Case 1 C 1.22 (19 January 2023).

398 Common European Asylum System perform military service qualifies as an act of persecution in the absence of war crimes. Article 9(2)(c) and (e) can be read to answer that question to the negative, but alternative views can also be defended, also considering that UNHCR recommends otherwise.433

13.5.6 Exclusion and cessation The refugee definition comprises ‘inclusion’ standards, discussed on the previous pages, together with criteria for the ‘exclusion’ of persons having committed war crimes or serious non-political crimes prior to their arrival in accordance with Article 1F Refugee Convention. In addition, those receiving assistance from the United Nations Relief and Works Agency for Palestine Refugees in the Near East (UNRWA) are exempted under Article 1D. This twofold exclusion standard is integrated into Union law by means of Article 12 Qualification Directive 2011/95/EU. One step further, Article 11 covers ‘cessation’ which concerns any significant and non-temporary change of circumstance in countries of origin, in line with Article 1C Refugee Convention. Exclusion and cessation have generated more than a dozen preliminary references.

13.5.6.1 Palestinian refugees UNRWA was established in response to the displacement crisis which accompanied the creation of the State of Israel. To date, more than five million Palestinians are registered with the Agency, more than half of them in Jordan, Lebanon, and Syria. They have mostly not been naturalised and often live in huge camps, with the support of UNRWA—despite the recent focus of the policy debate on local integration as a durable solution. In this overall context, Article 12(1)(a) requires Member States to reject an asylum claim if the applicant was registered with the Agency, or received its assistance otherwise.434 The requirement, in Article 1D Refugee Convention, that protection must be available ‘at present’ does not hinder an application of the exclusion clause once the applicant had travelled to Europe.435 Rather, the exclusion clause stops applying whenever UNRWA services are no longer available or when Palestinians leave for reasons beyond their control, in particular fear of their personal safety.436 Domestic authorities must assess whether the reasons for departure meet this threshold and whether return would be possible, either to an earlier place of registration or anywhere else in UNRWA’s area of operation.437 National authorities perform an ex nunc assessment, meaning that return can be expected if the situation improves after departure.438 The adequacy of support must be assessed objectively, in light of the factual situation on the ground, and take into account special needs, for instance regarding disabilities.439 Support of civil society and NGOs is immaterial in the absence of a formal relationship of

433 See UNHCR, ‘Guidelines on International Protection No 10: Claims to Refugee Status related to Military Service’ (HCR/GIP/13/10/Corr 1, 12 November 2014). 434 See Case C-31/09 Bolbol EU:C:2010:351, paras 45–54. 435 See Case C-364/11 El Karem El Kott and others EU:C:2012:286, paras 45–52. 436 ibid paras 53–64. 437 See Case C-507/19 Bundesrepublik Deutschland EU:C:2021:3, paras 51–80. 438 See Secretary of State for the Home Department (n 370) paras 54–57. 439 ibid paras 67–72.

Refugee Status 399 cooperation with UNRWA.440 Effective protection by state actors, instead of UNRWA, may be covered by the provision on first countries of asylum.441 Whenever the exclusion clause applies, refugee status will not be granted, although subsidiary protection may still be available, since Article 12(1)(a) concerns refugee status only.442 Moreover, the provision explicitly foresees ipso facto recognition as a refugee when an applicant was registered with UNRWA, left the area of operation for legitimate reasons, and cannot be expected to return. Ipso facto recognition entails that Palestinians are exempted from the need to demonstrate a well-founded fear of persecution; they must still file an asylum application, so that domestic authorities can assess whether the conditions of the provision are met.443 When circumstances change, Member States may revoke the protection status.

13.5.6.2 Terrorism and other serious crime Article 12(2) Qualification Directive corresponds to Article 1F Refugee Convention and excludes war criminals and terrorists from the rights and benefits associated with refugee status, although they may still be covered by the absolute prohibition of refoulement under human rights law.444 According to Article 17, war criminals and other serious offenders are excluded from subsidiary protection as well, under less onerous circumstances: serious crime is defined broader and exclusion is not—like in the case of refugees—subject to a geographic and temporal limitation to acts committed outside the territory and before arrival.445 State authorities must very the facts of each individual case, even though that factual assessment stays short of the higher legal standard of individualised public policy threats.446 While the notion of ‘war crimes’ has to be interpreted in line with international criminal law, ‘serious non-political crime’ covers terrorism. The Court confirmed that membership in a terrorist organisation was not sufficient in itself, even for the leadership; the applicant’s function and activities have to be assessed on an individual basis.447 Criminal sanctions may serve as an indicator of the seriousness of the crime, although such an abstract criterion cannot be applied mechanically; additional aspects have to be taken into account, by examining personal circumstances.448 No further conditions have to be fulfilled, if the crime qualifies as serious; judges did not require—contrary to UNHCR’s viewpoint—a proportionality test weighing the gravity of the offence against the consequences of refusal, nor does the applicant have to pose a present threat.449 Another instance of divergence concerned exclusion on grounds of acts contrary to the UN principles under Article 12(2)(c), which judges did not restrict to persons who have held a position of power.450 Such broad interpretation reflects the evolution of international law, which 440 ibid paras 73–84. 441 See Alheto (n 346) paras 131–43. 442 See El Karem El Kott and others (n 435) paras 67–68. 443 ibid paras 69–77; and Alheto (n 346) paras 85–89. 444 See ch 13.2.2; and KI v France App no 5560/19 (ECtHR, 15 April 2021). 445 See Case C-369/17 Ahmed EU:C:2018:713, paras 38–47. 446 See Országos Idegenrendeszeti Főigazgatóság and others (n 312) para 92; and ch 10.3. 447 See Joined Cases C-57/09 and 101/09 B and D EU:C:2010:661, paras 79–99; and further Ingo Kraft, ‘Asylum Qualification Directive 2011/95/EU’ in Thym and Hailbronner (n 65) Article 12 MN 29–76. 448 See Ahmed (n 445) paras 48–57. 449 See B and D (n 447) paras 100–11; contra UNHCR, ‘Guidelines on International Protection No 5: Application of the Exclusion Clauses’ (HCR/GIP/03/05, 4 September 2003) No 24. 450 See Case C-573/14 Lounani EU:C:2017:71, paras 62–79; contra UNHCR, Handbook on Procedures and Criteria for Determining Refugee Status (2nd edn, UNHCR 1992) No 163.

400 Common European Asylum System generally took a tough line against terrorism in the new millennium. Article 14(4)–(6) on the refusal and revocation of ‘refugee status’ for applicants posing a security threat will be mentioned in the context of the status after recognition.

13.5.6.3 Change of circumstance in the home state In a judgment of principle, the Court emphasised the parallelism between the original examination of protection against persecution and the opposite conclusion of cessation.451 Standards of proof are the same, and cessation also necessitates careful assessment of the individual case, not an abstract evaluation of the situation in a country.452 In line with previous comments, Union law recognises home states and international organisations, including military missions, as actors of protection, whereas support by private networks or family members does not equal protection by public authorities.453 In accordance with the parallelism yardstick, additional criteria, such as adequate living conditions, are irrelevant.454 Judges did not require—in contrast to UNHCR—a proportionality assessment.455 Consideration of broader humanitarian reasons appears possible, in light of Article 11(3) Qualification Directive, and evaluation reports show that this is regularly done in practice.456 Other instances of cessation, in Article 11(1)(a)–(d), have to be assessed in accordance with international practice and the case law of domestic courts, for instance as to consequences of short visits to family and friends residing in home states.457

13.6 Subsidiary and Complementary Protection A simple statistic illustrates the importance of subsidiary protection: more than 600 thousand people received subsidiary protection status in today’s twenty-seven Member States during 2015–20. Introduction of the status was a major achievement of the Qualification Directive, filling a conceptual lacuna that had plagued refugee law ever since its inception (13.6.1). Abstract definitions left judges with the task of elucidating their meaning (13.6.2). Discrepancies between subsidiary protection and human rights law with regard to living conditions in countries of origin (13.6.3) are one reason why domestic complementary protection regimes are relevant in practice (13.6.4). The idea behind temporary protection is discussed below.

13.6.1 Added value of Europeanisation It was not lost to the drafters of the Refugee Convention that the refugee definition with its focus on persecution did not embrace all instances of forced displacement. People escaping civil wars, in particular, have not traditionally been covered.458 The conscious choice for a 451 See Abdulla (n 369) paras 65–59; and Secretary of State for the Home Department (n 399) paras 33–39. 452 Abdulla (n 369) paras 81–87; and Secretary of State for the Home Department (n 399) paras 30–32. 453 See Secretary of State for the Home Department (n 399) paras 40–47. 454 ibid paras 48–62. 455 cf UNHCR (n 450) para 136. 456 See Commission, Evaluation of the Application of the Recast Qualification Directive Final Report, Publications Office (2019) 104–105, 184–87 https://data.europa.eu/doi/10.2837/59251 (accessed 1 March 2023). 457 See Kraft (n 447) Article 11 MN 8–15. 458 See Jane McAdam, Complementary Protection in International Refugee Law (OUP 2007) 35–40.

Subsidiary and Complementary Protection 401 narrow definition was reaffirmed on the occasion of the New York Protocol of 1967—in contrast to the OAU Convention, adopted two years later—echoing our introductory comments on the history of half-hearted commitment. Debates about the situation of ‘de facto refugees’ gathered pace in Europe from the 1970s onwards, domestically and within the Council of Europe.459 Memories of the wars in the former Yugoslavia in the 1990s were still fresh when the EU institutions started negotiating the former Qualification Directive 2004/83/EC, which introduced subsidiary protection in accordance with the Tampere Conclusions.460 Subsidiary protection was widely hailed as an important novelty,461 building upon diverse forms of protection at the national level.462 Recital 33 Qualification Directive 2011/95/EU emphasises that subsidiary protection ‘should be complementary and additional to’ refugee status, hence the description as ‘subsidiary’. Earlier concerns that Member States would use subsidiary protection as a substitute to refugee status to lower protection standards did not materialise. Such practices would be illegal anyway, since the Qualification Directive requires domestic authorities, as we have seen, to examine refugee status first. The rights and benefits for beneficiaries were at least as important as the introduction of the concept as such; national practices had varied considerably and often stayed behind today’s Articles 20–35 Qualification Directive.463 Moreover, the Directive introduced an individual right to receive subsidiary protection, replacing discretionary national practices.464 Article 78(2)(b) TFEU mentions ‘subsidiary protection’ but does not prevent, as we have seen, the legislature from changing the relevant criteria for status acquisition in the future. Indeed, there may be room for further harmonisation, especially with regard to national ‘complementary’ protection schemes. Having said this, the potential of further harmonisation does not call into question the added value of subsidiary protection in a historic perspective. Protracted debates, throughout the 1990s, about whether and how to protect people fleeing civil wars have subsided. Subsidiary protection is an essential element of the CEAS. Contemporary debates concentrate on access to the territory, not about protection after arrival.

13.6.2 Protection against indiscriminate violence Article 15 Qualification Directive was loosely modelled on human rights law, but the Court emphasised the autonomy of the notion of ‘serious harm’ early on.465 Distinguishing subsidiary protection from human rights law prevented Member States from lowering the level of protection for those fleeing civil wars.466 ECtHR judgments embracing situations of general violence were not yet firmly established when the CJEU delivered its first judgment 459 See PACE, ‘Situation of de facto refugees’ (Recommendation 773 (1976), 26 January 1976). 460 On the drafting process see Jane McAdam, ‘The European Union Qualification Directive’ (2005) 17 IJRL 461, 462–68. 461 By way of example see James Hathaway, ‘What’s in a Label?’ (2003) 5 EJML 1, 6–8. 462 See Lambert (n 13) ch 6; and Bouteillet-Paquet, Subsidiary Protection (n 410). 463 See McAdam, Complementary Protection (n 458) 198–214. 464 See ch 6.2.1; and Helene Lambert, ‘The Next Frontier. Expanding Protection in Europe for Victims of Armed Conflict and Indiscriminate Violence’ (2013) 25 IJRL 207, 209–10. 465 See ch 6.4.4; and Case C-465/07 Elgafaji EU:C:2009:94, para 28. 466 See Roger Errera, ‘The CJEU and Subsidiary Protection’ (2011) 23 IJRL 93, 96–97.

402 Common European Asylum System ever on the substantive provisions of EU asylum law: the Elgafaji ruling. During the negotiations on Article 15(c), the legislature had agreed on abstract language as a result of controversies over the precise scope of the provision.467 This left the Court with the task of elucidating its meaning, thus ideally overcoming the stark discrepancies in the application of the Directive. A first uncertainty concerned the notion of ‘international or internal armed conflict’. Autonomous interpretation independent of international humanitarian law ensured that subsidiary protection may cover situations of general violence below the threshold of (civil) war.468 More specifically, the Court called for a ‘specific assessment of the intensity of such confrontations’, while emphasising that violent protest alone will usually be insufficient.469 Anyone reading this abstract formulae wonders how they should be applied coherently in practice. A second question confronts a conundrum enshrined in Article 15(c). On the one hand, the wording offers protection against ‘indiscriminate violence’, in the sense of a general danger not addressing specific individuals.470 On the other hand, the legislature required an ‘individual threat’, which, according to today’s Recital 35, may not normally result from risks the population is generally exposed to. This tension between generalisation and individualisation complicates the identification of practicable standards guiding application. Judges tried to square the circle by means of a compromise: ‘the more the applicant is able to show that he is specifically affected by reason of factors particular to his personal circumstances, the lower the level of indiscriminate violence required for him to be eligible for subsidiary protection’.471 Combining generalisation and individualisation results in a ‘sliding scale’.472 General risks in a third state will give rise to protection in ‘exceptional situation[s]‌’ only. Civil wars do not cross the threshold automatically, and we may have to distinguish between different regions of a country.473 Whenever the level of indiscriminate violence does not cross that high threshold, applicants will only receive subsidiary protection if they can demonstrate personal factors increasing the risk of exposure. This solution is conform with later ECtHR case law, mentioned previously. Application of the ‘sliding scale’ raises conceptual and practical difficulties.474 Judges specified, on a later occasion, that any application of the Elgafaji test requires a combination of quantitative and qualitative factors, transcending simple ‘body counts’ in terms of the number of casualties.475 Rather, a variety of factors has to be considered, ranging from quantitative elements to the intensity, duration, and geographical scope of the confrontations, the level of organisation of the armed forces, intentional targeting of civilians, etc.476 467 See McAdam (n 460) 480–90. 468 See Case C-285/12 Diakité EU:C:2014:39, paras 20–26; and Charlotte Lülf, Conflict Displacement and Legal Protection (Routledge 2019) 151–60. 469 Diakité (n 468) paras 9, 28–34; and Elgafaji (n 465) paras 35–37. 470 See Elgafaji (n 465) paras 32–34. 471 ibid para 39. 472 UNHCR, ‘Safe at Last? Law and Practice in Selected EU Member States with Respect to Asylum-Seekers Fleeing Indiscriminate Violence’ (July 2011) 49. 473 See Elgafaji (n 465) paras 35–40. 474 See Hugo Storey, ‘Asylum Qualification Directive 2011/95/EU’ in Thym and Hailbronner (n 65) Article 15 MN 14–26. 475 See Case C-901/19 Bundesrepublik Deutschland EU:C:2021:472, paras 31–36. 476 ibid paras 35, 40–43; UNHCR (n 472) 41–48, 71–74; and Lambert (n 464) 215–28.

Subsidiary and Complementary Protection 403 Anyone familiar with everyday decision-making understands that the conglomerate of criteria for assessing the level of general violence, together with the ‘sliding scale’ regarding personal circumstances, will hardly translate into uniform practices on the ground.477 Discrepancies in the recognition rates between and within states cannot be prevented on the basis of such abstract formulae.

13.6.3 Other scenarios: divergence from human rights Article 15(a) and (b) Qualification Directive concern scenarios of individual risk and echo ECtHR case law on non-refoulement obligations with regard to the death penalty and inhuman or degrading treatment. An explicit reference to treatment ‘in the country of origin’ was introduced to signal distance from ECtHR judgments on living conditions and health- based non-refoulement obligations.478 The wording of Article 15(a) and (b) would have been open enough to give judges leeway for synchronisation, but they chose not to do so. In a judgment of principle, the Court distinguished subsidiary protection from human rights, relying, amongst others, on the preference of the legislature and the focus of the Qualification Directive on acts and actors of persecution.479 Applicants covered by the prohibition of refoulement on socio-economic grounds under human rights law will not usually receive international protection, even though states are obliged not to return them.480 National law defines the rights and benefits in these scenarios.

13.6.4 ‘Complementary’ protection under national laws In the EU context, the notion of ‘complementary’ protection can be used as an umbrella term for domestic schemes transcending the scope of ‘subsidiary’ protection under the Qualification Directive. Distinction between subsidiary and complementary protection is warranted for legal and practical reasons. Legally, complementary statuses are not harmonised at the EU level; the Qualification Directive and the Charter do not apply, and the Court has no jurisdiction. Practically, national laws vary considerably, both regarding the conditions under which someone may receive complementary protection and concerning the rights and benefits associated with domestic statuses. At the international plane, ‘complementary protection’ is often used more broadly, for all mechanisms available to foreigners without refugee status.481 Such definition covers subsidiary and complementary protection in the above sense. Chapter 10.6.2 explained that the Court recognised that Member States remain free to introduce national permits under conditions which are more generous than the ones in Qualification Directive 2011/95/EU, provided that they do not undermine the effet utile of the directive. National legislation has to distinguish clearly whether a status is governed 477 See Nora Markard, Kriegsflüchtlinge (Mohr Siebeck 2012) 341–47. 478 See Storey, ‘Brave New World?’ (n 394) 31; and McAdam, ‘Qualification Directive’ (n 460) 490–93. 479 See M’Bodj (n 71) paras 32–41; and MP (n 402) paras 46–56. 480 See ch 13.2.2; and MP (n 402) paras 36–43. 481 See UNCHR ExCom, ‘Complementary Forms of Protection’ (EC/50/SC/CRP.18, 9 June 2000); and McAdam, Complementary Protection (n 458) 19–52.

404 Common European Asylum System by supranational legislation or not. States can be obliged, as we have seen, to introduce such schemes where human rights reach further than the definition of serious harm in the Qualification Directive. Note that the ECHR protects third country nationals against return and does not comprise prescriptions as to the nature of the residence permit, let alone the rights and benefits during the stay of deportation.482 Comparative studies illustrate that national practices vary to a great extent.483 While some domestic schemes concern specific situations, others are formulated in an open manner and cover diverse scenarios. States may operate several regimes in parallel. Some countries have guarantees at constitutional level, others rely on legislation or administrative practices alone; all these different complementary protection statuses can be assessed as an integral part of the asylum procedure or be subject to distinct procedural avenues. In the design of complementary protection schemes, human rights requirements often coalesce with humanitarian considerations. Besides medical cases, protection for children and family members may be influenced by human rights law. Moreover, Article 3 ECHR may require states not to return terrorist suspects covered by the exclusion clauses in Articles 1D and 1F Refugee Convention. One step further, national schemes often cover situations where return is not possible or warranted for humanitarian or practical reasons, which are unrelated to human rights. Such ‘compassionate’ provisions usually cover broad categories of humanitarian grounds and may be linked to the level of integration in the host society.484 During the Covid-19 pandemic, the Commission encouraged Member States not to qualify as illegal periods of stay caused by travel interruptions.485 Natural disasters are an area where Italy introduced an innovative complementary protection scheme, thus advancing a category that might serve as a political model in debates about climate change.486 Multiple other criteria may exist elsewhere. One may ideal-typically distinguish between humanitarian grounds in response to the situation in home states and compassionate concerns related to the undesirability or impracticability of return, although both motivations may overlap in practice. While the former are covered by Article 78(2)(c) and (d) TFEU, the latter come within the ambit of Article 79(2)(a) TFEU. Programmes related to ‘non-removability’ are often discussed under the heading of ‘regularisation’, including in Chapter 16.8 on return. Arguments supporting further harmonisation at the supranational level comprise legal certainty and coherence of the CEAS legislation. At the same time, practical considerations argue against legislative action: national schemes are flexible and common standards would probably not exceed the lowest common denominator.487 In any case, the Commission does not seem eager to proceed along the path of harmonisation.

482 See ch 15.2.2. 483 See EMN, Comparative Overview of National Protection Statuses in the European Union (EU) and Norway: Synthesis Report (May 2020); building upon EMN, The Different National Practices Granting Non- Harmonised Protection Statuses: Synthesis Report (December 2010); see also Schieber (n 70) ch 3. 484 Schieber (n 70) chs 2–11. 485 See Commission Guidance on the use of flexibility provisions in asylum and return procedures [2020] OJ C126/12. 486 See Fornalé (n 408) 533–39. 487 See Schieber (n 70) 317–40.

Temporary Protection 405

13.7 Temporary Protection The Temporary Protection Directive 2001/55/EC was the first supranational instrument on asylum to be adopted under the Treaty of Amsterdam. The memory of mass displacement during the wars in the former Yugoslavia informed the debate at a time when states perceived temporary protection as an aliud for refugee protection (13.7.1). After two decades of oblivion, the Council activated the directive in response to the war of aggression against Ukraine. Doing so was an ideal-typical example of ‘mass influx’ for which the scheme had been established (13.7.2). The temporary character raises the question about how it interacts with asylum legislation and what happens thereafter (13.7.3).

13.7.1 Rationale behind blanket recognition Discussions about the Temporary Protection Directive responded to the variety of national responses to the wars in the former Yugoslavia. The main destination countries introduced temporary protection complementing refugee status at a time when domestic legislation did not know the concept of subsidiary protection, which covers movements caused by indiscriminate violence.488 Anticipating contemporary debates about solidarity, the Council failed to agree on burden sharing during the 1990s, as mentioned in the context of the Dublin system. The original European debate focused on differences between ‘temporary’ protection and refugee status.489 In contrast to this European usage, the notion of temporary protection is commonly perceived as an addition or prelude to refugee status, not as an aliud, in Africa, Latin America, and southeast Asia.490 The UNHCR adopted guidelines in 2014, which are somewhat abstract.491 Temporary protection had originally been a double-edged sword: while enthusiasts present it as a ‘magic gift’, the status simultaneously ‘serves as a magic mirror of its observers’ fears’.492 Proponents emphasised the added value for people fleeing civil wars and the benefit of procedural simplicity, which helps saving scarce administrative resources during a ‘mass influx’; the option of burden sharing was another advantage. Critics, by contrast, lamented the element of state discretion most domestic temporary protection schemes had traditionally involved and the lesser degree of socioeconomic rights in comparison with the Refugee Convention.493 States focused on the ‘temporary’ character of protection, including the promise of swift return after the war.494 It served migration control purposes as much as it was about humanitarian assistance. An activation of the Temporary Protection Directive was discussed on several occasions but discarded for different reasons. The arrival of several tens of thousands of Tunisians 488 See Joanne Thorburn, ‘Transcending Boundaries’ (1995) 7 IJRL 459, 469–77; and Matthew J Gibney, ‘Between Control and Humanitarianism’ (1999) 14 Georgetown Imm LJ 689, 693–96. 489 See Meltem İneli Ciğer, Temporary Protection in Law and Practice (Brill/Nijhoff 2018) 198–203; and Joanne Van Selm, ‘Temporarily Protecting Displaced Persons or Offering the Possibility to Start a New Life in the European Union’ (2001) 3 EJML 23. 490 İneli Ciğer (n 489) ch 4; and Goodwin-Gill and McAdam, International Law (n 243) 292–98. 491 See UNHCR, ‘Guidelines on Temporary Protection or Stay Arrangements’ (February 2014); and İneli Ciğer, Temporary Protection (n 489) ch 7. 492 Joan Fitzpatrick, ‘Temporary Protection of Refugees’ (2000) 94 Am J Intl L 279, 280. 493 ibid 287–91. 494 See Gibney (n 488) 689–707.

406 Common European Asylum System during 2011 arguably did not qualify as a ‘mass influx’, and most new arrivals did not have protection needs. Syrians entering during 2015/16 might have been given temporary protection.495 However, toxic political debates about solidarity and preference for protection in the region of origin explain why the Directive was not activated; people fleeing the civil war in Syria who arrived in Greece were to be returned in accordance with the EU–Turkey Statement.496 Lengthy asylum procedures and dire living conditions in the ‘hotspots’ on the Greek islands were the opposite of the unbureaucratic delivery of temporary protection status to Ukrainians seven years later. In the case of Ukraine, the objective of preventing a collapse of the asylum system and the desire of swift and generous reception explain the activation at a moment when most experts had thought that Directive 2001/55/EC was obsolete. ‘Welcome’ was the motto of the day. Even the idea of ‘free choice’ enjoyed an unexpected renaissance. Ukrainians were allowed to circulate freely in the Schengen area (the need for a biometric passport was waived generously), and Member States promised not to insist on returning beneficiaries of temporary protection moving elsewhere.497 Free choice had the side effect of permitting Ukrainians to go where family and friends were staying already, thus reducing the pressure on reception facilities. EU institutions mobilised significance funds,498 in addition to the tremendous efforts of governments, private citizens, and civil society. The defensive and ad hoc character of the response to the wars in the former Yugoslavia gave way to a proactive and open-hearted temporary protection regime. The contrast to the mostly restrictive response to movements from non-European countries was manifest. Think of the standoff with Belarus a few months earlier, the control component of the EU–Turkey Statement, and ongoing disputes about search and rescue. In the case of Ukraine, obstacles were replaced by open arms. The sense of discrimination was reinforced by border guards hindering the entry of African students from Ukraine and some politicians highlighting cultural similarities.499 In contrast to permanent residents who were unable to return to countries of origin safely, the application of the Implementing Decision to foreign students who had studied in Ukraine remained optional.500 They had to apply for asylum instead. Chapter 10.4 on equality before the law illustrated that it remains a formidable challenge to maintain that these distinctions amounted to illegal discrimination, possibly even on grounds of race.

13.7.2 Conditions and contents of temporary protection The notion of ‘mass influx’ in the official title of Directive 2001/55/EC reaffirms that the idea of temporary protection was closely associated with perceptions of urgency and crisis, 495 See Meltem İneli Ciğer, ‘Time to Activate the Temporary Protection Directive’ (2016) 18 EJML 1, 13–32. 496 See Commission/ICF, ‘Study on the Temporary Protection Directive: Final Report’ (January 2016) 27–36. 497 See Temporary Protection Implementing Decision (EU) 2022/382, recitals 15, 16; and Daniel Thym, ‘Temporary Protection for Ukrainians’ EU Immigration and Asylum Law and Policy (7 March 2022). 498 See ch 7.4. 499 See ‘UNHCR Chief Condemns “Discrimination, Violence and Racism” against Some Fleeing Ukraine’ UN News (21 March 2022) https://news.un.org/en/story/2022/03/1114282 (accessed 1 March 2023). 500 See Temporary Protection Implementing Decision (EU) 2022/382, art 2; and Julia Kienast, Nikolas Feith Tan, Jens Vedsted-Hansen, ‘Preferential, Differential or Discriminatory?’ in Sergio Carrera and Meltem İneli Ciğer (eds), EU Responses to the Large-Scale Refugee Displacement from Ukraine (Migration Policy Centre 2023) 383–99.

Temporary Protection 407 regarding both the number of arrivals and administrative capacities. These two parameters will take centre stage when the Council determines whether a situation of ‘mass influx’ exists—a necessary condition for any activation of the Directive.501 Like in the case of Article 78(3) TFEU, the Council benefits from a margin of appreciation.502 It shall decide by qualified majority on a proposal by the Commission, without the involvement of the Parliament.503 Without the adoption of an implementing decision, the Temporary Protection Directive remains dead letter. The implementing decision defines the scope of the temporary protection regime ratione personae and materiae for an initial period of validity of one year, subject to automatic and discretionary extensions to up to three years altogether.504 To prolong the protection regime further would require an amendment of Directive 2001/55/EC, since the simple reactivation after three-year period would undermine the effet utile of the legislation.505 Majority requirements in the Council for an official amendment of the Directive are the same as for a second activation. EU institutions have much leeway when determining the scope ratione personae. Rudimentary benchmarks may be derived from the rationale behind international and temporary protection, enshrined at Treaty level, which refers, as we have seen, to risks in countries of origin or transit. The very idea of a pragmatic response to a ‘mass influx’ indicates that it is primarily about general risks, not an individual fear of persecution. Indeed, the definition of ‘displaced persons’ in the Directive echoes criteria for subsidiary protection, while being broader.506 In the case of Ukraine, all nationals residing there were covered; it was irrelevant whether someone came from war-torn Mariupol near the Russian border or from comparatively safe Lviv in the far West of the country—differences which would inform the application of the ‘gliding scale’ approach to subsidiary protection in line with previous comments. Having said this, Temporary Protection Directive 2001/55/EC gives the Council much leeway. Reference to ‘specific groups of persons’507 indicates that abstract characteristics can be used. The Council may confine, for instance, temporary protection to members of certain religious groups or regions of origin. Article 7 authorises Member States generally to extend temporary protection to additional categories of displaced persons, whenever the Directive was activated. To say that they cannot introduce domestic temporary protection regimes in the absence of activation ignores the optional character of Article 7.508 Of course, the ‘privileges’ enshrined in the Directive, such as the suspension of asylum procedures, do not apply without activation. Several countries are operating national temporary protection schemes as a specific form of complementary protection.509 Article 28 mirrors the exclusion clause in Article 1F Refugee Convention. It is the only justification for refusing temporary protection. Member States 501 See Achilles Skordas, ‘Temporary Protection Directive 2001/55/EC’ in Thym and Hailbronner (n 65) Article 1 MN 8–28. 502 See Slovakia and Hungary v Council (n 75) paras 113–24. 503 See Temporary Protection Directive 2001/55/EC, art 5(1). 504 ibid art 4. 505 Contra Violeta Moreno-Lax and Elspeth Guild, ‘Temporary Protection’ in Peers and others (n 110) 571, 577. 506 See Temporary Protection Directive 2001/55/EC, arts 2(c), 5(3); and Skordas (n 501) Art 2, MN 8–20. 507 Temporary Protection Directive 2001/55/EC, arts 5(3)(a). 508 See ch 10.6.2; and Karoline Kerber, ‘The Temporary Protection Directive’ (2002) 4 EJML 193, 198; contra Moreno-Lax and Guild (n 505) 574. 509 See ECRE, ‘Five Years on Europe Is Still Ignoring Its Responsibilities towards Iraqi Refugees’ (AD1/03/2008/ ext/ADC, March 2008).

408 Common European Asylum System cannot maintain that someone does not face a real risk of ill-treatment, despite being covered by the implementing decision ratione personae.510 Temporary protection is an unconditional status, not a rebuttable presumption. Directive 2001/55/EC assumes that beneficiaries receive protection in a specific country, not the territory of the Union in the abstract. This raises the follow-up question of how and when individuals obtain the status: the moment they cross an external border or whenever state authorities have verified whether they qualify? Chapter 10.1.2 on general features explained that EU migration law habitually opts for the second solution; individuals may have an individual right to be issued a residence permit, but status acquisition is not automatic. Somewhat surprisingly, Directive 2001/55/EC remains largely silent on this matter; in particular, it does not foresee an application procedure; national laws differ as well.511 While the Commission considers status acquisition to be automatic, there are counter-arguments as well,512 also considering that the Directive was carefully drafted not to establish individual rights.513 Socioeconomic guarantees for status holders were crucial at the time of adoption, establishing a minimum level of protection. From today’s perspective, however, Articles 12–16 Temporary Protection Directive 2001/55/EC are unspectacular and may even stay behind the guarantees for subsidiary protection in Articles 20–35 Qualification Directive 2011/95/EU, for instance when limiting health care to emergency assistance. Better treatment can be one reason why beneficiaries of temporary protection may wish to apply for asylum in parallel.

13.7.3 Interaction with EU migration law Studying temporary protection can resemble an archaeological excavation. Legal experts have to dig out materials that had disappeared from the radar, since few would have thought that the instrument would ever be activated. That is one reason why some other supranational pieces of legislation on asylum, in particular, remain silent on how they interact with Directive 2001/55/EC. Generally speaking, the Temporary Protection Directive arguably serves as a lex specialis superseding provisions adopted later.514 That abstract finding becomes relevant whenever individuals apply for asylum before or after receiving temporary protection. Articles 3 and 17 Temporary Protection Directive emphasise that the right to apply for asylum remains intact, although Member States are free to suspend examination for up to three years.515 States not using this option may suspend asylum processing de facto, in light of the flexibility enshrined in Articles 31(3)–(6) Asylum Procedures Directive 2013/32/EU. Remember that procedures occasionally take years to complete, even in ordinary situations. 510 See Moreno-Lax and Guild (n 505) 577. 511 See Gregor Noll and Markus Gunneflo, ‘Directive 2001/55: Temporary Protection—Synthesis Report’ (Odysseus Academic Network 2007) 16, 28. 512 See No 4 Operational guidelines for the implementation of Council Implementing Decision 2022/382 [2022] OJ C126I/1; and Thym (n 498). 513 See Kerber (n 508) 201. 514 It depends on the individual case whether the conflict rule of lex posterior takes precedence over leges speciales adopted earlier. 515 See Skordas, ‘Directive 2001/55/EC’ (n 505) Articles 17–19.

Temporary Protection 409 In any case, applicants cannot rely on reception conditions for asylum seekers in the meantime. Article 3(3) Reception Conditions Directive 2013/33/EU contains an explicit exception to this end. Debates about burden-sharing defined the negotiations on the Temporary Protection Directive, as mentioned in the context of the Dublin system. Articles 24–26 embody a compromise that makes relocation subject to the double consent of the individual and the receiving country in the sense of ‘double voluntariness’.516 Return to other Member States in situations of secondary movement, under Article 11, would have to be realised on the basis of bilateral arrangements described in Chapter 16.6.3 on return. The alternative recourse to take back procedures under the Dublin III Regulation in accordance with Article 18 Temporary Protection Directive 2001/55/EC are rendered void by judgments on time limits under the Dublin system. Legislation on legal migration often excludes beneficiaries of temporary protection, for instance with regard to resident permits as students or blue card holders.517 One may argue, however, that the voluntary renunciation of temporary protection opens the door for status change. Notwithstanding the widespread emphasis on return, in the policy debate prior to the adoption of the Directive, half of those who had left Bosnia-Hercegovina and Kosovo during the 1990s stayed in receiving countries when the fighting stopped.518 State support for voluntary return, discussed in Chapter 16.5.4, gained prominence as an alternative to forced removal, which governments struggled to enforce. In the absence of return, residence permits and the prospect of settlement shall be made available, although Articles 20–23 Temporary Protection Directive leave Member States much freedom in this respect. In the case of Ukraine, the number of returns will depend on how long the war lasts, what the outcome will be, and whether the homes are still intact. The fact that most refugees are women and children might increase the likelihood of return, to reunite with the men who had defended the country. Before the first ever activation in 2022, the Commission had revived the idea of temporary protection under the heading of ‘immediate protection’ in the New Pact on Migration and Asylum. It proposed to replace Directive 2001/55/EC with a provision which applies in vaguely defined ‘crisis’ scenarios to people fleeing indiscriminate violence in situations of armed conflict; asylum procedures would be suspended officially, with the small print being determined by means of an implementing decision.519 The initiative was a building stone of the new focus on border procedures. Suspension of asylum procedures in scenarios of mass influx would be an instrument, amongst others, to prevent huge camps at the external borders. We can expect the ongoing negotiations about the reform proposals to leave the Temporary Protection Directive intact in light of the recent experience with Ukraine.

516 ibid art 23 MN 3; and Commission/ICF (n 496) 23–24. 517 See Students and Researchers Directive (EU) 2016/801, art 2(2)(a); and Blue Card Directive (EU) 2021/1883, art 3(2)(a). 518 Fitzpatrick (n 492) 300–302; and John Koo, ‘Mass Influxes and Protection in Europe: A Reflection on a Temporary Episode of an Enduring Problem’ (2018) 20 EJML 157, 169–71. 519 See Proposal for a Crisis and Force Majeure Regulation (n 45) art 11; and Meltem İneli Ciğer, ‘Immediate Protection in the New Pact on Migration and Asylum’ in Thym and Odysseus Academic Network (n 40) 149.

410 Common European Asylum System

13.8 Reception Conditions for Asylum Applicants Anyone studying living conditions must distinguish two sets of rules: standards for the treatment of asylum seekers can be found in the Reception Conditions Directive 2013/ 33/EU, whereas beneficiaries of international protection are covered by Articles 20–35 Qualification Directive 2011/95/EU; they will be discussed hereafter. Miserable reception conditions for asylum seekers at the external borders, in particular, are a stigma for EU asylum policy, epitomising structural difficulties in ensuring adequate living conditions (13.8.1). Closer inspection of the Reception Conditions Directive demonstrates the prevalence of vague formulations, leaving much legal and practical leeway to Member States (13.8.2). Human rights may be relied upon to challenge reduction or withdrawal (13.8.3), while restrictions on free movement within the territory would be fortified if Commission proposals for a ‘fiction of non-entry’ were realised (13.8.4). At the time of writing, a provisional political agreement had been reached between the Council and the Parliament on a minor revision, which looked likely to be adopted before the European elections in 2024. Its contents will be mentioned whenever appropriate.

13.8.1 Protracted differences between the Member States Twenty years after the adoption of the first generation of asylum legislation, enormous variations persist across the Union. These differences have to do with the lack of administrative resources and political will, but they also reflect deeper challenges. Permanent and resilient structures are not available in all Member States.520 Structural deficits persist in diverse locations, ranging from miserable living conditions in Greek hotspots to the absence of adequate housing facilities in French cities.521 These shortfalls violate the Reception Conditions Directive more or less openly, but its wording and general scheme arguably contribute to widespread asymmetries in other respects. The Directive often contains, as we shall see, abstract provisions which leave much leeway to the Member States as to how to implement common standards. It remains to be seen whether proposals for enhanced contingency planning by and supervisory functions of the Asylum Agency will help to improve the situation on the ground.522 Recitals 11 and 12 Reception Conditions Directive 2013/33/EU mention a double rationale for harmonisation: ensuring a dignified standard of living, and preventing secondary movements. Interdisciplinary studies, traced in Chapter 4.1, show that reception conditions may influence migratory patterns, even though other elements tend to be more significant. Family and ethnic networks are crucial driving forces, as are labour market prospects. EU institutions have limited regulatory leverage in this respect. Many factors influencing the choice of destination are beyond their direct reach. When it comes to reception conditions, the CEAS is bound to stay, much more than in other domains, a conglomerate of national systems.523 520 See Madeline Garlick, ‘Reception of Asylum-Seekers in the EU in Times of “Crisis”’ in Paul Minderhoud and Karin Zwaan (eds), The Recast Reception Conditions Directive (Wolf 2016) 141, 146–50. 521 See European Committee for the Prevention of Torture (CPT), ‘Report to the Greek Government on the visit to Greece’ (CPT/Inf (2019) 4, 19 February 2019) 9–31; and NH and others v France (n 90) §§ 165–88. 522 See ch 8.4.3; and RCD Proposal (n 238) arts 27–28. 523 See generally ch 7.1.1.

Reception Conditions for Asylum Applicants 411 The Commission Proposal for a recast of the Reception Conditions Directive, put forward in 2016, openly admits the persisting deficits, but fails to spell out tangible parameters for core guarantees such as living conditions.524 Those studying the reform proposal should work with the provisional political compromise, even though it was not officially adopted as a result of the ‘package approach’ and a rare scenario of the Council plenary objecting to the compromise agreed upon in the informal trilogue format.525 Instead of changing the basic contours of the legislation, political debates concentrate on—albeit important—side aspects. Whereas Member States fancy stricter sanctions, the Parliament promotes better protection of vulnerable groups.526

13.8.2 Living conditions and other guarantees Today’s Reception Conditions Directive 2013/33/EU covers anyone who has ‘made’ an application for asylum in accordance with previous comments.527 Member States must provide reception conditions for as long as asylum applicants have the right to remain. A special category are Dublin transferees for whom the state of factual residence shall be responsible until the completion of the transfer.528 Other applicants loose the benefits once their legal remedy against a negative decision has been rejected.529 After recognition, beneficiaries of international protection are covered by the Qualification Directive. Uncertainties as to the impact of the Proposal for a Screening Regulation, mentioned previously, would have to be settled in the legislative process. Article 17(2) Directive 2013/33/EU defines ‘material reception conditions’ as ‘an adequate standard of living for applicants, which guarantees their subsistence and protects their physical and mental health’, provided that they do not have means of subsistence of their own. Article 2(g) does not add much precision when specifying that the definition includes ‘housing, food and clothing’. Structural difficulties in ensuring a level playing are highlighted by the absence of further clarification. The Commission proposes continuing the vague definition, while adding ‘other essential non-food’, notably ‘sanitary items’—a formulation that goes too far for some.530 That is not the sort of clarity a ‘common’ transnational asylum system requires, even if we figure in the potential of non-binding guidelines of the Asylum Agency. Modalities of provision amplify the room for disparities. Reception conditions may be provided in kind, either by public authorities or private subcontractors, or in the form of financial allowances, with vouchers as an intermediate category. The level of support for nationals serves as a benchmark, followed by a caveat on the option of ‘less favourable’ treatment.531 In the only judgment considering the modalities of provision, the Court 524 See RCD Proposal (n 238) arts 16–18; and Lieneke Slingenberg, ‘Political Agreement on a Recast Asylum Reception Conditions Directive’ in Thym and Odysseus Academic Network (n 40) 257, 262–64. 525 See Presidency (n 238); and ch 2.3.2. 526 See Nicola Rogers, ‘Minimum Standards for Reception’ (2002) 4 EJML 215; and Evangelia (Lilian) Tsourdi, ‘EU Reception Conditions’ in Chetail and others (n 27) 271. 527 See ch 13.4.1; and Case C-179/11 Cimade and GISTI EU:C:2012:594, paras 37–39. 528 ibid paras 51–61; and Joined Cases C-322/19 and C-385/19 The International Protection Appeals Tribunal and others EU:C:2021:11, paras 62–72. 529 Cimade and GISTI (n 527) paras 40–49; and Gnandi (n 290) para 63. 530 See RCD Proposal (n 238) art 2(7); and Presidency (n 525). 531 See Reception Conditions Directive 2013/33/EU, arts 2(g), 17(5).

412 Common European Asylum System emphasised that Member States have a ‘certain margin of discretion’ as regards the methods for the provision of material reception conditions.532 The case concerned the former Reception Conditions Directive 2003/9/EC but the basic contours remain: vague definitions are combined with legal and practical leeway of the Member States regarding the level of protection and the modalities of provision. Article 18(1)–(9) add important guarantees for accommodation, which focus on special needs, protection of vulnerable groups, and family unity. At the same time, the provision does not contain guidance as regards, by way of example, room sizes, private and communal areas, cooking and washing facilities, furniture, or internet access. Article 19 on health care is replete with open-ended formulations like ‘necessary’, ‘emergency care’, ‘essential treatment’, and ‘serious mental illness’, which can be interpreted in diverse way.533 Only vulnerable persons and minors, both unaccompanied and with families, are given special attention. Articles 21–25 lay down substantial guarantees, reinforced by leges speciales throughout the Directive. A distinct ‘vulnerability assessment’ must be performed, which can be combined with the corresponding examination under the Asylum Procedures Directive.534 Unfortunately, these standards are not always complied with in practice, especially in challenging circumstances at the external borders. Children must have access to education, which may be provided in accommodation centres according to Article 14(1). This has been criticised in light of the best interests of the child, together with optional postponement for a three-month period for preparatory classes, in particular for acquiring language skills.535 Labour market access has repeatedly caused controversies during the negotiations, and it is currently available after no later than nine months. Debates about labour market access are classic examples of the clash of interests between, on the one side, migrant self-sufficiency and, on the other side, the minimisation of ‘pull’ factors.

13.8.3 Restrictions and human rights compliance Member States may reduce or withdraw material receptions conditions in accordance with Article 20(1). Other guarantees in the Reception Conditions Directive cannot, as a result, be withheld on the basis of Article 20, although Member States may possibly rely on the unwritten general principle of abuse to justify a reduction. The Court recognised that labour market access may exceptionally be delayed beyond the statutory nine-month limit if the applicant fails to cooperate, even though secondary movements and legal remedies do not qualify as a reproachable conduct as such.536 In any case, decisions for reduction or withdrawal must be taken on an individual basis and provide for a legal remedy in accordance with Articles 20(5) and 26.

532 See Case C-79/13 Saciri and others EU:C:2014:103, paras 47–51. 533 See Evangelia (Lilian) Tsourdi, ‘Asylum Reception Conditions Directive 2013/ 33/ EU’ in Thym and Hailbronner (n 65) Articles 18–19. 534 See Reception Conditions Directive 2013/33/EU, art 22; and Asylum Procedures Directive 2013/32/EU, art 24(1), (2). 535 See Tsourdi (n 533) Article 14 MN 13–14, 19–23; and Anna-Maria Konsta, ‘Is There a Right to Human Dignity? The Example of the Right to Education of Refugees’ (2019) 21 EJML 261. 536 See ch 10.8; and The International Protection Appeals Tribunal and others (n 528) paras 74–93.

Reception Conditions for Asylum Applicants 413 Reasons for reduction or withdrawal in Article 20(1)–(4) must be read as being complementary.537 Member States have ‘some latitude’ when determining the character of sanctions, which can take different forms and relate to behaviour outside reception centres.538 At the same time, judges highlighted that, according to Article 20(5), access to health care and a ‘dignified standard of living’ shall always be guaranteed, together with respect for the best interests of the child.539 This conclusion contrasts with the legislative option of complete withdrawal ‘in exceptional and duly justified cases’. A declaration, on the part of judges, that sanctions cannot result in a ‘situation of extreme material poverty that does not allow that person to meet his or her most basic needs such as a place to live, food, clothing, and personal hygiene’540 can be read to declare complete withdrawal to be contrary to the Charter. Nevertheless, the ruling was not crystal clear, since judges reference to human dignity replicated the wording of the Directive; a follow-up judgment stressed the wording and did not mentioned Article 1 CFR as an additional argument any more.541 Judges might possibly defend, therefore, a different conclusion for other scenarios. These uncertainties reiterate our earlier finding, on the constitutional foundations, regarding the scope of Article 3 ECHR and Article 4 CFR. They are particularly relevant for the ongoing reform debate, since the Council proposes, as we have seen, sanctioning secondary movements by means of a complete withdrawal of reception conditions, subject to a human rights caveat. To require Member States to ensure ‘basic needs such as a place to live, food, clothing, and personal hygiene’ would establish a minimum level of protection. What this means exactly would have to be hammered out, either by the legislature or by judges. For instance, it would have to be determined to what extent ‘basic needs’ transcend an ‘adequate’ standard of living under Article 17(2). In practical terms, much would depend on how national legislation defines ‘regular’ reception standards in their national laws and administrative practices. Room for restrictions would exist whenever the ‘regular’ standards are higher than the minimum standard of ‘basic needs’. Other judgments on reception conditions similarly invoked human rights and the notion of a ‘dignified’ standard of living.542 Having said this, the references remained abstract and were employed as one argument amongst others to interpret secondary legislation. It remained uncertain, in other words, whether they merely reinforced statutory guarantees or established an autonomous level of protection under human rights law.543 ‘Hard’ cases that cannot be resolved by means of interpretative finesse would require the Court to abandon such ambivalence by correcting the outcome of the legislative process in light of human rights. To do so remains challenging given the fragility of the social guarantees in the Charter, mentioned in Chapter 5.4.1. Legislative proposals on the instrumentalisation of migration by third states foresee a reduction to ‘basic needs’, as opposed to ‘adequate’ standards in regular circumstances. Food, water, clothing, health care, and ‘temporary shelter adapted to seasonal weather conditions’ 537 See Haqbin (n 91) paras 42–44. 538 ibid para 41; Case C-422/21 Ministero dell’Interno EU:C:2022:616, paras 26–35; and ch 6.4.4. 539 Haqbin (n 91) paras 45–46, 53–55. 540 ibid para 46. 541 ibid paras 46, 51; and Ministero dell’Interno (n 538) paras 37–39. 542 See Saciri and others (n 532) para 65; Cimade and GISTI (n 531) paras 42, 56; and The International Protection Appeals Tribunal and others (n 528) para 69. 543 See ch 6.5.3.2; Nika Bačić Selanec and Davor Petrić, ‘Migrating with Dignity’ (2021) 17 EuConst 498, 502– 507; and Tsourdi (n 526) 311–16.

414 Common European Asylum System are mentioned in what appears to be support for makeshift accommodation in tents.544 Additional reference to the ‘full respect of human dignity’ reiterates that the human rights will be an essential yardstick for anyone challenging insufficient reception conditions.

13.8.4 Right to remain and ‘fiction of non-entry’ Asylum seekers shall be allowed to remain on the territory, with the exception of subsequent applicants. This right to remain endures from the moment the application has been ‘made’ in the sense of the initial expression of the desire to request asylum, not formal lodging or registration. It extends to Dublin cases up until the completion of the transfer.545 The right to remain during legal remedies against a negative asylum decision was mentioned in the comments on asylum procedures. Asylum seekers will receive a ‘document’ certifying their status as an applicant.546 Note that the right to remain is ‘for the sole purpose of the procedure’, according to Article 9(1) Asylum Procedures Directive, thus staying short of a ‘residence permit’ authorising cross-border mobility, as explained in Chapter 12.4.1.2 on the Schengen area. Explicitly stating that secondary movements are illegal, as the Commission suggests,547 would be declaratory in character. To recognise a right to remain does not specify the degree of free movement within the territory. Two caveats must be discussed in this respect. First, limitations are possible under Article 7 Reception Conditions Directive 2013/33/EU, which the Commission proposes extending. The provision is based on the assumption that free movement shall usually be authorised, unless applicants are obliged to stay in an assigned area. State practices vary greatly, also considering that the Directive does not give guidance as to the size of the area.548 Assignment to a specific area involves a prohibition to leave the assigned area, except by authorisation. The additional option of prescribing the location of ‘actual residence’ concerns the place where someone resides, not mobility during daytime. Provision of accommodation in reception centres does not automatically involve a restriction on the freedom of movement, although an obligation not to leave the centre can amount to detention, which is subject to stricter conditions under Article 8 Reception Conditions Directive, even if domestic legislation speaks of a restriction of liberty.549 The second caveat remains hypothetical for the time being, since it concerns an amendment the Commission proposes. Both the screening mechanism and the extended border procedure shall go hand in hand with a ‘fiction of non-entry’, which the incoming German Presidency had put forward in line with national legislation.550 Legally, this concerns the absence of authorisation to enter.551 The fiction does not entail formal rightlessness, since 544 See Proposal for an Instrumentalisation Regulation (n 51) art 3. 545 See Cimade and GISTI (n 527) paras 45–48. 546 See Reception Conditions Directive 2013/33/EU, art 6. 547 See AMMR Proposal (n 44) art 9(4) 548 See Tsourdi (n 533) Article 7 MN 8–21. 549 See Országos Idegenrendészeti Főigazgatóság Dél-alföldi Regionális Igazgatóság (n 107) paras 253–54; on detention Valstybės sienos apsaugos tarnyba (n 109) paras 37–42; and ch 16.7. 550 See German Government, ‘Outline for Reorienting the Common European Asylum System’ (Food for thought, 13 November 2019) www.statewatch.org/media/documents/news/2019/dec/eu-asylum-FoodForThou ght-GermanNoPaper.pdf; and Residence Act (Aufenthaltsgesetz), s 13(2) www.gesetze-im-internet.de/aufenthg_2 004/__13.html (accessed 1 March 2023). 551 See Proposal for a Proposal for a Screening Regulation (n 248) art 4(1); and Amended APR Proposal (n 181) art 41(6).

Rights of Beneficiaries of International Protection 415 human rights apply unreservedly on the territory; the fiction is administrative, not constitutional, in character.552 Non-entry is fictitious insofar as factual presence and legal authorisation diverge. Somewhat surprisingly the Commission refrains from spelling out the detail of what the fiction of non-entry entails, thus leaving Member States leeway to define the parameters in their national laws. In any case, restrictions on movement appear to be the core element of the fiction of non-entry, mirroring airport procedures where applicants are not allowed to leave the space where border procedures and accommodation take place, within or adjacent to the transit area of international airports. From a human rights perspective, such limitations to mobility during periods of entry clearance may be qualified, for some time at least, as a ‘restriction’ on liberty, staying short of a ’deprivation’ in the sense of detention. As a result, the fiction need not contradict the Commission’s insistence that detention shall not be automatic and require an individualised assessment, also during border procedures. The ensuing conundrum between generalised ‘restrictions’ on liberty and further ‘deprivation’ (detention) will be discussed in Chapter 16.2.2 on return. Article 5 ECHR and Article 6 CFR can be relied upon to challenge restrictions. By contrast, Article 2 Additional Protocol No 2 does not seem to apply, since asylum applicants are not ‘lawfully staying’. Article 45(2) CFR is formulated in an even narrower manner.553 At the time of writing, an agreement on the Screening Regulation appeared to be a likely scenario, but we can expect the institutions to define the conditions and the legal effects of the ‘fiction of non-entry’ during the negotiations.

13.9 Rights of Beneficiaries of International Protection The notion of ‘international protection’ is firmly established in EU legislation as an overarching category for refugee status and subsidiary protection. Nevertheless, the choice of terminology remains ambiguous from an international perspective, as the term is equally used for protection afforded by international actors, especially UNHCR.554 Articles 20–35 Qualification Directive 2011/95/EU cover the rights and benefits associated with refugee status and subsidiary protection, building upon Articles 12–34 Refugee Convention. Although the Qualification Directive does not bring about a level playing field (13.9.1), it contains important guarantees (13.9.2), including free movement within the territory (13.9.3). Difficult to understand are scenarios when refugees receive neither a protection status nor a residence permit (13.9.4).

13.9.1 Limits of equal treatment Most provisions in the Refugee Convention concentrate on the treatment of refugees after recognition, not access to the territory or the asylum procedure. The Convention was originally primarily about empowering refugees residing in the host state to be

552 See ch 12.2.2. 553 cf AG Villalón, Opinion in Joined Cases C-443/14 and C-444/14 Alo and Osso EU:C:2015:665, point 46. 554 UNHCR’s mandate covers a broad range of ‘mandate refugees’, not only those covered by the Refugee Convention.

416 Common European Asylum System economically active and to pursue an independent life. The CJEU has confirmed that Articles 20–35 Qualification Directive 2011/95/EU must be interpreted in light of the Refugee Convention.555 The pervasive focus on equal treatment with either nationals or other foreigners reflects the regulatory technique of the former ‘aliens law’, described in Chapter 5.1.1 on state sovereignty. To concentrate on equal treatment may be of intuitive appeal, although it will not turn the ‘common’ European asylum system in a level playing field. Equal treatment with nationals in the domain of social welfare (Article 29 Qualification Directive) entails that the level of benefits depends on the welfare state of the host state: refugees will receive less support in Bulgaria than in Belgium. In some countries, beneficiaries of international protection may even be worse-off than asylum applicants during the procedure, for which the Reception Conditions Directive establishes a set of—albeit vague—guarantees, for instance with regard to housing. Beneficiaries often live in precarious situations, especially in countries where widespread unemployment hinders the pursuit of an independent life without state support. Chapter 15.3.4 on integration will explain that structural differences between national welfare systems amplify the degree of factual divergence, if important benefits are confined to employment-based social security systems. EU institutions are confronted with a regulatory dilemma in this respect, since they can hardly require governments to treat refugees better than nationals. Article 34 Qualification Directive on access to integration facilities remains vague, and financial support for integration projects, mentioned in Chapter 7.4, cannot fundamentally alter the situation either.

13.9.2 Specific guarantees Article 20(1) Qualification Directive starts with an important disclaimer: additional benefits may be available under the Refugee Convention, with regard to rights not covered by Articles 20–34 Qualification Directive. Member States are obliged to respect these guarantees qua international law, as made clear in Chapter 5.6. An important innovation of the Qualification Directive were assurances for unaccompanied minors and other vulnerable people in Articles 20(3)–(5) and 31.556 Article 23(2)–(5) lay down basic guarantees for family members who do not qualify for international protection themselves. Entry from abroad is governed by the Family Reunification Directive and will be discussed in Chapter 14.3.5 on legal migration. Leges speciales exist, as we have seen, for family unity in the Dublin system. Directive 2011/95/EU approximated the status of refugees and beneficiaries of subsidiary protection, with some exceptions. The Commission wants to reverse this trend by increasing the difference—a project that can be challenged in light of human rights, as explained in Chapter 15.3.5 on integration. A different question is whether Member States are allowed to treat beneficiaries of international protection unlike nationals or foreigners legally resident, although Articles 20–34 Qualification Directive prescribe equal treatment.

555 556

See Joined Cases C-443/14 and C-444/14 Alo and Osso EU:C:2016:127, paras 28–30, 51. See Moreno-Lax and Garlick (n 110) 159–60, 176–77.

Rights of Beneficiaries of International Protection 417 Chapter 15.3.2 will illustrate that the Court recognises the option of justification whenever their situation is not comparable. Articles 26–28 on access to employment, education, and procedures for the recognition of qualifications are of great practical relevance. Note, however, that they do not exempt beneficiaries from compliance with domestic rules on the exercise of specific professions. Whenever they require formal qualifications, individuals may be excluded de facto if the education they have received at home does not correspond to the host country’s requirements. Again, further comments on the practical limitations of abstract equal treatment can be found in Chapter 15.3 on integration. The prohibition of inhumane and degrading treatment and the social guarantees in the Charter may possibly be invoked to request additional state support, subject to the above-mentioned uncertainties about the implications of Court judgments on a ‘dignified standard of living’.

13.9.3 Free movement and option of ‘uniform status’ Beneficiaries of international protection may move freely within the territory of the country that issued the status. Notwithstanding divergent language versions, Article 33 Qualification Directive covers the choice of residence if read in light of Article 26 Refugee Convention.557 Judges held that equal treatment with third country nationals implies the option to justify differential outcomes whenever their situation is not comparable. German courts were asked to assess whether beneficiaries of subsidiary protection are confronted with specific integration challenges which justify mandatory assignment to a place of residence if they were unable to find a job.558 Similar questions may arise in other Member States. Travel documents must be issued under Article 25 Qualification Directive 2011/95/EU, in accordance with Article 28 and the Schedule attached to the Refugee Convention. The idea behind issuing travel documents takes up the historic example of the ‘Nansen passport’, which was meant to facilitate cross-border movements. Yet, the travel document does not give rise to an authorisation to enter another state, as described in Chapter 12.4.1.2 on the Schengen area. The Council of Europe Agreement on Transfer of Responsibility for Refugees, which is binding upon ten Member States, does not bring about transnational mobility rights either.559 However, it might possibly be used when the Dublin system does not foresee, as a result of the ongoing reform debate, the transfer of responsibility after onward movements of beneficiaries of international protection. To say so would be challenging, however, since the Convention requires ‘two years of actual and continuous stay in the second State with the agreement of its authorities’.560 What appears to be an automatic transfer of responsibility under Article 2(3) might harbour more potential. Having said this, the Convention covers movements between state parties only and is not subject to the CJEU’s jurisdiction.

557 Alo and Osso (n 555) paras 24–40. 558 ibid paras 54, 57–61. 559 Council of Europe Agreement on Transfer of Responsibility for Refugees (adopted 16 October 1980, entered into force 1 December 1982) ETS No 107; and for the state of ratification https://www.coe.int/en/web/conventi ons/full-list?module=treaty-detail&treatynum=107 (accessed 1 March 2023). 560 ibid art 2(1).

418 Common European Asylum System Domestic courts may interpret it not to confer individual rights, in accordance with the classic characteristics of international law.561 The vision of a ‘uniform status of asylum, valid throughout the Union’, enshrined in Article 78(2)(a) TFEU, need not result in enhanced mobility either. The provision entrusts the legislature with deciding whether and if so how to design a truly uniform status, which could take alternative forms. On the one hand, it might entail mutual recognition of positive asylum decisions. The practical effects would be the opposite of free movement. Member States would be authorised to refer refugees to the country which had issued the status.562 Political debates in Brussels about how to respond to secondary movements are moving in this direction, as we have seen, although they are not commonly framed in terms of establishing a ‘uniform status of asylum’. On the other hand, a ‘uniform status of asylum’ can hint at enhanced mobility, possibly even mirroring the free movement of Union citizens. Article 45 CFR leaves the legislature much leeway in this respect, with existing rules being limited to short stays abroad.563 An official authorisation to work and reside elsewhere is available after five years under the conditions laid down in Articles 14–23 Long-Term Residents Directive 2003/109/EC, although additional entry options may be available under domestic laws. The Commission proposes lowering the threshold to three years.564 Directives on legal migration, with the exception of blue card, exclude beneficiaries of international protection ratione personae,565 or require residence abroad.566 The EU legislature could change this, or establish distinct standards, requiring, by way of example, a job offer and sufficient resources.567 For the time being, such an amendment remains elusive.

13.9.4 Refugees without protection status Recitals 21 and 40 Qualification Directive 2011/95/EU embody a tension inherent in refugee status determination: whereas the outcome of the procedure is declaratory in character, access to important benefits presupposes a residence permit. What might appear to be an unresolvable puzzle can be elucidated in light of our previous comments that refugee law does not embrace an individual right to be granted asylum. Respect for non-refoulement is mandatory, not, however, access to formal asylum status. Rights in the Refugee Convention are categorised in groups, and the scope ratione personae of each provision is distinguished semantically through the words ‘present’, ‘lawfully present’, or ‘lawfully staying’. While presence covers anybody on the territory, ‘lawful stay’ is habitually understood to presuppose a formal status.568 561 On the German practice see Heiko Habbe, ‘Übergang der Verantwortung für anerkannte GFK-Flüchtlinge’ [2021] Anwaltsnachrichten Ausländer-und Asylrecht (ANA-ZAR) 49. 562 See Peers (n 119) 243–44; and note that the Refugee Convention does not require otherwise. 563 See ch 12.4.1.2. 564 See AMMR Proposal (n 44) art 71; and ch 15.6.2. 565 See Blue Card Directive (EU) 2021/1883, art 2(1), 2(a), (b); and Students and Researchers Directive (EU) 2016/801, art 2(2)(a). 566 See ICT Directive 2014/66/EU, art 2(1); and Seasonal Workers Directive 2014/36/EU, art 2(3). 567 For regulatory options see Expert Council on Integration and Migration (Sachverständigenrat für Integration und Migration), Chancen in der Krise. Jahresgutachten 2017 (May 2017) 41–45, as whose member and vice- chairperson the author of this book served from 2016 to 2022. 568 See Grahl-Madsen (n 98) §§ 209–27; and Hathaway (n 334) ch 3.

Rights of Beneficiaries of International Protection 419 Against this backdrop, the status of asylum seekers remains subject to debate; some maintain that their presence qualifies as ‘lawful’ for the purposes of the Refugee Convention.569 In practical terms, the question has little significance, since the prohibition of refoulement applies anyway and the other guarantees associated with lawful presence (free movement and self-employment) foresee equal treatment with foreigners ‘generally in the same circumstances’. Third country nationals presenting themselves at the border or within the territory without authorisation will not, however, usually be authorised to work or move freely. Additional guarantees, notably welfare benefits and paid work, are confined to those ‘lawfully staying’. To say that this covers asylum applicants would negate the graded approach underlying the Refugee Convention and would have little chance to convince the Court, which interprets the Convention authoritatively within the EU legal order. Judges at the ECtHR confirmed that outcome with regard to free movement.570 Judges in Luxembourg had to confront the conundrum of refugees without formal status in a judgment on Article 14(4)–(6) Qualification Directive, which defy easy qualification as a result of unsatisfactory drafting. The provision can be read to introduce an additional exclusion clause that is wider than Article 1F Refugee Convention, since the wording indicates, in some language versions at least, that the denial of ‘refugee status’ implies that the applicant ceases to be a refugee.571 The Court rejected that interpretation by focusing on Article 14(6) which requires Member States to ensure that refugees without a formal status benefit from some minimum guarantees under the Refugee Convention.572 Vague wording covering rights ‘similar to’ the ones mentioned explicitly prepared the ground for an outcome covering at least the rights requiring ‘presence’ and ‘lawful presence’; additional benefits for ‘lawful stay’ are not available, unless domestic authorities authorise the stay on other grounds than international protection.573 A related question concerns Article 24(1) which authorises Member States not to issue a residence permit to refugees for compelling reasons of national security or public security—a lower threshold than the exception from the prohibition of refoulement under Article 21(2) Qualification Directive and Article 33(2) Refugee Convention. Nevertheless, judges confirmed that Member States may refuse or withdraw residence permits under more generous conditions.574 The remainder of the judgment concentrated on the definition of the standards for refusal or withdrawal, including the public policy and security caveat, discussed in Chapter 10.3 on general features. For our purposes, the difference to the preceding judgment stands out: refusal of residence permits need not coincide with withdrawal of refugee status under Article 14. As long as refugee status is not revoked, refugees without a residence permit will benefit from Articles 20–35 Qualification Directive.575 Absence of residence permits may, however, have disadvantages under domestic laws.

569

See Hathaway (n 334) 196–211; and Battjes (n 69) 448–69; contra Grahl-Madsen (n 98) 361–64.

570 See Omwenyeke v Germany App no 44294/04 (ECtHR, 20 November 2008) § 1.

571 See M and others (n 108) paras 83–92; and Moreno-Lax and Garlick (n 110) 129–32. 572 See M and others (n 108) paras 94–102. 573

ibid paras 103–109. See Case C-373/13 T EU:C:2015:413, paras 47–55. 575 ibid paras 95–98. 574

420 Common European Asylum System

13.10 Resettlement and Other Legal Pathways Member States and EU institutions have invested considerable energy in erecting hurdles for access to the territory by means of international cooperation, visa policy, and border controls. Corresponding accusations of ‘fortress Europe’ are often refuted by a rhetoric emphasis on legal pathways. Resettlement and other protected entry procedures feature prominently in policy papers—in striking contrast to constrained national practices (13.10.1). An upgrade of the EU’s legal framework for resettlement is in the making; its adoption would maintain the voluntary character and the absence of elaborate procedures (13.10.2). The toolbox of legal pathways includes community sponsorship and other initiatives, which may give a greater role to private parties (13.10.3). Humanitarian visas featured in Chapter 11.5 on visa policy.

13.10.1 Soft side of asylum policy Resettlement and other legal pathways serve important humanitarian objectives by allowing individuals to travel safely to a destination where they will benefit from a secure status.576 A classic example are Somalis nominated by UNHCR in a refugee camp in Kenya to fly to Sweden where they will receive international protection swiftly. Resettlement is often available for vulnerable groups who might not have the capacities or resources to reach Europe irregularly, even if they wished to do so. Notwithstanding these undeniable humanitarian credentials, critical scholars highlight the potential of ulterior motives behind the rhetorical emphasis on legal pathways in policy debates. They may, implicitly or explicitly, distinguish between ‘good’ refugees (waiting to be resettled) and ‘bad’ refugees (entering irregularly), thus undermining territorial asylum; moreover, states may sponsor resettlement in parallel with enhanced control practices by means of externalisation.577 The danger is particularly pronounced when low numbers indicate that legal pathways are little more than a humanitarian fig leaf. Recital 4 Refugee Convention signals that the desire for inter-state solidarity is as old as the Convention. Multiple initiatives for ‘burden-sharing’ have been promoted over the years, including ‘sharing people’ via resettlement as one option amongst others (echoing the debate about the meaning of Article 80 TFEU). Its heyday was the Hungarian uprising of 1956 and the ‘Comprehensive Plan of Action’ regarding Indo-Chinese boat people. Western states welcomed those fleeing Communism with open arms, in contrast to other crisis scenarios.578 Large resettlement schemes in the US, Canada, and Australia build upon this tradition, whereas resettlement never gained much traction in Europe outside Scandinavia and the UK. EU institutions started dealing with resettlement in earnest in response to the British proposal for external processing, mentioned in the context of asylum procedures, which

576 See UNHCR, Resettlement Handbook (revised edn, July 2011) 3. 577 See Susan Kneebone and Audrey Macklin, ‘Resettlement’ in Costello and others (n 264) 1080, 1080–82; and Ziebritzki (n 72) 305–16. 578 See the introductory section to this chapter; and Penelope Mathew and Tristan Harley, Refugees, Regionalism and Responsibility (Edward Elgar Publishing 2016).

Resettlement and Other Legal Pathways 421 generally paved the way for a robust external migration policy.579 Debates within the Council about national initiatives to resettle vulnerable minorities from Iraq during 2008, mostly Christians and Yazidis, proved decisive in spreading the idea to countries with little or no previous track record; the Commission supported the initiative through funding as an incentive to increase numbers.580 Similar projects were pursued in the years to come, until the European Agenda on Migration and the policy crisis of 2015/16 accelerated things.581 Member States agreed to resettle more than 20,000 persons over a two year period, while insisting on the voluntary nature.582 Another 50,000 places were made available during 2017–19, followed by almost 30,000 during 2020/21.583 Implementation proved reasonably successful, until the Covid-19 pandemic slowed things down.584 Additional places were made available under the EU–Turkey Statement. Member States committed to resettle one Syrian from Turkey for each Syrian returned there from the Greek islands under the ‘one-for-one mechanism’, which was scheduled to be replaced by a humanitarian admission scheme in the medium run.585 The Statement produced the intended results insofar as the number of arrivals decreased dramatically (partly as a result of the parallel closure of land borders on the Western Balkans). By contrast, the one-for-one mechanism was hardly implemented. Asylum procedures on the Greek islands took ages, Greek courts stopped many returns, and resettlement remained modest.586 Notwithstanding these caveats, the example reiterates that debates about legal pathways may be intricately linked to international cooperation. The same can be said of the emergency transfer mechanism for resettling refugees from Libya via Niger,587 in the broader context of cooperation with Libya to be traced throughout Chapter 18.3 on the external dimension.

13.10.2 EU resettlement framework Resettlement is an outlier among the legislative instruments discussed in this chapter. None of the measures constitutes legislation in the sense of being legally binding (with the exception of budgetary support). Commission recommendations coalesce with voluntary pledges of the Member States, which are coordinated in an intergovernmental framework. At the time of writing, the EU’s resettlement framework was a prime example of governance by soft 579 See Commission, ‘Communication on improving access to durable solutions’ COM(2004) 410 final, 7–11; and Joanne Van Selm, Erin Patrick, and Tamara Woroby, Study on the Feasibility of Setting Up Resettlement Schemes in EU Member States or at the EU level (Publications Office, May 2004). 580 See ‘Press Release’ (Council doc 16325/ 1/ 08 REV 1, 28 November 2008) 23– 26; Commission, ‘Communication on the establishment of a joint EU resettlement programme’ COM(2009) 447 final; and Moreno- Lax (n 353) 619–33. 581 See Meltem İneli Ciğer, ‘Resettlement Framework Regulation (Proposal)’ in Thym and Hailbronner (n 65) Article 1 MN 20–31; and Tom de Boer and Marjoleine Zieck, ‘The Legal Abyss of Discretion in the Resettlement of Refugees’ (2020) 32 IJRL 63. 582 See ‘Conclusions of the Representatives of the Governments of the Member States meeting within the Council’ (Council doc 11130/15, 22 July 2015); and Commission Recommendation (EU) 2015/914 on a European Resettlement Scheme [2015] OJ L148/32. 583 See Commission Recommendation (EU) 2017/1803 on enhancing legal pathways [2017] OJ L259/21; and Commission Recommendation (EU) 2020/1364 [2020] OJ C317/13. 584 See Commission (n 158) 19–20, Annex V; and EASO Asylum Report 2022 (EASO 2022) 250. 585 ‘EU–Turkey Statement’ (n 32) Nos 1–4; and Commission Recommendation for a voluntary humanitarian admission scheme with Turkey, C(2015) 9490 final. 586 See Commission (n 328) 9. 587 See Altai Consulting, ‘Case Study. Emergency Transit Mechanism’ EUTF (July 2021).

422 Common European Asylum System law, mentioned in Chapter 2.3.4 on the institutional practice. To be sure, the Commission tabled a Proposal for a Resettlement Framework Regulation in 2016, on which provisional political agreement had been reached between the Council and the Parliament, although adopted way postponed as a result of the ‘package approach’ to asylum reform.588 It was likely to be formally adopted before the European elects in the spring of 2024. Article 78(2) (d) TFEU may serve, as we have seen, as a legal basis. As it stands, the new instrument is a genuine ‘framework’, not an extensive rulebook like the Asylum Procedures Directive. It comprises two elements: identification of priority countries and annual plans setting out the contributions of each Member State, as well as eligibility criteria and rudimentary procedural standards for the identification of beneficiaries. Regarding political programming, the Council and the Parliament agree that the voluntary character of national pledges should be maintained; adoption of a legally binding regulation would not, in other words, bring about an obligation to resettle refugees to a country that chooses not to participate.589 NGOs criticise that the selection of partner countries would formalise previous practices that used resettlement as a bargaining chip to improve cooperation on migration control.590 Eligibility criteria are modelled on the refugee definition and subsidiary protection, together with priority for vulnerable groups and family members of people residing in Europe already; extended grounds for exclusion respond to the security concerns of national governments.591 While orientation at the criteria for international protection is politically warranted and corresponds to the international practice, this is not legally mandatory, under neither international nor Union law. Resettlement remains a ‘sovereign’ prerogative, since non-refoulement obligations presuppose presence on the territory, or effective state control by other means.592 Along similar lines, international procedural standards remain abstract and non-binding, with varying degrees of UNHCR involvement and accountability deficits.593 Articles 10–11 Proposal for a Resettlement Framework Regulation replicate this model, in obvious contrast to the Asylum Procedures Directive. Recital 19 reaffirms that individuals shall not have a ‘subjective right to be resettled’, and Article 10(6) does not foresee legal remedies. To say so can be compatible with EU constitutional law if we conclude that the application of Article 47 CFR presupposes an individual right, enshrined in secondary legislation, as explained in Chapter 6.5.4 on the doctrinal foundations.594 Additional political controversies concern the authorisation to take into account ‘social or cultural links . . . that can facilitate integration’, reflecting previous practices in the selection of beneficiaries.595 Those resettled will receive international protection. 588 See Proposal for a Resettlement Framework Regulation COM(2016) 468 final; and İneli Ciğer (n 581) ch 24. 589 Resettlement Proposal (n 588) arts 7–8, 13–15; and LIBE Committee, ‘Report on the Proposal for a Resettlement Framework Regulation’ (A8-0316/2017, 19 October 2017) Amendment 75. 590 Resettlement Proposal (n 588) art 4(d), (e); and ECRE, ‘Untying the EU Resettlement Framework’ (Policy Note 01/2016) 2. 591 Resettlement Proposal (n 588) arts 5–6. 592 See ch 12.2.2; and Hathaway (n 334) 1189–206. 593 See UNHCR (n 576) ch 7; and Kneebone and Macklin (n 577) 1091–93. 594 Moreover, the extraterritorial reach of the Charter would have to be ascertained; see de Boer and Zieck (n 581) 74–83. 595 See Resettlement Proposal (n 588) art 10(1)(b); de Boer and Zieck (n 581) 70–73; and Nathalie Welfens, ‘Selection Categories in European Resettlement and Humanitarian Admission Programmes’ in Moritz Jesse (ed), European Societies, Migration, and the Law (CUP 2020) 81–103.

Summary 423 Doing so corresponds to the basic idea behind resettlement that legal entry shall be combined with the prospect of settlement in the host country, defining resettlement as a ‘durable solution’.596

13.10.3 Other protracted entry procedures Resettlement is firmly established on the international plane and presents the prototype of a protected entry procedure (PEP), as legal pathways are often referred to in EU jargon. Other prominent examples are humanitarian visas, discussed in Chapter 11.5, and temporary protection, portrayed previously. ‘Humanitarian admission schemes’ complement the toolbox. They serve as an umbrella concept for national schemes which do not involve UNHCR in the selection process and which may foresee other legal statuses than international protection.597 Differences between resettlement and humanitarian admission will often be gradual, and the latter does not constitute a ‘subordinate’ legal pathway. A common denominator of resettlement and humanitarian admission is that the selection of those to be admitted rests with public authorities, either states or the UNHCR, not refugees as in the case of humanitarian visas. A sub-category are ‘community sponsorship schemes’, which constitute an intermediate category. They involve private actors in the identification and reception of beneficiaries, including financial and other responsibilities during the first years of residence. Community sponsorship builds upon the widely celebrated ‘humanitarian corridors’ in Italy, as well as the example of Canada.598 The Commission promotes them as best practice.599 Finally, refugees are not only victims of persecution. They are individuals who may have valuable skills that can be beneficial for host societies in times of demographic change. Legal pathways for economic purposes can be an alternative avenue of entry, mirroring the original idea behind the Refugee Convention to empower refugees to pursue an independent life by means of socioeconomic incorporation. Russian and Turkish nationals, for example, may use legal migration channels to leave their home state without applying for asylum. Similar phenomena can be observed elsewhere.600 In such scenarios, the decision to leave will often rest on mixed motives, with political oppression and the lack of freedom as one element amongst others. In the EU context, the focus of the economic migration instruments on skilled workers limits the potential of the ‘labour route’.

596 Resettlement Proposal (n 588) art 10(7); and UNHCR (n 576) 28–41. 597 See FRA, ‘Legal Entry Channels to the EU for Persons in Need of International Protection. A Toolbox’ (FRA Focus 02/2015); and EMN, Resettlement and Humanitarian Admission Programmes in Europe: What Works? Study (9 November 2016). 598 See ICF/MPI, ‘Study on the Feasibility and Added Value of Sponsorship Schemes as a Possible Pathway to Safe Channels for Admission to the EU’ (October 2018); and Carola Ricci, ‘The Necessity for Alternative Legal Pathways’ (2020) 21 German LJ 265. 599 See Commission Recommendation (EU) 2020/1364 on legal pathways to protection in the EU [2020] OJ C317/13, Nos 11–18. 600 See Katy Long and Sarah Rosengaertner, ‘Protection through Mobility: Opening Labor and Study Migration Channels to Refugees’ (Migration Policy Initiative, October 2016).

424 Common European Asylum System

13.11 Summary Europeanisation was triggered by the desire of the main asylum destination countries to promote new legislative tools and to share responsibilities with other countries. Restrictive approaches dominated the early years of intergovernmental cooperation, as exemplified by the provisions on asylum jurisdiction and the exchange of ‘best practices’ on safe countries. The malfunctioning of the Dublin system draws our attention to structural deficits that have plagued European asylum policy from the beginning. Implementation deficits and legislative design flaws continue to be widespread. These deficiencies were not fundamentally reversed during the first and second phase of legislative harmonisation after the millennium change. EU institutions agreed on complex directives and regulations, which rely on sophisticated procedures Member States struggle to implement in practice. Events during 2015/16 sparked a sense of crisis and increased politicisation, which was reinforced by disputes with neighbouring countries about migration. Proposals for legislative reform have been on the table since 2016 but adoption appears beyond reach. Comments in this chapter concentrated on the legislation harmonising domestic laws, with the objective of establishing a Common European Asylum System (CEAS). Measures dealing with access to the territory by means of visa policy, border controls, and cooperation with third states are discussed elsewhere in this volume. An essential instrument is the Dublin III Regulation (EU) No 604/2013 which assigns jurisdiction for the examination of each application to one Member State, thus connecting national asylum systems horizontally. The futile quest for inter-state solidarity and debates about the ‘first entry rule’, according to which states at the external borders should take back those entering irregularly, demonstrate how the path-dependent continuation of earlier legislation complicates policy reform. Highly relevant for administrative practices are numerous Court judgments on take back procedures, legal remedies, and the substantive criteria for asylum jurisdiction. Human rights limits to Dublin transfers caused controversies during the 2010s. By contrast, secondary movements often receive less attention, although they feature prominently in the ongoing reform debate. EU institutions consider discontinuing the transfer of jurisdiction, which requires destination countries officially to assume jurisdiction whenever an applicant is not transferred to the state responsible. Provisions in the Asylum Procedures Directive 2013/32/EU support domestic authorities and courts in the identification of those with protection needs recognised by Union law. They apply whenever an application has been ‘made’ in the sense of the initial expression of a desire to request asylum, not formal lodging or registration. Procedures revolve around the personal interview, which prepares the ground for an individualised assessment by the competent authorities. A defining feature of the Directive is procedural differentiation, establishing distinct—and stricter—rules for admissibility checks, accelerated procedures, and subsequent applications. Border procedures are a modality more than a distinct procedural avenue, involving, amongst others, restrictions on free movement; they take centre stage, controversially so, in the latest reform proposals. EU legislation concentrates on criteria for safe countries of origin, safe third countries, and first countries of asylum, while Member States retain the responsibility for designating specific countries as safe. Legal remedies against negative asylum decisions are mandatory and several Court judgments have extended the scope and intensity of judicial oversight. All these procedural safeguards could possibly be evaded if the EU institutions reverted to external processing,

Summary 425 which is extremely controversial. Doing so would require an amendment of the Asylum Procedures Directive and raise myriad legal and practical challenges. EU legislation proved decisive in promoting a dynamic interpretation of the refugee definition, including persecution by non-state actors and gender-specific considerations. Abstract criteria in the Refugee Convention were translated into detailed prescriptions in today’s Qualification Directive 2011/95/EU. Notwithstanding the common set of criteria, which have given rise to important judgments of principle, recognition quotas continue to diverge substantially among and within Member States. This divergence results from the high level of abstraction of some rules and the absence of supranational guidance on how to assess the factual situation. To define persecution as a ‘severe violation of basic human rights’ may be intuitively convincing, yet this hardly translates into clear-cut criteria when the severity threshold is crossed, let alone how to assess the situation in specific countries. Judges opted for a generous reading regarding homosexuality and religious freedom, while economic hardship will not usually be covered. Doing so renders it difficult to cover the effects of climate change and to support a generous reading of the internal protection alternative. The Court confirmed that international actors, including military operations, may provide protection, and it rendered several rulings on exclusion and cessation. Judges were criticised for a narrow reading of the notion of ‘particular social group’, but the criteria put forward might prove flexible in practice. Subsidiary protection was an innovation of the first version of the Qualification Directive, especially for those fleeing civil wars. The relevant benchmark of ‘serious harm’ constitutes an autonomous concept of Union law, distinct from human rights. Judges were confronted with the difficult task of clarifying the conditions under which indiscriminate violence may result in an individual threat. The ‘gliding scale’ approach requires national authorities to consider a wealth of factors; that may be a reasonable solution in the abstract but remains notoriously difficult to put into practice. Incoherent outcomes are an almost inevitable consequence. Member States remain free to establish domestic ‘complementary’ protection schemes in diverse scenarios, ranging from human rights grounds to humanitarian or compassionate concerns. For many years, the Temporary Protection Directive 2001/55/EC appeared to remain dead latter, until it was activated, for the first time ever, in response to the arrival of several million Ukrainians during 2022. To do so saved scarce administrative resources, by not requiring an individualised assessment, notwithstanding the marked contrast to the prevalence of restrictions towards people coming from other countries. The activation was legally unproblematic and can remain in place for up to three years. We can expect the EU institutions to introduce new legislation to deal with the legal status of Ukrainians thereafter, in case the war continues. This would avoid complex questions about the interaction between temporary protection and other instruments. Substandard reception conditions for asylum seekers at the external borders and elsewhere are a stigma for EU asylum policy. Vague definitions in the Reception Conditions Directive 2013/33/EU are one reason for the pronounced discrepancies; they leave much legal and practical leeway to states when providing material support, accommodation, and health care. Several judgments hint at a minimum level of ‘basic needs’ under the Charter, although it remains unclear what Member States are obliged to provide exactly. A controversial component of the reform debate is the ‘fiction of non-entry’ during border procedures, which restricts free movement even in the absence of detention. Beneficiaries of

426 Common European Asylum System international protection are subject to a different set of rules: Articles 20–35 Qualification Directive define their status in accordance with the Refugee Convention. Equal treatment with nationals or other foreigners serves as the gold standard after recognition, thus effectively perpetuating differences between national welfare systems. Difficult to understand are special scenarios when refugees do not receive a formal protection status or are not issued a residence permit on security grounds. Resettlement and other legal pathways feature prominently in the policy debate, and recent years have witnessed the proliferation of resettlement programmes to Member States without a previous track record. Notwithstanding the humanitarian credentials of these initiatives, the overall numbers involved remain low, and the policy framework is grounded in soft law instruments. Adoption of the Resettlement Framework Regulation would not substantially alter the picture, since it reiterates the voluntary character of resettlement and does not embrace extensive procedural safeguards. Comparison with the elaborate prescriptions on asylum jurisdiction, procedures, and qualification criteria demonstrates how important the directives and regulations discussed throughout this chapter are. The sheer number of Court judgments on these matters illustrates that administrative practices across Europe are increasingly guided by a common rulebook, which—despite its flaws—is an essential component of European migration law.

14

Legal Migration Migration policy is defined by a mismatch between numbers and thematic attention. Most policy debates revolve around asylum, although most third country nationals enter and stay legally as workers, family members, or students. More than six million first permits were issued during 2018–20 before the Covid-19 pandemic: three million for remunerated economic activities, two million for family reasons, and one million for education.1 Six directives cover entry and stay for these purposes, supplemented by additional entry channels under national legislation. Political circles and legal research pay little attention to these themes. This discrepancy reflects the salience of asylum and a lesser degree of practical relevance of the instruments discussed hereafter. Legal migration often unfolds without controversies, as illustrated by the low number of preliminary references, especially on economic migration. These contextual factors are bound to stay, and yet more political and academic attention is warranted: the number of people involved is impressive and will increase further as a result of demographic change. Our comments will concentrate on entry for purposes of family unity, economic activity, and education. Integration and settlement, including long-term residence, will be discussed in the chapter that follows. Notwithstanding the seemingly clear-cut distinction between legal migration and other entry channels, there is an overlap. By way of example, unsuccessful asylum applicants are authorised to stay legally after having married, and refugees may be given a blue card after having completed university. Both situations do not concern first admission to the territory but are covered by the legal migration instruments nonetheless. Indeed, most directives on legal migration cover first admission and the change in migration status on the territory. Some additional entry channels will be presented in Chapter 17 on association agreements. Europeanisation of legal migration was no foregone conclusion. Protracted disputes hampered the legislative debate, resulting in a patchwork of sectoral instruments on economic migration that lack innovative appeal (14.1). Despite these disputes, EU Treaties lay down a broad range of competences subject to few constraints under human rights law, with the notable exception of family reunification (14.2). The first legal migration instrument to be adopted was the Family Reunification Directive 2003/86/EC, which has not been amended a single time (14.3). A classic distinction defines the rules on economic migration: skilled and highly skilled migrants are usually authorised to enter with the prospect of settlement under the Blue Card Directive, the Students and Researchers Directive, and the Single Permit Directive (14.4), whereas seasonal workers, intra-corporate transferees, and posted workers stay temporarily (14.5). Our analysis will come across gaps and lacunae in the supranational rulebook, thus indicating the continued potential for innovation at the domestic level, especially on economic migration. 1 See the Eurostat dataset for ‘MIGR_RESFIRST’ for the EU-27 https://ec.europa.eu/eurostat (accessed 1 March 2023). European Migration Law. Daniel Thym, Oxford University Press. © Daniel Thym 2023. DOI: 10.1093/oso/9780192894274.003.0015

428 Legal Migration

14.1 Theory and Policy Design In contrast to visas, border controls, and asylum, the argument for common rules on legal migration had never been particularly strong. Nevertheless, EU institutions had made attempts at occupying the terrain ever since the end of the ‘guest worker’ programmes (14.1.1). Family bonds proved a critical entry channel in the following years. Important destination countries responded with restrictions, which informed the negotiations on the Family Reunification Directive (14.1.2). Economic migration caused heated disputes about the desirability of common action (14.1.3). Member States finally agreed to sectoral directives on specific segments of labour migration (14.1.4). The basic outline of the sectoral approach seems to be set in stone, leaving the initiative for innovative entry channels with the Member States (14.1.5). Empirical and theoretical studies assess how economic migration interacts with the welfare state (14.1.6).

14.1.1 ‘Guest workers’ outside the reach of EU institutions The period after the Second World War witnessed comparatively little migration to and from Europe. Two important exceptions were late colonial immigration, mentioned in Chapter 1.1.6, and the ‘guest worker’ programmes. France, Germany, Switzerland, Belgium, and Switzerland had negotiated bilateral recruitment arrangements with countries from the Mediterranean basin, including Italy, Spain, and Greece.2 These initiatives built upon an existing pattern of intra-European labour migration before and after the First World War, as explained in Chapter 1.1.4 on the prehistory of EU migration law. Work in the industrial heartland of the continent was a temporary project for many to earn money and acquire new skills; the majority of the ‘guest workers’ returned home.3 Nevertheless, political debates and the public memory concentrated, for legitimate reasons, on those who stayed after the abrupt end of recruitment during the economic shock during the ‘oil crisis’ which began in the autumn of 1973. Although decisions were taken on the basis of domestic laws, there was some intergovernmental coordination.4 EU institutions tried get a foot in the door when it became apparent that several million ‘guests’ ended up staying. The Commission put forward, at the bequest of the European Council, an Action Programme which focused on what is commonly called social integration nowadays: social services, housing, health, education, and language courses.5 Projects on these matters received funding from the EU budget, but the idea of coordinating immigration policies more generally was not pursued further at the time; national governments had little interest in interference from ‘Brussels’.6 Similarly, the creation of the Schengen area made little inroads into legal migration. There were only modest intra-European movements of the people who had entered as 2 See Stephen Castles and Godula Kosack, Immigrant Workers and Class Structure in Western Europe (OUP 1973) chs 1–2. 3 See Cord Pagenstecher, ‘Die “Illusion” der Rückkehr’ (1996) 47 Soziale Welt 149. 4 See Marcel Berlinghoff, Das Ende der ‘Gastarbeit’ (Schöningh 2013) 191–208. 5 See Commission, ‘Action Programme in Favour of Migrant Workers and their Families’ COM(74) 2250 final http://aei.pitt.edu/1278 (accessed 1 March 2023). 6 ibid 22; and Marcel Berlinghoff, ‘Between Emancipation and Defence’ in Centre international de formation européenne (ed), L’Europe en Formation (CAIRN 2009) 183.

Theory and Policy Design 429 ‘guest workers’ or from the former colonies. The Schengen rulebook confined itself to sort- term travel of third country nationals with domestic residence permits.7 Things changed under the Treaty of Maastricht. The Council agreed on two intergovernmental resolutions, which tellingly featured the words ‘limitations on admission’ in the title, since they were based upon the assumption ‘that the present restrictive measures should be continued and where necessary reinforced’.8 The political context was the politicisation of asylum policy which resulted in an increase in public scepticism in the main destination countries.9 The Commission signed up to the restrictive policy agenda. It accepted that states should have discretion when admitting someone for economic purposes, instead of replicating the freedom-enhancing rationale of Union citizenship.10

14.1.2 Family bonds as a critical gateway Family unity became a controversial topic after the termination of the ‘guest worker’ programmes and the suspension of late colonial migration. Many family members joined migrants living in Europe. This had the seemingly paradoxical effect that the end of liberal admission schemes resulted in more migration, not less. Host societies experienced a classic phenomenon of ‘chain migration’ along family links, as described in the interdisciplinary Chapter 4.1.3. Notwithstanding visible instances of restriction, governments by and large tolerated family reunification, since stopping it would have conflicted with the self-image of liberal democracies. Domestic and international courts influenced this outcome, although judges were not usually the decisive players.11 Remember that the ECtHR made the famous dictum that states may control migration ‘as a matter of well-established international law’ in a case concerning spouses from the former colonies.12 To date, human rights limit state discretion without, as we shall see, eliminating it. We may distinguish semantically between ‘family reunification’ and ‘family formation’. While the former concerns a pre-existing family being reunited across borders, the latter is for the purpose of establishing a new family. Classic examples are children joining the mother or father who had moved elsewhere (family reunification) and a Belgian national marrying her Canadian partner whom she had met during holidays abroad (family formation). The official title of the Family Reunification Directive is misleading insofar as the Directive covers both scenarios. That distinction also helps us to understand the policy debate. While family reunification is widely accepted, family formation became controversial in the new millennium. Public opinion and governments were critical of the second and third generation of former ‘guest workers’ and late colonial migrants who married partners living in the country of origin of their parents. Such ‘import of spouses’ was depicted 7 See ch 12.4.2. 8 Council Resolution on limitations on admission of third-country nationals for employment [1996] OJ C274/ 3, recital vi; see also Council Resolution relating to the limitations on the admission of third-country nationals for the purpose of pursuing activities as self-employed persons [1996] OJ C274/7; and Steve Peers, ‘Building Fortress Europe’ (1998) 35 CML Rev 1235, 1242–51. 9 See ch 4.2.3; and Andrew Geddes, Immigration and European Integration (Manchester UP 2000) 26–30. 10 See Proposal for a Convention on Rules for the Admission of Third-Country Nationals to the Member States, COM(97) 387 final, arts 7, 12, 15, 22; and Elspeth Guild, ‘Mechanisms of Exclusion’ in Joanna Apap (ed), Justice and Home Affairs Law in the EU (Edward Elgar Publishing 2004) 211, 218–20. 11 See ch 4.2.2; and Saskia Bonjour, ‘The Power and Morals of Policy Makers’ (2011) 45 IM Rev 89. 12 See chs 1.1.6 and 5.3.1.

430 Legal Migration as a lack of social integration and loyalty towards the host state. Instances of forced marriage, marriages of convenience, and child marriage dominated the public discourse, thus allowing political actors to frame policy restrictions as liberal achievements.13 These debates were still in the making when the EU institutions started dealing with family unity in the 1990s. An intergovernmental resolution, adopted prior to the Treaty of Maastricht, and a non-binding common accord on marriages of convenience anticipated the basic contours of today’s Family Reunification Directive.14 Somewhat more liberal was the initial Commission Proposal, which was amended later to take on board the non-binding views of the European Parliament.15 The Council disagreed with this generous approach, thus triggering a substantial revision.16 The Directive has not been changed a single time, purportedly not to give national governments an opportunity to insist on further restrictions.17 The background of the negotiations was domestic legislation which complicated family formation for the second and third generation. Pre-departure language tests are widely known, but other restrictions are equally important: income requirements, waiting periods, age limits, and higher fee levels. Unanimity meant that optional clauses, which one government had insisted upon, could be used by others later on. Like in the field of asylum, Europeanisation facilitated horizontal policy transfer. After having been acquainted with the domestic rules of others, governments promoted these ‘best practices’ domestically. In a sort of domino effect, the restrictions were introduced in countries with many ‘guest workers’ and late colonial migrants. The Directive did not command that outcome but served as a framework for informal policy transfer.18 These restrictions were mostly confined to the classic destination countries, whereas legislation in Southern and Eastern Europe was often more liberal, at least initially.19 Moreover, the Family Reunification Directive 2003/86/EC contains important guarantees, including an individual statutory right to be admitted. Interpretation held the potential of further benefits. We shall see that the Court accepted waiting times, age limits, and integration requirements in light of human rights but required Member States to introduce hardship clauses; judges also reined in excessive income requirements and fee levels. Like most other instruments, the Family Reunification Directive combines restrictive components with protective elements.

13 See Arjen Leerkes and Isik Kulu-Glasgow, ‘Playing Hard(er) to Get’ (2011) 13 EJML 95, 96–102; and Sara R Farris, In the Name of Women′s Rights (Duke UP 2017). 14 See Ad Hoc Group on Immigration, ‘Resolution on the harmonization of national policies on family reunification’ (1 June 1993); Council Resolution on measures to be adopted on the combating of marriages of convenience [1997] OJ C382/1; and Peers (n 8) 1239–42, 1258–61. 15 See Proposal for a Family Reunification Directive, COM(1999) 638 final; Amended (Second) Proposal for a Family Reunification Directive, COM(2000) 624 final; and, previously, Proposal for a Convention (n 10) arts 24–31. 16 See Revised (Third) Proposal for a Family Reunification Directive, COM(2002) 225 final; and Philippe De Bruycker, ‘L’émergence d’une politique européenne d’immigration’ in Philippe De Bruycker (ed), The Emergence of a European Immigration Policy (Bruylant 2003) 1, 35–40. 17 See ch 2.1.4. 18 See ch 13.1.2; Anne Staver, ‘Free Movement and the Fragmentation of Family Reunification Rights’ (2013) 15 EJML 69, 81–85; and Laura Block and Saskia Bonjour, ‘Fortress Europe or Europe of Rights?’ (2013) 15 EJML 203, 211–19. 19 See Commission, ‘Implementing Report’ COM(2008) 610 final; and Helena Wray, Agnes Agoston, and Jocelyn Hutton, ‘A Family Resemblance?’ (2014) 16 EJML 209, 221–43.

Theory and Policy Design 431

14.1.3 Globalisation and the ‘battle’ over labour migration The Treaty of Amsterdam and the Tampere Conclusions marked the beginning of a new era. EU institutions acquired supranational decision-making powers, and the Commission was eager to use them. In doing so, it promoted a liberal agenda that broke with the past: ‘the existing “zero” immigration policies which have dominated thinking over the past 30 years are no longer appropriate’.20 This about-face reflected the widespread enthusiasm about economic neoliberalism and globalisation at the millennium change, which culminated in the adoption of the ‘Lisbon agenda’ with the strategic goal, for the EU, ‘to become the most competitive and dynamic knowledge-based economy in the world’.21 Labour migration was presented, by the Commission, as an integral part of this venture. IT specialists and highly skilled workers were the centre of attention, in contrast to the widespread focus of previous debates on the asylum system and the lesser qualified. Some predicted that demographic change in aging societies would result in a ‘race for talent’ among OCED countries, and several Member States relaxed entry rules.22 Tellingly, the only two instruments on economic migration adopted during that period concerned the highly skilled: students and researchers. Adoption of the former Students Directive 2004/114/EC proved uncontroversial, since it essentially reiterated domestic laws.23 Admission of researchers under the former Researchers Directive 2005/ 71/EC suited the idea behind the ‘race for talent’ particularly well. Nevertheless, the non- participation of the United Kingdom curtailed the practical effects of both instruments. The UK was an attractive destination for international students, and its economy was booming in the early 2000s. London was an epicentre of optimism regarding the economic benefits of globalisation. Germany, by contrast, was depicted as the ‘sick man of Europe’ at the time, with mass unemployment and economic stagnation.24 That mattered for the European debate, since the government of Europe’s largest economy vehemently opposed the harmonisation of labour migration, under domestic pressure from powerful regional leaders.25 Germany took the lead among a group of countries that rejected the Commission Proposal for an Economic Migration Directive.26 It questioned supranational competences and rejected an extension of the informal open method of coordination, which was highly popular at the time, to legal migration.27 To be sure, the Commission proposals would have left room for

20 Commission, ‘Communication on a Community immigration policy’ COM(2000) 757 final, 6. 21 European Council, ‘Presidency Conclusions’ (24 March 2000) No 5. 22 See Ayelet Shachar, ‘The Race for Talent’ (2006) 81 NYU L Rev 148; and Philip Martin, Manolo Abella, and Christiane Kuptsch, Managing Labour Migration in the Twenty-First Century (Yale UP 2006) ch 3. 23 See Steve Peers, ‘Admission of Students and Others’ in Steve Peers and others (eds), EU Immigration and Asylum Law (Text and Commentary), vol 2 (2nd edn, Martinus Nijhoff 2012) 195. 24 See ‘Germany on the Mend’ Economist (17 November 2004) http://www.economist.com/node/3352024 (accessed 1 March 2023). 25 See Matthias M Mayer, ‘Governmental Preferences on Liberalising Economic Migration Policies at the EU Level’ (PhD thesis, LSE 2011) ch 6 http://etheses.lse.ac.uk/id/eprint/370 (accessed 1 March 2023). 26 See Proposal for a Council Directive on the conditions of entry and residence of third-country nationals for the purpose of paid employment and self-employed economic activities, COM(2001) 386 final; which was withdrawn later [2006] OJ C64/3; and Julia Mourão Permoser, ‘Redefining Membership’ (2017) 43 JEMS 2536, 2547–49. 27 See Commission, ‘Communication on an open method of coordination for the Community immigration policy’ COM(2001) 387 final 4–6; and De Bruycker (n 16) 48–59.

432 Legal Migration accommodating governmental concerns,28 but Member States decided to pick a fight—and won. Chapter 2.2.1.2 explained that this was not the only occasion when the Commission adapted to governmental preferences with technocratic pragmatism. Various reasons help rationalise failure. Most importantly, the Commission did not present convincing arguments for the desirability of Europeanisation. Spillover effects from the abolition of internal border controls and the single market programme were limited; labour markets were fragmented along national borders, with few third country nationals applying for jobs abroad.29 Furthermore, interior ministries defended their prerogatives and rejected the use of migration policy for broader economic objectives, let alone the portrayal of Union citizenship as a model.30 To highlight the diversity of national legislation is tautological, since it refrains from giving positive reasons for joint action.31 Demographic change was a distant prospect at the time, which, moreover, contrasted with enduring unemployment. Finally, the accession of Central and Eastern Europe states opened a sizable reservoir of potential workers. Uncertainties about the effects of enlargements reduced the appetite for additional entry channels.

14.1.4 Sectoral approach to economic migration Failure of the first economic migration initiative resulted in a lengthy consultation process, during which the Commission explored areas where harmonisation might be achieved.32 The result was what became to be known as the sectoral approach: adoption of five pieces of legislation on common procedures, the highly skilled, seasonal work, and intra-corporate transfers, in addition to the existing directives for students and researchers.33 That was a critical departure from the horizontal approach the Commission had put forward a few years earlier. Member States agreed to consider common rules for selected economic activities but were not willing, however, to generally cede regulatory control over economic migration. The sectoral approach delivered. EU institutions agreed on several directives, albeit after years of debates which were complicated by the absence of a widely shared policy rationale behind Europeanisation.34 Whereas agreement on the former Blue Card Directive 2009/50/EC and the ICT Directive 2014/66/EU reiterated the preoccupation of the policy debate with the entry of the highly skilled, the Single Permit Directive 2011/98/EU was a face-saving exercise for the Commission. The Directive reiterates the horizontal outlook of earlier proposals but

28 See also Working Party on Migration and Expulsion, ‘Proposal for a Council Directive’ (Council doc 13954/ 03, 25 November 2003). 29 Note that the five-year deadline for harmonisation did not embrace legal migration according to EC Treaty, art 63(3)(a), last sentence, as amended by the Treaty of Amsterdam. 30 See ch 2.1.1; and Daniel Wilsher, ‘Economic Migration into the European Union’ (2002) 21 Yearb Eur L 163; and Bernard Ryan, ‘The European Union and Labour Migration’ in Helen Toner and others (eds), Whose Freedom, Security and Justice? (Hart Publishing 2007) 489. 31 See Jean-Baptiste Farcy, ‘Labour Immigration Policy in the European Union’ (2020) 22 EJML 198, 203. 32 See Commission, ‘Green Paper on an EU approach to managing economic migration’ COM(2004) 811 final. 33 See Commission, ‘Communication: Policy plan on legal migration’ COM(2005) 669 final; training was another component of the initial proposal, which was later integrated into the Students and Researchers Directive (EU) 2016/801. 34 See Jean-Baptiste Farcy, L’Union européenne et l’immigration économique (Anthemis 2021) 151–257.

Theory and Policy Design 433 refrains from harmonising entry conditions; it applies, as we shall see, once states have authorised labour market access. Common rules on seasonal work seem to contradict the widespread scepticism, among national governments, about the desirability of low skilled labour migration. After all, the Seasonal Workers Directive 2014/36/EC could be presented as a remake of the guest worker programmes. However, Chapter 18.4.4 on the external dimension will explain that several Member States had agreed on bilateral entry channels with neighbours in Northern African and Eastern European for temporary admission. These schemes received widespread attention under the heading of ‘circular migration’ when the Commission initiated the sectoral approach. They satisfied labour market needs and were often used as an incentive to cooperate in the prevention of irregular migration.

14.1.5 Limits of Europeanisation A holistic assessment of the labour migration instruments is bound to notice the lack of innovative appeal. Supranational legislation often confirmed existing national practices, instead of promoting further liberalisation. A ‘fitness check’ found in 2019 that ‘the current legal migration framework had a limited impact’.35 That is not to say, crucially, that Europeanisation was irrelevant: state discretion was replaced by individual rights to enter under the conditions set out in supranational legislation; and several Court judgments on the equal treatment provisions extended the rights of third country nationals.36 At the same time, the sectoral approach introduced a highly fragmented legislative patchwork. Chapter 6.2.2 explained that the idea of an ‘immigration code’ seems to have been abandoned. The revision of the Blue Card Directive and the Students and Researchers Directive increased complexity, instead of advancing coherence. Not even the flagship project changed much: most countries issued between a few dozen and several hundred blue cards per year throughout the 2010s.37 Germany was the exception to the rule, after the country had used the transposition of the Directive—somewhat ironically—as an occasion for liberalisation, by not activating the optional clauses on which it had insisted during the negotiations. Throughout the 2010s, Europe’s biggest economy was booming and gradually introduced one of the most liberal immigration regimes for the highly skilled.38 Other countries, such as Austria or the Netherlands, similarly launched new admission schemes. The Commission followed these developments with scepticism and proposed to outlaw parallel national entry channels for the highly skilled.39 Member States rejected monopolisation, arguably for convincing reasons. Labour migration exemplifies that ‘more Europe’ will not always add value. Experience with lengthy inter-institutional negotiations and practical difficulties in amending directives indicate that the EU level is badly placed to experiment with innovative

35 Commission, ‘Fitness Check on the EU Legislation on Legal Migration (Part 1)’ SWD(2019) 1055 final, 105. 36 See chs 10.1.2 and 15.3. 37 See the Eurostat dataset (n 1) for ‘MIGR_RESBC1’. 38 See Roman Lehner, ‘Gesetzgebung. Vor-und Nachteile der Reformschritte in der letzten Dekade’ [2022] Zeitschrift für Ausländerrecht 144. 39 Contrast Blue Card Directive (EU) 2021/1883, art 3(3) with Proposal for a Revised Blue Card Directive, COM(2016) 378 final, art 3(4).

434 Legal Migration solutions.40 Entry channels for start-ups and entrepreneurs in several countries show the room for regulatory innovation at the national level.41 The same applies to legal pathways for the lesser qualified, discussed at the end of this chapter, which are often presented as an alternative to irregular migration and may require frequent adaption to lessons learned. Tardy decision-making was one factor, as we shall see, why EU institutions opted against a points-based admission scheme for the highly skilled. Instead of embarking on deep harmonisation, the Commission may wish to focus on sponsoring and facilitating innovation at the national level. A fine example of how that could be achieved are new initiatives to encourage national programmes to attract care-workers and to admit start-ups by means of forward-looking policy coordination.42

14.1.6 Effects on the welfare state Chapter 4.1 explained that diverse drivers at the micro-, meso-, and macro-level influence the volume and composition of migratory movements. A different but related question concerns the impact of migration on the welfare state. Like in the case of drivers, the overall picture is complex and multifaceted. Answers depend on the characteristics of migratory movements and the situation in the receiving country; both can change over time.43 Moldovan women caring for elderly relatives in private households, Egyptian physicians in public hospitals, Moroccan fruit pickers, and refugees delivering parcels all perform valuable work. Nevertheless, the economic and social effects of these activities can differ markedly. Admission of qualified workers is generally perceived to buttress economic growth, innovation, and public finances. Moreover, skilled migrants tend to pay more taxes and social security contributions than they consume in the form of public services and social benefits.44 Demographic change renders their economic and fiscal contribution more significant, in line with the basic assumption behind the ‘race for talent’ hypothesis. Of course, highly skilled migration can have side effects as well. Think of the gentrification of city centres and strains on public services in the absence of state investment compensating for the increase in demand.45 On the whole, however, positive economic and fiscal effects outweigh possible downsides. The impact of lower skilled labour migration is discussed controversially. Diverse studies have shown that the overall socioeconomic impact tends to be positive, albeit marginally so. Crucially, costs and benefits tend to be distributed unevenly: employers and higher income groups reap the advantages, unlike domestic workers with lesser skills. An increase in the offer of unskilled labour can hold back wage increases and an improvement of working conditions; many fear competition for public services and an indirect downward spiral of social

40 See Farcy (n 34) 312–38. 41 See EMN, Migratory Pathways for Start-Ups and Innovative Entrepreneurs in the EU and Norway (Focussed Study, EMN 2019). 42 See Commission, ‘Communication: Attracting skills and talent to the EU’ COM(2022) 657 final, 17–21. 43 See Hein de Haas, Stephen Castles, and Mark J Miller, The Age of Migration (6th edn, Palgrave 2020) ch 12. 44 See the classic position of Julian L Simon, The Economic Consequences of Immigration (2nd edn, University of Michigan Press 1999) chs 4–8; and IMF, World Economic Outlook (April 2020) ch 4. 45 See also ch 4.2.3.

Constitutional Foundations 435 protection.46 At the economic macro-level, the availability of inexpensive manpower can slow mechanisation and productivity growth, and some claim, controversially so, that diversity can undermine social cohesion and solidarity.47 The example of posted workers will highlight that policy debates are often defined by a confrontation of countervailing interests, both domestically and transnationally. Supranational legislation prescribes the equal treatment of migrant workers with nationals. Excluding migrant workers from social benefits (greater numbers, fewer rights) has not generally been an option on the European continent. Doing so prevented immediate downward pressure on wage levels and working conditions (notwithstanding the potential of indirect effects). Even so, legal requirements are not always complied with, with seasonal work being a classic example of widespread informality and exploitation. Contextual factors may complicate the assessment further: whereas some countries restrict benefits to specific categories of workers, others have universal support schemes. As a result, the legislative equal treatment provisions, which will be discussed in Chapter 15.3, can have diverse consequences in practice. Most of these contextual factors can be difficult to understand. That may be one reason why the public debate often concentrates on a rather crude regulatory tool: labour market tests. Experience of the author indicates that their effects depend very much on how they are handled in practice. Even if they do not ultimately prevent employment, they may have chilling effects in terms of prolonging the procedure and deterring employers. In the EU context, labour market tests have a history that is not always remembered. Italy insisted, with limited success, on ‘Community preference’ early on to prevent third country nationals from taking jobs Italians might exercise.48 The topic regained prominence when enlargement was high on the political agenda.49 Today’s legislation entrusts Member States with deciding whether they want to use the tool.

14.2 Constitutional Foundations Earlier debates about the scope of Union competences lost relevance when the Treaty of Lisbon introduced Article 79 of the Treaty on the Functioning of the European Union (TFEU). EU institutions have far-reaching powers whose activation depends on a political decision about the desirability of joint action (14.2.1). National prerogatives to ‘determine volumes of admission’ for labour migration, enshrined in Article 79(5) TFEU, do not fundamentally reverse the overall picture of a wide range of competences (14.2.2). Human rights are of limited relevance when it comes to entry for economic purposes, whereas they may, under certain conditions, oblige states to authorise family reunification (14.2.3). Finally, readers are reminded of the Irish and Danish opt-outs (14.2.5). 46 See Christiane Kuptsch and Philip Martin, ‘Low-Skilled Labour Migration’ in Alexander Betts (ed), Global Migration Governance (OUP 2011) 34, 48–54; and Christian Joppke, Neoliberal Nationalism (CUP 2021). 47 See Gary P Freeman and Nikola Mirilovic (eds), Handbook on Migration and Social Policy (Edward Elgar Publishing 2016). 48 See ch 1.1.5–6; and Kenneth A Dahlberg, ‘The EEC Commission and the Politics of the Free Movement of Labour’ (1968) 6 JCMS 310, 324–29. 49 See Resolution on limitations (n 8) recital (iii); Proposal for paid employment (n 26) art 6; and, by way of example, Act concerning the conditions of accession of the Republic of Croatia [2012] OJ L112/21, Annex V, No 5(12).

436 Legal Migration

14.2.1 Supranational competences EU Treaties allow for the seamless regulation of migration statuses. While short-term stays are covered by the provision on visas, Article 79(2)(a) TFEU concerns entry and stay for longer periods. Family reunification is mentioned by way of illustration (‘including’); the legislature may adopt rules for other purposes as well. Supranational competences embrace both entry from abroad and permits for those residing on the territory already, as indicated by the reference to both ‘long-term visas’, issued by consulates, and ‘residence permits’, delivered by domestic authorities.50 Few areas are left to Member States as a reserved power. Nationality law is among them, as described in Chapter 15.7 on integration. Any activation of the Union competences must comply with the principles of subsidiarity and proportionality. Chapter 6.3.3 on the doctrinal foundations explained that doing so can be an uphill struggle in the field of legal migration, even though the Court affords the institutions much discretion. Use of the open formula ‘measures’, in the introductory sentence of Article 79(2) TFEU, indicates that legislation can be supplemented by other activities. EU institutions may, for instance, introduce legal pathways for specific categories of economic migrants. By way of example, the provision could be used for a seasonal agricultural migration scheme with Morocco. Legally, the EU Treaties would cover such initiatives, even though country- specific quotas may fall foul of Article 79(5) TFEU. Having said this, any activation of shared competences remains a political choice, in compliance with principle of subsidiarity. Whenever EU institutions do not act, they can join forces with Member States in the form of mixed agreements. Chapter 2.3.5 explained that informal soft law arrangements require a foundation in the Treaties, which Article 79 TFEU may provide. Legal bases for the regularisation of illegal stay will feature in Chapter 16.8 on return. Article 79(2)(b) TFEU covers common standard for the treatment of migrants during periods of legal residence. The unspecific wording signals that the legislature has broad discretion when deciding which rights to harmonise. The general scheme of the Treaties indicates that voting in elections is, like nationality law, not covered. Specific problems may arise if supranational legislation, or association agreements, align the status of third country nationals with Union citizenship, through the extension of single market instruments on social security coordination or the recognition of professional qualifications. Judges held that Treaty articles on the single market apply in these scenarios, provided they replicate the single market model.51 Presumably, the extension of the Social Security Coordination Regulation should have been based on Article 48 TFEU, which is not subject to county- specific opt-outs.52 When the rules in question differ from single market model, Article 79(2)(b) TFEU may be used.53 The same applies for corollary rules on social security and

50 See also ch 10.1.1. 51 See Case C-431/11 United Kingdom v Council EU:C:2013:589, paras 49–64; Case C-656/11 United Kingdom v Council EU:C:2014:97; and Paula García Andrade, ‘EU External Competences in the Field of Migration’ (2018) 55 CML Rev 157, 185–91. 52 Regulation (EU) No 1231/2010 extending Regulation (EC) No 883/2004 to nationals of third countries [2010] OJ L344/1 was adopted prior to the above-mentioned judgments. 53 See Case C-81/13 United Kingdom v Council EU:C:2014:2449, paras 40–46; and Daniel Thym, ‘Legal Framework for Immigration Policy’ in Daniel Thym and Kay Hailbronner (eds), EU Immigration and Asylum Law: Article-by-Article Commentary (3rd edn, CH Beck/Hart Publishing/Nomos 2022) MN 17a.

Constitutional Foundations 437 the recognition of diplomas in the directives on legal migration, whenever the centre of gravity of the instruments concerns migration. References to ‘third country nationals’ mean that Article 79(2) TFEU can be used for family members of Union citizens who do not have Member State passports, including the families of ‘immobile’ Union citizens living in the country of nationality. The final version of the Family Reunification Directive opted against this possibility, whereas other directives on legal migration cover family members from third states.54 Note, however, that the application of migration law to family members of Union citizens is subject to an important constitutional caveat. Settled case law gives, as we shall see, enhanced guarantees to family member of Union citizens who have exercised free movements rights. They will usually invoke the migration law instruments only if they offer a higher level of protection than the derived rights stemming from Union citizenship.

14.2.2 National prerogatives for labour migration Article 79(5) TFEU confirms, argumentum e contrario, that the Union may harmonise the admission of third country nationals for economic purposes. The introduction of that provision was based on the assumption that the EU had acquired a legislative competence for labour migration.55 Prior to the Treaty of Lisbon, competences for labour migration had been discussed controversially. While some had rejected a supranational competence,56 the Legal Service of the Council had proposed that social policy should be used for certain aspects.57 The idea of using social policy can be traced back to legal battles over the scope of Community competences in the 1980s, when the Court had held that they may cover labour migration.58 For our purposes, however, the argument has lost relevance. Article 79(2) (a) TFEU covers the admission of third country nationals for economic purposes as a lex specialis supplanting social policy. Article 79(5) TFEU covers migrants ‘coming from third countries’ and does not encompass, therefore, labour market access of those admitted for other purposes, such as family unity. Notwithstanding ambiguous wording, the provision should be interpreted to cover any admission for work, including with a binding job offer. Some argue that the caveat covers quota schemes only.59 Indeed, the provision was introduced to appease German concerns, in response to heated domestic disputes about labour migration and the idea of a quota.60 Nevertheless, a comparison of language versions shows that the exemption concerns ‘volumes of admission’ (French fixer les volumes d’entrée; German wie viele . . . einreisen dürfen),

54 By way of example see Blue Card Directive (EU) 2021/1883, art 3. 55 See Convention Presidium, ‘Draft Constitution, Vol II’ (CONV 847/03, 8 July 2003); and Convention Secretariat, ‘Reactions to draft text CONV 802/03’ (CONV 821/03, 27 June 2003) 83. 56 See Kay Hailbronner, ‘European Immigration and Asylum Law under the Amsterdam Treaty’ (1998) 35 CML Rev 1047, 1049–52. 57 See ‘Provisions on access to employment: Legal basis’ (Council doc 9077/02, 22 May 2002); and ‘Contribution of the Legal Service’ (Council doc 14150/02, 12 November 2002). 58 See Joined Cases 281, 283, 284, 285 and 287/85 Germany and others v Commission EU:C:1987:351, paras 9–31. 59 See Steve Peers, EU Justice and Home Affairs Law, vol I (4th edn, OUP 2016) 328–31. 60 See Arne Niemann, Explaining Decisions in the European Union (CUP 2006) 252, 257–59; the background concerned negotiations about a domestic Immigration Act (Zuwanderungsgesetz), whose revised second version abandoned the quota project.

438 Legal Migration without indicating how they are determined at the national level. Quotas are an option but not necessarily the only one. For the contemporary debate, Article 79(5) TFEU indicates that EU institutions may not promise a certain number of legal pathways in negotiations with third states. Saying that 10,000 trainees should be admitted from Nigeria each year, in return for a readmission agreement, would be incompatible with Article 79(5) TFEU. Member States have to sign up to such a pledge separately. Moreover, the provision could be interpreted to require a certain flexibility for national legislatures, for instance via labour market tests, so as to not render it meaningless for countries without a quota.61 In any case, the caveat serves as a symbolic reminder that EU institutions should proceed carefully when regulating labour migration.

14.2.3 Human rights, family life, and rights of the child Admission for economic purposes is not subject to far-reaching human rights constraints. Chapter 1.3.2 explained that Articles 15(2) and 45(2) Charter of Fundamental Rights (CFR) leave much leeway, since both articles distinguish third country nationals from Union citizenship. Chapter 10.4 illustrated that sector-specific admission criteria, for instance with regard to research and seasonal work, will not usually amount to illegitimate unequal treatment if they favour the admission of nationals from certain countries de jure or de facto. Procedural safeguards were discussed in Chapter 7.2 on the administrative dimension. Human rights gain relevance in the field of legal migration, once someone has been admitted to the territory. They will be discussed in the following chapter on integration, including protection against expulsion under Article 8 ECHR. Enhanced guarantees exists for family unity, although the ECtHR emphasises that Article 8 ECHR ‘cannot be considered as extending to a general obligation . . . to respect the choice by married couples of the country of their matrimonial residence’.62 In its first judgment on migration law, the CJEU confirmed this position when rejecting an action of annulment the Parliament had brought against some provisions of the Family Reunification Directive.63 To date, only a handful of applications for family reunification have succeeded before the ECtHR.64 Many more claims are settled domestically, in accordance with supranational legislation or national laws. International conventions do not generally establish higher standards, nor does Article 19 Revised European Social Charter of 1996. Moreover, the ECtHR confines Article 8 ECHR to the nuclear family of spouses and minor children in regular circumstances.65 Judges have developed a set of criteria guiding family reunification cases; their relative weight will depend on the individual circumstances of the persons involved.66 In applying these standards, states benefit from a ‘margin of appreciation’, whose scope depends on the 61 See Thym (n 53) MN 27. 62 This has been settled case law ever since Abdulaziz and others v United Kingdom App nos 9214/80, 9473/81 and 9474/81 (ECtHR, 28 May 1985) § 68. 63 See Case C-540/03 Parliament v Council EU:C:2006:429, paras 53–59. 64 See the summary in MA v Denmark App no 6697/18 (ECtHR [GC], 9 July 2021) §§ 134–35. 65 See Slivenko and others v Latvia App no 48321/99 (ECtHR [GC], 9 October 2003) § 94. 66 See Jeunesse v Netherlands App no 12738/10 (ECtHR [GC], 3 October 2014) §§ 100–105; IAA and others v United Kingdom App no 25960/13 (ECtHR, 8 March 2016) §§ 38–41; and Pieter Boeles and others, European Migration Law (2nd edn, Intersentia 2014) 223–29.

Constitutional Foundations 439 quality of domestic decision-making and whether an international consensus exists on how to deal with specific scenarios of family reunification.67 More specifically, the Court emphasises connections with the countries where family members are currently residing and whether they had made a conscious choice to discontinue family unity; public interests to be considered include financial independence and criminal behaviour.68 In situations of ‘insurmountable obstacles’ or ‘major impediments’ to unite the family in a third state, the ECtHR is more inclined to support admission, especially to the benefit of refugees.69 Nevertheless, it accepted a two-year waiting period for beneficiaries of subsidiary protection.70 On the whole, it is difficult to discern clear patterns from the case law, besides the trend to find a human rights violation in exceptional scenarios. Family life intersects with the best interests of the child, which the ECtHR integrates into the application of Article 8 ECHR.71 Chapter 5.4.2 explained that the best interests of the child can be balanced with countervailing interests and that the child’s best interests ‘cannot be a “trump card” which requires the admission of all children’.72 Any assessment of the individual case must consider the age of the children, the situation in their country of origin, and the degree of dependency on parents.73 The ‘trump card’ metaphor has tangible effects when judges conclude that families may be separated indefinitely, also when young children are involved. The Court of Justice follows this balancing requirement as a matter of principle.74 Furthermore, judges in Luxembourg have mentioned the rights of the child in Article 24 CFR on several occasions when they interpreted the Family Reunification Directive dynamically. However, these verdicts have not, as we shall see, reversed basic legislative choices in light of human rights, so far at least. Future research may want to pay more attention to direct contacts and personal relationships with both parents, in accordance with Article 24(3) CFR.75 It may harbour the potential of tilting the balance of the balancing exercise slightly further in the direction of migrants. Judgments on the Brussels II Regulation (EU) 2019/1111 on jurisdiction on parental responsibility and on child abduction may serve as a point of reference.

14.2.4 Territorial scope Ireland and Denmark are not obliged to participate in new legislation. Despite the original intent of using the opt-out as little as possible, Ireland abstained from all legal migration instruments, with the exception of the former Researchers Directive 2005/71/EC. The Directive continues to apply in the Republic, which chose not to join the successor instrument. Chapter 2.4 explained that Denmark cannot participate on a case-by-case basis. 67 Jeunesse v Netherlands (n 66) § 106; and MA v Denmark (n 64) §§ 140–63. 68 See the summary in El Ghatet v Switzerland App no 56971/10 (ECtHR, 8 November 2016) §§ 43–47. 69 See Tanda-Muzinga v France App no 2260/10 (ECtHR, 10 July 2014) §§ 69, 74–76. 70 See MA v Denmark (n 64) § 162. 71 See Nunez v Norway App no 55597/09 (ECtHR, 28 June 2011) § 84. 72 IAA and Others v the United Kingdom (n 66) § 46. 73 ibid §§ 41–48; and Ciara Smyth, ‘The Best Interests of the Child in the Expulsion and First-Entry Jurisprudence of the European Court of Human Rights’ (2015) 17 EJML 70. 74 See Parliament v Council (n 63) para 59; and Joined Cases C-356/11 and C-357/11 O and S EU:C:2012:776, para 79. 75 See also Case C-635/17 E EU:C:2019:192, paras 55–57.

440 Legal Migration Table 14.1 Participation in the legislation on legal migration Instrument

Ireland /UK (before Brexit)

Denmark

Family Reunification Directive 2003/86/EC

no

no

Single Permit Directive 2011/98/EU

no

no

Seasonal Workers Directive 2014/36/EU

no

no

Intra-Corporate Transfer Directive 2014/66/EU

no

no

Students and Researchers Directive (EU) 2016/801

no (yesa) /no

no

aIreland continues to be bound by the former Researchers Directive 2005/71/EC, as stated in recital 27, which was

not repealed for the Republic under Students and Researchers Directive (EU) 2016/801, art 41(1).

None of the measures belongs to the Schengen acquis. As a result, new Member States are bound immediately after accession, and no third states are associated with the legal migration rulebook. Table 14.1 summarises the degree of participation.

14.3 Family Reunification Almost one million third country nationals are issued an entry visa or residence permit for family reasons each year, mainly on the basis of the Family Reunification Directive 2003/ 86/EC. Additional guarantees exist for spouses of Union citizens who have exercised the right to free movement (14.3.1). Provisions on the scope ratione personae of the Family Reunification Directive give Member States leeway not to apply the Directive to certain categories of family members (14.3.2), even if the conditions for entry and stay are met (14.3.3). Procedural safeguards have featured in a considerable number of Court judgments (14.3.4), as have the rules governing family reunification with refugees, especially unaccompanied minors (14.3.5).

14.3.1 EU citizens and their family members EU Treaties do not mention family members of Union citizens. Chapter 1.1.5 explained that the legislature agreed on generous entry and residence rights nonetheless, which were reinforced in numerous judgments. Unfortunately, the overall picture is complex: different strands of judicial innovations complement the provisions of the Free Movement Directive 2004/38/EC. On the whole, we may distinguish five scenarios. Their analysis requires an advanced knowledge of free movement law and Union citizenship. They are presented here as a supplement and counterpoint to the legislation building the area of freedom, security, and justice. Our comments will concentrate on family members with the nationality of a third

Family Reunification 441 state, not the legal status of spouses and children from another Member State. Nationals of neighbouring countries may benefit from the association agreements which will be described in Chapter 17. First, family members are authorised to ‘accompany or join’76 a Union citizen living abroad, for instance when an Italian is married to a Tunisian with whom she lives in France. Family unity is a corollary of free movement and not subject to any further conditions, although family members from third states may be required to apply for a visa free of charge.77 There is no need to be economically self-sufficient or to learn the language, provided the sponsor stays legally. EU legislation goes beyond the nuclear family: children are covered up to the age of twenty-one (the former age of majority) and may even be married; parents and grandparents of a Union citizen and of his or her spouse, irrespective of the nationality of the spouse, are admitted in situations of dependency; admission of other relatives must be ‘facilitated’.78 Judgments reinforced the level of protection on numerous occasions.79 In so doing, they reiterated the basic idea behind free movement law that the promise of family unity will support the integration of Union citizens into host societies.80 Secondly, children with the nationality of a third state are allowed to continue school education, vocational training, or university classes after the departure of the Union citizen.81 Crucially, these autonomous residence rights of children may—irrespective of the nationality of the child—serve as the basis for derived residence rights of parents who are third country nationals. By way of example, a Russian mother may reside legally with her school child in Latvia, even though the Estonian father had returned home when the marriage fell apart. Sufficient resources are not required, as long as the child proceeds with the education.82 Thirdly, the Free Movement Directive covers only residence in countries whose nationality the Union citizen does not possess. This raises the question of whether home states can refuse residence rights for family members from third states upon return. By way of example, Dutch authorities may wish to refuse a residence permit to the Indian spouse of a Dutch national who had lived in Dublin with his family previously. Judges address such return scenarios from the angle of the fundamental freedoms. In a remarkable about-turn, the Court overturned an earlier judgment which had found the return rationale not to embrace family members who had previously resided irregularly.83 This opened the door for the ‘Europe route’ to regularise illegal stay. A Danish national who is married to an unsuccessful asylum applicant and does not have sufficient resources, as a part time worker, to sponsor family reunification under domestic Danish legislation may relocate to the Swedish city of Malmö, where her partner will reside legally.84 After a while, they can move back across 76 See Free Movement Directive 2004/38/EC, art 3(1). 77 ibid art 5(2). 78 ibid arts 2(2), 3(2). 79 See generally Elspeth Guild, Steve Peers, and Jonathan Tomkins, The EU Citizenship Directive. A Commentary (2nd edn, OUP 2019); and Anastasia Iliopolou-Penot (ed), Directive 2004/338 relative au droit de séjour des citoyens de l’Union européenne et des membres de leur famille (Bruylant 2020). 80 See Regulation (EU) No 492/2011 on freedom of movement for workers within the Union [2011] OJ L141/1, recital 6. 81 ibid art 10; and Free Movement Directive 2004/38/EC, arts 12–14. 82 See Case C-413/99 Baumbast and R EU:C:2002:493, paras 47–75; and Case C-310/08, Ibrahim EU:C:2010:80. 83 See Case C-109/01 Akrich EU:C:2003:491, paras 47–54; Case C-127/08 Metock EU:C:2008:449, paras 48–80; and Chiara Berneri, Family Reunification in the EU (Hart Publishing 2017)ch 3. 84 Part-time work is covered by the fundamental freedoms.

442 Legal Migration the Øresund to Copenhagen. Using free movement for these purposes does not amount to abuse.85 Judges reined in the outer limits by concluding that stays of less than three months will not usually give rise to protection upon return.86 Fourthly, free movement in the single market is a transnational activity by definition. Settled case law requires a cross-border element, as ‘purely internal situations’ are beyond the reach of the fundamental freedoms.87 Special scenarios are Union citizens with the passports of two Member States, or after naturalisation in another Member State which requires abandoning the former nationality; they may invoke the fundamental freedoms for purposes of family unity.88 By contrast, immobile Union citizens residing in home states are not covered. This can result in a phenomenon dubbed ‘reverse discrimination’ if Member States restrict family reunification for nationals, whereas mobile Union citizens benefit from enhanced protection qua Union law. Notwithstanding fierce criticism, the Court has kept the cross-border element intact.89 Note that the purely internal rule concerns the fundamental freedoms: EU migration law usually covers domestic scenarios as well. Finally, one of the most tantalising judicial developments in recent years has been the Ruiz Zambrano judgment and follow-up rulings, which concerned residence rights of family members of immobile Union citizens residing in the home state.90 Judges seemed to enhance the level of protection considerably at first, although the Court explained later that its findings mainly concerned parents of minor Union citizens. The substance of rights associated with Union citizenship was found to be at risk when minors may be obliged to leave Union territory if parents are under an obligation to return home.91 In practical terms, the case law is relevant mainly for the parents of children who obtain nationality by birth, as well as for binational couples with minor children who do not meet the requirements for regularising the stay of one parent. In all other scenarios, domestic courts are asked to apply human rights.92 Unlike the rulings mentioned previously, protection is confined to residence in the country of nationality.93 Single market rules prevail over secondary legislation in cases of conflict, since the fundamental freedoms belong to primary law. In practical terms, the room for overlap with EU migration law is reduced, since the Family Reunification Directive does not cover family members of Union citizens. Extending its scope to family members of immobile Union citizens was rejected by the Council.94 Legislation discussed elsewhere in this volume foresees that it is without prejudice to the rights associated with free movement, for instance with regard to entry conditions and visa requirements.95 Chapter 10.4.1 explained that the 85 See ch 10.8; and Case C-200/02 Zhu and Chen EU:C:2004:639, paras 34–41. 86 See Case C-456/12 O and B EU:C:2014:135, paras 44–54. 87 See Joined Cases C-64/96 and C-65/96 Uecker and Jacquet EU:C:1997:285, paras 23–24. 88 See Case C-148/02 Garcia Avello EU:C:2003:539, paras 24–28; and Case C-165/16 Lounes EU:C:2017:862, paras 45–62. 89 cf Alina Tryfonidou, Reverse Discrimination in EC Law (Kluwer 2009) ch 4; and Niamh Nic Shuibhne, ‘Free Movement of Persons and the Wholly Internal Rule’ (2002) 39 CML Rev 731, 741–60. 90 See Case C-34/09 Ruiz Zambrano EU:C:2011:124, paras 37–45; and Kay Hailbronner and Daniel Thym, ‘Case C-34/09, Gerardo Ruiz Zambrano v Office national de l’emploi’ (2011) 48 CML Rev 1253. 91 See Case C-256/11 Dereci and others EU:C:2011:734, paras 59– 69; Case C- 82/ 16, KA and others EU:C:2018:308, paras 66ff; and Berneri (n 83) ch 4. 92 Dereci and others (n 91) paras 70–74. 93 See Case C-86/12 Alopka and Moudoulou EU:C:2013:645, para 34. 94 See the initial FRD Proposal (n 15), art 3(1)(c). 95 See Schengen Borders Code Regulation (EU) 2016/399, art 3(a); and Visa Code Regulation (EC) No 810/ 2009, art 2(2)(a).

Family Reunification 443 Court has refused to qualify the better treatment of Union citizens as illegitimate unequal treatment.

14.3.2 Scope of the Family Reunification Directive Directive 2003/86/EC covers family reunification when both the sponsor and the family member are third country nationals. Dual nationals are excluded, in light of Article 2(a) defining ‘term third country national’, for the purposes of the Directive, as anyone who is not a Union citizen. An important bottleneck is Article 3(1): the sponsor must have a residence permit for a period of validity of at least one year and show ‘reasonable prospects of obtaining the right of permanent residence’. This vague wording excludes temporary stays, for instance by seasonal workers; in cases of doubt, domestic authorities must perform a prognosis.96 Such prognosis will consider the factual and legal context. By way of example, students arguably have a prospect of settlement nowadays, provided they can demonstrate that they might find a job after completion of their studies in accordance with Article 25 Students and Researchers Directive (EU) 2016/801.

14.3.2.1 Limitation to the nuclear family Despite the official title, the Directive covers both ‘family reunification’ and ‘family formation’ in the ordinary meaning of the terms.97 We may have recourse to the conflict of laws, which is also called private international law, to determine whether a marriage has been entered into. By way of exception, Article 4(4) excludes more than one spouse of a polygamous marriage. It is a different matter whether human rights would authorise states to deport a second spouse residing on their territory. Same-sex couples were discussed under the heading of ‘unmarried partners’ originally, but they may count as ‘spouses’ nowadays, provided they entered into a regular marriage.98 Entry for purposes of marriage is not covered, although Member States may grant a national visa for that purpose. Children must be minor (commonly eighteen years) and be unmarried. Adoption requires formal registration as a child and does not cover, therefore, foster children or the Islamic kafala system.99 Article 4(1)(a)–(c) require the sponsor or the spouse to have sole custody of the child, thus excluding situations of contact, usually after divorce. The provision may possibly be challenged in light of Article 24(3) CFR, in line with previous comments. Member States may decide to apply the Directive to situations of shared custody, as well as to parents and unmarried adult children under certain conditions. An activation means that the remainder of the Directive applies; domestic legislation cannot foresee stricter conditions. Other relatives, such as siblings, can be admitted on the basis of more favourable domestic laws in accordance with Article 3(5). Chapter 10.6.2 on general features explained that such favourable entry schemes are not usually covered by Union law. Critical scholars may challenge the focus on the nuclear family in the design of the Family Reunification

96

See Steve Peers, ‘Family Reunion’ in Peers and others (n 23) 250–53. See Family Reunification Directive 2003/86/EC, art 2(d); and O and S (n 74) para 62. 98 By analogy see Case C-673/16 Coman and others EU:C:2018:385. 99 See Case C-129/18 SM EU:C:2019:248, paras 53–56. 97

444 Legal Migration Directive, which reiterates the predominant social norms in contemporary Europe, but contrasts with the persistence of wider family bonds elsewhere.100 This critique cannot be channelled into human rights law easily, considering that the ECHR confines Article 8 ECHR to the nuclear family in regular circumstances.

14.3.2.2 Optional clauses Member States insisted on the introduction of several optional clauses. To start with, Article 8(1) acknowledges the possibility of a waiting period of two years of prior residence. Note that the additional three-year period in the following subparagraph, introduced at the request of Austria, was designed as a standstill provision: it covers only national legislation which foresaw a waiting period on the date when the Directive was adopted. Waiting periods were among the optional clauses whose validity the Parliament challenged. In a seminal ruling, the Court found Article 8 to comply with fundamental rights, provided that domestic authorities retain some leeway to authorise family reunification in exceptional scenarios.101 This outcome goes conform, as we have seen, with the position of the ECtHR about the absence of a generic human right to family unity and can be extended to other optional clauses throughout the directive. Several academics assert that the Noorzia judgment defended a different outcome with regard to the optional minimum age of twenty-one years for spouses.102 The introduction of Article 4(5) was motivated by resentment towards child marriage and anxieties about young age increasing the risk of forced marriage. Judges reproduced this view without indicating that human rights might necessitate an individualised assessment.103 Nevertheless, the very short judgment need not necessarily be interpreted to contain a definite statement about human rights. The reasoning was essentially limited to the specification of relevant moment for the minimum age: the lodging of the application or the decision by the domestic authority or court. Judges confirmed that Austria may use the former. Silence on human rights can be reconstructed, in accordance with Advocate General Mengozzi, as having deferred that matter to later rulings.104 That is not to say that the minimum age violates Article 7 CFR, especially if states retain some leeway to authorise entry nonetheless. Along these lines, the Court recognised explicitly that the optional clauses in Article 4(1) (3) and 4(6) are subject to a human rights caveat. States may exceptionally have to authorise the entry or stay of minor children who do not comply with integration conditions for those aged over twelve; this conclusion extends to the seemingly rigid exclusion clause for those aged above 15.105 Note that both provisions are optional and must be activated by states on the occasion of implementation at the latest, thus limiting their future significance. When assessing the human rights caveat, they must consider the best interests of the child in the balancing exercise.

100 See Maria de Lourdes Peroni, ‘Challenging Culturally Dominant Conceptions in Human Rights Law’ (2010) 4 Hum Rts Int Leg Disc 241, 260–64. 101 See Parliament v Council (n 63) paras 97–107. 102 See Peers (n 59) 406; and Georgios Milios, ‘Family Reunification for Third-Country Nationals’ (2015) 17 EJML 127, 131–33. 103 See Case C-338/13 Noorzia EU:C:2014:2092, para 18. 104 See AG Paolo Mengozzi, Opinion in Case C-338/13 Noorzia EU:C:2014:288, points 64–67. 105 See Parliament v Council (n 63) paras 61–71, 84–90.

Family Reunification 445

14.3.3 Conditions under the Directive Of great practical relevance are the conditions in Articles 6 and 7 Family Reunification Directive 2003/86/EC. Where these conditions are not satisfied, an application ‘may’ be rejected; in contrast to the Long-Term Residents Directive, compliance is not mandatory. Where it turns out that the conditions are no longer fulfilled, domestic authorities may withdraw or not renew a residence permit.106 The conditions must be complied with at the point of application and during periods of legal residence, until the acquisition of long- term residence. The bearing of the different conditions was discussed in Chapter 10 on general features, which bound together insights from the case law on various migration law instruments. We saw that the ‘public policy’ and ‘public security’ exception must be distinguished from the judgments on Union citizenship, with the objective of family unity carrying more weight in the overall assessment than private interests in other scenarios. Along similar lines, the Court banned prohibitively high financial resources requirements, while leaving Member States some leeway as to the precise definition of this important condition. Judgments on fee levels may similarly be relied upon to challenge excessive national practices. The need for ‘accommodation regarded as normal’, in Article 7(1)(a) Family Reunification Directive, raises questions of interpretation and has to be assessed in light of local circumstances. This condition is symbolically important, since parallel negotiations on the Free Movement Directive 2004/38/EC chose not to expect Union citizens to have appropriate accommodation. On the whole, the Court’s approach displays a desire to occupy a middle ground. That outcome may have disappointed those who had hoped for a series of dynamic rulings, which would effectively reverse the restrictions agreed upon during the legislative process.107 At the same time, however, the judgments do not exhibit a bias towards state preferences either. The Court reined in critical restrictions the destination countries of former ‘guest workers’ and late colonial migration had introduced after the millennium change: prohibitive income requirements, excessive fee levels, and public policy exceptions. Chapter 15.4 on integration will explain that the Court accepted the symbolically important pre-departure language tests as a precondition for family reunification, albeit subject to a hardship clauses.

14.3.4 Procedure and rights of family members Member States determine whether applications are submitted by the sponsor or family members.108 This freedom reiterates our previous conclusion about the significance of national procedural autonomy. Procedural rules differ markedly across the Union, and the Family Reunification Directive refrains from deep harmonisation. Chapter 7.2 on the administrative dimension explained that abstract procedural provisions are complemented by unwritten general principles and procedural safeguards in the Charter. By way of example,

106 See Family Reunification Directive 2003/86/EC, art 16(1)(a). 107 See ch 1.3; Kees Groenendijk, ‘Family Reunification as a Right under Community Law’ (2006) 8 EJML 215, 220–27; and Ryszard Cholewinski, ‘Family Reunification as a Constitutional Right?’ in Apap (n 10) 259, 271–72. 108 See Family Reunification Directive 2003/86/EC, art 5(1).

446 Legal Migration the expiration of the time limit of nine months for examining the application does not mean that family reunification must be authorised.109 Moreover, the abstract reference to ‘legal challenges’ must be interpreted to prescribe judicial oversight by courts, as required by Article 47 CFR.110 Practitioners are well aware of the significance of domestic laws and practices on documentary evidence. Article 5(2) remains abstract in this respect, thus leaving the competent authorities ‘a margin of discretion’.111 Applicants may rely on the concept of effet utile to challenge restrictive practices, but it does not reverse the leeway of the Member States for regulating the procedure and assessing individual applications. In doing so, they can oblige applicants to cooperate, and the lack of cooperation may be sanctioned.112 Marriages of conveniences are a telling example of how important rules and practices on documentary evidence can be. Allegations of convenience concern situations where authorities question the motivation of the marriage. To reject an application, the marriage must have been contracted ‘for the sole purpose of enabling’113 legal entry or stay; an examination of such scenarios habitually involves asking intimate questions.114 Judgments on the asylum procedure, mentioned in Chapter 13.5.2, may be relied upon to censure overly intrusive questioning. Directive 2003/86/EC does not contain an elaborate equal treatment provision. Article 14 simply states that family members will usually be entitled to education, labour market access, and training in the same way as the sponsor. These rudimentary safeguards were supplanted by the later adoption of the Single Permit Directive 2011/98/EU, whose equal treatment guarantee is much broader and applies whenever a third country national is authorised to work. Family members are covered by this provision, which will be presented in Chapter 15 on integration, provided the narrow restriction on labour market access does not apply. Residence permits for family members shall be valid for at least one year and can be renewed indefinitely. States may withdraw or refuse to renew the permit when the conditions are no longer fulfilled, when the marital relationship ceases (irrespective of divorce), or when they discover fraudulent behaviour.115 Fraud does not presume knowledge on the part of all family members, and withdrawal requires an individual assessment of the relevant circumstances.116 Withdrawal probably extends to errors by the authorities, like in the case of subsidiary protection.117 Autonomous residence permits for family members, which are no longer contingent on compliance with the original conditions for entry and stay, shall be available after five years at the latest, in accordance with Article 15. Deference to national legislation means that states may establish further conditions, including, as we shall see, compliance with integration requirements.

109

ibid art 5(4); and ch 7.2.2. ibid art 18; and chs 3.1.3 and 7.2.3.1. 111 E (n 75) para 52. 112 ibid paras 60–62, 67. 113 Family Reunification Directive 2003/86/EC, art 16(2)(b). 114 See Betty de Hart, ‘The Europeanization of Love’ (2017) 19 EJML 281. 115 See Family Reunification Directive 2003/86/EC, art 16(1)(a)–(c), (2)(a). 116 See Case C-557/17 Y.Z. and others EU:C:2019:203, paras 42–56. 117 See Case C-720/17 Bilali EU:C:2019:448, paras 44–52. 110

Family Reunification 447

14.3.5 Beneficiaries of international protection Family reunification with third country nationals enjoying refugee status under the Geneva Convention is governed by the Family Reunification Directive. Its application requires the successful completion of the asylum procedure; asylum seekers may invoke the Dublin III Regulation to be reunited with family members residing in another Member State.118 Moreover, beneficiaries of subsidiary or complementary protection are not covered by Articles 9–12 Family Reunification Directive 2003/86/EC. Nevertheless, the Court will respond to preliminary references on these provisions whenever states voluntarily extend them to subsidiary or temporary protection.119 Guarantees in the Directive were reinforced by a series of judgments that display a tendency for generous interpretation, especially for unaccompanied minors. Refugees do not have to demonstrate accommodation, sickness insurance, and stable and regular resources when sponsoring the entry or stay of spouses and minor children.120 Optional clauses on waiting periods and children above the age of twelve do not apply, in contrast to the optional minimum age of twenty-one years for spouses.121 States may confine these benefits to family relationships which predate entry, and they can require the submission of the application within a period of three months after a positive asylum decision.122 Procedural consequences of late applications are subject to domestic laws, including legitimate reasons not to comply with the time limit.123 Submissions after the time limit and family relationships entered into after entry are subject to the regular conditions.124 The Family Reunification Directive mainly concerns family members residing abroad, since Article 23(2) Qualification Directive 2011/95/EU covers those on the territory already, even in scenarios of irregular secondary movements.125 Exceptions from the entry conditions under Article 7 are confined to the nuclear family. Member States have ‘significant latitude’126 when regulating the admission of other relatives, such as parents or siblings. The abstract provision of Article 10(2) covers only members of the extended family who are dependent on the refugee.127 Admission in other scenarios transcends the scope of the Family Reunification Directive and is not, as a result, subject to the jurisdiction of the Court. Family unity in scenarios of temporary protection by and large follows the model for refugees, including a reasonably generous definition of family membership.128 Unaccompanied minors have an individual right to sponsor the admission of their parents, or of other family members if the parents or grandparents are dead or cannot be traced.129 That is relevant as teenagers often try to reach Europe on their own, thus serving as a sort of ‘anchor’ to support the subsequent entry of parents and siblings. Against this

118

See ch 13.3.3.

119 See E (n 75) paras 34–42. 120

See Family Reunification Directive 2003/86/EC, art 12(1). ibid arts 10(1), 12(2). 122 ibid arts 9(2), 12(1)(3). 123 See Case C-380/17 K and B EU:C:2018:877, paras 43–50, 56–63; and ch 7.2. 124 K and B (n 123) paras 64–65. 125 See also Case C-483/20 Commissaire général aux réfugiés et aux apatrides EU:C:2022:103, paras 38–43. 126 Case C-519/18 Bevándorlási és Menekültügyi Hivatal EU:C:2019:1070, para 40. 127 ibid paras 41–59. 128 See Temporary Protection Directive 2001/55/EC, art 15. 129 See Family Reunification Directive 2003/86/EC, art 10(3). 121

448 Legal Migration background, the moment for the determination of the age of teenagers can be of critical practical relevance. Judges require domestic authorities to rely on the age at the time of the asylum application. Family reunification has to be authorised, if the minor turns eighteen during the asylum procedure or before the application is lodged officially.130 Similarly, children residing abroad can join their parent if they turn eighteen during the asylum procedure of the parent.131 Sponsorship presupposes a ‘real family relationship’ in the meaning of Article 16(1)(b) based on personal and emotional bonds beyond the abstract legal status of family member, even though joint residence or financial support are not required.132 Such generous interpretation confirms our earlier finding, in Chapter 5.4.2 on human rights, about the dynamic interpretation in light of the rights of the child.

14.4 Skilled Labour Migration Few aspects of migration law are, by and large, uncontroversial. Admission of the highly skilled is among them. States generally welcome them, as indicated by the low number of preliminary references. Nevertheless, absence of legal disputes should not be confused with practical irrelevance or theoretical dullness. Demographic change will transform European societies profoundly over the next two decades, and migration can help to offset negative repercussions. The regulatory tool book contains several methods of how states may try to attract the highly skilled, which will not always choose Europe as their destination (14.4.1). Analysed against this backdrop, the defining features—and limitations—of the Blue Card Directive stand out (14.4.2). Students are an entry channel that is often underestimated (14.4.3), whereas the Single Permit Directive gains practical importance mainly through its equal treatment provision (14.4.4). Widespread attention to transnational mobility within the single market in the supranational debate contrasts with the limited practical relevance of the highly complex legislative provisions (14.4.5).

14.4.1 Regulatory toolbox and driving forces beyond the law Economists disagree about how to measure the ‘skills’ of workers. Indicators include formal education, professional experience, other human capital aspects (such as language skills), and success on the labour market.133 These differences have repercussions for the regulation of economic migration. A classic dichotomy is the distinction between a ‘human capital model’ focusing on someone’s potential and ‘demand-driven systems’ relying on labour market needs. Points systems in the Canadian tradition are commonly associated with the human capital model: they use points to identify high potentials to be admitted without a work contract. By contrast, European states are commonly categorised as relying on

130 See Case C-768/19 Bundesrepublik Deutschland EU:C:2021:709, paras 33–42; Joined Cases C-133/19, C-136/ 19 and C-137/19 État belge EU:C:2020:577, paras 24–47; Case C-550/16 A and S EU:C:2018:248, paras 39–63; and Joined Cases C-273/20 and C-355/20 Bundesrepublik Deutschland EU:C:2022:617, paras 36–46. 131 See Case C-279/20 Bundesrepublik Deutschland EU:C:2022:618, paras 27–54. 132 ibid paras 63–68. 133 See Anna Katherine Boucher, ‘How “Skill” Definition Affects the Diversity of Skilled Immigration Policies’ (2020) 46 JEMS 2533, 2356–58.

Skilled Labour Migration 449 demand-driven labour recruitment. Instead of admitting workers on the basis of their potential, they require a work contract and other indicators of labour market needs. In practice, however, state policies overlap to varying degrees: many points systems award credit for work contracts and shortage occupations, whereas the Blue Card Directive relies on formal education and wage levels as embodiments of human capital.134 Notwithstanding overlap in practice, we may distinguish four approaches to attracting skills: points systems and other entry channels for job-seeking; demand in the form of work contracts, labour market tests, and shortage occupations; incentives for businesses to encourage trade and to support start-ups; and drawing talent from the pool of foreign students.135 Other options, such as a ‘migration tax’ as a fiscal allocation mechanism, are not feasible realistically.136 Seen against the backdrop of this ideal-typical classification, the limitations of the EU’s labour migration regime stand out. Secondary legislation covers labour market needs and the academic route but remains largely silent on alternative mechanisms. Introduction of a points system was discarded, in light of internal economic disparities and the need for constantly updating selection criteria.137 These difficulties do not exist at the national levels. Austria and the UK have introduced points systems,138 and the German government wants to follow suit.139 A recurring feature of the European debate are complaints about the low ratio of highly skilled migration in comparison to North America. Lowering the legal thresholds for entry is presented as the natural response.140 There is much truth in this, given that most Member States had closed their labour markets in the 1970s (with the important exception of mobility within the single market). At the same time, however, an inspection of the drivers of migration shows that the highly skilled do not necessarily wait en masse to be allowed entry. Economic cost-benefit calculations and family-related reasons, presented in the interdisciplinary Chapter 4.1, are supplemented by other factors: career prospects in dynamic business environments, the difficulty of the local language, openness towards cultural diversity, the weather, leisure options, and the tax level for higher incomes.141 These contextual factors will not always convince highly skilled Indians, Nigerians, or Chinese to prefer a European destination over emigration to North America. Moreover, legal and practical factors beyond the direct reach of EU legislation are of great practical importance. Application procedures can be cumbersome and frustrating, for businesses and migrants alike. EU legislation requiring swift decision-making may be useful, but it cannot change administrative practices in the absence of meaningful sanctions. Fast-trek avenues for approved sponsors are an option to guarantee swift processing,

134 ibid 2538–45; and Kay Hailbronner and Rey Koslowski, Models for Immigration Management Schemes (German Marshall Fund 2008). 135 See Manolo Abella, ‘Global Competition for Skilled Workers and Consequences’ in Christiane Kuptsch and Eng Fong Pang (eds), Competing for Global Talent (ILO 2006) 11, 17–19. 136 See Jagdish Bhagwati and Martin Partington (eds), Taxing the Brain Drain (North Holland Publishing 1977) 3. 137 See Commission, ‘Impact Assessment’ SEC(2007) 1403 final 202–05. 138 See Kelly Buchanan and others, Points-Based and Family Immigration (The Law Library of Congress 2020); and Commission, ‘Impact Assessment’ SWD(2016) 193 final, Annex 6. 139 See Daniel Thym, ‘Komplexität als Chance’ [2022] Zeitschrift für Ausländerrecht 139. 140 See Proposal for a Blue Card Directive, COM(2007) 637 final, 3; and Commission, ‘Communication: Enhancing legal pathways to Europe’ COM(2018) 365 final, 2–5. 141 See Demetrios G Papademetriou and Madeleine Sumption, Attracting and Selecting from the Global Talent Pool (Migration Policy Institute 2013) 5–8.

450 Legal Migration even though they can be criticised as a privilege of the rich.142 The German term ‘welcome culture’, widely used during the refugee policy crisis of 2015/16, had originally been developed for economic migration. Recognition of professional qualifications is another factor. Comments on the practical effects of equal treatment in Chapter 15.3.4 will illustrate that equal access to recognition procedures can be useful for highly skilled migrants with a university degree. By contrast, professional qualifications will rarely be recognised as equivalent whenever the home state does not operate formalised apprenticeships and vocational education—something few countries outside Europe do. Other incentives include labour market access for family members, tax benefits after entry, and transparent migration laws. Websites explaining entry channels can be a pull factor, in the same way as an attractive brand name. Visibility is a major advantage of points systems, and the idea behind the ‘Blue Card’ was to establish a symbol of openness.

14.4.2 Blue Card Directive The ‘Blue Card’ is the flagship of the EU’s labour migration policy. Its name was loosely modelled on the US ‘Green Card’, although the latter mirrors long-term resident status more than a distinct entry channel for the highly skilled.143 Germany had experimented with an ‘IT Green Card’ in the early 2000s, and a draft report of the LIBE Committee, prepared by a conservative German politician who had studied in the US, called upon the Commission to ‘come forward with a proposal for a European green card system’.144 The Commission did not explain the choice of terminology when it tabled its proposal for the former Blue Card Directive 2009/50/EC two years later, but the semantic parallelism stands out—with the adjective ‘blue’ echoing the colour of the European flag. The Blue Card Directive (EU) 2021/ 1883 substantially revised the rules applicable; it has be transposed by national parliaments until 18 November 2023. The final text was adopted after the Commission had consented to Member States experimenting with parallel domestic schemes, such as points systems in line with previous comments.

14.4.2.1 Personal and thematic scope Highly qualified third country nationals residing on the territory are covered by the blue card scheme, despite the widespread perception of this and other legal migration instruments as entry channels. Students are the prime example of such a switch to a blue card within the territory. In Germany, which issues the overwhelming majority of blue cards, more than half of the applicants receive the status as a second or third residence permit.145 In a symbolically important amendment, the Directive was opened for beneficiaries of 142 See ICT Directive 2014/66/EU, art 11(6); Students and Researchers Directive (EU) 2016/801, arts 9, 15; Blue Card Directive (EU) 2021/1883, art 13; and Tesseltje de Lange, ‘A “Guildian” Analysis of the Equivocal Trusted Sponsorship under EU Labour Migration Law’ in Paul Minderhoud and others (eds), Caught in Between Borders (Wolf 2019) 209. 143 The H-1B visa would be the appropriate comparator. 144 LIBE Committee, ‘Report on an EU approach to managing economic migration (Rapporteur: Ewa Klamt)’ (A6-0286/2005, 7 October 2005) No 25; the idea was not replicated in the final text, adopted by the plenary. 145 See Federal Office for Migration and Refugees, Educational and Labour Migration Monitoring (Annual Report, BAMF 2021) 14, 24–25, which is available in English.

Skilled Labour Migration 451 international protection, thus authorising legal onward movements of refugees, whereas asylum seekers, ‘non-removable’ migrants, and those with complementary protection under domestic laws are excluded.146 Third country nationals with purely temporary residence statuses are not covered either, with the exception of seasonal workers.147 They may theoretically prolong their stay as blue card holders, provided they find a job as a highly skilled worker (a hurdle they will rarely clear). As an instrument of demand-driven recruitment, the Blue Card presupposes a valid work contract or a binding job offer. Applicants must demonstrate that they fulfil the conditions for exercising the profession, for instance admission to the bar as a solicitor.148 As stated previously, the recognition of professional qualifications can impede admission. By way of example, authorities may recognise the university degree of a medical doctor from Georgia, while concluding that the contents of the education stays behind domestic standards in Poland (an important destination for blue card holders). Notoriously difficult to measure are ‘higher professional skills’149 attested by work experience comparable to tertiary education, albeit without the acquisition of formal certificates. Much depends on how domestic authorities apply this standard in practice. Consideration of informal skills is mandatory, under the revised version of the Blue Card Directive, for two distinct categories of IT specialists, while Member States remain free whether to accept informal skills for other professions enumerated in the Directive.150 This list can be amended, according to the revised text, although this will have to be done in the ordinary legislative procedure; the Commission was not given the authority to adopt delegated acts.151 The example of informal skills reiterates our previous comment on the limits of Europeanisation. Conditions for exercising certain professions vary considerably within the EU, as do labour shortages in these sectors. As a result, the supranational institutions should leave national parliaments breathing space to experiment with new admission schemes in the domain of informal skills, which might possibly integrate indicators of human capital in the tradition of points systems. Informal skills and salary levels proved highly controversial during the negotiations.152 Salary thresholds were lowered considerably on the occasion of revision and will be at least 1.0 times, but not higher than 1.6 times, the average gross annual salary in the host state. Lower benchmarks can be introduced for shortage occupations and graduates during the first three years after the completion of their studies.153 Salary thresholds render an additional sufficient resources requirement superfluous; usually, the work contract will automatically give access to a sickness insurance as well.154 Differences in average income mean that the factual requirements will differ considerably within the EU. Several states have used the option to introduce labour market tests or to limit the volumes of admission through

146 See Blue Card Directive (EU) 2021/1883, art 3(2)(a), (b), (f). 147 ibid art 3(1)(e)–(g). 148 ibid art 5(1)(a)–(c). 149 ibid art 2(9). 150 ibid arts 5(1)(1), 2(7), (9)(a), (b), Annex I. 151 ibid art 26(2); and ch 2.3.3. 152 See ‘Proposal for a Council Directive’ (Council doc 11512/08, 24 July 2008) 3–4, 9; and ‘Proposal for a Directive’ (Council doc 5336/17, 18 January 2017) 26–27, 33–34. 153 See Blue Card Directive (EU) 2021/1883, art 5(3)–(5). 154 ibid art 5(1)(e).

452 Legal Migration quotas.155 Ethical recruitment is supposed to prevent ‘brain drain’ to the detriment of lesser developed countries, in particular in the health sector.156

14.4.2.2 Procedure and rights of blue card holders Member States decide whether third country nationals or employers submit the application, and they can foresee fast trek procedures for trusted sponsors.157 Procedural safeguards and the objective of a response no later than ninety days, mentioned in Article 11, should be read against the backdrop of the case law mentioned in Chapter 7.2 on the administrative dimension. Furthermore, Chapter 10 explained how the articles on fee levels, public policy, and more favourable national provisions have to be interpreted.158 Notwithstanding previous uncertainties, the revised Directive should be read to embrace an individual right to be admitted, under the conditions set out by the legislature.159 Chapter 3.3.4.1 on the court architecture explained why legal remedies are of secondary importance in the field of labour migration. Unemployment and loss of status is not widespread among the highly skilled, which mostly navigate on a wave of labour market demand. Nevertheless, the loss of job—or income below the salary threshold—may justify status withdrawal or refusal of renewal.160 Depending on the length of stay, unemployment not exceeding three or six months will not entail loss of status.161 Exceptionally, the birth of several children may give rise to withdrawal if resources are not sufficient to sustain the family.162 Blue card holders can change employers, provided the new occupation meets the original entry conditions, whereas the admission of blue card holders to self-employed activities remains optional.163 Many Member States allow the highly skilled to start businesses, including start-ups, albeit subject to caveats.164 Long-term resident status can be obtained somewhat easier, since previous residence as a student counts fully towards the five-year threshold.165 Extended periods of absence are meant to support ‘circular migration’ to promote home state development, as described in Chapter 18.4.4 on international cooperation. Family members will usually be admitted together with the blue card holder. Article 17 exempts them from several provisions of the Family Reunification Directive 2003/86/EC, not least with regard to pre-departure language tests. Empirical studies show that residence security and unconditional labour market access for spouses, irrespective of their level of qualification, are an important incentive for highly skilled migrants.166 Chapter 10.4.4 explained that settled case law indicates that these exemptions, which often concern entrants from industrialised countries, will not usually amount to unequal treatment on grounds of 155 ibid arts 6, 7(2)(a); and Commission, ‘Impact Assessment’ (n 138) 26; Germany is wrongly mentioned to have operated a labour market test for the blue card in 2016. 156 Blue Card Directive (EU) 2021/1883, recital 41, art 7(2)(e). 157 ibid arts 10, 13. 158 ibid arts 3(3), 4(2), 7(1)(c), 12. 159 ibid art 7(1); ch 10.1.2; and Steve Peers, ‘Legislative Update’ (2009) 11 EJML 387, 394. 160 Blue Card Directive (EU) 2021/1883, art 8(1)(b), (d). 161 ibid art 8(5). 162 ibid art 8(2)(c). 163 ibid art 15. 164 See EMN, Innovative Entrepreneurs (n 41). 165 See Blue Card Directive (EU) 2021/1883, art 18(2)(a); and Long-Term Residents Directive 2003/109/EC, art 4(2). 166 See Barbara Heß, Zuwanderung von Hochqualifizierten aus Drittstaaten nach Deutschland (BAMF 2009) 58–59.

Skilled Labour Migration 453 racial or ethnic origin. The equal treatment provision will be examined in Chapter 15.3 on integration.

14.4.3 Students and Researchers Directive Whereas Member States vehemently opposed, as we have seen, the Commission’s initial approach to labour migration, they agreed to two sectoral directives that suited the orientation of the early debate at the highly skilled. Today’s Students and Researchers Directive (EU) 2016/801 substantially revised and merged the former Students Directive 2004/114/ EC and the former Researchers Directive 2005/71/EC in a single document of considerable size and complexity. Its practical impact is asymmetric: whereas several hundred thousand students from third states are enrolled at European universities (excluding Ireland, as a result of the opt-out), few researchers reside on the basis of the Directive. One may question the added value of a distinct admission scheme for researchers, which could possibly be transformed into a sub-category of the Blue Card.

14.4.3.1 Students as ‘ideal immigrants’ The EU’s approach to student mobility has changed fundamentally. An intergovernmental resolution, adopted in 1996, followed that assumption that successful ‘students must in principle return’167 to support the development of home states. Today, graduates will receive a residence permit for the purpose of job searching or entrepreneurship for a period of at least nine months, subject to procedural caveats.168 States have understood that students come close to what might be called an ‘ideal immigrant’.169 They are by definition highly skilled and are trained in accordance with local traditions; students usually speak the language and are familiar with social and cultural practices; and their recruitment reduces criticism of ‘brain drain’, with the industrialised North hiring workers who have been trained at the cost of home states. Once graduates have found a job, they may receive a blue card, or another domestic residence permit, provided they have a job offer and meet the salary threshold. Doing so replicates the ‘academic route’, which Canada and the US, in particular, employ with great success to attract the highly skilled.170 The new European openness towards students reaffirms our earlier conclusion that the conceptual distinction between the Canadian points system and the demand-driven European model loses relevance. A substantial number of foreigners admitted under the Canada points system are former students. Member States will achieve the same result through the combination of a residence permit for job-seeking, to be followed by a blue card. This system is even more generous than the Canadian one, since it covers all graduates, not only those with enough points.

167 Council Resolution on the admission of third-country nationals for study purposes [1996] OJ C274/10, Part B No 3. 168 See Students and Researchers Directive (EU) 2016/801, art 25. 169 Expert Council on Integration and Migration (Sachverständigenrat für Integration und Migration), Steuern, was zu steuern ist. Jahresgutachten 2018 (May 2018) 58, as whose member and vice-chairperson the author served from 2016 to 2022. 170 See OECD, ‘Untapped Skills. Realising the Potential of Immigrant Students’ (July 2012).

454 Legal Migration Having said this, the Students and Researchers Directive is not limited to first admission. It also covers third country nationals who have attended school in the country.171 While the Directive gains practical relevance mainly through students, it similarly covers school pupils, trainees, volunteers, and au pairs. These distinct entry channels are no longer optional, although Member States are given leeway to regulate and restrict entry and stay for these purposes.172 All applicants must have sufficient resources, together with a sickness insurance, thus effectively limiting admission for these scarcely paid activities to citizens of the OECD countries and some other rich youngsters.173 The option of part-time work, which can only be restricted under narrow conditions,174 alleviates the practical effects of the sufficient resources condition. Part-time work can be of great practical importance, given that states habitually exclude foreign students from study and maintenance grants, especially if they have not resided in the country previously.175 Other conditions can be found in the horizontal Article 7, which must be read jointly with purpose-specific rules. For students, enrolment at a higher education institution, private or public, is particularly relevant, together with sufficient knowledge of the language of the course (not the national language), whenever national law so requires.176 Judges interpret these conditions to be enumerative, and applicants benefit from an individual right to be admitted if they fulfil them.177 This raised the question of whether national authorities may exceptionally refuse an application if enrolment was a pretext for irregular full-time work.178 Today’s Article 20(2)(f) introduced a ground for rejection to this end, thus advancing legal certainty. Domestic authorities have factual leeway when activating this clause, subject to legal remedies.179 Chapter 10.3 explained that administrative discretion in applying entry conditions extends to public security. Procedural safeguards must be analysed in light of the general principles discussed in Chapter 7.2 on the administrative dimension. Equal treatment with nationals will be assessed in the following chapter on integration, and the admission of family members of students is governed by the Family Reunification Directive. Its requirement of a ‘reasonable prospect’ of permanent residence can be interpreted, as we have seen, to cover students in light of the new job-searching option. Of course, all other entry conditions have to be complied with, including sufficient and stable resources for the whole family, which part-time work as a student will rarely be able to guarantee. Access to state-financed family benefits may be restricted by national laws.180

14.4.3.2 Delegation of responsibility to research organisations Admission of researchers deserves our attention for two reasons. First, supranational legislation introduced an example of ‘public private partnership’ to an area of the law which

171 See Students and Researchers Directive (EU) 2016/801, art 2, also for exceptions. 172 ibid arts 12(2), 13(2)–(3), 14(2), 16(2)–(6). 173 ibid art7(1), (c), (3), (3). 174 ibid art 24; and Case C-15/11 Sommer EU:C:2012:371, paras 37–45. 175 See Single Permit Directive 2011/98/EU, arts 3(3), 12(1)(c), 12(2)(a)(ii), (iii). 176 See Students and Researchers Directive (EU) 2016/801, art 11(1)(a), (c). 177 See Case C-491/13 Ben Alaya EU:C:2014:2187, paras 23–32; and ch 10.1.2. 178 Ben Alaya (n 177) paras 34–35. 179 ibid para 33; using the German term Beurteilungsspielraum in the official language of the proceedings, which—unlike Ermessen—does not preclude full judicial oversight. 180 See Single Permit Directive 2011/98/EU, arts 3(3), 12(1)(e), (2)(b)(2).

Skilled Labour Migration 455 is mostly defined by classic forms of hierarchical administration.181 The Directive defines ‘researcher’ as anyone who would have access to doctoral programmes, irrespective of whether she in enrolled in a doctoral programme or not. This entry channel may be used for academic research at universities and public research bodies, as well as for industrial research and development in private companies. The employer will sign a hosting agreement with individual researchers, thus certifying that the person fulfils the research-specific entry conditions; state authorities will limit their assessment to general considerations, such as the public policy caveat.182 Member States may introduce an accreditation procedure for reliable sponsors, together with the threat of sanctions.183 In practice, few countries seem to operate such approval schemes on a larger scale. Secondly, the research-specific entry channel adds value mainly for academic research projects. They are often confined to the analysis of specific idea which will be realised in a pre-defined time frame, for instance doctoral research or post-doctoral projects with third party funding. By contrast, private companies will usually want flexibility to modify the work assignment. The need to specify the purpose and duration of the research project meticulously, under the former legal framework, was unattractive as a result.184 Today’s Article 10(2)(a) gives more flexibility in terms of designating ‘the research activity or the research area’. Nevertheless, the amendment is unlikely to reinvigorate the entry channel. Researchers at private organisations will regularly meet the conditions to be awarded a blue card, with the exception of short stays.185 The situation may be different for universities and research bodies, where wage levels tend to be lower and which do not always give researchers regular work contracts.

14.4.4 Single Permit Directive Political battles over labour migration policy resulted, as we have seen, in the sectoral approach. One horizontal instrument survived, however, in the form of ‘a “light” framework directive’,186 which is nowadays widely known as the Single Permit Directive. The Directive essentially comprises two elements: a single application procedure for residence and work and a horizontal equality clause. It is the equal treatment provision which receives most attention and has resulted in several Court judgments. They will be presented in the chapter on integration that follows, together with the contents of similar provisions in the sectoral directives. The horizontal character of the Single Permit Directive 2011/98/EU comes to the fore when we focus on the scope ratione personae. The Directive applies not only to third country nationals admitted as workers but to anyone with a residence permit for other 181 See Daniel Thym, Migrationsverwaltungsrecht (Mohr Siebeck 2010) 155–62; and Tesseltje de Lange, ‘A Low Level of Trust in EU Labour Migration Regulation’ in Carolus Grütters and Tineke Strik (eds), The Blue Card Directive (Wolf 2013) 17, 18–20. 182 See Students and Researchers Directive (EU) 2016/801, arts 8(3), 10. 183 ibid arts 8(2), 9, 20(2). 184 The author has served as chairperson of the advisory board on research migration at the German Federal Office for Migration and Refugees since 2016 and learnt about the deficits of the former Researchers Directive 2005/71/EC, art 6(2)(a) in this function. 185 Blue Card Directive (EU) 2021/1883, recitals 19, arts 3(2)(c), 5(1)(a), and 29 require a stay of more than six months and prevent parallel applications under different directives. 186 Commission, ‘Impact Assessment’ SEC(2005) 1680 final.

456 Legal Migration purposes, such as family unity, for as long as the person stays legally and is authorised to work.187 Only temporary permits and various humanitarian statuses are excluded in accordance with Article 3(2), in the same vein as irregular migrants. The abstract requirement that beneficiaries must be ‘allowed to work’ indicates that is irrelevant whether they are actually working or not.188 Article 1(2) reaffirms that the Directive refrains from harmonising entry conditions or labour market access. Note that the scope rationale materiae is uneven: Articles 4–11 cover all applications for residence permits, both first time applications and submissions for renewal, whereas the equality clause in Article 12 presupposes that residence has been authorised. Both the scope rationale materiae and the equality provision were discussed controversially during the negotiations, which took four years to complete.189 By contrast, the single application procedure caused fewer headaches. One reason may be that Germany, the harshest critic of EU labour migration policy at the time, had introduced a similar scheme a few years earlier.190 The idea behind a single application procedure and a single permit for residence and work can only be understood against the background of previous state practice. Several countries had two sets of rules: residence permits were the realm of migration authorities, whereas work permits were issued by the labour administration. This division could result in conflicting outcomes: foreigners with a work permit but without a residence permit.191 The Single Permit Directive rendered this problem obsolete. Other procedural guarantees in the Single Permit Directive remain abstract. States are largely free how to organise the procedure; they may foresee an internal coordination of the views of different public authorities.192 Procedural safeguards on the obligation to state reasons, the possibility to mount a ‘legal challenge’, and fee levels are meagre.193 They must be interpreted in light of general principles and fundamental rights described in Chapters 7.2 and 10.1.3. Labour market access and mobility within the state territory can be restricted by national laws.194 A draft amendment does not foresee major changes, apart from an authorisation to switch employers and to retain the residence permit during short periods of unemployment of up to three months.195 This limited ambition, on the part of the Commission, to sponsor further harmonisation signals that the Single Permit Directive is bound to remain a ‘light’ instrument indeed.

14.4.5 Vexed issue of intra-European mobility A paradox seems to lie at the heart of the EU’s labour migration policy: although a single market would ideally require full factor mobility, third country nationals residing legally are not usually authorised to work freely in other Member States. As explained in the 187 See Single Permit Directive 2011/98/EU, arts 2(b), 3(1)(b). 188 See Sara Iglesias Sánchez, ‘Single Permit Directive 2011/98/EU’ in Thym and Hailbronner (n 53) Article 2 MN 5–10. 189 See Gisbert Brinkmann, ‘Opinion of Germany on the Single Permit Proposal’ (2012) 14 EJML 351, 357–64. 190 See Thym (n 181) 147–55. 191 See also Case C-416/96 El-Yassini EU:C:1999:107, paras 32–46, 63–66. 192 See Single Permit Directive 2011/98/EU, recital 12; and Commission, ‘Report on the Implementation of the Single Permit Directive’ COM(2019) 160 final, 4–8. 193 Single Permit Directive 2011/98/EU, arts 8(1), (2), 10. 194 ibid arts 11(a), (b); and Iglesias Sánchez (n 188) Article 11 MN 6–8. 195 See Proposal for a Recast of the Single Permit Directive, COM(2022) 655 final, arts 11(2)–(4).

Skilled Labour Migration 457 introductory chapter, this dichotomy is deeply ingrained in the history of European integration. The free movement of workers was never intended to be for third country nationals. From today’s perspective, Article 45(2) CFR and Article 79(2)(b) TFEU reaffirm that the legislature decides on the desirability and degree of approximation.196 Similarly, the creation of the Schengen area made little inroads, as we have seen, on legal migration, since it confined itself to short-term travel. For third country nationals, labour markets remain segmented along state borders, legally and factually. Few companies would consider actively looking for workers from third states residing in another Member State and few third country nationals apply for such jobs in real life. Political debates in Brussels tend to pay much attention to intra-European mobility. To do so is partly an end in itself, since transnational mobility serves as a justification, in light of the principle of subsidiarity, why the EU should become active in the first place. Unlike in the fields of asylum, border controls, or short stays, the added value of common action is not self-evident.197 Incredibly small numbers of transnational movements reiterate this point.198 To blame statutory restrictions for this outcome gives only part of the explanation. Imagine an Indian IT specialist who has relocated to Lyon after having learnt some basic French; moving on to another country with a different language would require a considerable effort, especially if the family came along. A direct comparison with the US is asymmetric. Empirical studies show that many workers make a deliberate choice to emigrate to a specific country, whereas others perceive their stay to be temporary to gain experience.199 The situation can be different in border regions, for instance in the Benelux. Moreover, some economic sectors are more mobile than others. Academic research, for instance, is an exception to the rule of little transnational mobility. Several instruments discussed in this chapter, as well as the Long-Term Residents Directive, contain provisions on transnational mobility. They are highly fragmented and do not generally allow third country nationals to choose freely whether to work elsewhere. Provisions in the Long-Term Residents Directive 2003/109/EC are particularly narrow: they apply after five years of legal residence at the earliest and allow the second Member State to prescribe economic self-sufficiency, labour market tests, and integration requirements; third country nationals must apply for admission before taking up employment abroad.200 The Commission proposes lowering that threshold, building upon similar changes in the revised Blue Card Directive.201 It remains to be seen whether they will increase the numbers of mobile workers significantly. Provisions on long-term mobility also exist for students, researchers, and intra-corporate transferees.202 All other third country nationals with residence permits may travel freely for up to ninety days in accordance with the Schengen acquis, whereas Union law does not govern their entry and stay for longer periods.

196 See ch 1.1.5–7 and 1.3.2; see also Sara Iglesias Sánchez, ‘Free Movement of Third Country Nationals in the European Union?’ (2009) 15 ELJ 791. 197 See Farcy (n 31) 215. 198 See EMN, Intra-EU Mobility of Third-Country Nationals: Synthesis Report (2013) 47–55, while mentioning that some more cross borders as frontier workers, family members, or on the basis of domestic laws. 199 See Heß (n 166) 70–73; and Siew-Ean Khoo, Graeme Hugo, and Peter McDonald, ‘Which Skilled Temporary Migrants Become Permanent Residents and Why?’ (2008) 42 IM Rev 193. 200 See Long-Term Residents Directive 2003/109/EC, arts 14(3), (4), 15(2), (3), 16(4)(c), 19. 201 See Proposal for a Recast of the Long-Term Residents Directive, COM(2022) 650 final, arts 16–25; and Blue Card Directive (EU) 2021/1883, arts 21–22. 202 See ICT Directive 2014/66/EU, art 22; and Students and Researchers Directive (EU) 2016/801, arts 29–32.

458 Legal Migration An innovation, introduced first by the ICT Directive, concerns simplified procedures for short-term mobility for up to 90 days in any 180-day period. The idea behind short- term mobility is to allow intra-corporate transferees, researchers, and blue card holders to work elsewhere on a temporary basis. A symbolically and practically relevant novelty is the application procedure. It is essentially limited to a notification, without the need of prior authorisation and with limited options of the destination country to object; such enhanced transnational effects come close to the mutual recognition of residence permits for economic purposes.203 Unfortunately, the provisions are quite complex and support our overall conclusion of legislative fragmentation, presented in Chapter 6.2. Statistics will show whether short-term mobility will be used much. Talking to practitioners, the author learnt repeatedly that one of the reasons for the limited practical relevance is the easy availability of alternatives. Short-term mobility for professional reasons may often be performed on the basis of the Schengen rulebook de facto, even if free travel for up to ninety days cannot officially be used for economic purposes. Long-term mobility can be realised on the basis of existing laws as well, by simply applying for another residence permit in the second country, instead of using the cumbersome intra-EU mobility scheme. By way of example, a computer specialist may receive another blue card, or a domestic permit, when relocating from Antwerp to Rotterdam.204 Remember that most states have little interest in obstructing the entry and stay of the highly skilled. High earners may ignore the risk of negative repercussions, in case of illness or unemployment, if they do not acquire long-term resident status quickly by accumulating periods of residence in several Member States.205

14.5 Temporary Economic Activities EU migration law does not generally distinguish between ‘temporary’ statuses and ‘immigration’ with the prospect of settlement. Most residence permits can be renewed, until migrants acquire long-term resident status. The exception to the rule are distinct avenues for temporary labour migration. These temporary admission schemes concern both qualified specialists and workers with lower skills. A first pattern is the employer-driven secondment of personnel. Service provision in the single market covers ‘posted’ workers from third states at all levels of qualification (14.5.1), whereas international trade law mostly concerns managers and other experts (14.5.2). The ICT Directive builds on this tradition, albeit for longer periods and flexible work assignments (14.5.3). Seasonal work is a classic phenomenon, for which the Seasonal Workers Directive establishes common framework rules (14.5.4). Legal pathways feature prominently in policy debates about asylum and can come in various forms (14.5.5).

203 See ICT Directive 2014/66/EU, art 21; Students and Researchers Directive (EU) 2016/801, art 28; and Blue Card Directive (EU) 2021/1883, art 20; they serve as leges speciales to the Schengen acquis; see Fabian Lutz, ‘Transposition of the ICT Directive 2014/66/EU’ in Paul Minderhoud and Tesseltje de Lange (eds), The Intra Corporate Transferee Directive (Wolf 2018) 19, 27–29. 204 The head of a migration authority in a big German city told me that it would require extra work effort to apply highly complex transnational mobility rules, which the officers do not know well. 205 See Blue Card Directive (EU) 2021/1883, art 18(2).

Temporary Economic Activities 459

14.5.1 Posted workers in the single market Article 56 TFEU allows companies to relocate their workforce within the single market for purposes of service provision, for instance if an Italian construction company wins a contract to build a hotel on the French Riviera. Admission of workers serves an ulterior purpose: the freedom to provide services would be ineffective if companies were not allowed to transfer their personnel; the workers benefit from a derived right.206 In Vander Elst, the Court confirmed that the rationale behind the posting of workers extends to third country nationals.207 By way of example, the Italian construction company may move Tunisians workers with an Italian work contract to the French construction site. In the specific context of enlargement, posting usually becomes available from day one after accession, irrespective of transitional periods for the free movement of workers.208 Posting of workers became highly controversial in this context. Posted workers exemplify the political, conceptual, and practical dynamics which can complicate the labour migration of the low skilled. While the service providers, their workers, and home states value the economic opportunity, feelings in host states are often mixed. Customers appreciate the cost advantage, but trade unions often complain about the danger of downward pressure on wage levels and working conditions. ‘Social dumping’ became a slogan in policy debates about intra-European labour migration after the fall of the iron curtain, as did the widely publicised example of the ‘polish plumbers’.209 Legal and political controversies resulted in the adoption of the Posted Workers Directive 96/71/ EC, which was accepted by the Court to be compatible with the fundamental freedoms.210 Another Commission initiative to formalise the legal status of workers from third states failed; until today, the migration-specific elements of the posting of workers are essentially governed by the supranational case law.211 At the heart of the Posted Workers Directive 96/71/EC lies a delicate compromise between the countervailing interests of the multiple actors involved. Temporary posting means that workers are not fully integrated in the labour market of the host state, since they remain covered, in accordance with single market rules, by the social security system of the home state for stays of up to 24 months.212 Directive 96/71/EC departed from this rule by requiring companies to comply with the host state legislation on remuneration, amongst others.213 Whereas host states in Western Europe praised ‘equal treatment’, home states further East deplored ‘protectionism’ in the form of social standards. Indeed, cheaper labour costs were a main driver behind the expansion of posting over the past three decades, if, by way of example, Ukrainian workers undertake plumbing work for a Polish company on a construction site in Sweden. An amendment adopted in 2018, again after intense political 206 See Case C-113/89 Rush Portuguesa EU:C:1990:142, paras 12–18. 207 See Case C-43/93 Vander Elst EU:C:1994:310, paras 18–26. 208 The freedom to provide services is not usually restricted; for a rare and limited exception see Croatian Act of Accession (n 49) Annex V, No 2(13). 209 See https://en.wikipedia.org/wiki/Polish_plumber (accessed 1 March 2023); and Nicola Countouris and Samuel Engblom, ‘Protection or Protectionism’ (2015) 6 Eur Labour LJ 19. 210 On the early debate see Paul Davies, ‘Posted Workers’ (1997) 34 CML Rev 571. 211 See Proposal for a Directive on the posting of workers who are third-country nationals, COM(99) 3 final, which was withdrawn later; and Kay Hailbronner, Immigration and Asylum Law and Policy of the European Union (Kluwer 2000) 201–12. 212 See Social Security Coordination Regulation (EC) No 883/2004, arts 11(1), 12(1). 213 See Posted Workers Directive 96/71/EC, art 3(1)(c).

460 Legal Migration debates, specified what counts as remuneration and required companies to respect additional rules of the host state after twelve months.214 The example of posted workers illustrates that compliance and enforcement can be as important as legislation. An important strand of the case law concerns the control powers of host states. Companies may be obliged to report the names and numbers of workers from third states beforehand, thus establishing a simplified de facto visa procedure, subject to quasi-automatic authorisation and without administrative discretion.215 Most importantly, however, host states retain primary responsibility for checking social security contributions other than remuneration. Posted workers are issued an A1 certificate, formerly known as the E101 form, which confirms that social security contributions are paid for. Host states cannot usually question the veracity of these documents, raising concerns about abuse.216 Recourse to subcontracting, temporary employment undertakings, and placement agencies further complicates enforcement.

14.5.2 GATS and other trade agreements Chapter 5 explained that migration law generally lacks a coherent international legal framework besides human rights law and the Refugee Convention. During the initial euphoria about globalisation after the end of the cold war, much hope was placed on the dynamic evolution of trade law. For a while, it seemed as if cross-border movements of people might be liberalised in the slipstream of international trade negotiations.217 Political scientists had predicted early on that this hope would be frustrated for contextual reasons: ‘people are not bananas’.218 Dynamism in the field of goods cannot be projected upon people easily. Direct and visible consequences of migration for societies undermine domestic support, and conflicts of interests between home states and receiving countries complicate international negotiations.219 Notwithstanding these differences, service provision has a certain impact on migration law. The rationale behind trade-related entry channels is the same as for posted workers. The General Agreement on Trade in Services (GATS), which is one of the pillars of the World Trade Organization (WTO), distinguishes four modes of service provision. Mode IV concerns cross-border service delivery ‘through presence of natural persons . . . in the territory of any other [State Party]’.220 There is, however, no general obligation to admit natural persons for purposes of service provision. Rather, the option has to be activated in the form of country-specific ‘schedules’, which enumerate the degree of liberalisation in greatest detail. Commitments distinguish meticulously between sectors and for which of the four modes

214 ibid art 3(1), (1a), as amended by Directive (EU) 2018/957 [2018] OJ L173/16. 215 On the unresolved ambiguity underlying Case C-445/03 Commission v Luxemburg EU:C:2004:655, paras 30– 46 see Commission (n 35) 113. 216 See Case C-370/17 CRPNPAC EU:C:2020:260, paras 48–86. 217 See Joel P Trachtman, The International Law of Economic Migration (WE Upjohn Institute for Employment Research 2009) chs 8–9. 218 Jennifer Gordon, ‘People Are not Bananas’ (2010) 104 NW U L Rev 1109. 219 ibid 1130–39; and James F Hollifield, ‘Migration and the “New” International Order’ in Bimal Ghosh (ed), Managing Migration (OUP 2000) 75, 94–100. 220 General Agreement on Trade in Services (adopted 15 April 1994, entered into force 1 January 1995) [1994] OJ L336/191, art 1(2)(d).

Temporary Economic Activities 461 market access and/or national treatment shall be guaranteed.221 The original schedule of the EU and the Member States covered no fewer than ninety-six pages, with numerous country-specific commitments. Inspection of the small print unearths how limited their practical impact on the admission of workers is.222 Few people foresaw twenty years ago that attempts at further liberalisation would falter. Trade negotiations on additional commitments in the WTO have effectively come to a standstill.223 Even if the early dynamics had continued, the legal design of the GATS would have impeded transformative effects. Country-specific commitments mean that governments are in full control of the degree of openness. Moreover, a legally binding annex specifies that countries are neither obliged to authorise labour market access beyond service provision nor prevented from applying measures to regulate the entry of natural persons, including through visa requirements.224 The absence of direct effect of world trade law in the EU legal order, in accordance with settled case law, further diminished the potential of surprise outcomes. Mode IV under the GATS remains a far cry from the liberalising effects of the posting of workers in the single market. Chapter 17 on association agreements will explain that the Trade and Cooperation Agreement with the UK, as well as some agreements with the Western Balkans countries and under the Eastern Partnership, contain provisions on service provision in the single market. Moreover, countries further afield have occasionally tried to use trade negotiations as an opportunity to break open European labour markets. Caribbean states were reasonably successful in receiving concessions beyond the level of commitments under the GATS for key personnel, business visitors, and contractual service supplies, albeit subject to a generic caveat on labour market access and entry clearance.225 Other third states, notably India, have insisted on similar concessions during negotiations, only to be rebutted by the EU side. Countries with weaker economies would not usually benefit much from commitments on services anyway, although they might be interested, as we shall see, in other forms of legal pathways for economic purposes.

14.5.3 Inter-Corporate Transfers Directive Unlike the Posted Workers Directive, the ICT Directive 2014/66/EU applies to third country nationals residing abroad. It is about first admission to the Member States, not secondment within the single market or status change for those residing on the territory already. In that respect, the ICT Directive overlaps with transnational service provision under 221 See https://www.wto.org/english/tratop_e/serv_e/serv_commitments_e.htm (accessed 1 March 2023); and Joel P Trachtman, ‘Economic Migration and Mode 4 of GATS’ in Vincent Chetail and Céline Bauloz (eds), Research Handbook on International Law and Migration (Edward Elgar Publishing 2014) 346. 222 See Simon Tans, Service Provision and Migration (Brill/Nijhoff 2017)chs 5–6. 223 On the state of play before negotiations lost vigour see Michele Klein Solomon, ‘GATS Mode 4 and the Mobility of Labour’ in Ryszard Cholewinski and others (eds), International Migration Law (TMC Asser Press 2007) 107, 115–26. 224 See Annex on Movement of Natural Persons supplying Services under the Agreement [1994] OJ L336/217, Nos 2, 4. 225 See Economic Partnership Agreement between the CARIFORUM States and the EC and its Member States (signed 15 October 2008, provisionally applied since 29 December 2008) [2008] OJ L289/3, arts 60(5), 80–84, Annex IV; and Natasha Ward, ‘Facilitating the Temporary Movement of Natural Persons’ in Rahel Kunz and others (eds), Multilayered Migration Governance (Routledge 2011) 143.

462 Legal Migration international trade law, although its scope is much broader ratione loci, materiae, and personae. Geographically, it covers transfers all third states, not only WTO members or associated countries. Thematically, all economic sectors benefit from the Directive, in contrast to the sector-specific commitments under the GATS. Finally, the personal scope includes anyone who qualifies as a manager, specialist, or trainee. Several Member States had similar entry channels beforehand, and Europeanisation meant that others had to introduce them as well; nevertheless, the practical impact remains muted.226 One reason may be that beneficiaries are mainly multinational undertakings. Small and medium-sized companies are covered as well, but they will rarely have larger subsidiaries outside the single market. That is a prerequisite for the use of the ICT Directive, however, which applies to employees who are being transferred between different branches of the same company. States set a minimum length of previous employment outside the Union of between three and twelve months, in order to avoid concealed recruitment from abroad.227 Only certain categories of employees may be transferred on the basis of the Directive, which covers ‘managers’, ‘specialists’, and ‘trainees’. The vague definition of the notions of managers and specialists are an open flank of the regulatory design: they are defined in terms of significance for the company, not formal qualification like in the case of the blue card.228 Only personnel working in regulated professions are required to have the relevant diplomas.229 Moreover, ‘trainees’ must be university graduates, in addition to the imprecise requirement of transfer for the purposes of career development to obtain specific skills.230 These ambiguities leave domestic authorities much leeway. Intra-corporate transfers can last in-between ninety days and three years. Extension is not possibly, although Member States remain free to authorise work on other legal bases.231 Transferees are not allowed to switch employers in the absence of general labour market access, but companies are guaranteed entry clearance if the conditions are met.232 Procedural safeguards and grounds for refusal must be interpreted in light of the horizontal comments in Chapters 7.2 and 10 on the administrative dimension and general features of secondary legislation. Equal treatment will be discussed in the next chapter on integration. Family members may accompany the transferee in accordance with the Family Reunification Directive, despite the absence of the prospect of longer stay and subject to privileges mirroring the situation of blue card holders.233 Intra-European mobility was mentioned previously, in the context of skilled labour migration. The temporary character of the transfer means that the design of the employment contract determines whether the labour laws and social security legislation of the home state or the host country apply. In any case, Member States shall require remuneration to follow the host state practices. Orientation at the rules applicable to posted workers within the single market is strengthened by an additional caveat: states may require compliance with collective agreements and comparable practices which are widely but not universally

226

See the Eurostat dataset (n 1) for ‘migr_resict1_1’. See ICT Directive 2014/66/EU, arts 3(d), (l), 5(1)(a)–(b). 228 ibid arts 3(e), (f), 5(1)(b), (d); and Tesseltje de Lange, Welcoming Talent (Radboud University 2021) 12–13. 229 ICT Directive 2014/66/EU, arts 3(o), 5(1)(e). 230 ibid arts 3(g), 5(6). 231 ibid arts 1(a), 12. 232 ibid recital 21, arts 6, 13(1). 233 ibid art 19. 227

Temporary Economic Activities 463 applied.234 Doing so effectively erects higher thresholds than the level of mandatory protection under domestic laws. Admission will be rejected if remuneration is insufficient to sustain the transferee and their families.235 Resolution of further questions about the applicable social standards, including the transfer of social benefits to home states after or during the stay, will be discussed in the context of the equal treatment provisions.

14.5.4 Seasonal Workers Directive Seasonal workers must come from outside Union territory. The Seasonal Workers Directive 2014/36/EC was designed as a vehicle for first admission, thus preventing the legalisation of the irregular stay. To offer a legal pathway for migrants with low skills builds upon entry channels several Member States had agreed upon with countries in Northern Africa and Eastern Europe, amongst others as an incentive to foster cooperation on the prevention of irregular movements. Seasonal work remains statistically relevant, although the financial crisis resulted in a decrease in admissions to Southern Europe during the 2010s. Narrow definition of ‘seasonal work’ as an activity dependent on the passing of the seasons implies that the Directive does not cover other short-term work, for instance construction or in-house care for family members.236 Parallel domestic admission schemes may cover such non-seasonal temporary work. To be allowed entry, third country nationals must sign a work contract with a local employer, including temporary work agencies and subsidiaries.237 Local employment entails that the labour laws and social policies of the host state apply, although some states have special rules for short-term employment or seasonal work.238 Insistence upon sufficient resources can be relevant for destination countries where low-income earners may receive supplementary in-work benefits if they do not earn enough money; the additional obligation to apply collective agreements and prevalent practices implies that pay levels and working conditions may transcend mandatory minimum requirements under domestic laws.239 Directive 2014/66/EU distinguishes between stays not exceeding ninety days and seasonal work for up to nine months during any twelve- month period, with extension being generally possible within these upper limits. The right to switch employers is meant to reduce the danger of exploitation.240 Equal treatment with nationals will be discussed in the chapter on integration that follows, whereas procedural safeguards and grounds for rejection must be interpreted in accordance with the horizontal comments in Chapters 7.2 and 10 on the administrative dimension and general features. Member States may apply a labour market test or numerical caps in the form of quotas.241 Family members are not allowed entry: neither under the Seasonal Workers Directive nor under the Family Reunification Directive, in the absence of a prospect of longer stay. The option of return for seasonal work in the following year 234 ibid art 5(4)(a), (b); and ch 15.3.4. 235 ICT Directive 2014/66/EU, art 5(5); and ch 10.2. 236 See Seasonal Workers Directive 2014/36/EC, arts 2(2), 3(b), (c). 237 ibid recital 12, arts 5(1), 6(1). 238 ibid art 23(2)–(3); and EMN, Attracting and Protecting Seasonal Workers from Third Countries in the EU: Synthesis Report (2020). 239 Seasonal Workers Directive 2014/36/EC, arts 5(1)–(3), 6(1)–(3); and ch 15.3.4. 240 ibid arts 5–6, 14–15. 241 ibid arts 7, 8(3).

464 Legal Migration reminds us of the prominence of circular migration in policy circles at the time.242 In contemporary debates, the migration-development nexus usually features under the heading of legal pathways as an alternative to irregular entry. Any analysis of seasonal work will have to consider the reality of informal employment, not least in Southern Europe.243 Empirical and critical studies demonstrate that structural legal and economic factors can facilitate abuse and exploitation, especially in agriculture.244 The situation may be better for tourism, another classic area of seasonal work, even though the predominance of smaller companies means that much depends on the context. The legislature responded to the challenge of exploitation by reaffirming the right to strike and by prescribing information provision, a complaints mechanism, monitoring, inspections, and employer sanctions in case of abuse.245 The Commission proposes extending several of these guarantees to all workers covered by the Single Permit Directive.246

14.5.5 Legal pathways for economic purposes Reaching agreement on the directives on economic migration was a long and cumbersome process. They focus on the highly skilled, with the exception of seasonal work. Admission of unskilled or lesser qualified workers remains the prerogative of the Member States, and the arguments against common action cannot be brushed aside easily. Labour market projections indicate that the demand for unskilled workers varies significantly across the Union and between economic sectors; retail and distribution are commonly designated as areas of future demand.247 Even so, governments hesitate to promote inward migration as a policy response. Activation of the domestic population often takes priority over recruitment from abroad, in accordance with previous comments on the effects on the welfare state. Nevertheless, several countries have statistically significant temporary migration regimes for the lesser skilled. Poland offered circular migration opportunities for hundreds of thousands of Ukrainians even before Russia attacked the country, and Germany operates the ‘Western Balkans System’ as a prelude to EU accession.248 In light of labour market differentials and political reluctance, the Parliament’s request for a common supranational rulebook on workers with lesser skills will probably not be successful.249 Legislative harmonisation can be expected to remain confined to the horizontal provisions in the Single Permit Directive, which covers all workers admitted for a period exceeding six months.250 However, absence of legislative harmonisation need not

242 ibid art 16; and ch 18.4.4. 243 See Ferruccio Pastore, ‘The Governance of Migrant Labour Supply in Europe’ (2014) 2 Comp Migr Stud 385. 244 See Lydia Medland, ‘Misconceiving “Seasons” in the Global Food Systems’ (2017) 23 ELJ 157; and Letizia Palumbo, ‘Exploitation in the Agri-Food Sector in Europe’ (2022) 24 EJML 287. 245 See Seasonal Workers Directive 2014/36/EC, arts 10–11, 17, 20, 23(1)(b), (c), 24–25; and Judy Fudge and Petra Herzfeld Olsson, ‘The EU Seasonal Workers Directive’ (2014) 16 EJML 439, 459–63. 246 See Proposal for a Recast of the Single Permit Directive (n 195) arts 13–14. 247 See Commission (n 35) 164–68. 248 See Zvezda Vankova, Circular Migration and the Rights of Migrant Workers in Central and Eastern Europe (Springer 2020) chs 5–8; and Expert Council on Integration and Migration/Migration Policy Institute, Legal Migration for Work and Training (Study 2019-2, November 2019) 24–43. 249 cf European Parliament, ‘Resolution with recommendations to the Commission on legal migration policy and law’ (P9_TA(2021)0472, 25 November 2021) Nos 2, 6–8. 250 See Single Permit Directive 2011/98/EU, art 3(3).

Temporary Economic Activities 465 imply inactivity on the part of the EU institutions. Article 79(2) TFEU can be interpreted, as we have seen, to cover admission schemes for specific purposes, subject to the principle of subsidiarity. Moreover, EU institutions may support domestic initiatives through the exchange of best practices, operational backing, financial incentives, and bilateral cooperation frameworks with third states. The supranational toolbox in support of legal pathways comprises at least four components. They have been proclaimed with much fanfare repeatedly over the past years, although the track record of output delivery remains meagre.251 Limited success in practice need not, however, invalidate the idea behind legal pathways for all skill levels. First, much political energy has been invested in the idea of tailor-made ‘mobility partnerships’, ‘common agendas’, or ‘talent partnerships’ with selected third states. Chapter 18.4.4 will explain that most of these initiatives have failed to deliver much substance, so far at least, even though the Commission promises to breathe life into a longstanding idea in the form of pilot project for future talent partnerships. We shall learn whether the initiative is successful. Secondly, operational support includes the project of a ‘talent pool’ to bring together employers in Europe with potential workers residing abroad, building upon similar schemes in Australia, Canada, and New Zealand.252 Such match-making facilitates the use of existing entry channels under supranational legislation or national laws. A designated website is scheduled to be set up by mid-2023 for applications from outside Europe. In a smart strategic move, the Commission plans to initiate a pilot phase for the labour market integration of Ukrainians residing within Europe already, which may ideally build momentum for future expansion.253 At the end of the day, practical effects will depend on whether the ‘talent pool’ will be used by national employment agencies, aspiring workers, and future employers. Cooperation of national support agencies within the European Network of Employment Services (EURES) may sustain some momentum behind this venture in the medium run.254 Practical relevance is far from guaranteed, however. Thirdly, EU institutions may use the financial muscle of the justice and home affairs budget, presented in Chapter 7.4, to convince Member States to become active. Support for classic entry channels, such as the Erasmus +Programme for international students, can be combined with innovative entry routes. In this respect, the international debate has moved away from the earlier focus on the circular migration of the highly skilled, discussed in Chapter 18.4.4 on mobility partnerships, to ‘global skill partnerships’ to be implemented on the territory of third states. The idea is to set up training capacities in third states to serve a double objective: the economy of the home state benefits through a better trained workforce, and some trainees will use the skills they acquire as a springboard for migration to Europe.255 Doing so may ideally foster the economic development of home states, also considering that skilled workers—as opposed to highly skilled university graduates—are often rare outside OECD countries. Fourthly, capacity building may have to transcend financial support for pilot projects. Effective cooperation with countries of origin requires reasonably well-functioning labour

251

See Farcy, L’immigration économique (n 34) 237–51.

252 See Building an EU Talent Pool (OECD, March 2019). 253

See Commission (n 42) 14–17. See Regulation (EU) 2016/589 on a European network of employment services (EURES) [2016] OJ L107/1. 255 See generally Michael A Clemens, ‘Global Skill Partnerships’ (2015) 4 IZA J Labor Pol 1. 254

466 Legal Migration and migration authorities, which do not exist everywhere. Germany, for instance, conducted a screening exercise to identify partner countries to set up pilot projects on legal pathways for health and care workers. Future labour migrants are to obtain basic preparation at home, before entering Germany for training and eventual employment. Screening of potential partners meant that African states, besides mid-income countries like Egypt and Tunisia, were excluded, since they lacked state capacities to run such cooperative projects.256 Northern Africa is also the setting of another pilot scheme, which receives EUR 20 million during 2019–23 and will be realised by different partners.257 These practical measures may be less visible than new legislation, but they are, arguably, the kind of instruments that may deliver in practice. Legal pathways for economic purposes are an end in themselves to attract workers. At the same time, they often serve an ulterior purpose: to help reduce irregular migration via the asylum system by offering legal migration channels. Notwithstanding the political appeal of this objective, inspection of the contextual factors advises caution. Such ‘diversion’ of irregular movements into legal migration channels presupposes that individual migrants perceive a realistic chance to be admitted to the Member States legally for work purposes. That might be the case, to some extent at least, in specific scenarios, such as temporary work permits for Moroccans or Eastern neighbours. By contrast, a direct causal linkage between legal pathways and irregular movements is less plausible for other countries of origin. Around 25 thousand Nigerians applied for asylum in the EU during 2020, many of them without success. Member States will hardly agree to open an equivalent number of legal migration opportunities; even if they did, work permits might be given to other applicants than those trying to reach Europe irregularly.258 Absence of a direct causal connection between irregular arrivals and the availability of legal pathways need not undo the political rationale behind the linkage. European politicians present legal pathways as an integral part of a balanced policy mix approach embracing restrictive elements and legal migration opportunities at the same time. Moreover, legal entry channels for lower qualified workers can be used to convince third states to sign up to collaborative migration management, to be analysed in Chapter 18 on the external dimension. Such cooperation may come in the form of ‘mobility partnerships’ and ‘talent partnerships’, or more informal arrangements with a give-and-take in the mutual interest.

14.6 Summary Legal migration habitually receives little political and academic attention, despite being highly relevant. Several million third country nationals enter the Member States for purposes of work, family unity, and education each year. EU institutions benefit from a broad range of competences and habitually emphasise the need for an overarching supranational

256 See Expert Council on Integration and Migration (Sachverständigenrat für Integration und Migration), Migration aus Afrika nach Europa. Jahresgutachten 2020 (May 2020), 103–04, in which the author participated as Vice-Chairperson. 257 The official project name is ‘Towards a Holistic Approach to Labour Migration Governance and Labour Mobility in North Africa’ (THAMM); see https://www.ilo.org/africa/technical-cooperation/WCMS_741974 (accessed 1 March 2023). 258 Remember that the overall population of Nigeria stood at roughly 200 million in 2020.

Summary 467 migration policy. Nevertheless, the rationale for common rules on legal migration is less evident than in the fields of asylum and border controls. Limited transnational mobility, fragmentation of national labour market, and the absence of shared policy objectives hinder the emergence of a truly pan-European outlook. Labour migration is an area where national initiatives harbour the potential of regulatory innovation, such as points systems for the highly skilled or legal pathways for the lower skilled, which cumbersome supranational law-making procedures have failed to deliver. Policy debates in Brussels witnessed an interinstitutional battle over labour migration after the millennium change. Member States rejected Commission proposals for a horizontal and moderately generous entry regime for labour migration, which would have departed from the ‘zero immigration’ policies most governments and intergovernmental measures had pursued until then. The dispute was about the desirability of Europeanisation as much as it was about the contents of the harmonisation proposals. EU institutions finally agreed on a ‘sectoral approach’, which combines five pieces of legislation on specific aspects of economic migration with a horizontal framework instrument. This sectoral approach continues to frame the debate and implies that important aspects of economic migration are governed by national laws. Member States had traditionally allowed family members to accompany or join third country nationals. However, family unity became a bone of contention when the destination countries of ‘guest workers’ and late colonial migration in Western Europe made it harder to sponsor the entry and stay of spouses from countries of origin. Optional clauses in the Family Reunification Directive 2003/86/EC authorise states to maintain or introduce restrictions. Judges generally accepted these conditions to be compatible with human rights, subject to hardship clauses. Doing so reflects settled case law that the human right to family life does not usually bring about a guarantee of admission. The official title of the Family Reunification Directive is misleading insofar as it covers both the reunification of pre-existing families and the formation of new ones. More generous entry schemes exist for refugees and beneficiaries of temporary protection, as well as for spouses and children of Union citizens with the nationality of third states. A widespread consensus holds that industrialised economies benefit from the admission of the highly skilled, which will gain relevance as a result of demographic change. A comparison with other options of regulating labour migration demonstrates that the revised Blue Card Directive (EU) 2021/1883 principally continues a conventional model of demand-driven recruitment, based on formal qualification and work contracts. Blue card holders may be admitted from abroad or be former students staying after the successful completion of their studies. Enhanced opportunities for job-searching were an innovation of the revised Students and Researchers Directive (EU) 2016/801. Statistically speaking, the academic route is one of the most important entry channels. By contrast, the provisions on intra-European mobility in the diverse legal migration instruments are rarely used in practice, reflecting regulatory hurdles, a lack of interest, and the easy availability of alternatives. Procedural safeguards in the Single Permit Directive 2011/98/EU remain abstract, whereas the equal treatment provision has gained practical relevance. It is the only instrument that applies horizontally to all third country nationals with labour market access, irrespective of whether entry or stay is governed by supranational legislation. EU migration law does not generally distinguish between temporary migration and permanent settlement. Nevertheless, distinct entry channels exist for temporary employment.

468 Legal Migration Of great relevance is the posting of workers in the single market. By way of example, a Polish construction company may bring Ukrainian workers to Sweden to build a new hospital. Corresponding rules on cross-border service provision under the GATS or bilateral trade agreements never gained much traction. The ICT Directive 2014/66/EU gives multinational companies more flexibility in terms of sending managers and other specialists who are employed in a third state to work for the same company in Europe for up to three years. Whereas such transfers concern skilled employees, seasonal workers are mostly unskilled, especially in agriculture and tourism. Empirical studies show that abuse and exploitation are widespread, notwithstanding protection clauses in Seasonal Workers Directive 2014/ 36/EC. States decide to which sectors the Directive shall apply, and non-seasonal temporary employment is governed by domestic laws. Pilot projects for legal pathways for economic purposes at the national level remind us that a holistic assessment of legal migration policies requires the combined analysis of supranational and national legislation.

15

Integration and Settlement Entry onto the territory is the beginning of a process whose endpoint is not always predetermined. While some migrants will return home, often perfectly voluntarily, others will stay. European societies increasingly recognise this reality of permanent settlement. ‘Zero immigration’ policies have given way to acceptance. In doing so, governments often try to influence the outcome, reflecting the European social model of active state involvement. These policy measures are usually discussed under the heading of ‘integration’ in contemporary Europe. The success of this venture transcends the realm of migration law. Other policy areas and factors beyond the direct control of state authorities are at least as important as the legislation governing entry and stay. Think of the welfare state, the labour market, education, local communities, and everyday societal contacts. A holistic analysis of migrant integration has to put the distinct outlook of European migration law into perspective by assessing how it interacts with other instruments and factors. Integration is a flexible concept that presents itself in different forms, ranging from the legislative conditions for entry and stay to social support services. The policy response comprises ‘hard’ measures that deliver tangible results, such as access to social benefits, as well as ‘soft’ components regarding the self-image of European societies. Moreover, integration is much more than an abstract label for state policies. Sociology and political theory have discussed the integration concept for decades, thereby unearthing different approaches to migrant integration. A problematic aspect of the notion of ‘integration’ concerns the underlying dichotomy between migrants and host states, even though migrants belong to the societies they live in. The notion of ‘settlement’ in the chapter title hints at this ideal of equal membership, which EU migration law supports through gradual status upgrade. Long- term residence serves as the endpoint of settlement for the purposes of EU legislation, since naturalisation and the acquisition of Union citizenship remain, by and large, the responsibility of the Member States. National and supranational approaches to migrant integration have evolved over time. The emergence of European migration law witnessed political disputes about the vision to be pursued between denial of permanent settlement, the distinct outlook of Union citizenship, and civic integration (15.1). EU Treaties limit the scope of the supranational intervention, while complementing migration law by the contribution of other policy areas (15.2). Equality clauses in secondary legislation are complex and subject to numerous exceptions, but they embody a baseline of equal treatment with nationals nonetheless (15.3). The protective thrust of equal treatment contrasts with the conditionality of mandatory integration requirements, notably language skills (15.4). Soft law measures and other policies, such as anti-discrimination legislation, complete the EU’s contribution (15.5). Long-term resident status presents a sort of ‘citizenship’ light, with enhanced residence security independent of the purpose of first admission (15.6), whereas the acquisition of nationality is governed by domestic laws predominantly (15.7). European Migration Law. Daniel Thym, Oxford University Press. © Daniel Thym 2023. DOI: 10.1093/oso/9780192894274.003.0016

470 Integration and Settlement

15.1 Theory and Policy Design The notion of ‘guest worker’ stands paradigmatically for the prevalent mode of thinking in the main destination countries in Europe after the Second World War. Migrants were not considered to be equal members of society, despite being given enhanced rights in many domains (15.1.1). Increasing acceptance of permanent settlement involves that different normative ideals of how ‘integration’ should unfold have gained relevance and will continue to cause political controversies (15.1.2). EU institutions were engulfed in a political fight over the road to be taken (15.1.3). The outcome is a mixture of measures and policies which combine various aspects of the theoretical integration models. Most migrants have the option of gradual status upgrade, culminating in long-term residence or naturalisation (15.1.4). Our analysis of the legislative framework should recognise ‘explicit’ and ‘implicit’ integration requirements, which, moreover, interact with diverse policies beyond the thematic scope of migration law (15.1.5).

15.1.1 From ‘denizenship’ to citizenship The self-image of most European societies was defined by demographic stability during the decades after the Second World War. The iron curtain and practical hurdles for intercontinental migration meant that comparatively few people entered Europe at the time. ‘Guest workers’ and late colonial migration were presented as an exceptional occurrence during that period, while earlier phases of enhanced mobility, traced in Chapter 1.1, were largely forgotten. Within Europe, emigration from the South to the industrial heartland further North gradually lost relevance. Demographic stability was particularly pronounced in Eastern Europe. Things changed, in the main destination countries at least, when many ‘guest workers’ stayed and brought their families along. ‘Zero immigration’ policies were officially maintained, but practical measures in support of integration were put in place nonetheless. The highly stratified character of modern societies entailed that migrants benefited from path- dependent inclusion. Social policies or education, for instance, projected their enabling and supportive characteristics onto migration.1 This trend was reinforced by the plurality of state actors. Local communities may actively support migrants, irrespective of the central government’s policy objectives.2 One need not be a system theorist to recognise that the state response to migration is multifaceted. Migration law is one element amongst others. Changes in migration laws supported the turn towards pragmatic inclusion. Behind the curtain of official zero immigration policies, domestic legislation quietly increased residence security and extended equal treatment in socioeconomic domains. Many migrants obtained enhanced rights, including permanent residence, without being treated as equal members of the societies they lived in. Nationality laws and societal self-perceptions were not altered fundamentally. Social scientists coined the term ‘denizenship’ to describe this

1 See ch 4.2.2. 2 See Ilke Adam and Tiziana Caponio, ‘Research on the Multi-Level Governance of Migration and Migrant Integration’ in Agnieszka Weinar and others (eds), The Routledge Handbook of the Politics of Migration in Europe (Routledge 2019) 26.

Theory and Policy Design 471 intermediate stage between ‘alienage’ and ‘citizenship’.3 A combination of official exclusion and pragmatic inclusion was widespread, notwithstanding differences: France granted nationality to many, while emphasising the ‘otherness’ of migrants indirectly;4 the UK and the Netherlands pursued multicultural policies, thereby excluding migrant communities from the societal mainstream;5 courts and legislation enhanced the legal status of migrants in Germany, short of naturalisation.6 The main destination countries experienced a simultaneity of status improvement and non-membership. Fierce debates after the millennium change can be rationalised, in part at least, as the response to a growing awareness that inward migration would continue, via the asylum system and as a result of demographic change. Moreover, the second and third generation started claiming equal membership. Social scientists speak of a ‘paradox of integration’: educational attainment and inclusion result in more—not less—opposition to discrimination and social hierarchies.7 The resurgence of cultural categories in debates about migrant integration responds, partly at least, to the transformation of ‘denizens’ into ‘citizens’. Designating the new compatriots as ‘Muslims’ undermined their normative claim to equal citizenship.8 Note that our use of the term ‘citizenship’ goes beyond nationality law. The English and French understanding of citizenship is broader and embraces equal membership, for instance during the civil rights movement.9 Indeed, contemporary debates about ‘migration’ increasingly transcend the remit of migration law. Think of political disputes and court judgments about the prohibition of the burka or the use of the headscarf, as well as mounting controversies about racism. From a legal perspective, the individuals concerned are often not migrants in the sense of people with a different passport. Their frequent designation as ‘migrants’ in the public discourse, or subtle forms of semantic othering through labels such as ‘migration background’, undermine equal membership. These debates leave their mark on EU migration law, in particular with regard to civic integration, but they are not intrinsically linked to it. Experts in migration law should recognise that it can even be counterproductive to highlight the significance of their area of expertise, since doing so unintentionally reinforces the conceptual distinction between ‘host societies’ and ‘migrants’.10 The same can be said about ‘integration’, especially when applied to the second and third generation.11 This book employs the term nonetheless, also considering that it is deeply ingrained in the public discourse at the national and EU levels. Moreover, our analysis is primarily concerned with the legal rules governing first admission and the initial period of residence. Many third country nationals residing for prolonged periods often have a comparatively secure residence status with enhanced equal treatment. Their claim to equal citizenship will oftentimes transcend questions of migration law. 3 See Tomas Hammar, Democracy and the Nation-State (Avebury 1990) 9–25. 4 See Patrick Weil, How to Be French (Duke UP 2008) ch 9. 5 See Will Kymlicka, Multicultural Odysseys (OUP 2007) chs 3–5. 6 See Daniel Thym, ‘Vom “Fremdenrecht” über die “Denizenship” zur “Bürgerschaft”’ (2018) 57 Der Staat 77, 85–98. 7 See Irene ten Teije, Marcel Coenders, and Maykel Verkuyten, ‘The Paradox of Integration’ (2013) 44 Soc Psych 278. 8 See generally Andreas Wimmer, Ethnic Boundary Making (OUP 2014). 9 See Richard Bellamy, Citizenship. A Very Short Introduction (OUP 2008). 10 See ch 4.1.2; and, mutatis mutandis, Adrian Favell, ‘Integration Nations’ in Grete Brochmann (ed), Multicultural Challenge (Emerald 2003) 13, 20–32. 11 See Willem Schinkel, Imagined Societies (CUP 2017).

472 Integration and Settlement

15.1.2 Models of migrant integration Integration is a basic concept of social theory which is often broken down into at least three distinct axes of analysis. Structural integration refers to equal participation in essential segments of modern society: the labour market, housing, education, the health system, and the welfare state. Academic and policy debates often use the notions of ‘inclusion’ and ‘incorporation’ to designate structural integration. Corresponding analyses habitually transcend formal access on the basis of legislative equal treatment by asking whether migrants, especially among the second and third generation, achieve equal results.12 Social integration involves establishing social contacts, within or across ethnic boundaries, and identificatory patterns, including an emotional bond.13 Cultural integration concerns learning the norms and language of the receiving society; it fluctuates between cognitive knowledge and shared values.14 Such analytical distinction prepares the ground for sector-specific analyses and the contextualisation of EU migration law throughout this chapter. The objective of structural integration is habitually widely shared. Policy debates revolve about appropriate mechanisms of inclusion and exceptions for those whose claim to membership is disputed, for instance irregular presence or temporary work. The degree of social integration is mostly beyond the direct reach of state measures and legal obligations, whereas cultural integration has been the focal point of recent controversies. In this respect, we may distinguish three conceptions of integration which, as ideal-typical categorisations, accentuate distinct features, while overlapping in practice.15 First, the historic model of migrant assimilation to the structural specificities and the way of life of the receiving society dominated the US literature until the civil rights movement.16 In Europe, it often coincided with statist conceptions of ethno-cultural closure, for example in the German debate until reunification and the principled rejection of migration from outside Europe in some Visegrád countries. In a symbolic move against such views, the Court rejected the argument of the Polish government that relocating asylum seekers would be incompatible with the national identity of a ‘virtually ethnically homogeneous’ nation.17 Secondly, integration as equality concentrates on structural integration and perceives of cultural integration primarily as a means to an end. Language skills, for instance, are accepted insofar as they are a precondition for equal access to the labour market. Integration as equality often goes hand in hand with multicultural visions of cultural diversity on the basis of universal human rights.18 Most experts in migration law tend to support such equality-based normative ideals intuitively. We shall see to what extent integration as equality informs the supranational case law. 12 See Alejandro Portes and Rubén G Rumbaut, Legacies (UC Press 2001). 13 See ch 4.2.3; and Alastair Ager and Alison Strang, ‘Understanding Integration’ (2008) 21 J Refug Stud 166. 14 See Richard Alba and Victor Nee, Remaking the American Mainstream (Harvard UP 2003). 15 See generally Hiroshi Motomura, Americans in Waiting (OUP 2006); and Sarah Song, ‘Three Models of Civic Solidarity’ in Rogers M Smith (ed), Citizenship, Borders, and Human Needs (University of Pennsylvania Press 2011) 192. 16 See Milton M Gordon, Assimilation in American Life (OUP 1964). 17 See Joined Cases C-643/15 and C-647/15 Slovakia and Hungary v Council EU:C:2017:631, paras 302–5; and Ségolène Barbou des Places, ‘L’empreinte des nationalismes sur le droit de l’immigration de l’Union européenne’ [2018] Revue trimestrielle de droit européen 725, 730–38. 18 See Clíodhna Murphy, Immigration, Integration and the Law (Ashgate 2013).

Theory and Policy Design 473 A third position emphasises the significance of social affiliation beyond equal treatment. Social contacts are held in high regard, as are shared values and a feeling of ‘togetherness’. This third pattern differs from assimilation, since shared values can be construed in a forward-looking manner through the normative reconfiguration of societal self- perceptions. It would be reductionist to present ethno-cultural closure as the only alternative to integration as equality.19 Defining the—necessarily relative—degree of ‘unity in diversity’ remains notoriously challenging, and the example of mandatory integration requirements will demonstrate that motivations can overlap. Nevertheless, these difficulties do not unmake the theoretical soundness of an argument which recognises migrants as equal members in a society, which is more than the sum of the individualistic parts.20 Chapter 4.1 referred to the rich normative debate on these matters.

15.1.3 Disputes over the direction of EU migration law Contemporary controversies came after decades of broad agreement that structural integration should be encouraged, at least for those staying legally. Member States did not object to the idea of pragmatic inclusion as such but opposed supranational intervention. They rejected several initiatives the Commission had put forward ever since the end of the ‘guest worker’ programmes.21 By way of example, third country nationals were excluded from a directive on schooling, which, tellingly, fostered teaching of the mother tongue and culture of the country of origin.22 Migrants were integrated as ‘denizens’, in line with previous comments, without becoming equal members. Intergovernmental debates similarly addressed the ‘challenge of integration’, albeit in abstract terms and without initiatives for common action.23 Things changed when the Treaty of Amsterdam prepared the ground for legislative harmonisation, and the timing helps to explain why the discussion turned controversial. To start with, the seminal Tampere Conclusion promoted a ‘more vigorous integration policy’ for third country nationals staying legally that ‘should aim at granting them rights and obligations comparable to those of EU citizens’.24 Commission took this declaration of intent seriously and made ambitious proposals for family reunification and long-term residence. Residence security, equal treatment, and family reunification were presented as a means to an end, to foster integration.25 Union citizenship appeared as the prototype of post- national membership.26 In a symbolically important move, the Commission pondered the 19 cfMoritz Jesse, The Civic Citizens of Europe (Brill/Nijhoff 2017) ch 2; and Anuscheh Farahat, Progressive Inklusion (Springer 2014) ch 4. 20 See Christian Joppke, Citizenship and Immigration (Polity Press 2010) ch 4. 21 See ch 14.1.1; and Sarah Ganty, L’intégration des citoyens européens et des ressortissants de pays tiers en droit de l’Union européenne (Bruylant 2021) 532–80. 22 See Council Directive 77/486/EEC on the education of the children of migrant workers [1977] OJ L199/32, arts 1–3. 23 See Commission, ‘Report: The social integration of third-country migrants residing on a permanent and lawful basis’ SEC(89) 924 final; and Commission, ‘Communication on immigration and asylum policies’ COM(1994) 23 final, 32–41. 24 European Council, ‘Presidency Conclusions of the Meeting in Tampere’ (15–16 October 1999) No 18. 25 See Kees Groenendijk, ‘Legal Concepts of Integration in EU Migration Law’ (2004) 6 EJML 111, 113–17. 26 See ch 4.2.2; Joseph HH Weiler, ‘To Be a European Citizen. Eros and Civilization’ (1997) 4 JEPP 495; and Jo Shaw, ‘Citizenship of the Union’ in European University Institute (ed), Collected Courses of the Academy of European Law 1995, vol VI-1 (Kluwer 1998) 237, 278–96.

474 Integration and Settlement ‘development of the concept of civic citizenship’; the idea rested upon human rights, cultural plurality, and the option of extending political rights to long-term residents.27 For a moment, it seemed as if nationality was an idea of the past. Member States disagreed. Integration had become a salient issue in the main destination countries after the millennium change, when it became evident that migration would transform societies permanently.28 This meant that Union citizenship was not regarded as a model any longer, also considering that it was often perceived of, among the elite in continental Europe, as a proto-federal status that should be distinguished from third country nationals.29 The unanimity requirement entailed that the main destination countries could effectively engrave their views on the final text of the Family Reunification Directive 2003/ 86/EC and the Long-Term Residents Directive 2003/109/EC.30 The restrictive elements of both directives embraced a vision of integration as a precondition for status security and equal treatment.31 Individual rights were the reward for integration, not a means of fostering inclusion; the earlier idea of civic citizenship was abandoned. Acquisition of nationality appeared as the ultimate trophy.

15.1.4 Status change in multiple directions EU migration law construes the legal dimension of migrants’ biographies as a sequence of events involving the acquisition of different migration statuses. Primary law accepts this gradual approach, as indicated by the objective of managing migration flows ‘at all stages’.32 Legislation may provide for a ‘short-stay residence permit’,33 which may lead towards ‘long- term residence’34 or ‘unauthorised residence’ and ‘removal’.35 Careful distinction between these scenarios in Articles 77–79 TFEU specifies that migration law is not about either one- time settlement or indefinite rejection but about a differentiated and selective admission process.36 For many, the initial temporary residence permit is the first step in the direction of status upgrade, culminating in the acquisition of long-term residence or naturalisation. Chapter 4.3.2 explained that the element of conditionality inherent in phased admission serves statal control imperatives. The gradual approach of EU migration law contrasts with the classic position of the United States, Canada, and Australia. These long-established settler societies had traditionally distinguished categorically between ‘immigrants’ with the prospect of naturalisation and ‘non-immigrants’ with a temporary status, even though that binary model has

27 See Commission, ‘Communication on a Community immigration policy’ COM(2000) 757 final 19; and the shift towards national prerogatives and naturalisation by Commission, ‘Communication on immigration, integration and employment’ COM(2003) 336 final, 30. 28 See ch 14.1.2; and Christian Joppke, Immigration and the Nation State (OUP 1999). 29 See ch 1.1.7–8; and Neil Walker, ‘Denizenship and Deterritorialisation in the European Union’ in Hans Lindahl (ed), A Right to Inclusion and Exclusion? (Hart Publishing 2009) 261. 30 See Sergio Carrera, In Search of the Perfect Citizen? (Martinus Nijhoff 2009) ch 4. 31 See Groenendijk (n 25) 117–24; and Karin de Vries, ‘The Integration Exception’ in Daniel Thym (ed), Questioning EU Citizenship (Bloomsbury/Hart Publishing 2017) 267. 32 TFEU, art 79(1). 33 ibid art 77(2)(a). 34 ibid art 79(2)(a). 35 ibid art 79(2)(c). 36 See Daniel Thym, Migrationsverwaltungsrecht (Mohr Siebeck 2010) 18–24.

Theory and Policy Design 475 increasingly given way to a gradual system in the European tradition.37 Against this background, the use of the term ‘settlement’ in the chapter title does not imply that EU migration law follows the earlier Anglo-Saxon example. Rather, the word settlement was chosen as an overarching linguistic category that can cover status improvement in the form of long-term residence and naturalisation. While the notion of settlement is not widely used in Europe, it has occasionally been employed by the CJEU and the ECtHR to designate a foreigner who has taken roots in a country.38 Most residence permits are issued for short periods first and can be renewed, if need be repeatedly, provided applicants continue fulfilling entry conditions. EU legislation recognises that individual rights to be granted a residence permit cover renewal.39 Even seasonal work and Schengen visas, which are designed for short-term stays, do not prevent Member States from extending residence.40 Moreover, third country nationals may switch statuses. Students, for instance, are given a residence permit for job-searching after completion of their studies.41 Provisions in legislation on the scope rationale personae define who can apply for the permit in question. Purpose-specific statuses converge into long-term residence as an overarching permit, which is no longer dependent, as we shall see, on sectoral entry conditions. It presents as a sort of ‘citizenship light’42 that brings about enhanced residence security and socioeconomic opportunity, mirroring earlier forms of ‘denizenship’ in many European countries. One step further, third country nationals may acquire the nationality of a Member State; doing so turns them into Union citizens. From a legal perspective, acquisition of nationality is usually the endpoint of status change, notwithstanding the potential of deprivation in some countries. While some foreigners are naturalised comparatively quickly, others will retain different residence permits for life. Factual challenges of integration and discrimination may persist, as we have seen, after naturalisation.

15.1.5 Mixed output in migration law and beyond Legal debates habitually concentrate on questions that give rise to legal disputes, for instance pre-departure language tests as a precondition for family reunification. Use of the term ‘integration’ in the legislative text reinforced the impression that the EU’s migrant integration policy essentially revolves around restrictions. However, such a conclusion would be too simple. In light of the broad understanding of integration, we may distinguish ‘implicit’ from ‘explicit’ provisions dealing with migrant integration.43 Implicit provisions do not use the term integration and rely on proxies for structural integration instead, such as sufficient resources, criminal convictions, or accommodation. Moreover, third country 37 See Catherine Dauvergne, The New Politics of Immigration and the End of Settler Societies (CUP 2016) 125–34. 38 See Üner v Netherlands App no 46410/99 (ECtHR [GC], 18 October 2006) § 59; and Case C-432/20 Landeshauptmann von Wien EU:C:2022:39, paras 36–37. 39 See ch 10.1.2; and, by way of example, Blue Card Directive (EU) 2021/1883, art 8(2); and Family Reunification Directive 2003/86/EC, art 13(2). 40 See Seasonal Workers Directive 2014/36/EC, art 14(1); and Visa Code Regulation (EC) No 810/2009, art 25(1) (b), with later amendments. 41 See ch 14.4.3. 42 Joppke (n 20) ch 5. 43 See Ilke Adam and Daniel Thym, ‘Integration’ in Philippe De Bruycker and others (eds), From Tampere 20 to Tampere 2.0 (EPC 2019) 73, 75; and Jesse (n 19) chs 4–9.

476 Integration and Settlement nationals benefit from equal treatment clauses. A holistic account of the legislation will find a combination of restrictive integration requirements and supportive guarantees. In doing so, EU legislation routinely combines diverse elements of the theoretical conceptions of integration in a mixed approach. Inspection of the case law indicates that the ECtHR and the CJEU similarly produce mixed results.44 The Court in Strasbourg considers the ‘solidity of social and cultural ties’ to be an important factor when assessing the proportionality of expulsion.45 Judges in Luxembourg found, as we shall see, language requirements to ‘encourage interaction and the development of social relations’. Rulings on the Long-Term Residents Directive embody a baffling elasticity of the implications of the assertion that the ‘the principal purpose’46 is integration.47 Depending on the outcome, judges present individual rights as a means of integration, for instance in the form of access to family benefits, or as a reward for successful integration, notably in the form of mandatory language classes.48 On one occasion, the Court found the acquisition of the nationality to represent ‘the most accomplished level of integration’49. Judges blend elements of the theoretical models into a seemingly uniform idea of integration, without reflecting the differences. A holistic analysis of the EU’s approach has to take into account the contribution of other policy domains. Social policy and anti-discrimination directives contrast with the conditionality of migration law. The Common Basic Principles and the Commission’s Action Plans mention migration law as one elements amongst others besides classic manifestations of structural integration, such as inclusion in the labour market, housing, or education.50 As a result, any evaluation of the EU’s approach essentially depends on the thematic scope of the measures analysed. When focusing exclusively on explicit integration conditions in secondary legislation, a one-sided focus on civic integration becomes apparent. By contrast, a more balanced conclusion will be reached by those considering implicit integration provision and other policy areas.

15.2 Constitutional Foundations The broad scope of migrant integration policies, ranging from education to non- discrimination, entails that migration law is one element amongst many others. Supranational competences are confined to coordination and support, the design of the migration law instruments, and recourse to other Union policies (15.2.1). The ECtHR has established human rights boundaries for expulsion, which are particularly strict for settled migrants (15.2.2). Readers are also reminded of the special status of Ireland and Denmark (15.2.3). 44 See Daniel Thym, ‘Supranational Courts in Europe. A Moderately Communitarian Turn in the Case Law on Immigration and Citizenship’ (2021) 47 JEMS 4534. 45 See Üner v Netherlands (n 38) § 58. 46 Case C-508/10 Commission v Netherlands EU:C:2012:243, para 66. 47 See also Alessandra Lang, ‘Social Integration’ (2018) 3 European Papers 663. 48 See ch 6.4.3; Case C-303/19 Istituto Nazionale della Previdenza Sociale EU:C:2020:958, paras 28–29; and Case C-579/13 P and S EU:C:2015:369, paras 44–47. 49 Case C-171/13 Demirci and others EU:C:2015:8, para 54. 50 See ‘Common Basic Principles for Immigrant Integration Policy in the European Union’ (Council doc 14615/ 04, 19 November 2004); and Commission, ‘Communication: Action plan on integration and inclusion 2021–2027’ COM(2020) 758 final.

Constitutional Foundations 477

15.2.1 Scope of supranational competences Member States have long rejected extensive supranational intervention in integration policies. Instead, the Treaty of Lisbon established a power ‘to provide incentives and support for the action of Member States . . . excluding any harmonisation’. The exclusion of harmonisation entails that Article 79(4) TFEU classifies as a support and coordination competence in the meaning of Article 2(5) TFEU. The provision responded to heated disputes about EU intervention. A Commission initiative to apply the open method of coordination was rejected.51 In an act of symbolic opposition, the Common Basic Principles were adopted by the Council acting as an EU institution and as an intergovernmental setting.52 Today, such complex construction would no longer be necessary. Coordination under Article 79(4) TFEU may cover all aspects of integration policy. Nevertheless, the follow-up resolution was adopted as a hybrid act as well.53 Other measures the Union may adopt include statistical indicators, national contact points, and funding. The prohibition of harmonisation is confined to Article 79(4) TFEU. Settled case law that the institutions can adopt legally binding measures under recourse to other legal bases, provided that centre of gravity justifies the activation of other Treaty articles.54 By way of example, Article 79(2)(a) and (b) TFEU cover explicit and implicit integration provisions in legislation on entry and stay, such as the need for sufficient resources or integration conditions. When using this competence, subsidiarity compels the EU to refrain from an in-depth intervention, such as common curricula for integration courses. Supplementary legal bases include non-discrimination and social policy under Articles 19 and 153 TFEU. Recourse to alternative legal bases does not render the limitation enshrined in Article 79(4) TFEU superfluous. The provision excludes any harmonisation of those facets of structural, social, and cultural integration which are not covered by an explicit authorisation elsewhere in the Treaties.

15.2.2 Human rights protection against expulsion Chapter 14.2.3 explained that neither the right to family life nor the best interests of the child usually bring about a generic human rights guarantee to be admitted to the territory. By contrast, states have less leeway when expelling people, especially after years of residence. In a case concerning the separation of a Moroccan ‘guest worker’ from his Dutch daughter, the ECtHR first declared an expulsion to be incompatible with Article 8 ECHR.55 Ever since, human rights have provided a crucial safety net against expulsion. When justifying expulsion, states have to strike a fair balance by considering multiple factors: the nature of potential criminal offences; time elapsed since an offence, and conduct thereafter; length 51 See Commission, ‘Communication on an open method of coordination for the community immigration policy’ COM(2001) 387 final 4–6; Carrera (n 30) 55–58; and Samantha Velluti, ‘What European Union Strategy for Integrating Migrants?’ (2007) 9 EJML 53, 68–80. 52 See ‘Common Basic Principles’ (n 50). 53 See ‘Conclusions of the Council and the Representatives of the Governments of the Member States on the Integration of Third-Country Nationals Legally Residing in the EU’ (Council doc 9905/1/14, 26 May 2014). 54 See Joined Cases C-620/18 and C-626/18 Hungary and Poland v Parliament and Council EU:C:2020:1001, paras 65–69, 77–85. 55 See Berrehab v Netherlands App no 10730/84 (ECtHR, 21 June 1988).

478 Integration and Settlement of stay; solidity of social, cultural, and family ties with countries of residence and origin; age of the applicant; the best interests of children; the applicant’s family situation, and the nationalities of the persons involved; whether spouses knew about the instability of the migration status when starting the relationship; and difficulties they might encounter upon return.56 Whenever someone was born in a country, or moved there in early childhood, only the first four criteria shall be considered, thereby effectively increasing the level of protection.57 States benefit from a margin of appreciation when assessing specific scenarios,58 and the best interests of the child do not categorically prohibit deportation of one parent.59 International judges will not micromanage state practices. Although residence security was a driving force behind the case law, the second and third generation is not immune from forced removal; only naturalisation brings about an absolute protection against expulsion.60 Judges also confined the protective reach of Article 8 ECHR to the nuclear family in regular circumstances, while simultaneously protecting settled migrants under the heading of ‘private life’. They defined private life as ‘the network of personal, social and economic relations that make up the private life of every human being’, irrespective of relations with family members.61 Doing so obliges states to justify the expulsion of settled migrants in light of the above-mentioned criteria.62 There are no clear-cut quantitative and qualitative standards (length of stay, nature of personal contacts) to determine when expulsion amounts to an interference with private life, thus triggering the need to justify expulsion under human rights law. Four our purposes, the threshold for the application of the private life benchmark has limited relevance, since the ‘public policy’ exception, discussed in Chapter 10.4 on general features of secondary legislation, will often bring about a higher level of protection. It can be found in most EU migration law instruments as a precondition for refusing, withdrawing, or revoking residence permits. However, Article 8 ECHR cannot usually be relied upon to obtain a ‘better’ residence permit.63 Private and family life serve as a safety net against expulsion and do not oblige states to guarantee additional rights, for as long as they refrain from removal.

15.2.3 Territorial scope Chapter 2.4 gave details about the special status of Denmark and Ireland. Neither country decided to join the Long- Term Residents Directive 2003/ 109/ EC. Non- participation of Denmark in the extension of social security coordination is problematic from a 56 See Boultif v Switzerland App no 54273/00 (ECtHR, 2 August 2001) § 48; and Üner v Netherlands (n 38) §§ 57–58. 57 See Maslov v Austria App no 1638/03 (ECtHR [GC], 23 June 2008) § 71. 58 See Savran v Denmark App no 57467/15 (ECtHR [GC], 7 December 2021) §§ 187–89. 59 See ch 5.4.2; and, by way of example, Antwi and others v Norway App no 26940/10 (ECtHR, 14 February 2012) §§ 94ff. 60 See Üner v Netherlands (n 38) §§ 55–56; and Maslov v Austria (n 56) §§ 74–75. 61 See Slivenko and others v Latvia App no 48321/99 (ECtHR [GC], 9 October 2003) § 97. 62 Those relating to family unity are usually irrelevant, thus giving greater weight to the other benchmarks; see Daniel Thym, ‘Residence as de facto Citizenship?’ in Ruth Rubio Marín (ed), Human Rights and Immigration (OUP 2014) 106, 113–17, 125–26. 63 See Sisojeva v Latvia App no 60654/00 (ECtHR [GC], 15 January 2007) § 91; and further Daniel Thym, ‘Respect for Private and Family Life under Article 8 ECHR in Immigration Cases’ (2008) 57 ICLQ 87, 98–99.

Equality Provisions in Secondary Legislation 479 Table 15.1 Participation in the most important legislation on migrant integration Instrument

Ireland /UK (before Brexit)

Denmark

Racial Equality Directive 2000/43/EC

yes

yes

Long-Term Residents Directive 2003/ no 109/EC

no

Social Security Coordination Regulation (EC) No 883/2004, as extended to legally resident third country nationals by Regulation (EU) No 1231/2010

no

yes /no

constitutional perspective. Chapter 14.2.1 on legal migration explained why the regulation extending the single market rulebook to third country nationals should have been adopted on a different legal basis. Since the opt-outs are confined to measures building the area of freedom, security, and justice, all the Member States participate equally in Union legislation dealing with other aspects of migrant integration beyond the scope of EU migration law, as demonstrated by Table 15.1.

15.3 Equality Provisions in Secondary Legislation Equal treatment is a central axis of the theoretical models of migrant integration. It features prominently in the EU’s migration law instruments and has given rise to several Court judgments. Our analysis of the equality clauses is complicated by excessive fragmentation (15.3.1). Court rulings have highlighted the protective credentials of these provisions, thus mitigating the focus of the legislative debate on exceptions (15.3.2). Most directives cover social benefits to a limited degree only, thus leaving full integration to national legislation (15.3.3). Furthermore, the practical effects of equal treatment can be reduced by the structure of the welfare state and in cross-border scenarios (15.3.4). Distinctions between refugees and beneficiaries of subsidiary protection exemplify the potential of human rights law to deliver further equality (15.3.5). Case law on the human right to equality before the law was presented in Chapter 10.5.

15.3.1 Complexity of the legislative framework Each equality provision is unique and reflects the outcome of the legislative negotiations, which were conducted by changing protagonists at different moments in time.64 The end

64 See Temporary Protection Directive 2001/55/EC, arts 12–14; Long-Term Residents Directive 2003/109/EC, art 11; Family Reunification Directive 2003/86/EC, art 14; Qualification Directive 2011/95/EU, arts 20–34; Single Permit Directive 2011/98/EU, art 12; Seasonal Workers Directive 2014/36/EC, art 23; ICT Directive 2014/66/EU, art 18; Students and Researchers Directive (EU) 2016/801, art 22; and Blue Card Directive (EU) 2021/1883, art 16.

480 Integration and Settlement product presents us with a complex and fragmented overall picture of sector-specific guarantees. Differences concern the subject areas covered, the scope of optional derogations, and the definition of the terminology.65 By way of example, tax benefits are covered by many instruments but are not mandatory, qua Union law, for intra-corporate transferees and blue card holders. Equal treatment in access to housing services is available for long-term residents and blue card holders, not, however, for seasonal workers and intra-corporate transferees, whereas an optional derogation exists under the Students and Researchers Directive and the Single Permit Directive. A certain degree of cross-sectoral harmonisation was brought about by Article 12 Single Permit Directive 2011/98/EU. The previous chapter explained that this provision applies to all third country nationals with labour market access, irrespective of the purpose of first admission, although some migration statuses are excluded, for instance beneficiaries of international or temporary protection.66 In its recast proposal, the Commission refrained from recommending an extension.67 For students, the original option not to apply the Single Permit Directive may have been supplanted by the explicit reference to the instrument in Article 22(3) Students and Researchers Directive (EU) 2016/801.68 In other scenarios, sectoral legislation remains relevant beside the Single Permit Directive, for the simple reason that it offers a higher level of protection. Think of unconditional access to housing services under the Blue Card Directive. Equality provisions cover subject matters that are enumerated in the articles. The legislature opted against an overarching equality guarantee, in contrast to the treatment of Union citizens under Article 18 TFEU. While the complexity of the legislative framework can be frustrating, some divergences are based on persuasive arguments, reflecting the factual and legal context. Family benefits or study grants, for instance, are often financed from the general budget and are not means-tested. To limit such redistributive payments to those staying for longer periods, as some migration law instruments do, can be a legitimate social policy choice.69 First admission need not necessarily bring about full equal treatment immediately. Having said this, some discrepancies can only be rationalised by the time factor. Even though the long-term residence aimed at enhancing protection, long-term residents receive less equal treatment than other migrants under Articles 3(2)(i) and 12 Single Permit Directive 2011/98/EU. In its recast proposal, the Commission suggests aligning the level of protection, not least with regard to the export of family benefits for family members residing abroad.70 The amendment would effectively reverse a Court judgment, which had highlighted that the Single Permit Directive offers a higher level of protection.71 At the same time, the Commission does not propose extending the equal treatment catalogue to new subject matters. 65 See Bjarney Friðriksdóttir, What Happened to Equality? (Brill/Nijhoff 2017) 146–54, 210–22, 252–59, 292–308, 315–20; and Herwig Verschueren, ‘Employment and Social Security Rights of Third-Country Labour Migrants under EU Law’ (2016) 18 EJML 373, 376–98. 66 See Single Permit Directive 2011/98/EU, art 3(2)(d)–(g). 67 See Proposal for a Recast of the Single Permit Directive COM(2022) 655 final, arts 4, 12. 68 The provision refers to Single Permit Directive 2011/98/EU, art 12(1), (4), without indication how this references relates to the exclusion of students under Article 3(3) of the said Directive; this exception can be construed as a lex anterior, supplanted by the newer instrument, or as a lex specialis not affected by later legislation. 69 See Joseph H Carens, ‘Live-in Domestics, Seasonal Workers, Foreign Students, and Other Hard to Locate on the Map of Democracy’ (2008) 16 J Pol Phil 419; and Floris de Witte, Justice in the EU (OUP 2015). 70 See Proposal for a Recast of the Long-Term Residents Directive, COM(2022) 650 final, art 12(2). 71 See Case C-302/19 Istituto Nazionale della Previdenza Sociale EU:C:2020:957, paras 38–39.

Equality Provisions in Secondary Legislation 481

15.3.2 Overarching doctrinal characteristics Anyone reading the equal treatment provisions is bound to notice numerous exceptions allowing Member States to withhold or restrict equal treatment in specific scenarios. While some derogations are widely used, for instance on the access to study grants, others are less relevant. I am unaware of any study assessing systematically whether Member States use the derogations in practice. With regard to Article 12 Single Permit Directive 2011/98/EU, the Commission found a mixed picture: while some countries had activated not a single optional clause, other derogations were relied upon by roughly half of the Member States.72 Optional derogations are not, therefore, applied universally. Moreover, domestic legislation will often grant additional equal treatment. Chapter 17 will mention that enhanced protection may be available under association agreements. One step further, judges have interpreted the derogations strictly. The terminology must be interpreted uniformly,73 and an activation of the derogations presupposes that ‘the authorities in the Member State concerned . . . have stated clearly that they intended to rely on them’.74 While applying this standard to national legislation adopted after the entry into force of a Directive may be comparatively straightforward, it also covers pre-existing laws. On two occasions, judges censured Italy for national laws which do not give equal treatment and which had existed before the adoption of a Directive for the sole reason that the national legislature had not actively restated the desire to maintain the exception.75 Silence is not enough. Chapter 10.5 explained that the human right to equality before the law does not require equal outcomes whenever two situations are not comparable. Moreover, unequal treatment can justified on the basis of legitimate reasons and a proportionality assessment. In two judgments, the CJEU projected these doctrinal characteristics on the equal treatment provisions in Articles 29 and 33 Qualification Directive 2011/95/EU. Restrictions on the free choice of the place of residence for beneficiaries of subsidiary protection may be justified by the objective to facilitate integration.76 A reduction in social benefits for refugees during the first three years of residence was found to be illegal. Nevertheless, the Court reiterated that a different outcome could have been defended if legitimate reasons had required so in a proportionate manner.77 To accept an option of justification assumes that the legislature perceived of equal treatment as an obligation of result. It might be relevant that both rulings concerned legislative provisions that replicate equal treatment guarantees under the Refugee Convention, which, in accordance with human rights law, must probably be interpreted to authorise states to justify unequal treatment.78 It is a different matter whether this conclusion extends to equal treatment clauses in the EU’s legal migration instruments, which explicitly allow Member 72 Commission, ‘Report on the Implementation of the Single Permit Directive’ COM(2019) 160 final, 9–11. 73 See ch 6.4.4. 74 Istituto Nazionale della Previdenza Sociale (n 48) para 23; building upon Case C- 571/ 10, Kamberaj EU:C:2012:233, para 87. 75 Istituto Nazionale della Previdenza Sociale (n 48) paras 37–38; and Case C-350/20 INPS EU:C:2021:659, para 64. 76 See Joined Cases C-443/14 and C-444/14 Alo and Osso EU:C:2016:127, paras 54–64. 77 See Case C-713/17 Ayubi EU:C:2018:929, paras 30–34. 78 See James Hathaway, The Rights of Refugees under International Law (2nd edn, CUP 2021) 1018–23; and also Jean-Yves Carlier, ‘Droit d’asile et des réfugiés: de la protection aux droits’ (2007) 332 Recueil des cours 9, 270–87.

482 Integration and Settlement States to derogate in specific scenarios. Judges have not addressed this question explicitly, even though the case law could be interpreted to have rejected the option of unwritten justification standards implicitly.79 Doing so would effectively conceive of Article 12(2) Single Permit Directive 2011/98/EU as a lex specialis that cuts off options for justification that would exist under human rights law. Case law on Union citizens points in a different direction: judges interpreted legislative equal treatment guarantees to embrace an unwritten option of justification.80 Be it as it may, there is nothing in the wording or the case law indicating that secondary legislation forbids factual outcome differentials. Equality provisions require formal equality, not the substantive convergence of outcomes on the ground.81 They cannot be relied upon by migrants who do not succeed in the labour market, since their diplomas were not recognise as equivalent.

15.3.3 Social security other benefits An essential guarantee is equal treatment with regard to the ‘branches of social security, as defined in Article 3 of Regulation (EC) No 883/2004’.82 Reference to the Social Security Coordination Regulation No 883/2004 supports legal certainty. Settled, albeit complex, case law on the definition of social security informs the scope of the equality provisions. Social security in this sense concerns access to benefits on the basis of a legally defined position without any assessment of personal needs, beside the option of taking into account the recipient’s resources.83 It is irrelevant whether the social security benefit is financed from the general budget or an insurance schemes, as long as it concerns one of the risks enumerated in Article 3(1) Social Security Coordination Regulation. That list is broad and embraces, amongst others, sickness, maternity, invalidity, old-age benefits, unemployment, death grants, and family benefits.

15.3.3.1 Exclusion from social assistance and advantages Notwithstanding the broad scope of ‘social security’, benefits who not qualify as such are not covered by the equality provisions. A comparison with judgments on the notion of ‘social and tax advantages’ under Article 7(2) Regulation (EU) No 492/2011 on freedom of movement for workers shows what kind of benefits are beyond the scope of the legal migration instruments: concessionary public transport fares, reduced admission charges for cultural events, or subsidised meals for school children.84 The Council rejected giving long- term residents the same treatment.85 Article 6(1) ILO Convention No 97 does not go much

79 See Istituto Nazionale della Previdenza Sociale (n 48) paras 32–38; and Case C-462/20 ASGI and others EU:C:2021:894, paras 36–38. 80 See Case C-20/12 Giersch and others EU:C:2013:411, para 35. 81 See also Emanuela Pistoia, ‘Social Integration of Refugees and Asylum Seekers through the Exercise of Socio- economic Rights’ (2018) 3 European Papers 781, 800–04. 82 By way of example see Seasonal Workers Directive 2014/36/EC, art 23(1)(d). 83 See INPS (n 75) paras 52–56; Case C-449/16 Martinez Silva EU:C:2017:485, paras 19–22; and Maximilian Fuchs and Roberts Cornelissen (eds), EU Social Security Law (CH Beck/Nomos/Hart Publishing 2015). 84 cf the Commission’s Memorandum accompanying the Proposal for a Long-Term Residents Directive, COM(2001) 127 final, 22. 85 ibid art 12(1)(f).

Equality Provisions in Secondary Legislation 483 further if we interpret the prohibition of discrimination on grounds of nationality in analogy to settled ECtHR case law.86 A comparison with Article 11(1)(d) Long-Term Residents Directive 2003/109/EC reveals another limitation of the equal treatment provisions in many other instruments: ‘social assistance’ is mentioned beside social security. To be sure, both categories are to be ‘defined by national law’, thus authorising Member States to introduce their own, and potentially narrower definition.87 Notwithstanding this caveat, which the Commission wants to abolish,88 equal treatment in the domain of social assistance is unique. All the other directives on legal migration do not cover ‘social assistance’ in the form of minimum income support to meet essential living costs. Non-contributory cash benefits under Article 70 Social Security Coordination Regulation (EC) No 883/2004 are probably excluded from the notion of ‘social security’ as well, since they are—notwithstanding their hybrid character—a distinct category of their own.89 Excluding equal treatment in the domain of social assistance can be presented as the mirror image of the sufficient resources requirement, which the directives on legal migration habitually foresee. Indeed, beneficiaries of international protection benefit from enhanced guarantees: refugees receive equal treatment in social assistance, while support to beneficiaries of subsidiary protection can be limited to ‘core benefits’.90 The meaning of ‘assistance in terms of social welfare and means of subsistence’ in situations of temporary protection remains unclear, since the wording does not take up established supranational legal categories.91 Member States may also limit social assistance, in contrast to social security payments, for long-term residents to ‘core benefits’. Judges interpreted core benefits to cover support which ‘enable[s]‌individuals to meet their basic needs such as food, accommodation and health’.92 Although the notion of ‘core’ benefits can be read to designate a minimum threshold available to anyone,93 judges concluded that state support transcending minimum guarantees may still qualify as a core benefit, thus preventing Member States from derogating from it.94 Doing so reined in the option of reduction under Article 11(4) Long- Term Residents Directive 2003/109/EC, which the Commission proposes to discontinue. Some benefits are covered by leges speciales. Equal treatment with regard to tax benefits, study grants, and access to public goods and services, including housing, is mentioned in several directives, although Member States are often given leeway to derogate.95 In other domains, Member States remain free to grant or withhold equal treatment in accordance with domestic laws, which, generally speaking, have embraced residence as the door-opener for 86 See Migration for Employment Convention [No 97] (adopted 1 July 1949, entered into force 22 January 1952) 120 UNTS 1616; and Ryszard Cholewinski, Migrant Workers in International Human Rights Law (Clarendon Press 1997) 104–07. 87 See Kamberaj (n 74) para 77; and Daniel Thym, ‘Long-Term Residents Directive 2003/109/EC’ in Daniel Thym and Kay Hailbronner (eds), EU Immigration and Asylum Law: Article-by-Article Commentary (3rd edn, CH Beck/Hart Publishing/Nomos 2022) Article 11 MN 23–27. 88 See LTRD Recast Proposal (n 70), art 11(1)(d). 89 See Herwig Verschueren, ‘The Labyrinth of Employment and Social Rights in the EU Intra-Corporate Transfer Directive’ (2021) 12 Eur Labour LJ 280, 293–95. 90 See Qualification Directive 2011/95/EU, art 29(1), (2). 91 See Temporary Protection Directive 2001/55/EC, art 13. 92 Kamberaj (n 74) para 91. 93 It was explained in ch 13.2.2 that the ‘basic needs’ formula equally features in the judgments on CFR, art 4. 94 See Case C-94/20 Land Oberösterreich EU:C:2021:477, paras 40–44. 95 By way of example see Single Permit Directive 2011/98/EU, art 12(1)(c), (f), (g), (2)(a), (c), (d).

484 Integration and Settlement social benefits after the Second World War, at least for those staying legally.96 Silence on the part of the EU legislature need not translate, therefore, into exclusionary practices on the ground.

15.3.3.2 Human rights compliance The Court has confirmed that exemptions must comply with the Charter.97 Nevertheless, it remains a formidable challenge to challenge unequal treatment in light of human rights. Chapter 10.5 explained that the ECtHR exercises enhanced supervision in some domains, which mostly qualify as social security in the above sense and are subject, therefore, to equal treatment anyway. That is not to say that Article 20 Charter of Fundamental Rights (CFR) and Article 14 ECHR cannot be relied upon. They certainly can, but the judicial practice treads carefully, as we have seen, with regard to ‘resource-hungry’ services and in the early phase of residence. This makes it difficult to oppose unequal treatment, for instance with regard to study grants or temporary stay.98 Article 34 CFR may serve as an alternative yardstick, but activating this provision remains challenging as well. The doctrinal contours of the social rights and principles in the solidarity chapter are contested, irrespective of migration law.99 Article 34 CFR was mentioned by the Court on at least three occasions involving migration law, but these judgments remained largely silent on the added value of the Charter.100 The situation may be different for asylum applicants and beneficiaries of international or temporary protection, as explained in Chapter 13.8.3 on reception conditions. Subject matters not covered by the supranational equality provisions must be assessed in light of the ECHR and national constitutions.

15.3.4 Disparate effects of equal treatment Equal treatment need not result in a level playing field, neither between the Member States nor for migrants who move across state borders. Discrepancies within the European Union result from the relational character of equal treatment. States are obliged to treat migrants on an equal footing with nationals. Doing so effectively replicates cleavages between the Member States. Social security payments, for instance, will generally be higher in Finland than in Bulgaria. Furthermore, disparate practical effects transcend payment levels. The structure of the welfare state differs across Europe, as highlighted by the literature on the varieties of capitalism.101 For our purposes, five considerations are particularly important. First, important social benefits are connected to registered work in some countries, while others offer universal social protection schemes—Mediterranean and Scandinavian 96 See Gijsbert Vonk and Sarah Van Walsum, ‘Access Denied’ in Gijsbert Vonk (ed), Cross-Border Welfare State (Intersentia 2012) 3, 12–23; and Constanze Janda, Migranten im Sozialstaat (Mohr Siebeck 2012) ch 3. 97 See Kamberaj (n 74) paras 84–85; and generally Case C-236/09 Association belge des Consommateurs Test- Achats ASBL and others EU:C:2011:100. 98 See Single Permit Directive 2011/98/EU, art 12(2)(a), (i), (iii), (b); ICT Directive 2014/66/EU, art 18(2), (3); and Blue Card Directive (EU) 2021/1883, art 16(2)(1). 99 See ch 5.4.1. 100 See Kamberaj (n 74) paras 80, 92; INPS (n 75) paras 44–47; Land Oberösterreich (n 94) paras 39–40; and generally ch 6.5.3. 101 See Diane Sainsbury, Welfare States and Immigrant Rights (OUP 2012); and generally Gosta Esping- Andersen, The Three Worlds of Welfare Capitalism (Princeton UP 1990).

Equality Provisions in Secondary Legislation 485 countries are the classic antipodes. Both systems will work reasonably well for those with a regular job, as a result of full integration into the social security system. By contrast, beneficiaries of international or temporary protection, as well as other migrants in a precious situation, may struggle to meet basic needs in countries with weak universal support schemes. Equal treatment with nationals is literally not worth much, if they do not find a job, even though widespread informality can make it easier to earn a living in the informal sector.102 Secondly, all equal treatment provisions cover, in slightly varying formulations, ‘working conditions, including pay and dismissal as well as health and safety at the workplace’,103 in accordance with Article 15(3) CFR.104 That may sound like a straightforward guarantee, but practical effects will differ as well. Whereas working conditions are subject to state legislation in most countries, enhanced protection can be available under collective agreements (Scandinavian legal systems have traditionally relied on collective agreements primarily). Migrants working in jobs or with companies not subject to collective agreements may be much worse off in practice than those in other segments of the labour market. Thirdly, Member States usually insist on equivalent skills. Equal treatment in the recognition of professional qualifications under Article 12(1)(d) Single Permit Directive 2011/98/EU supports migrants in fulfilling this condition ‘in accordance with the relevant national procedures’. The provision applies, as we have seen, to anyone with labour market access, irrespective of whether the residence permit is subject to EU harmonisation. In practice, however, this guarantee is not worth much if home states do not operate formalised professional training, or where curricula differ.105 Migrants may have informal skills, but these are not usually recognised as being equivalent. Equality in the recognition of qualifications sounds good on paper but may result in what is occasionally called ‘brain waste’,106 nonetheless. EU institutions may consider extending the Professional Qualifications Directive, which only covers Union citizens at present.107 However, doing so would only cover qualifications obtained in another Member State. Fourthly, similar problems arise, albeit under sign reversal, whenever migrants return home after the termination of the work contract. Within the single market, Regulation (EC) No 883/2004 coordinates the interaction of national social security schemes—a guarantee that was extended to third country nationals residing legally, after several decades of debate.108 Again, social security coordination with third states is not covered by the single market rulebook, with the exception of association agreements. Member States have concluded more than 350 bilateral agreements with third states instead. Their contents differs and their geographic distribution reflects long-standing mobility patterns. Migrants coming from unusual destinations are not usually covered by a bilateral agreement, which,

102 See Claudia Finotelli and Irene Ponzo, ‘Migrant Inclusion in Southern European Societies’ (2018) 44 JEMS 2303. 103 Single Permit Directive 2011/98/EU, art 12(1)(a). 104 The term ‘equivalent’ (French équivalentes) may allow for some differentiation. 105 cfhttps://en.wikipedia.org/wiki/Dual_education_system (accessed 1 March 2023); and Commission, ‘Fitness Check on the EU Legislation on Legal Migration (Part 1)’ SWD(2019) 1055 final, 125–30. 106 Çağlar Özden, ‘Educated Migrants: Is There Brain Waste?’ in Çağlar Özden and Maurice Schiff (eds), International Migration, Remittances, and the Brain Drain (World Bank 2006) 227. 107 See Recognition of Professional Qualifications Directive 2005/36/EC, art 2(1), with later amendments. 108 See Regulation (EU) No 1231/2010 extending Regulation (EC) No 883/2004 to nationals of third countries [2010] OJ L344/1, art 1.

486 Integration and Settlement moreover, have mostly been concluded by the bigger Member States and the classic destination countries in Western Europe.109 Statutory pensions, based on previous employment, are the most relevant social security benefit migrants may wish to ‘export’ upon return. Several legal migration instruments foresee equal treatment in this respect, albeit under the same conditions as nationals.110 As a result, practical effects will depend on domestic laws: while most Member States allow nationals to transfer pensions, some require reciprocal bilateral agreements111—a requirement judges have found to justify unequal treatment in light of human rights.112 Equal treatment with nationals will not, as a result, translate into an automatic guarantee that pensions can be transferred. A related question concerns the payment of family benefits to workers who have family members residing abroad. Whereas Article 11(2) Long-Term Residents Directive and Article 23(2)(i) Seasonal Workers Directive 2014/36/EC contain optional derogations, judges interpreted the absence of such a derogation to mean that third country national covered by the Single Permit Directive must be treated equally with nationals.113 Remember, however, that equal treatment does not prevent states from generally restricting exportation, for nationals and migrants alike. Finally, European labour and social legislation generally rests on territorial presence as the entry gate into the welfare state. Doing so corresponds to the European social model of compassion and support for fellow citizens, oftentimes irrespective of nationality. Comparative studies illustrate that there are alternative models, which contrast with the intuitive preference for equal treatment among most experts in migration law.114 Other countries prefer to ‘build a wall around the welfare state, not around the country’;115 the Gulf States are classic examples of a political preference for more migrants with fewer rights.116 A variant of moderate welfare state enclosure is the restriction of important social benefits to registered work, instead of universal support schemes, which can be found in some Member States.

15.3.5 Distinctions between refugees and subsidiary protection Article 78(2)(a) and (b) TFEU list refugee status and subsidiary protection separately and do not require, therefore, equal treatment. Absence of Treaty-based uniformity did not, however, prevent the EU institutions from introducing the overarching concept of ‘international protection’ with identical rights as the general rule.117 Against this backdrop, 109 See Saskia Klosse, ‘External Aspects of Social Security Coordination’ MISSOC Analysis (November 2013). 110 See eg Single Permit Directive 2011/98/EU, art 12(4); and Seasonal Workers Directive 2014/36/EC, art 23(1). 111 See Klosse (n 109) 31–67. 112 See ch 10.4.2. 113 See Istituto Nazionale della Previdenza Sociale (n 71) paras 27–39. 114 See Lieneke Slingenberg, The Reception of Asylum Seekers under International Law (Hart Publishing 2014) 5–8; and Daniel Thym, ‘Migrationsfolgenrecht’ (2017) 76 Veröffentlichungen der Vereinigung Deutscher Staatsrechtslehrer 169, 176–85. 115 William A Niskanen, ‘Chairman‘s Message’ in Cato Polity Report (September/October 2006) 2; see also ‘A Way Forward on Immigration’ The Economist (25 August 2018) 10; and the position of Milton Friedman https:// openborders.info/friedman-immigration-welfare-state (accessed 1 March 2023). 116 See Martin Ruhs, The Price of Rights (Princeton UP 2013). 117 See Qualification Directive 2011/95/EU, recital 39, art 20(2).

Mandatory Integration Requirements 487 the Court concluded that the rights of refugees and beneficiaries of subsidiary protection should be approximated ‘so far as possible’, mirroring the case law on Turkish nationals.118 The protective reach of this verdict ultimately depends, like in the case of Turkish nationals, on the contents of secondary legislation. For many years, gradual approximation appeared as within reach, starting with equal access to employment under Article 29 Qualification Directive 2011/95/EU.119 However, the Commission changed course in 2016 when it emphasised the need to ‘to better clarify the difference’120. Ongoing negotiations on the revision of the Qualification Directive do not indicate a desire to reverse existing guarantees, but the trend towards approximation has stopped. Distinctions regarding travel documents, social assistance, and family reunification are here to stay.121 Legislative distinctions could possibly be challenged on human rights grounds, notwithstanding the reluctance of the ECJ and the ECtHR described in Chapter 10.5. In one judgment, the Court in Strasbourg concluded that the absence of ‘an element of choice’122 entailed that the situation of refugees may warrant stricter scrutiny; it found unequal treatment of refugees and students and other temporary migrants to be illegal.123 Nevertheless, a more recent ruling concluded that it was legitimate to lay down stricter standards for family reunification for beneficiaries of subsidiary protection. Judges did not decide whether their situation was not comparable in the first place or may be justified in light of overriding public interests. In doing so, they rebuffed, indirectly at least, the argument that challenges the widely held belief that those fleeing civil wars may be able to return home quickly.124 Nevertheless, the outcome of future court cases remains difficult to predict, also considering that the aspect of choice was only one factor amongst others in the judicial reasoning and that several rulings evaded dealing with the lesser degree of rights associated with subsidiary protection.125

15.4 Mandatory Integration Requirements An essential element of the legal dimension of migrant integration are explicit integration requirements, introduced first by the main destination countries after the millennium change. Pre-departure language tests as a precondition for family reunification are emblematic of the new approach, which has caused heated academic and political controversies (15.4.1). EU legislation authorises Member States to introduce such integration requirement on the basis of optional clauses in the Family Reunification Directive and the Long-Term 118 See ch 17.1.3; and Case C-369/17 Ahmed EU:C:2018:713, para 43. 119 See Stefan Salomon, ‘Constructing Equality in EU Asylum Law’ (2022) 34 IJRL 608, 612–22; and Jürgen Bast, ‘Vom subsidiären Schutz zum europäischen Flüchtlingsbegriff ’ [2018] Zeitschrift für Ausländerrecht 41. 120 Commission, ‘Communication: Towards a reform of the Common European Asylum System’ COM(2016) 197 final, 10. 121 See Qualification Directive 2011/95/EU, arts 24, 25, 29(2); and Family Reunification Directive 2003/86/EC, arts 3(2)(c), 9–12. 122 Hode and Abdi v United Kingdom App no 22341/09 (ECtHR, 6 November 2012) § 47; see also Tanda-Muzinga v France App no 2260/10 (ECtHR, 10 July 2014) § 74 in the context of ECHR, art 8. 123 Hode and Abdi v United Kingdom (n 122) §§ 48–55. 124 See MT and others v Sweden App no 22105/18 (ECtHR, 20 October 2022) §§ 98–117 (not final at the time of writing); Salomon (n 119) 633–9; and Emanuela Pistoia, ‘The Right of Refugees and Beneficiaries of Subsidiary Protection to National Treatment in Social Assistance’ (2020) 45 EL Rev 396, 404–407. 125 See Alo and Osso (n 76); Ayubi (n 77) para 20; and MA v Denmark App no 6697/18 (ECtHR [GC], 9 July 2021) §§ 177, 197; see also Case C-364/11 El Karem El Kott and others EU:C:2012:286, paras 78–80.

488 Integration and Settlement Residents Directive, thus directing legal discontent at the Court of Justice (15.4.2). Judges endorsed national integration conditions in two landmark judgments, delivered in 2015, while simultaneously obliging national legislatures to introduce a hardship clause (15.4.3). These judgments illustrate how important doctrinal arguments are for anyone trying to decipher—and influence—the evolution of EU migration law.

15.4.1 Proliferation and significance of domestic laws To emphasise the significance of language is not a new phenomenon. States have organised language classes for decades, and the US had introduced a literacy test in the 1920s after years of debate.126 On the eve of supranationalisation, Germany experimented with language tests to restrict first admission of ethnic Germans from the former Soviet Union, building upon two decades of experience with language skills as a precondition for permanent residence.127 After the millennium change, similar requirements spread in the main destination countries. Heated policy debates about family formation prepared the ground for a particularly controversial instrument: pre-departure language tests in the Netherlands, Germany, Denmark, France, the UK, and Austria.128 Even more widely used were different post-arrival measures, notably for permanent residence and naturalisation; ‘citizenship tests’ exist in many countries.129 These domestic reforms differed in terms of contents, policy context, and factual implications, but some overarching trends can be observed. Chapter 14.1.2 on family reunification explained that regular contact between national officials in Brussels triggered informal policy exchange. Moreover, explicit integration clauses often complement implicit conditions, such as sufficient resources or fee levels. These implicit requirements can be as important as language tests in practice, although the political and academic discourse will usually concentrate on the latter. Language tests and civic education are perceived, by supporters and critics alike, as symbols for the broader reorientation of migrant integration policy. Comparative surveys demonstrate that mandatory integration requirements have often been accompanied by supportive policies, in line with our introductory comments on mixed outcomes. Some argue that they are primarily ‘control signals’ to appease sceptical public opinion, instead of spearheading a general hardening of migration laws.130 Citizenship tests, in particular, habitually exhibit rather ‘thin’ visions of social togetherness and national identity, notwithstanding the political background music of xenophobic rhetoric.131 Conversely, absence of mandatory integration requirements need not signal greater 126 See Aristide R Zolberg, A Nation by Design (Harvard UP 2006) ch 7; and Kees Groenendijk, ‘Pre-Departure Integration Strategies in the European Union’ (2011) 13 EJML 1, 3–4. 127 See today’s Federal Expellees Act (Bundesvertriebenengesetz), ss 6(2), 9, which codified an earlier administrative practice; and No 7(4)(1)(b) Administrative Regulations on the Foreigners Act (Verwaltungsvorschrift AuslG) [1978] Common Ministerial Bulletin (Gemeinsames Ministerialblatt) No 23/368. 128 See Groenendijk (n 126) 9–20. 129 See Sara Wallace Goodman, Immigration and Membership Politics in Western Europe (CUP 2014); and Elspeth Guild, Kees Groenendijk, and Sergio Carrera (eds), Illiberal Liberal States (Ashgate 2009). 130 See ch 4.2.3; Julia Mourão Permoser, ‘Civic Integration as Symbolic Politics’ (2012) 14 EJML 173; and Stephen J Larin, ‘Is It Really about Values?’ (2020) 46 JEMS 127. 131 See Ines Michalowski, ‘Required to Assimilate? The Content of Citizenship Tests in Five Countries’ (2011) 15 Citizenship Stud 749; and Liav Orgad, The Cultural Defense of Nations (OUP 2015) ch 5.

Mandatory Integration Requirements 489 openness in terms of equal membership.132 Integration conditions accompanied the acceptance of migration as a defining feature of European societies, from ‘denizenship’ towards ‘citizenship’.133 In doing so, they indicated a focus on social affiliation and cultural integration in line with previous comments. Beside the symbolic dimension, integration conditions have tangible repercussions for migrants who are denied entry visas. Their introduction coincided with the rise of populist parties in many countries, which promote ethno-cultural closure, instead of forward- looking social affiliation or equality-based integration. At an intermediate level, integration requirements appear as part of a broader trend towards the ‘meritocratisation’ of membership, with first admission and status upgrade being contingent upon markers of neoliberal utility and sociocultural affiliation—a phenomenon that defines Union citizenship as well.134 In this overall context, language skills are a comparatively neutral indicator: they facilitate structural integration in the form of labour market access or educational achievement, support social integration through inter-group contacts, and may nourish identificatory patterns of cultural integration.

15.4.2 Supranational legal framework EU migration law does not oblige Member States to introduce explicit integration requirements. Instead, optional clauses leave national parliaments with the choice whether to introduce—or repeal135—them. Family reunification can be made conditional upon compliance with integration ‘measures’, together with ‘conditions’ for children above the age of 12 and to prevent forced marriages.136 Member States may also require applicants for long- term residence to fulfil integration ‘conditions’, also in situations of intra-EU mobility.137 Moreover, the term integration is mentioned oftentimes in preambular recitals and occasionally in the operational articles. By way of example, beneficiaries of international protection shall have access to ‘integration facilities’.138 Family members of blue card holders, intra-corporate transferees, and researchers can be subject to integration measures only after arrival139—an exception the Commission wants to extend to long-term residents.140 The Common Basic Principles emphasise respect for basic values, civic knowledge, and social interaction, albeit without endorsing mandatory requirements.141 Integration requirements in the Family Reunification Directive and the Long-Term Residents Directive were discussed controversially during the negotiations. Member States

132 See Christian Fernández, ‘The Unbearable Lightness of Being Swedish?’ (2019) 19 Ethnicities 674. 133 See Christian Joppke, ‘Comparative Citizenship: A Restrictive Turn in Europe?’ (2008) 2 Law Ethics Hum Rts 1; and Thym (n 6) 98–114. 134 See ch 4.3.2; Ganty (n 21) passim; and Daniel Thym, ‘The Evolution of Citizens’ Rights in Light of the EU’s Constitutional Development’ in Thym (n 31) 111. 135 The German government has committed to discontinuing pre-departure language tests during the 2021–25 parliamentary term. 136 See Family Reunification Directive 2003/86/EC, arts 4(1)(3), (5), 7(2). 137 See Long-Term Residents Directive 2003/109/EC, arts 7(2), 15(3). 138 See Qualification Directive 2011/95/EU, art 45. 139 See ICT Directive 2014/66/EU, art 19(3); Students and Researchers Directive (EU) 2016/801, art 3(2); and Blue Card Directive (EU) 2021/1883, art 17(3). 140 See LTRD Recast Proposal (n 70), art 15(2). 141 See ‘Common Basic Principles’ (n 50) Nos 2, 4, 7.

490 Integration and Settlement in the Council departed, as we have seen, from the political promise of the European Council in Tampere to approximate the status of third country nationals to that of Union citizens. Publication of the final text of the Directives in the Official Journal was the starting shot for legal disputes over whether national integration requirements complied with Union law.142 Some contended that the textual distinction between ‘conditions’ for status acquisition and ‘measures’ after arrival could be used to defeat pre-departure language tests as a precondition for family reunification143—an argument that did not convince the Court.144 Judges finally delivered a ruling on that question in 2015, after two previous attempts at receiving an answer had failed on procedural grounds, since the claimants in the domestic proceedings had received entry clearance.145 Judges endorsed integration requirements in two judgments delivered by the same chamber in the summer of 2015. The K and A case concerned a classic scenario of the Dutch pre-departure language and civic knowledge exam. Dutch authorities had rejected the argument of two applicants that their medical and psychological problems came under a hardship clause, which was available under national law. The Council of State referred the matter to Luxembourg. The legal background of the parallel P and S ruling was rather specific. The Netherlands had introduced a language requirement for acquiring long-term resident status in 2007, which did not apply immediately to those already residing in the country. They were obliged to learn the local language after status acquisition instead; non- compliance could be sanctioned through fines. This transitional arrangement entailed that the P and S case did not come directly under Article 5(2) Long-Term Residents Directive, raising the question as to whether the Directive applied at all. The sense of anticipation increased when two Advocates General delivered divergent opinions. AG Szpunar propagated an ideal-typical incarnation of the equality-based outlook on integration, including a reference to the Tampere Conclusions and the worth of ‘denizenship’. Szpunar considered integration tests to be instruments of migration control that frustrate the realisation of the integration objective.146 By contrast, AG Kokott emphasised the significance of language skills and civic knowledge for diverse aspects of migrant integration, ranging from labour market participation to familiarity with gender equality. Language classes after entry may be better suited to learn the language, but they would not be equally effective in promoting integration from day one, including to prevent forced marriages.147 Her additional proposal to mitigate the outcome through an individualised assessment and a criticism of excessive few levels prepared the ground for a compromise package judges could agree upon.

142 See Carrera (n 30) ch 7; and Diego Acosta Arcarazo, The Long-Term Residence Status as a Subsidiary Form of EU Citizenship (Martinus Nijhoff 2011) ch 9. 143 See Kees Groenendijk, ‘Family Reunification as a Right under Community Law’ (2006) 8 EJML 215, 223–25; and Commission, ‘Communication on guidance for application of the Family Reunification Directive’ COM(2014) 210 final, 14–15. 144 See Case C-153/14 K and A EU:C:2015:453, paras 45–48; and AG Juliane Kokott, Opinion in ibid EU:C:2015:186, points 19–32. 145 See the Dutch reference in Case C-155/11 PPU Mohammad Imran EU:C:2011:387; and the German reference in Case C-513/12 Ayalti EU:C:2013:210. 146 See AG Maciej Szpunar, Opinion in Case C-579/13 P and S EU:C:2015:39, points 28–30, 79–97. 147 See AG Kokott, K and A (n 144) points 33–37 and nn 16–19.

Mandatory Integration Requirements 491

15.4.3 Judicial endorsement, subject to a caveat The judgments on integration conditions illustrate how important the doctrinal foundations of the case law can be. Instead of human rights, secondary legislation took centre stage. Doing so required some hermeneutical finesse, so as to bring the corollary obligation of long-term residents to learn the local language into the ambit of Union law; judges relied on the effet utile, as presented in Chapter 6.4.5. Regarding the meaning of integration, the Court was brief and emphasised that ‘it cannot be disputed that the acquisition of knowledge of the language and society of the host Member State greatly facilitates communication’ between foreigners and nationals, thereby ‘encourag[ing] interaction and the development of social relations between them’, as well as labour market participation.148 For our purposes, the emphasis on social integration through communication and inter-group relations stands out, with only corollary remarks on structural inclusion in accordance with accounts of integration as equality. This description of integration was sufficient to support the overall conclusion that pre-departure language tests are compatible with Union law as a matter of principle. Yet it hardly translates into a precise definition, let alone a vindication of a specific theoretical vision of migrant integration in line with previous comments. Arguably, the notion of integration functions as an incompletely theorised agreement on how to deal with specific scenarios, without resolving the theoretical dissensus about the meaning in the abstract.149 The divergent usage of the formula that integration shall be the ‘the principal purpose’ of the Long-Term Residents Directive, mentioned previously, fits such an approach, as does the generic statement, on the part of the Court, ‘that the concept of integration is not defined’ in the Family Reunification Directive.150 That was relevant insofar as its optional clauses on integration requirements defer to national law, thus indicating the absence of an autonomous meaning in accordance with comments in Chapter 6.4.4. The supranational outlook on migrant integration builds on a conglomerate of bits and pieces in various legislative instruments and Court judgments which need not always be coherent. Silence on human rights was straightforward for long-term residents. Denial of status acquisition under Article 5(2) Long-Term Residents Directive does not affect human rights, since the person will not usually lose the ‘weaker’ status, which is a precondition for long-term residence. By contrast, the human rights implications are evident with regard to family reunification. Nevertheless, there were cogent reasons why judges focused on secondary legislation. Remember that the Grand Chamber had recognised that the Family Reunification Directive comprised ‘clearly defined individual rights’ to family formation, ‘[g]‌oing beyond’ obligations under human rights law.151 Secondary legislation transcends human rights, which do not generally oblige states to issue entry visas. The ensuing focus on secondary legislation made it difficult to ban integration measures which are explicitly permitted by Article 7(2) Family Reunification Directive. This difficulty was one reason why the alternative focus on terminological distinction between measures and conditions was a smart strategic move, even though judges did not 148 P and S (n 48) para 47; and K and A (n 144) para 53. 149 See Cass R Sunstein, Legal Reasoning and Political Conflict (OUP 1996) ch 2; and Daniel Thym, ‘Towards a Contextual Conception of Social Integration in EU Immigration Law’ (2016) 18 EJML 89, 105–10. 150 See Case C-540/03 Parliament v Council EU:C:2006:429, para 70. 151 ibid paras 59–60; and ch 14.2.3.

492 Integration and Settlement eventually go along. Moreover, one can question whether the privileges of blue card holders, amongst others, are compatible with Articles 20 and 21 CFR. Silence of the Court on this matter reiterates our previous finding about the limited impact of equal treatment.152 Having said this, judges were careful not to give Member States a free hand. The Court obliged them to introduce a hardship clause to take account of ‘specific individual circumstances, such as the age, illiteracy, level of education, economic situation or health’.153 Reading the judgments, there is little doubt that any applicant must be able to insist on an individualised assessment, despite judges refraining from guidance as to how the cases under consideration should have been decided. All we can deduct from the reasoning is that the Dutch hardship clause with its focus on medical and psychological reasons was too narrow.154 This leaves domestic legislatures and courts with the task of defining the way forward. Comments in Chapter 10.5.2 illustrated that the imprecision on the precise contours of the hardship clause reflects a built-in ambiguity of the proportionality test in-between abstract legislative standards and administrative fine-tuning. One question domestic courts will have to consider is the significance of empirical evidence on the suitability of state measures to achieve the objective.155 Such empirical analyses should ideally rest upon several studies applying mixed methods. Also, the judicial focus on social integration, including communication and contact, renders it difficult to invoke empirical studies, which mostly deal with structural integration.

15.5 Other Integration Measures Legislation is one instrument amongst others public authorities at the local, regional, national, and supranational levels employ to influence migrant integration, which, generally, cannot be realised by means of hierarchical categories of order and command. The Common Basic Principles famously define integration ‘a dynamic, two-way process of mutual accommodation by all immigrants and residents’.156 This has been taken up widely across Europe, although each society, or part thereof, will ultimately have to develop its own response, reflecting the national and local context.157 Although Member States had rejected, as we have seen, the introduction of the open method of coordination, soft law instruments effectively provide for an overarching and informal policy framework nevertheless.158 Beside the Common Basic Principles, two common agendas and two action plans have been adopted by the Commission.159 152 See ch 10.4; (German) Federal Administrative Court (Bundesverwaltungsgericht), Case 1 C 8.09 (judgment of 30 March 2010) paras 54–67; and Karin de Vries, Integration at the Border (Hart Publishing 2013)chs 8–9. 153 K and A (n 144) para 58; and AG Kokott in K and A (n 144) points 40–46. 154 See Thym (n 149) 100–03. 155 See Anita Böcker and Tineke Strik, ‘Language and Knowledge Tests for Permanent Residence Rights’ (2011) 13 EJML 157; and Pistoia (n 81) 797–800. 156 See ‘Common Basic Principles’ (n 50) No 1; and Thym (n 36) ch 5. 157 ‘Common Basic Principles’ (n 50) recitals 4–6. 158 See Helene Urth, ‘Building a Momentum for the Integration of Third-Country Nationals in the European Union’ (2005) 7 EJML 163. 159 See Commission (n 42); ‘Action plan on the integration of third-country nationals’ COM(2016) 377 final; ‘European agenda for the integration of third-country nationals’ COM(2011) 455 final; and ‘A common agenda for integration’ COM(2005) 389 final.

Other Integration Measures 493 The diversity of state authorities responsible means that it is notoriously difficult to determine to which extent EU guidance trickles down. Surprisingly, the ‘European Website on Integration’ is available in three official languages only, thus making it difficult for local officials without foreign language skills to consult the platform.160 Indeed, the approximation of national approaches is far from guaranteed. Social scientists have long discussed whether national integration models persist or converge, whose singularity might have been overstated in the first place.161 In order to promote transnational cooperation, contact points were established early on. The Commission also developed a handbook for policy-makers and practitioners and even considered the design of pan-European modules, in accordance with the Stockholm Programme.162 Such semi-formalised contributions have given way to indirect horizontal diffusion and the exchange of best practices, which might be effective precisely because they are based on personal contacts and local networks. To facilitate rational decision-making and cross-national comparisons, much energy has been invested in developing indicators and benchmarks of migrant integration, notably with regard to structural inclusion.163 Funding plays a critical role in activating national resources. Chapter 7.4 described that budgetary arrangements distinguish between short-term needs upon arrival, covered by migration- specific instruments, and support for integration thereafter, which forms an integral part of other funding opportunities. Such ‘mainstreaming’ suits the thematic breadth of migrant integration, for which migration law and policy are, as we have seen, one element amongst others. Mainstreaming considers integration as an integral part of wider policies. Specific measures for migrants are often appropriate during the early stages (eg language courses to support the inclusion of refugees). In the long run, however, it is neither desirable nor possible to disentangle the integration of migrants from general policies. In doing so, we should ask whether migration requires a reconfiguration of existing policies, to take account of the perspective of migrants, for instance in the areas of education, labour market, or healthcare.164 Mainstreaming is not, in other words, about the uncritical continuation of the status quo ante but about the ‘de-migranticisation’ of research and policy design to embed migration-specific considerations into broader social trends.165 Combating discrimination was a crucial component of the early debate, in line with the widespread focus on equality.166 EU non-discrimination legislation does not belong to the area of freedom, security, and justice. Nevertheless, it receives and deserves much 160 See https://ec.europa.eu/migrant-integration (accessed 1 March 2023). 161 See Gary P Freeman, ‘National Models, Policy Types, and the Politics of Immigration in Liberal Democracies’ (2006) 29 West Eur Politics 227; and Ruud Koopmans, Ines Michalowski, and Stine Waibel, ‘Citizenship Rights for Immigrants. National Political Processes and Cross-National Convergence in Western Europe, 1980–2008’ (2012) 117 Am J Soc 1202. 162 See Carrera (n 30) 62–107; and Commission (n 159) 11–12. 163 See Eurostat, Migrant Integration (January 2021); OECD/Commission, Settling in 2018. Indicators of Immigrant Integration (2nd edn, December 2018); and Ganty (n 21) 660–66. 164 Expert Council on Integration and Migration (Sachverständigenrat für Integration und Migration), Steuern, was zu steuern ist. Jahresgutachten 2018 (May 2018) 80, as whose member and vice-chairperson the author served from 2016 to 2022 https://www.svr-migration.de/wp-content/uploads/2022/10/SVR_Jahresgutachten_2018.pdf (accessed 1 March 2023). 165 See Stephen Castles, ‘Understanding Global Migration’ (2010) 36 JEMS 1565; Janine Dahinden, ‘A Plea for the “De-migranticization” of Research on Migration and Integration’ (2016) 39 Ethn Racial Stud 2207; and also Thym (n 114) 169–76. 166 See Elspeth Guild, The Legal Elements of European Identity (Kluwer 2004) ch 11.

494 Integration and Settlement attention as one of the few legally binding supranational contributions to migrant integration. Directive 2000/78/EC establishing a general framework for equal treatment in employment and occupation has gained more traction, in particular with regard to judgments on religion, which reiterate our earlier finding about competing theoretical visions of migrant integration.167 By contrast, judges are reticent to find a violation of the Racial Equality Directive 2000/43/EC.168 Neither Directive can be applied to distinctions between migration statuses, as stated in their common Article 3(2). Nevertheless, they inform other state policies affecting migrants, as well as the behaviour of private parties.

15.6 Long-Term Residents Directive Directive 2003/109/EC is of great practical and theoretical importance. More than ten million third country nationals are permanent residents, under supranational or national laws. They benefit from residence security and equal treatment, in line with our introductory remarks on ‘denizenship’. Comments hereafter can be comparatively brief, since most legal questions the Directive gives rise to have been mentioned in the horizontal chapters elsewhere in this volume. We will concentrate on the autonomy of long-term residence, irrespective of the original purpose of admission (15.6.1), its scope ratione personae, and the conditions for status acquisition (15.6.2), as well as procedural safeguards and the loss of long-term residence (15.6.3). Almost twenty years after adoption, the Commission tabled a reform in 2022, which refrained from proposing far-reaching amendments, which will be mentioned whenever appropriate. At the time of writing, adoption before the European elections in 2024 appeared a likely scenario.

15.6.1 Independence from the original purpose Most Member States were operating permit residence schemes, typically after five years of legal residence, when EU institutions started dealing with long-term residence.169 Intergovernmental cooperation under the Treaty of Maastricht and within the Council of Europe had anticipated core features of today’s Long-Term Residents Directive, albeit with longer waiting periods and without transnational mobility.170 Observes were disappointed that the Council departed from the original promise, in the Tampere Conclusions, to establish a ‘set of uniform rights which are as near as possible to those enjoyed by EU citizens’171—a political choice which cannot be undone by means of interpreting the Directive dynamically. Recital 2 echoes in the past tense what had been ‘stated’ by the European Council, and judgments identifying integration as the ‘principal objective’ do not reverse

167 See Adam and Thym (n 43) 84–85. 168 See ch 10.4.4. 169 See Kees Groenendijk and Elspeth Guild, ‘Converging Criteria’ (2001) 3 EJML 37. 170 See Council Resolution on the status of third-country nationals residing on a long-term basis [1996] OJ 80/2, No 3; and Committee of Ministers, ‘Recommendation concerning the security of residence of long-term migrants’ (Rec(2000)15, 13 September 2000) No 4(b). 171 European Council (n 24) No 21.

Long-Term Residents Directive 495 the situation.172 The judicial practice is sporadic, thus reflecting the open texture of the integration concept in line with previous comments. Negotiations were cumbersome and resulted in a compromise, agreed upon unanimously on the eve of enlargement.173 Refugees had not been covered originally, before the adoption of the Qualification Directive, but they were incorporated eight years later, in the form of a an amendment in the form of Directive 2011/51/EU, which has to be read jointly with the original text of Directive 2003/109/EC.174 Member States were given the authority to revoke long-term residence whenever international protection is withdrawn or not renewed.175 That is an exception to the general rule that long-term residence shall not be affected by a change of circumstance, as reaffirmed by the categorical statement that the status ‘shall be permanent’.176 This emphasis on permanence does not refer to the length of residence permits, dealt with separately, but means that the status is not lost if the conditions for acquisition are no longer fulfilled. In practical terms, this self-sufficient autonomy entails that status renewal is not— in contrast to the other migration law instruments—the mirror image of status acquisition. Loss of sufficient resources, for instance, cannot justify withdrawal or non-renewal, although Member States are obliged to reject the original application in such scenarios. Reference to ‘autonomic’ renewal, in Article 8(2), confirms this conclusion; optional application procedures for renewal serve the sole purpose of establishing contact with the administrative authorities.177 The purpose of first admission to the territory is irrelevant as well. Unemployment can entail withdrawal of a blue card, and divorce may result in loss of family reunification rights,178 but none of these occurrences affects long-term residence. Independence from the purpose of first admission is the hallmark of long-term residence as a sort of ‘citizenship light’, as described at the outset of this chapter. Statistics illustrate that practical relevance varies greatly between the Member States. Less than half of the more than 10 million persons with permanent residence held a status under the Long-Term Residents Directive 2003/109/EC.179 Parallel national schemes are the main reason; national permits may be available under easier conditions, for instance shorter waiting periods. Particularly important are the parallel domestic schemes in France, Germany, and Spain, which jointly cover more than five million persons, of which some receive both the national and the EU permits.180 Chapter 10.6.2 on more favourable national provisions explained that the Court interprets Article 13 to authorise parallel national schemes. British nationals who were living in another Member State at the end of the

172 See Thym (n 87) Article 1 MN 14–17a. 173 See Kees Groenendijk, ‘Long-Term Residents’ in Steve Peers and others (eds), EU Immigration and Asylum Law (Text and Commentary), vol 2 (2nd edn, Martinus Nijhoff 2012) 287, 291–93. 174 See Directive 2011/51/EU amending Council Directive 2003/109/EC to extend its scope to beneficiaries of international protection [2011] OJ L32/1. 175 See Long-Term Residents Directive 2003/109/EC, arts 4(1a), 9(3a), as amended by Directive 2011/51/EU. 176 ibid art 8(1). 177 See the Commission’s view according to ‘Proposal for a Council Directive’ (Council doc 10698/01, 2 August 2001) 9. 178 See Blue Card Directive (EU) 2021/1883, art 8(4), (5); and Family Reunification Directive 2003/86/EC, arts 15, 16(1)(b). 179 See the Eurostat datasets for ‘MIGR_RESLONG’ and ‘MIGR_RESSHARE’ https://ec.europa.eu/eurostat (accessed 1 March 2023). 180 Select ‘national legislation’ as the ‘reason’ in the Eurostat dataset for ‘MIGR_RESLONG’.

496 Integration and Settlement transitional period after Brexit are covered by leges speciales in the Withdrawal Agreement, including distinct rules on permanent residence.

15.6.2 Scope and conditions Five years of legal and continuous residence is the ‘main criterion’181 for acquiring long-term resident status. In addition, some third country nationals are excluded from the scope of the Directive ratione personae, independent of the length and legality of residence. Asylum applicants and those with temporary or complementary protection are among them, in contrast to beneficiaries of international protection.182 ‘Non-removable’ migrants and those staying irregularly are not excluded explicitly, but they are bound to fail the legal residency test. Anyone not covered can come within the scope as a result of status change, whenever he obtains a residence permit that qualifies for the calculation of the five-year period. By way of example, asylum seekers may receive subsidiary protection, and those staying irregularly may benefit from national regularisation schemes. For students and beneficiaries of international protection, at least half of the time of factual presence prior to the acquisition of the new title shall be taken into account for the five-year qualification period.183 To count such previous periods of residence fully towards the five-year threshold is one of the few amendments the Commission put forward.184 Article 3(2)(e) comprises an exception for those staying ‘solely on temporary grounds’ and whenever residence permits ‘have been formally limited’. While the exclusion of temporary stay can be based on legitimate considerations,185 open-ended wording and poor drafting generates considerable uncertainty as to the bearing of the exception. Notwithstanding ambiguities of the English language version, the Court interpreted the provision to comprise two autonomous exceptions: one for residence ‘solely on temporary grounds’, the other for formal status limitation.186 Limited periods of validity of residence permits do not to bring migrants within the scope of the second exception. Instead, judges required a qualitative assessment whether a permit ‘prima facie reflect[s]‌any intention on the part of such nationals to settle on a long-term basis’.187 This examination will not usually investigate the personal motivation but should be directed at the purpose of admission, like in the case of seasonal or posted workers mentioned by way of example.188 Note that this abstract examination of temporality concerns Article 3(2)(e) only. Students are excluded under Article 3(2)(a) even if we conclude that the Students and Researchers Directive embraces the idea of permanence in the sense of Article 3(1) Family Reunification Directive 2003/86/EC. Finally, family members of Union citizens are covered by the Long-Term Residents Directive.189

181 Case C-469/13 Tahir EU:C:2014:2094, para 31. 182 See Long-Term Residents Directive 2003/109/EC, art 3(2)(b)–(d), as amended by Directive 2011/51/EU. 183 ibid art 4(2). 184 See LTRD Recast Proposal (n 70) art 4(5). 185 See Achilles Skordas, ‘Immigration and the Market’ (2006) 13 Colum JEL 201, 216–17. 186 See Case C-502/10 Singh EU:C:2012:3076, paras 30–38, which are easier to understand in the French language version of the judgment. 187 ibid para 47. 188 ibid paras 39–56; and Thym (n 87) Article 3 MN 15–20. 189 See Case C-40/11 Iida EU:C:2012:2405, paras 40–41 passim.

Long-Term Residents Directive 497 The five-year requirement is mandatory,190 although Member States remain free to introduce more generous national schemes. Whether residence qualifies as legal or not depends on the supranational and national legislation applicable.191 Five years of legal residence are required ‘immediately prior’ to the application, and periods of absence can be irrelevant in the conditions set out in Article 4(3). Longer absence interrupts the qualification period, which starts anew upon return. Special rules apply to blue card holders, who may accumulate periods of residence in different Member States in accordance with Article 18 Blue Card Directive (EU) 2021/1883. The Commission proposes extending this option to all other third country nationals;192 for the time being, they must reside continuously in the same country for five years. Article 5(1) requires sufficient resources and sickness insurance, whereas the bearing of the additional reference to documentary evidence on appropriate accommodation, in Article 7(1)(2), remains uncertain.193 The sufficient resources requirement must be interpreted in line with the comments in the horizontal Chapter 10.2. Refusal on grounds of ‘public policy’ and ‘public security’ under Article 6 follows somewhat more lenient standards than the criteria for expulsion under Article 12, which, nevertheless, presents us with a compromise between the status of Union citizens and settled ECtHR case law.194 This graded approach reiterates our earlier conclusion, in Chapter 10.3, on the context- sensitive reading of the public policy caveat. Judgments on optional integration requirements were mentioned previously in this chapter, and rulings on fee levels were analysed in Chapter 10.1.3. Implications of ‘golden visa’ schemes, with Member States applying the conditions for long-term resident status negligently so as to grant rich foreigners visa free access to the Schengen area, were mentioned in Chapter 7.3.3.

15.6.3 Procedure and loss Procedural safeguards are loosely knit. Articles 7 and 10 oblige Member States normally to take a decision within no more than six months and to give reasons for rejection, against which applicants may mount a legal challenge. Chapter 7.2 on the administrative dimension explained that these abstract guarantees are complemented by unwritten general principles and procedural safeguards in the Charter, for instance regarding the right to be heard and the consequences of procedural deficits. Long-term residence permits shall be valid ‘at least for five years’, thus allowing longer time periods, including unlimited validity.195 Long- term residence shall be renewed ‘automatically’ after expiration, as we have seen, even if the conditions for status acquisition are no longer fulfilled. Grounds for withdrawal and loss are enumerated, despite the open wording, in Article 9. Member States cannot introduce additional criteria.196 190 See Tahir (n 181) paras 26–37. 191 See Singh (n 186) paras 39–40. 192 See LTRD Recast Proposal (n 70) art 4(3). 193 See Thym (n 87) Article 7 MN 10–12. 194 ibid Article 6 MN 2–23, Article 12 MN 1a–10. 195 See Long-Term Residents Directive 2003/109/EC, art 8(2); and Commission, ‘Report on the Implementation of the Long-Term Residents Directive’ COM(2019) 161 final, 8. 196 See also Sonja Boelaert-Suominen, ‘Non-EU Nationals and Council Directive 2003/109/EC on the Status of Third-Country Nationals Who are Long-Term Residents’ (2005) 42 CML Rev 1011, 1025.

498 Integration and Settlement Fraudulent acquisition justifies status loss, which happens automatically, or withdrawal, by means of an administrative decision, in accordance with Article 9(1)(a). The provision does not require positive knowledge of the applicant when someone else misleads the authorities, especially family members.197 Automatic loss is not subject to an individualised proportionality assessment; the migration status prior to status acquisition can usually be relied upon again.198 Withdrawal or loss may possibly extend to errors by the authorities, like in the case of subsidiary protection, although such conclusion contrasts with the idea of permanence.199 Periods of absence of at least 12 consecutive months will usually result in the termination of long-term residence, unless national legislation provides otherwise.200 Whereas the English and French text employs the generic term ‘absence’, the German version refers to ‘residence’, thus indicating that short return visits are insufficient. In a response to an Austrian reference, the Court concluded that short visits of a few days at irregular intervals are sufficient to keep the status.201 Visits to other Member States than the one which issued the status equally prevent withdrawal or loss, since Article 9(1)(c) refers to the territory of the Union. To say that the idea behind this provision is to cover only those migrants for whom ‘maintenance of that status no longer serves any purpose’202 reminds us of why long- term residence has been dubbed ‘citizenship light’. Rights associated with long-term residence are discussed elsewhere. Equal treatment in accordance with Article 11 has been covered on the previous pages, and intra-European mobility was mentioned in Chapter 14.4.5 on legal migration. The Commission proposes extending both sets of rules during the upcoming legislative negotiations. Complex rules govern the withdrawal and loss of long-term resident status of beneficiaries of international protection, as well as their intra-European mobility.203 Their principal purpose is to ensure that states comply with their obligations under the Refugee Convention and the Qualification Directive.

15.7 Acquisition of Nationality EU migration law is about the entry and stay of third country nationals. Hence, the acquisition of nationality serves as the endpoint of migration law, which more than half a million foreigners cross each year.204 EU institutions have emphasised the desirability of naturalisation for decades, albeit at a high level of abstraction and without interfering with domestic policy choices. Prevalent silence on nationality law has the side effect that long- term residence may be perceived to be an alternative (15.7.1). Notwithstanding the absence of legislative competences, the Court of Justice has made some inroads into nationality law 197 See Case C-557/17 YZ and others EU:C:2019:203, paras 61–67. 198 ibid para 73, distinguishing fraud in applications for family reunification. 199 See Case C-720/17 Bilali EU:C:2019:448, paras 44–52. 200 See Long-Term Residents Directive 2003/109/EC, art 9(1)(c), (2), (5). 201 See Landeshauptmann von Wien (n 38) paras 25–45. 202 ibid para 42. 203 See Long-Term Residents Directive 2003/109/EC, arts 8(4)–(6), 12(3a)–(3c), 19a, 22(3a), as amended by Directive 2011/51/EU; ch 13.9.3; and Steve Peers, ‘Transfer of International Protection and European Union Law’ (2012) 24 IJRL 527. 204 See the Eurostat dataset (n 179) for ‘MIGR_ACQ’ for naturalisation, not ius soli acquisition.

Acquisition of Nationality 499 by establishing outer limits on acquisition and loss (15.7.2). Disputes about golden passport schemes, with Member States effectively ‘selling’ nationality to rich people, have stirred controversies in recent years (15.7.3).

15.7.1 Conceptual lacuna of EU migration law Naturalisation has been mentioned in numerous policy papers over the years. By way of example, the Tampere Conclusions endorsed the objective that long-term residents ‘be offered the opportunity to obtain the nationality’.205 Similarly, the Court accepts, in a judgment on Union citizenship, that naturalisation indicates that applicants ‘have sought . . . to become more deeply integrated in the society’206. Nationality brings about full residence security and complete equal treatment; it can also improve cultural integration by fostering a sense of belonging. To be sure, structural deficits may persist, and the normative claim to equal membership is often disputed. Nevertheless, the acquisition of nationality is the final step in the process of status change. Migration law stops applying when foreigners turn into citizens. Against this background, there are cogent reasons why nationality laws should be coordinated, to prevent stark discrepancies between national practices.207 In spite of this, the political and legal odds are against decisive moves in this direction in the foreseeable future. To establish a supranational competence for harmonisation would require Treaty change, which could possibly even be vetoed by national constitutional courts, relying on the constitutional identity proviso.208 Nothing in the wording of Article 79(2) TFEU indicates that nationality laws may be harmonised, and informal coordination remains elusive as well. Nationality is treated as a symbol of national autonomy, in accordance with Article 4(2) TEU. Explicit confirmation, in Article 9 TEU, that Union citizenship ‘shall be additional to national citizenship and shall not replace it’ can be traced back to the first and negative Danish referendum on the Treaty of Maastricht, whose outcome not even the project of the Constitutional Treaty would have revisited.209 Any attempt at harmonising nationality laws could be blocked by numerous veto players. Conversely, silence on nationality law does not mean that EU migration law has nothing to say on rights which are reserved to members of a community. European societies witnessed the disaggregation of membership statuses after the Second World War. Rights and duties that had traditionally been concentrated in nationality have been made available to multiple non-nationals on the basis of migration laws; long-term residence is, together with Union citizenship, a pronounced expression of this trend.210 Disaggregation reversed the earlier concentration of membership rights in state nationality that had characterised the 205 European Council (n 24) No 21; see also, amongst others, Council Resolution (n 170) No 2; and ‘Common Basic Principles’ (n 50) No 6. 206 Case C-165/16 Lounes EU:C:2017:862, para 58. 207 See Sara Iglesias Sánchez, ‘Nationality’ in Elspeth Guild and others (eds), The Reconceptualization of European Union Citizenship (Brill/Nijhoff 2014) 68, 69–75. 208 See eg the Lisbon judgment of the Federal Constitutional Court (Bundesverfassungsgericht), Case 2 BvE 2/08 and others (judgment of 30 June 2009) paras 249, 348–50. 209 See Eva Ersbøll, ‘Let Third-Country Nationals Become Citizens in Host Member States and of the European Union’ Verfassungsblog (22 January 2019). 210 See generally Hammar (n 3) ch 1; and Seyla Benhabib, The Rights of Others (CUP 2004) 147–65.

500 Integration and Settlement long 19th century, as traced in our introductory Chapter 1.2. For our purposes, disaggregation entails that equal treatment provisions and long-term residence make some—not all—membership rights available to non-nationals. Inability to influence nationality laws can motivate EU institutions indirectly to get a foot in the door, by employing long-term residence as a proxy. The highpoint of this development was the promulgation of ‘civic citizenship’.211 Realisation of this project would effectively have replicated the Union citizenship model for third country nationals, including enhanced political participation—an idea that was quietly abandoned. While some countries had extended municipal voting during the 1990s, this tendency has essentially stopped.212 Similarly, the idea of autonomously acquiring Union citizenship, independent of nationality, is bound to fail for legal, political, and conceptual reasons.213 In the context of Brexit, the Court highlighted the ‘inseparable and exclusive link’ between nationality and Union citizenship.214 Abandoning civic citizenship does not prevent indirect feedback loops. In many countries, absence of the right to vote and full residence security is the main difference between permanent residence and nationality. These differences matter for some, especially those with an interest in politics and anyone posing a public policy threat. For many others, however, permanent residence will be just all right. German statistics show that several million third country nationals would satisfy the conditions for naturalisation but do not apply for many reasons, including lack of immediate benefits.215 This conclusion extends to Union citizens. Many migrants have many rights, as ‘denizens’, but the path to full membership lies in the acquisition of nationality.

15.7.2 Member State prerogatives and their limits under EU law Article 9 TEU underscores that Union citizenship shall be acquired by means of naturalisation at the national level. Settled case law confirms: ‘Under international law, it is for each Member State, having due regard to Community law, to lay down the conditions for the acquisition and loss of nationality’.216 Deference to domestic law meant, for instance, that the Court abstained from reviewing British practices, which had reserved nationality to citizens of the UK and Gibraltar, not, however, other subjects of the Commonwealth; the judgment echoed a cautious ECtHR ruling on family reunification from the former colonies.217 Judges considerably reinforced the rights associated with Union citizenship, while leaving

211 See n 27 and accompanying text. 212 See Federico Fabbrini, Fundamental Rights in Europe (OUP 2014) ch 3; and Anja Lansbergen and Jo Shaw, ‘National Membership Models in a Multilevel Europe’ (2010) 8 ICON 50. 213 See Rainer Bauböck, ‘Why European Citizenship?’ (2007) 8 Theor Inq Law 453; and Martijn van den Brink, ‘The Relationship between National and EU Citizenship’ in Dora Kostakopoulou and Daniel Thym (eds), Research Handbook on European Union Citizenship Law and Policy (Edward Elgar Publishing 2022) 100. 214 See Case C-673/20 Préfet du Gers and Institut National de la Statistique et des Études Économiques EU:C:2022:449, paras 47–49. 215 See Expert Council on Integration and Migration (Sachverständigenrat für Integration und Migration), ‘Normalfall Diversität? Jahresgutachten 2021’ (May 2021) 33–67, whose vice-chairperson the author was at the time of writing. 216 Case C-369/90 Micheletti EU:C:1992:295, para 10. 217 See ch 1.1.6; Case C-192/99 Kaur EU:C:2001:106, paras 19–25; and Nadine El-Enany, Bordering Britain. Law, Race and Empire (Manchester UP 2020) ch 3.

Acquisition of Nationality 501 status acquisition to domestic laws. That changed in a series of judgments, starting with the Rottmann ruling, which established outer limits for the acquisition and loss of nationality. Judges did not call into question the derivative character of Union citizenship, as a corollary of nationality, but required Member States to respect general principles of Union law when nationality is acquired or lost.218 More specifically, domestic authorities must perform a proportionality test, considering the consequences of statelessness for the rights associated with Union citizenship.219 Similarly, loss of nationality after ten years of residence abroad, subject to several conditions, was found to require an individualised assessment, at least in scenarios in which minor children might lose nationality otherwise.220 These inroads into nationality law may be conceptually significant but did not fundamentally reverse the overall conclusion of limited impact. Judges are careful not to interfere in domestic policy choices, while empowering domestic courts to exercise closer scrutiny. That will be practically relevant in countries where nationality law has traditionally been perceived to be an area of executive discretion. Compliance with general principles extends to the acquisition of nationality.221 The Court concluded, for the first time, that a specific outcome was disproportionate: withdrawal of an earlier assurance as to grant nationality, after the renunciation of home state nationality, on grounds of administrative offences.222 The judgment reaffirms that overtly restrictive practices can be corrected in light of Union law. At the same time, the Court highlighted that Member States can ‘protect the special relationship of solidarity and good faith’223 which nationality embodies. On several occasions, it referred to international law, including the European Convention on Nationality of the Council of Europe.224 Regulatory options enshrined therein appear to be legitimate, provided they comply with the proportionality test. Judges effectively obliged Austrian authorities to prevent statelessness in the above-mentioned scenario, but there is nothing in the reasoning indicating that the choice against multiple nationalities was illegitimate to start with.225 Nationality laws are diverse. A useful tool for comparative research is the Global Citizenship Observatory (GLOBALCIT), which contains reliable information on legislative developments in countries across Europe and worldwide.226 The genuine link criterion is commonly referred to as the baseline of international law for the acquisition and loss of nationality.227 In the practice of the Member States it comes in diverse forms. The well- known distinction between the French ‘republican’ model with its focus on ius soli and the German ‘ethno-cultural’ vision emphasising ius sanguinis has given way, in the historic analysis and in contemporary practices, to a pluriverse of overlapping trends which defy easy

218 See Case C-135/08 Rottmann EU:C:2010:104, paras 39–45. 219 ibid paras 50–58; and Gerard-René de Groot and Ngo Chun Luk, ‘Twenty Years of CJEU Jurisprudence on Citizenship’ (2014) 15 German LJ 821. 220 See ch 10.5; and Case C-221/17 Tjebbes and others EU:C:2019:189. 221 See Case C-118/20 Wiener Landesregierung EU:C:2022:34, paras 47–50. 222 ibid paras 51–73. 223 Tjebbes and others (n 220) paras 33. 224 ibid para 37; Rottmann (n 218) paras 53–54; Tjebbes and others (n 220) para 37; Wiener Landesregierung (n 221) para 55; and European Convention on Nationality (adopted 6 November 1997, entered into force 1 March 2000) ETS No 166. 225 See Wiener Landesregierung (n 221) para 47. 226 See https://globalcit.eu (accessed 1 March 2023). 227 See Nottebohm (Liechtenstein v Guatemala) (Judgment, Second Phase) [1955] ICJ Rep 4, 23; and Tjebbes and others (n 220) paras 35.

502 Integration and Settlement categorisation and do not feed into overarching patterns. In a symbolic recognition of the idea of ius soli, the Commission proposes to extend long-term resident status (not nationality) automatically to children born in the territory.228

15.7.3 ‘Golden passport’ schemes Globalisation has formed a class of super-rich individuals. They prefer to travel easily across state-borders and may wish to have a safe haven in times of trouble. Russian oligarchs and Chinese industrialists are prototypes of such rich foreigners who might find it attractive to have easy access to the Schengen area, both for short visits and longer stay. Several Member States jumped on the bandwagon by offering ‘golden passport’ or ‘golden visa’ schemes. These programmes come in diverse forms, ranging from residence permits to entrepreneurs, which are not usually criticised, to naturalisation in return for little more than direct payments, thus effectively ‘selling’ passports or visas. This open commodification has been subject to marked normative and political criticism.229 The Russian war of aggression against Ukraine gave further impetus to the debate. Malta and Cyprus are at the centre of attention in the EU context,230 although the Cypriot scheme was discontinued in response to criticism of corruption and privileges for Russians during 2022.231 Golden passports and golden visas are problematic from a European perspective, since many beneficiaries will consider Malta or Cyprus to be a springboard for visa-free travel and legal residence elsewhere. The option of quick onward movement is an important advantage of ‘golden passports’ over ‘golden visas’, which habitually require longer residence and bring about less transnational mobility, as described in Chapter 12.4.1 on the Schengen area. Long-term residence even requires, as we have seen, five years of continuous residence. Chapter 7.3.3 indicated that Member States may rely on the prohibition of abuse to dispute illegitimate long-term residence permits. Moreover, the Commission recommended to withdraw passports and visas from Russian nationals supporting the attack on Ukraine.232 Golden passport schemes are difficult to attack from a legal perspective as a result of state prerogatives for nationality law. It can be argued, however, that some practices amount to disloyal behaviour in violation of Article 4(3) TEU.233 Moreover, state practices can possibly be confronted indirectly, under recourse to EU rules on money laundering, tax evasion, and security checks.234 The Commission has started infringement proceedings against Malta and Cyprus and referred the Maltese case to the Court in October 2022. The Parliament called upon the Commission to advance these procedures and to present legislation limiting the use of ‘golden visa’ schemes.235 There is little doubt that Article 79 TFEU, together 228 See LTRD Recast Proposal (n 70) art 15(1). 229 See Rainer Bauböck (ed), Debating Transformations of National Citizenship (Springer 2018). 230 See Commission, ‘Staff Working Document’ SWD(2019) 5 of 23 January 2019. 231 See ‘Russia’s War Puts Europe’s “Golden Passport” under Microscope’ Bloomberg (15 March 2022). 232 See Commission Recommendation on immediate steps in the context of the Russian invasion of Ukraine, C(2022) 2028 final, Nos 3 and 4. 233 ibid No 1. 234 See Commission, ‘Report on investor citizenship and residence schemes in the EU’ COM(2019) 12 final, 10–18. 235 See European Parliament, ‘Resolution with proposals to the Commission on citizenship and residence by investment schemes’ (P9_TA(2022)0065, 9 March 2022) Nos 14, 21.

Summary 503 with single market competences, would cover such ‘negative harmonisation’, but it remains a challenge to design legislation.236 Common rules have the advantage of being predictable, instead of reliance on abstract criteria whose precise meaning would have to be determined during years of legal proceedings.

15.8 Summary Many state policies influence migrant integration, which is understood, in the interdisciplinary analysis, to embrace at least three components: structural integration in the sense of equal participation in the labour market or the health system; social integration through social contacts and patterns of identification; and cultural integration by means of learning the language or respect for shared values. While structural inclusion is generally accepted as desirable, underlying models of social togetherness often give rise to normative and political disputes; they range from ethno-cultural closure over multicultural diversity to moderately communitarian forms of cohesion. These debates have informed the evolution of European migration law insofar as they gained traction in the classic destination countries after the millennium change. These debates prompted the design of mandatory integration requirements, notably pre-departure language tests. Such ‘explicit’ integration conditions are one instrument amongst others with which states try to influence migrant integration. They coincide with ‘implicit’ integration provisions, in particular self-sufficiency requirements, equal treatment clauses, or criteria for expulsion. The perspective of migration law is complemented by other policies at the local, national, and supranational levels, which often follow a different trajectory than the focus of migration law on hierarchical order and command. Social policy and schooling, for instance, habitually follow a supportive approach, as do the funding opportunities for integration-related projects under the EU budget. State policies on migrant integration are diverse and produce mixed outputs. In this overall context, the contribution of EU migration law essentially concerns rules governing entry and stay, and the informal coordination of other state policies. EU Treaties do not establish a generic legislative competence for other aspects of migrant integration. Most Member States had embraced territorial presence as the door-opener for social benefits after the Second World War, albeit subject to exceptions. Equality provisions in the migration law instruments take up both tendencies: they provide for equal treatment in a variety of domains, while allowing Member States to introduce or maintain derogations in some respects. The contents and scope of the equality provisions differs between the instruments, although a certain degree of cross-sectoral harmonisation was achieved by the Single Permit Directive. Several judgments have established abstract criteria on the conditions which Member States must fulfil when invoking a derogation. Uncertainties persist as to when and how Member States can justify unequal treatment when secondary legislation does not provide for an express derogation. An essential guarantee concerns equal treatment in the ‘branches of social security’, in accordance with the broad definition in the Social Security Coordination Regulation (EC) 236 See Meenakshi Fernandes and others, ‘Avenues for EU Action on Citizenship and Residence by Investment Schemes: Study for the European Parliament’ (PE 694.217, October 2021).

504 Integration and Settlement No 883/2004. By contrast, means-tested ‘social assistance’ and other benefits are usually not covered by the equal treatment provisions, with the exception of international and temporary protection. The practical effects of equal treatment can be curtailed by the factual context of inter-state movements. Recognition of diplomas, for instance, works reasonably well for university graduates but can cause frictions with regard to professional qualifications whenever countries of origin do not operate formalised vocational and professional training. Migrants may fail the equivalence test. Similar problems concern the export of social benefits upon return. European migration law conceives of the legal component of migrants’ biographies as a process of status change. Temporary permits for specific purposes give way to permanent residence, or expulsion and return. Long-term residence and nationality give secure statuses and can be summarised under the heading of ‘settlement’. Adoption of the Long-Term Residents Directive 2003/109/EC was crucial in this respect; it severed the linkage to the purpose and the conditions of first admission. A blue card holder who obtains long-term residence can stop working without losing the permit. Grounds for loss and withdrawal are essentially confined to serious crime and fraud; dependence on social benefits does not justify non-renewal, as opposed to status acquisition. Judges held that absence from the territory of the Union does not result in the termination of long-term residence either, provided beneficiaries visit at least once a year. This generous interpretation underscores why long-term residence can be referred to as ‘citizenship light’. Naturalisation is the endpoint of migration law. Instruments discussed in this volume stop applying when third country nationals turn into citizens. Against this backdrop, silence on nationality law is a conceptual lacuna of EU migration law. Policy papers have mentioned the acquisition of nationality occasionally, albeit at a high level of abstraction and without interfering with domestic policy choices. The symbolic relevance of nationality renders it unlikely that the situation will change in the foreseeable future. Nevertheless, the Court has started controlling the outer limits of domestic rules on the acquisition and loss of nationality. Member States must apply a proportionality test, which may be used to censure extreme national practices. By contrast, nothing in the rulings indicates that national parliaments loose the freedom to choose between different options that are permitted under international law, including the prohibition of double nationality. ‘Golden passport’ schemes, with Member States effectively selling entry and free movement, have stirred controversies in recent years. They may soon end up before the Court of Justice.

16

Irregular Presence and Return Article 79(1) Treaty on the Functioning of the European Union (TFEU) calls upon the institutions to ensure the ‘prevention of and enhanced measures to combat illegal immigration’. Policy initiatives at the domestic and supranational levels have developed dynamically over the years and were buttressed by the use of combative language. The thematic scope of the supranational response comprises preventive and reactive instruments, ranging from cooperation with third states over visa policy and border controls to measures on the territory and return to home states. Within this overall context, this chapter will concentrate on the domestic component of the ‘combat’ against illegal migration. Visa policy, border controls, and international cooperation are discussed elsewhere. Hence, the choice for the term ‘presence’ in the chapter title. Language is a sensitive issue in any debate about ‘illegal migration’, as the Treaties put it. That is why our comments begin with a critical reflection on the meaning and implications of the language of illegality. Uncertainties reach beyond the linguistic domain and the critical impetus to use alternative wording. Migration law partakes in the definition of where lawfulness ends and illegality starts—a distinction that is not always precise (16.1). Human rights are crucial, in particular with respect to detention and the socioeconomic status of migrants in an irregular situation (16.2). EU legislation on ‘smuggling’, trafficking in human beings, the criminalisation of illegal stay, and employer sanctions constitute a complex collection of restrictive measures and protective instruments at the interface between domestic rules and international law (16.3). The Return Directive serves as the legal backbone of the supranational rulebook presented in this chapter. Entry bans are one important instrument covered by the Directive; they are currently being transformed into an overarching prohibition of entry and stay (16.4). The Return Directive focuses on procedural and substantive standards for return, instead of harmonising the reasons for illegal stay (16.5). These guarantees are complemented by several instruments on inter-state cooperation within the Union, as well as readmission agreements with third states (16.6). Rules on the administrative detention of asylum seekers and returnees rightly receive much political and legal attention and feature prominently in the case law (16.7). By contrast, EU institutions largely abstain from interfering with the regularisation of illegal stay, which most Member States pursue in one way or another (16.8). Databases and Frontex were discussed in Chapters 8 and 9.

16.1 Theory and Policy Design Most legal experts will intuitively conceive of ‘illegality’ as a straightforward concept. Migration law demonstrates that this need not be the case. Upon inspection, the binary contrast between legality and illegality gives way to a multifaceted overall picture, which is European Migration Law. Daniel Thym, Oxford University Press. © Daniel Thym 2023. DOI: 10.1093/oso/9780192894274.003.0017

506 Irregular Presence and Return one of the reasons why some prefer speaking of ‘irregular’ migration (16.1.1). EU institutions have produced important initiatives. Numerous instruments were adopted to improve the return rate, but the underlying quest for effectiveness remains (16.1.2). Social and political dynamics reveal a basic tension at the heart of contemporary return policy, which help to explain divergences between the rhetorical emphasis on fighting illegal migration and practices on the ground (16.1.3).

16.1.1 Beyond binary conceptions of (il)legality ‘No human being is illegal’ is a rallying call. Pro-migrant groups employ the slogan to challenge prevalent restrictions, thus countering widespread references to ‘illegal migration’ in the public debate. The slogan is commonly attributed to Holocaust survivor and Nobel Laureate Elie Wiesel and appears to have been popularised by activist artists.1 Irrespective of the origin, the criticism of the ‘illegality’ label reflects discourse analysis insofar as the use of language can influence our thinking and actions, as described in the interdisciplinary Chapter 4.3. These indirect effects valid if we recognise that European migration law does not negate personhood when describing specific behaviours to be in breach of the law: the verdict of unlawfulness under EU legislation concerns ‘illegal stay’2 and ‘illegal employment’,3 not migrants as such. One step further, critical studies emphasise that the law partakes in the construction of illegality. By way of example, the ‘white list’ under the Visa List Regulation (EU) 2018/ 1806 determines who may cross the external borders without a visa. The law is not, in other words, neutral in the production of illegality. The proliferation and sophistication of entry restrictions increased the relevance of illegal entry and stay.4 Some speak of ‘illegalised’ movements to highlight the element of political choice behind the definition of illegality.5 Such language coincides with core assumptions of critical studies in the tradition of deconstruction theory, traced in Chapter 4.3 on interdisciplinary perspectives. To speak of ‘illegalised’ migration rhetorically puts the blame on states, instead of designating cross- border movements as a potential security threat. Inspection of the legislative detail reveals further complexity. The legal status of migrants often embraces different facets of (il)legality simultaneously and can change over time: people entering lawfully with a Schengen visa do not have labour market access and can engage, therefore, in ‘illegal work’; ‘overstayers’ do not return after the expiry of the visa, as they are obliged to do; the term ‘non-removable’ is common in EU migration law for those who do not comply with a return decision and are not forcibly removed either. Yet even these apparently precise descriptions of illegality display elements of ‘semi-legality’.6 By 1 See ‘No Human Being Is Illegal and Elie Wiesel’ Long Island Wins (6 July 2016) https://longislandwins.com/ news/national/no-human-being-is-illegal-and-elie-wiesel>; and https://en.wikipedia.org/wiki/No_one_is_illegal (both accessed 1 March 2023). 2 Return Directive 2008/115/EC, art 3(2). 3 Employer Sanctions Directive 2009/52/EC, art 2(d). 4 See Catherine Dauvergne, Making People Illegal (CUP 2008); the argument goes back to Nicholas De Genova, ‘The Legal Production of Mexican/Migrant “Illegality”’ (2004) 2 Latino Stud 160. 5 See Harald Bauder, ‘Why We Should Use the Term “Illegalized” Refugee or Immigrant: A Commentary’ (2014) 26 IJRL 327; and Tobias Klarmann, Illegalisierte Migration (Nomos 2021) ch 1. 6 Agnieszka Kubal, ‘Conceptualizing Semi-Legality in Migration Research’ (2013) 47 L & Soc Rev 555.

Theory and Policy Design 507 way of example, the Employer Sanctions Directive guarantees the payment of outstanding remuneration in situations of illicit employment; ‘non-removable’ migrants receive a document certifying their presence; and several countries have introduced ‘firewalls’ to prevent migration law from impeding the use of education or health services. The binary logic of (il) legality gives way to multiple nuances.7 One step further, political actors and academics promote the use of different labels to indicate critical distance from the contents of the law. To use alternative wording presents a semantic stumbling block that obliges the reader to reflect the rationale behind the law.8 Most widely used is the notion of ‘irregular’ migration, which this volume employs on various occasions, although comments in this chapter, in particular, will mostly reproduce the terminology of the instruments under discussion. Throughout Europe various other descriptions exist, and their use is not uniform. Popular alternatives are sans papiers, clandestinos, and unauthorised.

16.1.2 Relentless search for effectiveness An intergovernmental resolution, adopted prior to the Treaty of Maastricht, set out core aspects that define the EU’s approach to irregular migration until today: swift removal of unsuccessful asylum applicants; restrictions on liberty; prosecution of facilitators; and cooperation with third states.9 The Tampere Conclusions even called upon the institutions to ‘stop’10 illegal immigration, through an emphasis on visas, border controls, and international cooperation. With regard to domestic measures, the early debate concentrated on inter-state cooperation as a spillover of the abolition of internal border controls, such as transfers within the Schengen area or the mutual recognition of expulsion decisions.11 Harmonisation of domestic rules on return entered the picture when justice and home affairs became an end in itself. A series of Council conclusions and Commission communications gradually carved out the supranational approach.12 In the new millennium, the ‘fight against terrorism’ and the Palermo Protocol against the smuggling of migrants presented the context for the proliferation of combative language and restrictive initiatives, as highlighted by interdisciplinary research on ‘securitisation’.13 The Parliamentary Assembly of the Council of Europe was active as well, emphasising vulnerability and human rights as an alternative frame of reference;14 these international developments influenced the negotiations on today’s Return 7 See generally Franck Düvell, ‘Paths into Irregularity’ (2011) 13 EJML 275; and Anna Triandafyllidou and Laura Bartolini, ‘Understanding Irregularity’ in Sarah Spencer and Anna Triandafyllidou (eds), Migrants with Irregular Status in Europe (Springer 2020) 11, 13–19. 8 See Cathryn Costello, The Human Rights of Migrants and Refugees in European Law (OUP 2016) 64–67. 9 See Ministers responsible for Immigration, ‘Recommendation regarding practices followed by Member States on expulsion’ (Ad hoc Group on Immigration, 30 November 1992). 10 European Council, ‘Presidency Conclusions of the Meeting in Tampere’ (15–16 October 1999) No 3. 11 See Kay Hailbronner, Immigration and Asylum Law and Policy of the European Union (Kluwer 2000) 467–81. 12 See ch 2.1.2; Commission, ‘Communication on a common policy on illegal immigration’ COM(2001) 672 final; ‘Proposal for a comprehensive plan to combat illegal immigration’ (Council doc 6621/1/02, 27 February 2002); Commission, ‘Communication on a community return policy on illegal residents’ COM(2002) 564 final; and ‘Proposal for a Return Action Programme’ (Council doc 14673/02, 25 November 2002). 13 See ch 4.3.4. 14 See PACE, ‘Expulsion procedures in conformity with human rights and enforced with respect for safety and dignity’ (Recommendation 1547, 22 January 2002); and Sara Iglesias Sánchez, ‘Irregular Migrants in Europe’ in

508 Irregular Presence and Return Directive.15 Adoption of this instrument under the co-decision procedure caused heated policy debates and worldwide media attention, with non-governmental organisations (NGOs) passionately branding the instrument as the ‘directive of shame’.16 Governments are unhappy with the state of play as well, because of ineffectiveness. Low ‘return rates’, which stand far below 50 per cent for most countries of origin,17 are widely invoked as a justification for restrictions. EU institutions have invested much political capital in delivering practical results when the arrival of more than one million asylum applicants propelled return towards the top of the political agenda during 2015/16.18 Critics emphasise insufficient or unreliable data underlying the seemingly objective return rate, including legitimate reasons why some are not returned.19 However, the factual compliance and enforcement deficit remains sizeable, even if we account for legitimate reasons. Among the multiple factors explaining this outcome are the lack of cooperation on the part of third states, practical difficulties like the availability of travel documents, and administrative resources.

16.1.3 Contrasting policy dynamics Return is influenced, more than many other subjects discussed in this book, by the confluence of diverse impulses of multiple actors. Besides international cooperation, highlighted in Chapter 18.1.4, the structure of the labour market, decentralised law enforcement, social affiliation, and migrant agency influence outcomes. All these factors hinder the realisation of the abstract policy objective of increasing return rates which defines the policy debate in Brussels and most national capitals. This marked discrepancy between the political declarations of intent and the lived experience of non-removal need not be a contraction in terms: political scientists and social psychologists speak of ‘control signals’ through which governments reassert their ability to manage migration at a symbolic level, in full awareness of practical difficulties.20 Interdisciplinary analyses highlight economic factors as one element sustaining irregularity. Agriculture, construction, and household services are known for informal work, which, generally speaking, is higher in the South than in the North of Europe. The economic boom in the Mediterranean countries helps to explain the rise of unlawful stay after the millennium change, and its decline during the financial crisis (moreover, EU Francesca Ippolito and Sara Iglesias Sánchez (eds), Protecting Vulnerable Groups (Hart Publishing 2015) 429, 432–36. 15 See Tamás Molnár, The Interplay between the EU’s Return Acquis and International Law (Edward Elgar Publishing 2021) 84–99. 16 See Fabian Lutz, The Negotiations on the Return Directive (Wolf 2011) 83–90; ch 2.2.4.2; and Diego Acosta Arcarazo, ‘The Good, the Bad and the Ugly in EU Migration Law’ (2009) 11 EJML 19. 17 See Court of Auditors, ‘EU readmission cooperation with third countries’ (Special Report 17/2021, 13 September 2021) 7–10. 18 See Commission, ‘Communication: EU action plan on return’ COM(2015) 453 final; and ‘Communication on a more effective return policy in the European Union: A renewed action plan’ COM(2017) 200 final. 19 See Izabella Majcher, ‘Evaluation of the Implementation of the Return Directive (Part II): European Implementation Assessment’ (PE 642.840, June 2020) 64–65; the entry into force of the Entry/Exit System will help to improve the dataset. 20 See ch 4.2.3; and Bastian A Vollmer, ‘Policy Discourses on Irregular Migration in the EU’ (2011) 13 EJML 317, 331–38.

Constitutional Foundations 509 enlargement legalised the stay of irregular workers from some countries); more recently, numbers have fluctuated depending on the number of asylum applications.21 Instruments discussed below endow migration authorities with additional leverage, without undoing these contextual economic factors. Sociolegal research in the ethnographic tradition, presented in Chapter 4.2.4 and 4.2.5, emphasises the factual leeway of ‘street-level bureaucrats’ and the agency of individual migrants who manage, more or less successfully, to build a life in the shadow of governmental restrictions. Legal debates about the ‘risk of absconding’ as a ground for detention respond to one element of the multiple facets of migrant agency: returnees develop strategies to evade or forestall forced removal. Factors influencing street-level implementation are similarly complex, ranging from the lack of administrative capacities to the complexity of return procedures.22 These practical difficulties are the other side of the political control signals. On the ground, the abstract debate about ‘illegal migration’ means the deportation of human beings with social contacts. Support groups, business interests, compassion among bureaucrats, and scepticism against the use of force can translate into localised dynamics which contradict the abstract control signals.23 Sanctuary practices are one expression of how abstract declarations of intent can be frustrated at the stage of implementation.24 Again, social philosophy offers an explanation: territorial presence often generates a process of social affiliation which sustains an ethical and social momentum against return, thus reiterating the abstract finding of a liberal constraint of Western democracies.25 Social affiliation prepares the ground for regularisation programmes. Moreover, states react to the built-in weaknesses of law enforcement within the territory by means of distance creation. The new focus on border procedures can be rationalised, from an interdisciplinary perspective, as an attempt at hindering social affiliation. When analysing these contrasting dynamics, we should be careful not to support binary interpretations, such as the (bad) central institutions versus the (good) local level. Contrasting dynamics for and against return define the views of most citizens and the substance of state policies to varying degrees.26 The situation on the ground ultimately depends on the specificities of the domestic context of each Member State, as well as local characteristics. A holistic assessment of return policy has to move beyond the abstract supranational policy debate and consider the distinct national setting, especially when assessing the real- life impact.

16.2 Constitutional Foundations EU institutions may rely on a broad range of competences to build a common policy on the prevention of irregular movements and return (16.2.1). Human rights play a critical 21 See Triandafyllidou and Bartolini (n 7) 16–25. 22 See Court of Auditors, ‘Asylum, relocation and return of migrants’ (Special Report No 24/2019, 13 November 2019). 23 See Antje Ellermann, States against Migrants (CUP 2009). 24 See Randy Lippert and Sean Rehaag (eds), Sanctuary Practices in International Perspectives (Routledge 2012). 25 See ch 4.2.2; Linda Bosniak, ‘Being Here. Ethical Territoriality and the Rights of Immigrants’ (2007) 8 Theor Inq L 389; and Ruth Rubio-Marín, Immigration as a Democratic Challenge (CUP 2000). 26 See Sébastien Chauvin and Blanca Garcés-Mascareñas, ‘Contradictions in the Moral Economy of Migrant Irregularity’ in Spencer and Triandafyllidou (n 7) 33.

510 Irregular Presence and Return role in the legal analysis. Particularly important are constraints for administrative detention (16.2.2) and socioeconomic guarantees for migrants in an irregular situation (16.2.3). Finally, readers are reminded of the special status of Ireland and Denmark (16.2.4).

16.2.1 Supranational competences Article 79(2)(c) TFEU covers a wide range of ‘measures’, which may include legislative harmonisation and other instruments in support of return. Databases, financial support, and operational assistance by Frontex are among the policy initiatives which are sustained by the Treaty competence. Internal measures in response to irregular presence are at the centre of the Union competence, paradigmatically in the form of employer sanctions and return procedures. Express reference to ‘removal and repatriation’ clarifies that pre-removal detention and forced returns are covered as well, whereas regularisation programmes could be harmonised on the basis of Article 79(2)(a) TFEU. There is nothing in the general wording of that provision on legal migration indicating that regularisation shall be excluded, although any activation remains a political choice, subject to the principle of subsidiarity. Subsidiarity generally argues for caution whenever the legal and factual context differs markedly, as it does with regard to regularisation. Primary law contains an element of linguistic ambivalence when speaking of ‘unauthorised residence’, thus taking up, indirectly at least, the language of ‘irregularity’ many critical observers prefer using. This formulation goes back to the drafting process of the erstwhile Constitutional Treaty, which departed from the uniform reference to ‘illegality’ in the English version of the Treaty of Amsterdam.27 No explanation was given, and a comparison with other language versions unearths a puzzling coexistence of different formulations. The French version, in which the proposals for the Constitutional Treaty had been drafted, referred to ‘immigration clandestine et le séjour irrégulier’ in the Treaty of Amsterdam and today’s Article 79(2)(c) TFEU, while the German text used the adjective ‘illegal’ instead of ‘unauthorised’ uninterruptedly, in contrast to the English version. Multilingualism explains the linguistic diversion. In light of Treaty objectives to prevent illegal migration and to manage migration flows ‘at all stages’, Article 79(2)(c) TFEU need not be confined to the geographical border area. It may cover support for third states to respond to irregular movements, in the form of return operations or otherwise, provided there is a linkage with migration towards the EU. Frontex may possibly be given that power in the future, especially in neighbouring states. The additional competence for combating trafficking in persons, in Article 79(2)(d) TFEU, does not go beyond what could be done on the basis of other provisions; its introduction mirrors the preoccupation of the policy debate in the early 2000s with trafficking.28 The same can be said about Article 79(3) TFEU, whose introduction responded to earlier debates about whether the EU possessed an implicit and shared external competence for readmission agreements. 27 The language was first proposed by the Presidium, ‘Area of Freedom, Security and Justice: Draft Article 31, Part One; Draft Articles from Part Two’ (CONV 614/03, 14 March 2003) 16; it departed from EC Treaty, art 63(3) (b), as amended by the Treaty of Amsterdam. 28 See Daniel Thym, ‘Legal Framework for EU Immigration Policy’ in Daniel Thym and Kay Hailbronner (eds), EU Immigration and Asylum Law: Article-by-Article Commentary (3rd edn, CH Beck/Hart Publishing/Nomos 2022) MN 21.

Constitutional Foundations 511 An express legal basis overcame these uncertainties.29 Note that the competence is shared (‘may’) and that the Return Directive deals with the internal aspects of removal, not cooperation with third states; its adoption did not, therefore, bring about an exclusive external competence for the cooperation with third states.30

16.2.2 Detention in conformity with human rights The right to liberty is firmly established as one of the origins of modern constitutionalism in the form of habeas corpus. Nevertheless, its impact on migration law remains contested, in particular in the admission context. Settled case law holds that not any ‘restriction’ of liberty will amount to a ‘deprivation’ (as ‘detention’ is commonly referred to in the human rights context). Judges asses the individual situation in light of a range of factors to determine whether the deprivation threshold was crossed: the type of the measure, its duration, practical effects, and potential safeguards. On that basis, the ECtHR concluded that critical instances of restrictions in the admission context will not always amount to detention: transit zones at land borders, closed reception facilities for boat arrivals, and border procedures at airports.31 Particularly controversial is the conclusion, as one factor amongst others, that migrants can leave a transit area anytime by going to neighbouring states.32 The CJEU embraced this position of the Court in Strasbourg as a matter of principle, signalling that it will apply these criteria strictly.33 These abstract comments are crucial for the debate about the ‘fiction of non-entry’ in the ongoing debate about asylum policy reform, which was taken up in Chapter 13.8.4 on reception conditions. Article 5(1)(f) of the European Convention on Human Rights (ECHR) establishes two grounds for justifying detention: first, to prevent unauthorised entry into the territory and, secondly, with a view to deportation. In a judgment of principle, the Grand Chamber reaffirmed—against the views of UNHCR, amongst others—that the deprivation of liberty in the context of entry controls under the first limb of Article 5(1)(f) ECHR covers asylum applicants.34 Doing so effectively lowered the standard for justification, with judges applying an arbitrariness yardstick instead of a proportionality test, thus allowing detention on the basis of generalised criteria and for longer periods.35 The Human Rights Committee similarly accepts the initial detention of asylum seekers, while insisting, in accordance with the ECtHR, that detention is neither automatic nor permanent.36

29 ibid MN 22. 30 See ch 18.2.1. 31 See Ilias and Ahmed v Hungary App no 47287/15 (ECtHR [GC], 21 November 2019) §§ 211–18; Amuur v France App no 19776/92 (ECtHR, 15 June 1996) §§ 38–49; JR and others v Greece App no 22696/16 (ECtHR, 25 January 2018) §§ 83–87. 32 Ilias and Ahmed v Hungary (n 31) §§ 241–43. 33 See Joined Cases C-924/19 PPU and C-925/19 PPU Országos Idegenrendészeti Főigazgatóság Dél-alföldi Regionális Igazgatóság EU:C:2020:367, paras 216–25. 34 See Saadi v United Kingdom App no 13229/03 (ECtHR, 29 January 2008) §§ 64–67; and Costello (n 8) 285–93. 35 Saadi v United Kingdom (n 34) §§ 67–75; and Galina Cornelisse, Immigration Detention and Human Rights (Martinus Nijhoff 2010)ch 8. 36 See MMM and others v Australia Communication no 2094/2011 (HRC, 26 July 2013) Nos 9.3–9.4 for detention lasting several years.

512 Irregular Presence and Return Interaction between the ECHR and Article 6 Charter of Fundamental Rights (CFR) remains uncertain, since Article 52(1) CFR can be read to require a proportionality assessment in admission scenarios as well. The CJEU has emphasised the autonomy of Union law.37 While the practical impact of that claim had originally remained ambiguous, with the outcome fluctuating between higher and lower standards in comparison to the ECtHR, judges have recently reaffirmed that the Convention establishes a ‘minimum threshold of protection’.38 That finding does not, however, explain whether the Charter establishes a higher level of protection for detention in transit areas. The CJEU did not explain whether its proportionality assessment concerned human rights or the legislative requirement of a proportionality test in Articles 8 and 9 Reception Conditions Directive 2013/33/EU.39 As long as these articles are not changed, the abstract distinction between the Charter and the European Convention is not practically relevant. A special case might be border procedures, provided we interpret existing or future legislation not to require an assessment of the individual case, in line with the comments towards the end of this chapter. Any justification of the deprivation of liberty requires compliance with a two-step test. The restriction must be provided for by law and be proportionate. The requirement of a sound legal basis allows the ECtHR to censure the absence of a distinct legal framework, which the CJEU has reaffirmed as an integral part of the proportionality test in scenarios of return. With regard to proportionality, states must also avoid indefinite detention, in particular when removal procedures are blocked de facto.40 Moreover, alternatives to detention must be considered, and state authorities shall take into account the specific needs of vulnerable groups. Judges do not generally forestall the detention of families with young children or unaccompanied minors, while insisting that detention conditions respond to specific needs.41 Regarding the detention of asylum seekers and children, European and international practice upholds that it shall take place as a ‘last resort’,42 thus indicating the need for a strict proportionality or arbitrariness test. Some international actors go further and recommend that states should never detain children.43

37 See Case C-601/15 PPU N EU:C:2016:84, paras 44–47; and Case C-18/16 K EU:C:2017:680, paras 32, 50–52. 38 Case C-241/21 Politsei-ja Piirivalveamet EU:C:2022:753, para 47; and generally Giuliana Monina, ‘“Judging” the Grounds for Detention of Asylum Seekers’ in Andrea Crescenzi and others (eds), Asylum and the EU Charter of Fundamental Rights (Editoriale Scientifica 2018) 151; and Joyce De Coninck, ‘Rétention de demandeurs d’asile dans l’Union européenne et instruments parallèles de protection des droits fondamentaux’ (2017) 52 Cahiers de droit européen 83, 101–11. 39 See Országos Idegenrendészeti Főigazgatóság Dél-alföldi Regionális Igazgatóság (n 33) paras 249–66; and N (n 37) paras 47, 54–57. 40 On clear legal basis Politsei-ja Piirivalveamet (n 38) paras 48–54; on the prohibition of indefinite detention Országos Idegenrendészeti Főigazgatóság Dél-alföldi Regionális Igazgatóság (n 33) para 264; and Amuur v France (n 31) § 43. 41 See Muskhadzhiyeva and others v Belgium App no 41442/07 (ECtHR, 19 January 2010) §§ 55–63; and M.H. and others v Croatia App nos 15670/18 and 43115/18 (ECtHR, 18 November 2021) §§ 183–204. 42 See Reception Conditions Directive 2013/33/EU, recital 20, art 11(2); Return Directive 2008/115/EC, art17(1); UNHCR, Guidelines on the Applicable Criteria and Standards relating to the Detention of Asylum-Seekers and Alternatives to Detention (UNHCR, 2012) Nos 28, 51; and Committee of Ministers, ‘Twenty Guidelines on Forced Return’ (Council of Europe, 4 May 2005) Nos 6, 11. 43 See ch 5.5.3; and Izabella Majcher, The European Union Returns Directive and its Compatibility with International Human Rights Law (Brill Nijhoff 2020) 406–15.

Constitutional Foundations 513

16.2.3 Human rights of those irregularly present Article 14 Return Directive 2008/115/EC comprises, as we shall see, a rudimentary set of rights for ‘non-removable’ migrants, including emergency health care and access to primary education. Additional guarantees may be available under domestic laws or human rights. Having said this, any application of fundamental rights is complicated by careful drafting and uncertainties about the doctrinal sway of the ‘second generation’ of socioeconomic rights. By way of example, Article 34(1) and (3) and Article 35 CFR defer to legislation to define the contents of the entitlements to social security and healthcare, whereas access to other benefits under Article 34(2) CFR is limited to everyone ‘residing legally’. These guarantees may, as principles in the sense of Article 52(5) CFR, not comprise directly enforceable judiciable standards independent of legislation.44 The European Social Charter, mentioned in Chapter 5.5.1, limits the scope of the guarantees to those ‘lawfully resident’ or ‘working regularly’.45 Notwithstanding this comparatively straightforward wording, proposals have been put forward to apply the Charter to migrants in an irregular situation by means of dynamic interpretation.46 Guarantees in the UN Migrant Workers Convention caused diplomatic controversies, and Member States refused to ratify the document, as explained in Chapter 5.5.3 on international law. Ratification would not have had much practical effects anyway, considering that the provisions on irregular workers were carefully drafted to concern only remuneration, emergency healthcare, primary education, and a vague reference to regularisation— obligations most Member States would not find difficult to comply with, on paper at least.47 More promising, for those trying to construct a legal argument to the benefit of irregular migrants, may be the idea of minimum core obligations under the International Covenant on Economic, Social and Cultural Rights, although it remains notoriously difficult to define judiciable standards.48 Chapter 10.4 illustrated that the ECtHR often accepts irregularity as a justification for unequal treatment, whereas Chapter 13.2.2 on asylum explained that the outer limits of the judgments on basic socioeconomic needs under Article 3 ECHR and Article 4 CFR remain to be specified. Judges in Strasbourg famously maintain that the protection of private and family life under Article 8 ECHR can bring about a human right to regularise illegal stay ‘in exceptional circumstances’.49 Only a handful of applications have been successful so far, including in scenarios related to the dissolution of the former Soviet Union and the former Yugoslavia which cannot be generalised easily.50 Nevertheless, the argument can be applied to other

44 See ch 5.4.1. 45 Revised European Social Charter (adopted 3 May 1996, entered into force 1 July 1999) ETS No 163, Appendix, No 1. 46 See Antonis Bredimas, ‘Le régime des migrants dans le cadre de la Charte Sociale Européenne et de la pratique du Comité européens de droits sociaux’ in Julia Iliopoulos-Strangas and others (eds), Migration –Migration – Migrations (Nomos/Stämpfli 2017) 191, 196–201. 47 See International Convention on the Protection of the Rights of All Migrant Workers and Members of their Families (adopted 18 December 1990, entered into force 1 July 2003) 2220 UNTS 3, arts 25(3), 28, 30, 69(2). 48 See Committee on Economic, Social and Cultural Rights, ‘General Comment No 3: The Nature of States Parties Obligations (Art. 2, par. 1)’ (UN doc E/1991/23, 14 December 1990) Nos 9, 10. 49 Nunez v Norway App no 55597/09 (ECtHR, 28 June 2011) § 83. 50 See Hoti v Croatia App no 63311/14 (ECtHR, 26 April 2018) §§ 109–17; and Sisojeva v Latvia App no 60654/ 00 (ECtHR [GC], 15 January 2007).

514 Irregular Presence and Return migrants, especially when contextual factors support the claim to regularisation.51 When assessing countervailing interests in the proportionality test, the Court emphasises that ‘factors of immigration control (for example, a history of breaches of immigration law)’52 must be considered to the detriment of migrants. Additional guarantees may be available, as we shall see, on the basis of recent judgments of the Court of Justice on the rights of the child in return proceedings. Any analysis of irregular migration should bear in mind discrepancies between the law in the books and practices on the ground, albeit with countervailing dynamics. On the one hand, local authorities may make public services available, irrespective of whether they have to do so.53 On the other hand, fear of detection can have a chilling effect on access to public services; those who are irregularly staying may not request rights they have on paper. This anxiety may stem from a general scepticism towards the state or be the result of indirect enforcement, with schools or hospitals being obliged to report irregular migrants.54 Some countries have introduced ‘firewalls’ separating social services from law enforcement, either formally or informally,55 thereby reaffirming our introductory comments on contrasting policy dynamics.

16.2.4 Territorial scope Whereas Ireland determines on a case-by-case basis whether to participate, the status of Denmark is more complex. Chapter 2.4 explained that it is usually bound by measures building upon the Schengen acquis as a matter of public international law, mirroring the status of the associated countries Norway, Iceland, Switzerland, and Liechtenstein. By contrast, Denmark cannot, under the current opt-out arrangements, participate in measures which do not build upon the Schengen rulebook. New Member States participate from day one after accession, since the measures on return are not usually connected to the abolition of internal border controls.56 Table 16.1 captures the geographic scope.

16.3 Criminalisation of Illegal Entry and Stay Criminal sanctions for illegal entry and stay are politically controversial, precisely because they convey a value judgment that transcends administrative illegality. EU migration law deals with the matter indirectly and from different angles. Building upon the UN’s Palermo Protocol, the ‘Facilitators Package’ requires harsh sanctions for migrant smuggling, thus 51 See Daniel Thym, ‘Residence as de facto Citizenship?’ in Ruth Rubio Marín (ed), Human Rights and Immigration (OUP 2014) 106, 117–20; and the summary in Pormes v Netherlands App no 25402/14 (ECtHR, 28 July 2020) §§ 51–58. 52 Jeunesse v the Netherlands App no 12738/10 (ECtHR [GC], 3 October 2014) § 107. 53 See EMN, ‘Responses to Long-Term Irregularly Staying Migrants’ (Study, July 2021) 23–28. 54 See Ryszard Cholewinski, Study on Obstacles to Effective Access of Irregular Migrants to Minimum Social Rights (Council of Europe 2005) Pt III. 55 See Francois Crépeau and Bethany Hastie, ‘The Case for “Firewall” Protections for Irregular Migrants’ (2015) 17 EJML 157; critical of firewalls had been Council Recommendation on harmonising means of combating illegal immigration and illegal employment [1996] OJ C5/1, No 4. 56 See eg Act concerning the conditions of accession of the Republic of Croatia [2012] OJ L1112/21, art 4(1) and Annex II, No 8.

Criminalisation of Illegal Entry and Stay 515 Table 16.1 Participation in important instruments on irregular presence and return Instrument

Ireland /UK (before Brexit)

Denmark

Schengen? (NO, IS, CH, LI)

Facilitators Directive 2002/90/EC

yes

no (int. law)

yes

Facilitators Framework Decision 2002/946/JHA

Yes

yesa

yes

Victims of Trafficking Directive 2004/81/EC

no

no

no

Return Directive 2008/ 115/EC

no

no (partly)

no (partly)b

Employer Sanctions Directive 2009/52/EC

no

no

no

aDenmark’s

special position did not embrace intergovernmental decision-making under the former ‘third pillar’. bSome apply in accordance with Decision 1999/437/EC on certain arrangements for the application of the Agreement [1999] OJ L176/31, Pt 1(C).

giving rise to controversies over domestic practices criminalising humanitarian support (16.3.1). Moreover, states employ administrative measures to intimidate search and rescue by private vessels (16.3.2). In a sort of legal paradox, irregular migrants can invoke the objective of effective return to challenge custodial sentences (16.3.3). A peculiar feature of limited practical importance are residence permits for victims of trafficking, to support the prosecution of traffickers (16.3.4). Sanctions for employers hiring people who are illegally staying aim at buttressing compliance with migration law, although the legislation contains protective elements as well (16.3.5).

16.3.1 Prosecution of ‘smuggling’ Penalties for people assisting illegal entry and stay in Article 27 Convention Implementing the Schengen Agreement had been instrumental in the proliferation of administrative and criminal sanctions for the ‘smuggling’ of migrants. The ‘Palermo Protocol’, adopted in the UN framework in the year 2000, goes one step further by obliging states to classify migrant smuggling as a criminal offence.57 The EU’s ‘facilitators package’ builds upon this obligation in the form of the Facilitators Directive 2002/90/EC, as a migration law instrument, and the Facilitators Framework Decision 2002/946/JHA, adopted under the former ‘third pillar’ on criminal matters.58 They are a critical point of reference for anyone studying the criminalisation of ‘smuggling’.

57 See Protocol against the Smuggling of Migrants by Land, Sea, and Air (adopted 15 November 2000, entered into force 28 January 2004) [2006] OJ L262/34, art 6. 58 See Sergio Carrera and others, Policing Humanitarianism (Hart Publishing 2019) 15–23; and Valsamis Mitsilegas, The Criminalisation of Migration in Europe (Springer 2015) 48–52.

516 Irregular Presence and Return In the eyes of prosecutors, one of the weaknesses of the facilitators package is the limitation of criminal jurisdiction to acts committed within the territory or by nationals.59 This hindered the prosecution of the ‘masterminds’ residing abroad, instead of going primarily after the ‘small fish’ steering vessels, guiding migrants through forests, or driving lorries. National legislatures and courts reacted by gradually extending jurisdiction, especially when smugglers leave ships with migrants outside the territorial waters to evade prosecution.60 To do so is not mandatory under the Facilitators Framework Decision, but Union law does not prevent Member States from extending jurisdiction. Moreover, Europol established a ‘European Migrant Smuggling Centre’ to support coordination transnationally, as well as with other EU agencies, including Frontex.61 Such law and order approach to migrant smuggling as an organised crime and a multimillion euro business contrasts with a nuanced outlook of critical scholars accentuating migrant agency, described in Chapter 4.2.5. A controversial aspect of the facilitators package, both during negotiations and in debates ever since, has been the criminalisation of support based on humanitarian or political considerations. The EU’s definition of assistance to illegal entry omits the criterion of ‘financial gain’, in contrast to the Palermo Protocol. It is mentioned in the Facilitators Directive for the facilitation of irregular residence only.62 Instead, Member States ‘may’ decide not to impose sanctions for assisting illegal entry ‘where the aim is to provide humanitarian assistance’.63 This option effectively relegates the decision whether to criminalise humanitarian assistance to the domestic level, although one may possibly argue that the Directive has to be interpreted in light of the Palermo Protocol.64 In spite of this, no preliminary reference has reached the Court so far. Legal disputes about the criminalisation of humanitarian support unfolded before domestic courts. Proceedings against mayors, pastors, activists, and citizens have repeatedly made headline news. While some were convicted, others were acquitted. However, even the fear of prosecution may have chilling effects on humanitarian support and fund-raising.65 In an important victory for humanitarian support groups, the farmer Cédric Herrou won a high-profile case before the French Constitutional Council which found the criminalisation of humanitarian support to violate the French constitutional principle of ‘fraternité’ in scenarios of illegal residence.66 Judges effectively went beyond the requirements of EU legislation that the facilitation of unlawful presence shall only be criminalised when it occurs ‘for financial gain’; they did not, however, oblige the French legislature to decriminalise facilitation of unauthorised entry, as opposed to residence. In response to the ongoing debate, the Commission presented guidance on the criminalisation of humanitarian assistance, which declares unambiguously: ‘the criminalisation 59 See Facilitators Framework Decision 2002/946/JHA, art 4(1). 60 See Simona Ragazzi, ‘New Experiences in Investigating and Prosecuting Migrant Smuggling’ in Valsamis Mitsilegas and others (eds), Securitising Asylum Flows (Brill Nijhoff 2020) 9, 14–20. 61 See https://www.europol.europa.eu/about-europol/european-migrant-smuggling-centre-emsc (accessed 1 March 2023); and David Fernández Rojo, EU Migration Agencies (Edward Elgar Publishing 2020) 77–86. 62 Contrast Facilitators Directive 2002/90/EC, art 1(1)(a), (b) with Palermo Protocol, art 6(1); Convention Implementing the Schengen Agreement(CISA), art 27(1), which had required ‘financial gain’ as well, was repealed by Facilitators Directive, art 5. 63 Facilitators Directive 2002/90/EC, art 1(2). 64 Absence of direct effect, whose repercussions were explained in ch 17.1.2, could possibly be overcome by relying on the intention, on the part of the Council, to implement the Palermo Protocol. 65 See Carrera and others (n 58) chs 4–8. 66 See Constitutional Council (Conseil constitutionnel), Case 2018-717/718 QPC (judgment of 6 July 2018).

Criminalisation of Illegal Entry and Stay 517 of NGOs or any other non-state actors that carry out search and rescue operations at sea . . . amounts to a breach of international law, and therefore is not permitted by EU law’.67 Notwithstanding this robust language, the small print leaves room for interpretation, provided domestic courts stand ready to consider the non-binding position in the first place. It will ultimately depend on the circumstances of the individual case when exactly humanitarian assistance ‘is mandated by law’.68 Moreover, the requirement to ‘comply with the relevant legal framework’ reaffirms state prerogatives for coordinating search and rescue. Arguably, the Commission had no realistic alternative other than to resort to non-binding guidance, since a legislative amendment would have stood little chance of being adopted.

16.3.2 Intimidation of search and rescue Restrictions placed on private organisations in the Central Mediterranean became widespread during the second half of the 2010s. Coastal states carried out rigorous inspections of private vessels to check compliance with safety standards, in search of a pretext to seize the ships or delay their departure.69 Such reasons were often not difficult to find, since many vessels had not originally been built as search and rescue ships. Prominent occurrences were a security decree by the Italian government, proclaimed in 2019; a code of conduct, presented in 2017; and an obligation of instant disembarkation at ports often far away, implemented in 2022/23.70 Such administrative sanctions may be described as a manifestation of ‘criminalisation’ sensu largo, in accordance with the interdisciplinary debate on securitisation traced in Chapter 4.3.4. The matter reached the Court in a case on Italian detention orders against the private vessels Sea Watch 3 and Sea Watch 4, operated by a German NGO. Both ships had rescued hundreds of people and had been involved in several high profile disputes with the Italian and Maltese authorities.71 Detention in the summer of 2020 was defended by technical and operational irregularities. Union law entered the picture in the form of a directive regulating the control powers of port states (Italy), as opposed to flag states (Germany).72 The preliminary reference revolved around the applicability of the Directive and the criteria under which port states may perform additional inspections and seize the ship. While the judgment emphasised the duty to search and rescue in the abstract, the implications were limited. Judges confined the reach of Article 98 UNCLOS to an ongoing mission, without curtailing state supervision in-between rescue operations.73 Crucially, the inspections by port states may relate to the actual use of private ships for high passenger numbers during rescue operations, even if the official registration concerned other purposes (port states cannot, however, require registration for these other purposes in the flag 67 Commission Guidance on the implementation of EU rules on definition and prevention of the facilitation of unauthorised entry [2020] OJ C323/1, No 4(ii). 68 ibid Nos 3, 4; and Violeta Moreno-Lax, ‘Towards a Thousand Little Morias’ in Daniel Thym and Odysseus Academic Network (eds), Reforming the Common European Asylum System (Nomos 2022) 161, 174–76. 69 See Violeta Moreno-Lax and others, ‘The EU Approach on Migration in the Mediterranean (Study for the European Parliament, PE 694.413, June 2021) 102–12. 70 See Francesca Pusterla, ‘Legal Perspectives on Solidarity Crime in Italy’ (2021) 59 Intl Migration 79. 71 See https://fr.wikipedia.org/wiki/Sea-Watch (accessed 1 March 2023); the French and German versions are much better than the English one. 72 See Directive 2009/16/EC on port state control [2009] OJ L131/62. 73 See Joined Cases C-14/21 and C-15/21 Sea Watch EU:C:2022:604, paras 115–19.

518 Irregular Presence and Return state).74 Domestic practices will have to prove whether the Court’s emphasis on the legal thresholds for inspections, remedial action, and seizure will restrain pressure on private vessels in practice. In a half-hearted attempt to rein in excessive state practices, the Commission tabled a Recommendation on search and rescue in 2020. The text remains abstract and will have little impact on the ground.75 Emphasis on the need to ensure safety at sea and to allow competent authorities to monitor and verify compliance, with the Italian code of conduct being mentioned as a model, lent some rhetoric support to restrictive measures.76 The open-ended character of the Recommendation is emblematic of the limited influence of the supranational institutions. Anyone trying to understand what is happening on the ground has to dig deep into what states require in domestic law and are doing in practice.

16.3.3 Limits for criminal sanctions Article 5(3) Schengen Borders Code Regulation (EU) 2016/399 calls for ‘effective, proportionate, and dissuasive’ penalties for irregular border crossings, which can be administrative or criminal in nature.77 The additional reference to ‘international protection obligations’ alludes to Article 31 Refugee Convention, according to which refugees ‘coming directly from a territory where their life or freedom was threatened’ shall not be subject to penalties. Implications of that provision are subject to debate, amongst others regarding the bearing of ‘coming directly’ and whether it covers facilitators.78 Judges in Luxembourg may have to resolve these uncertainties with regard to criminal sanctions for external border crossings under the Schengen Borders Code. By contrast, an interpretation of Article 31 Refugee Convention for the criminalisation of unlawful presence went beyond the Court’s jurisdiction, in the absence an obligation in the Schengen Borders Code Regulation.79 Of course, lack of supranational harmonisation does not absolve domestic courts from applying Article 31 Refugee Convention as international law.80 Against the background of the silence of Union law on the criminalisation of illegal presence, it came as a great surprise that the Court banned imprisonment under specific circumstances. No fewer than eight judgments, in response to preliminary references from different countries, demarcated the authority of the Member States to sanction illegal entry and stay. They originate in a lively Italian debate involving scholars, practitioners, and judges on the implications of the Return Directive 2008/115/EC. That debate focused on fundamental rights and whether imprisonment on accounts of illegal entry and stay had to comply with Article 15 Return Directive.81 Judges in Luxembourg answered that question 74 ibid paras 111–54. 75 See Commission Recommendation (EU) 2020/1365 concerning operations carried out by vessels owned or operated by private entities for the purpose of search and rescue activities [2020] OJ L317/23. 76 ibid recital 14, No 2; and Moreno-Lax (n 68) 169–73. 77 See Case C-47/15 Affum EU:C:2016:408, para 89; previously CISA (n 62) art 3(2); and generally Palermo Protocol, art 6(4). 78 See Gregor Noll, ‘Article 31’ in Andreas Zimmermann (ed), The 1951 Convention Relating to the Status of Refugees and its 1967 Protocol. A Commentary (OUP 2011) 1244. 79 See Case C-481/13 Qurbani EU:C:2014:2101, paras 24–29. 80 See Case C-290/14 Celaj EU:C:2015:640, para 32. 81 See Rosa Raffaelli, ‘Criminalizing Irregular Immigration and the Returns Directive’ (2011) 13 EJML 467, 478–79, 473–77.

Criminalisation of Illegal Entry and Stay 519 to the negative, instead focusing on an alternative line of reasoning: imposing custodial sentences may risk jeopardising the attainment of the objective pursued by the Directive, namely effective and swift removal.82 As often in the judicial practice, the consequences became apparent gradually, when follow-up judgments specified the implications of the short El Dridi ruling. Doing so was much more than an exercise in judicial law-making. Those familiar with the doctrinal foundations will find it easier to rationalise the outcome. Chapter 6.4.5 explained that the Court relies on the effet utile to assess national rules not subject to harmonisation. That proved an appropriate yardstick, as the Return Directive did not harmonise criminal sanctions for illegal entry and stay as such.83 This focus on the effectiveness of the Return Directive implied that criminal sanctions remain possible whenever their imposition does not hinder return proceedings. Member States may not, in other words, impose criminal sanctions instead of adopting and enforcing return decisions.84 Financial penalties, by contrast, were accepted by the Court as being compatible with the Directive, provided they accompany a return decision and do not replace it.85 Criminal sanctions for other offences than illegal entry or stay can be imposed as well.86 The outer limits of national prerogatives in the field of criminal law are defined by the scope of Articles 6(1) and 8(1) Return Directive, which obliges Member States to issue a return decision to anyone staying illegally and to proceed with enforcement thereafter. As a result, the obligation not to undermine removal by means of criminal sanctions applies even before a return decision has been issued, as long as third country nationals are staying illegally.87 While some had understood early judgments to forbid criminal proceedings outright, careful reading reveals that negative repercussions for the return procedure had always been the decisive factor.88 In addition to financial penalties, home detention remains possible, provided that it comes to an end whenever removal may proceed.89 Similarly, custodial sanctions for non-compliance with identification efforts can be legal, as well as imprisonment after the failure of return.90 Considering all these judgments, one is bound to notice an ambiguity that is inherent in any doctrinal argument building upon the effet utile. The practical impact ultimately depends on how criminal sanctions relate to return proceedings. If both are pursued diligently in parallel, Member States may arguably impose criminal sanctions before and after issuing a return decision. By contrast, return proceedings take priority whenever criminal sanctions undermine or substitute removal. In two distinct scenarios, however, criminal sanctions remain possible irrespective of their impact on the return process: firstly, whenever someone comes back in violation of an entry ban after having been removed previously;91 secondly, whenever national legislation is incompatible with the Return Directive 82 See Case C-61/11 PPU El Dridi EU:C:2011:268, paras 55–59; and Celaj (n 80) paras 21–23. 83 El Dridi (n 82) paras 52–54; and Case C-329/11 Achughbabian EU:C:2011:807, paras 28–31. 84 See Case C-38/14 Zaizoune EU:C:2015:260. 85 ibid paras 31–34; and Case C-430/11 Sagor EU:C:2012:777, paras 31–36. 86 See Achughbabian (n 83) paras 40–41. 87 ibid paras 29–32. 88 See Mario Savino, ‘Irregular Migration at the Crossroads’ (2016) 53 CML Rev 1419, 1432–36. 89 See Sagor (n 85) paras 45–46; contra Steve Peers, EU Justice and Home Affairs Law, vol I (4th edn, OUP 2016) 520. 90 See Case C-806/18 JZ EU:C:2020:723, paras 27–29; Affum (n 77) paras 53–54; and Achughbabian (n 83) paras 29–31, 50. 91 See Celaj (n 80) paras 26–30, highlighting dissuasive effects.

520 Irregular Presence and Return but beneficial to individuals, thus putting the onus on national parliaments to change the law.92 These scenarios are the exception to the rule that criminal sanctions for illegal entry and stay must not undermine the effectiveness of return proceedings.

16.3.4 Victims of trafficking Speaking of ‘trafficking’ in a volume on migration law may leave the impression that the activity is transnational by definition, involving movement across borders. That is not the case. Contemporary terminology builds upon the prohibition of slavery and forced labour under Article 4 ECHR. Several international instruments, adopted in the early 2000s, fortified state practices under the headline of preventing and combating ‘trafficking in human beings’. While migrants are, as a consequence of their vulnerability, often victims of trafficking, the concept is not intrinsically linked to migration.93 For our purposes, trafficking is relevant as the EU institutions complemented sweeping criminalisation with a migration- specific instrument: the Victims of Trafficking Directive 2004/81/EC defines conditions for temporary residence permits. As in the case of the facilitators package on smuggling, EU institutions adopted two separate instruments. Whereas the Victims of Trafficking Directive dealt with residence permits, criminal sanctions were covered by a framework decision, which has been replaced by the Anti-Trafficking Directive 2011/36/EU in the meantime. This complex construction was chosen in response to political disputes and Court judgments about the scope of supranational competences for criminal matters. While trafficking in human beings concerns diverse activities on a global scale, it is mainly associated with forced prostitution, exploitative labour, criminal activities, and, controversially, begging in the EU context.94 Critical commentators highlight that the victimhood of women and children played a prominent role in the public discourse, allowing political actors to present criminalisation in a positive light.95 Member States must refrain from prosecuting ‘victims’ of ‘traffickers’ under Article 8 Anti-Trafficking Directive, whereas the ‘customers’ of ‘smugglers’ can be punished for illegal entry and presence in accordance with previous comments on the Facilitators Package. This distinction between the victims of trafficking, to be protected, and the customers of traffickers, to be sanctioned, is illustrative of our previous conclusion, in the interdisciplinary Chapter 4.3.3, that the discursive framing can have an impact on the policy substance. The humanitarian appeal of protecting victims gave criminal sanctions for others a positive touch. While the distinction between smuggling and trafficking is firmly established in the legal material at the supra-and international levels, it can be challenging to demarcate both phenomena in practice.96 92 See Case C-568/19 Subdelegación del Gobierno en Toledo EU:C:2020:807, paras 35–36. 93 See Directive 2011/36/EU on preventing and combating trafficking in human beings and protecting its victims [2011] OJ L101/1, art 2; building upon Protocol to Prevent, Suppress and Punish Trafficking in Persons, especially Women and Children (adopted 15 November 2000, entered into force 25 December 2003) [2006] OJ L262/ 51, art 3(a). 94 See Directive 2011/36/EU (n 93) art 2(3); and Susan Martin and Amber Callaway, ‘Human Trafficking und Smuggling’ in Alexander Betts (ed), Global Migration Governance (OUP 2011) 224. 95 See Elina Pirjatanniemi, ‘Victims of Trafficking in the Migration Discourse’ in Rita Haverkamp and others (eds), What is Wrong with Human Trafficking? (Hart Publishing 2019) 77, 91–94; and Sarah Krieg, Multilevel Regulation against Trafficking in Human Beings (Nomos 2014). 96 Pirjatanniemi (n 95) 84–91.

Criminalisation of Illegal Entry and Stay 521 EU legislation on trafficking had originally been criticised for focusing on law enforcement, instead of prevention and the protection of victims. Nowadays, further safeguards are available on the basis of instruments adopted later.97 Moreover, the ECtHR interprets Article 4 ECHR to cover trafficking and deducts procedural and other obligations from that provision.98 Granting residence permits to victims of trafficking was part of the original criticism.99 Why? What might sound like an act of compassionate benevolence at first turns into instrumentalisation upon inspection. Migrants do not receive a residence permit qua victimhood but for purposes of prosecuting perpetrators. Residence permits are conditional upon cooperation with investigations for a period of up to six months if the presence of the victim will advance prosecution; renewal is contingent upon these conditions still being fulfilled.100 There is no obligation to grant the residence permit, which remains discretionary.101 These limitations are an important reason why few resident permits are granted on the basis of the Victims of Trafficking Directive.102 Limited practical relevance does not mean that victims will necessarily have to leave the country. Some may have regular residence permits for other purposes (remember that trafficking is not intrinsically linked to new arrivals), and Member States remain free to introduce more favourable provisions.103 Moreover, victims may apply for asylum anytime, thus triggering the guarantees for vulnerable people in the Asylum Procedures Directive and the Reception Conditions Directive. While there is no guarantee that they will receive a positive asylum decision, victimhood can coincide with a well-founded fear of persecution or ill-treatment.104 In comparison to asylum legislation, the legal framework for victims of trafficking is week. Migrants cannot be deported while they hold a temporary residence permit for purposes of cooperating in criminal proceedings, as well as during the ‘reflection period’ during which they decide whether to accept the ‘offer’ of temporary regularisation; this reflection period is entrusted to national laws by EU legislation, whereas the Council of Europe requires at least thirty days.105 Rights associated with the status stay behind of what is available for asylum applicants.106 An optional guarantee of free interpretation and legal aid is supposed to support the victim in contributing to the criminal proceedings against traffickers. It reminds us of the purpose of the residence permit of ‘securing optimum cooperation from the victims’.107

97 See Council of Europe Convention on Action against Trafficking in Human Beings (adopted 16 May 2005, entered into force 1 February 2008) ETS No 197, arts 10–12; Directive 2011/36/EU (n 93) arts 11–19; and Directive 2012/29/EU establishing minimum standards on the rights, support and protection of victims of crime [2012] OJ L315/57. 98 See SM v Croatia App no 60561/14 (ECtHR [GC], 25 June 2020) §§ 286–97, 304–20. 99 See Gert Vermeulen, ‘Mutual Instrumentalization of Criminal and Migration Law from an EU Perspective’ (2007) 9 EJML 347, 354–57; and Klarmann (n 5) 252–56. 100 See Victims of Trafficking Directive 2004/81/EC, arts 6, 13–14. 101 See Marcel Kau, ‘Human Trafficking Directive 2004/81/EC’ in Thym and Hailbronner (n 28) Article 8 MN 8; and Peers (n 89) 496. 102 See Commission, ‘Report on the application of Directive 2004/81’ COM(2010) 493 final, 9–10. 103 See Victims of Trafficking Directive 2004/81/EC, art 4; and Council of Europe Convention (n 97) art 14. 104 See Vladislava Stoyanova, ‘Human Trafficking and Refugee Law’ in Satvinder Singh Juss (ed), Research Handbook on International Refugee Law (Edward Elgar Publishing 2019) 324, 328–32. 105 See Victims of Trafficking Directive 2004/81/EC, art 6(2); Council of Europe Convention (n 97) art 13(1); and Case C-66/21 Staatssecretaris van Justitie en Veiligheid EU:C:2022:809, paras 74–76. 106 Victims of Trafficking Directive 2004/81/EC, arts 7, 9–12. 107 Proposal for a Victims of Trafficking Directive, COM(2002) 71 final, 14.

522 Irregular Presence and Return Somewhat unexpectedly, the instrument has gained indirect relevance as a leverage to object to transfer under the Dublin III Regulation. In response to a Dutch reference, the Court qualified the actual transfer to the state responsible, as opposed to the preparation of return, as an enforcement measure that is not permitted during the reflection period under Article 6(2) Victims of Trafficking Directive 2004/81/EC.108 Such postponement renders it more difficult for Member States to meet with the strict time limits for Dublin transfers; non-compliance will result in the transfer of jurisdiction to the state where the applicant is currently staying.109 We can expect the debate about the practical impact of the judgment on the criteria for the reflection period starts running; judges required applicants to make it sufficiently plausible that they are a victim. Doing so requires the submission of information which state authorities will check, without being able to require certainty; AG de la Tour argues that a legal remedy against the refusal of a residence permit under the Directive entails that Dublin proceedings must be stayed.110 Similar questions may arise as to the end of the reflection phase before the expiration of the maximum time period in accordance with Article 6(4).111

16.3.5 Employer sanctions Fighting illegal employment is the main objective of the Employer Sanctions Directive 2009/52/EC. Doing so is supposed to reduce a ‘key pull factor for illegal immigration’.112 That assumption cannot be brushed aside easily, given that labour market prospect is an important driver of migratory movements.113 Debates about illegal work gained momentum during the 1990s when many nationals of the states in Central and Eastern Europe states were working irregularly in Western Europe. The Council recommended employer sanctions as a best practice,114 in accordance with ILO Convention No 143.115 Sanctions against private employers build upon the rationale of the Carrier Sanctions Directive 2001/51/ EC, discussed in Chapter 11.6.1 on visa policy. They were criticised as one-sided insofar as the Commission had not originally proposed legal pathways for unskilled workers in parallel.116 The Seasonal Workers Directive 2014/36/EC responded to this asymmetry, thus complementing pre-existing domestic entry channels. Harmonisation revolves around the prohibition of illegal employment of third country nationals staying illegally.117 This definition entails that unlawful employment of migrants 108 See Staatssecretaris van Justitie en Veiligheid (n 105) paras 53–71. 109 ibid paras 72–76; and ch 13.3.5. 110 ibid, paras 47–50; and AG Jean Richard de la Tour, Opinion in Case C-338/21 Staatssecretaris van Justitie en Veiligheid EU:C:2022:900; the judgment had not been delivered at the time of writing. 111 See AG Jean Richard de la Tour, Opinion in Case C-66/21 Staatssecretaris van Justitie en Veiligheid EU:C:2022:434, points 103–11. 112 Employer Sanctions Directive 2009/52/EC, recital 2, art 1. 113 See ch 4.1. 114 See Council Recommendation on illegal immigration (n 55); Part III Council Recommendation on combating the illegal employment [1996] OJ C304/1, Nos 5 and 6; and Steve Peers, ‘Undercutting Integration’ (1997) 22 EL Rev 76, 81–83. 115 See Migrant Workers (Supplementary Provisions) Convention [No 143] (adopted 24 June 1975, entered into force 9 December 1978) 1120 UNTS 323, arts 2–8; see also Proposal for a Directive to combat illegal migration and illegal employment [1976] OJ C277/2, which was not adopted for lack of competence. 116 See Elspeth Guild, ‘Sanctions for Employers of Irregular Migrants’ in Steve Peers and others (eds), EU Immigration and Asylum Law (Text and Commentary), vol 2 (2nd edn, Martinus Nijhoff 2012) 431, 432–35. 117 See Employer Sanctions Directive 2009/52/EC, recital 5, arts 2(d), 3(1).

Criminalisation of Illegal Entry and Stay 523 with a residence permit is not covered; extending the scope was not pursued, since it might have required unanimity in the Council at the time.118 The abstract definition of ‘illegally staying’ should be interpreted with due regard to other migration law instruments, thus excluding, by way of example, asylum seekers awaiting a final decision. Moreover, Member States may exempt returnees whose removal has been postponed.119 Private households are covered, although they may be subject to less onerous treatment.120 Irregular work in private household is a phenomenon that receives little attention, despite being widespread; think of cleaning or care for children and the elderly. Sanctions can be imposed whenever employers fail to comply with the obligation to check the residence status, keep records, and notify the competent authorities.121 Sanctions include financial payments, administrative measures, and criminal penalties, also for legal persons—a novelty in some countries.122 While the threat of sanctions can have dissuasive effects, practical implementation depends on the willingness of state authorities to inspect employers and to react to misconduct. The proposal of a quantitative 10 per cent inspection target was rejected, arguably rightly so, for lack of flexibility and considerations of subsidiarity.123 The Commission reports that employer sanction appear to have limited impact.124 Measures for protecting the rights of irregular migrants are equally ineffective in practice, partly as a result of the limited suitability of formal mechanisms of legal redress for improving the lot of migrants in an irregular situation.125 On paper, however, the situation looks reasonably well: employers must make back payments; contractors are liable for outstanding obligations of subcontractors; migrants can sue employers, including after return; third parties, such as trade unions and NGOs, may initiate proceedings; and procedures to recover debts may even start automatically.126 Back payments are available irrespective of whether the employment contract was legal under domestic labour law, and a double presumption regarding the length of employment and the level of remuneration facilitates compensation, considering that irregular workers often do not have written contracts.127 By contrast, enrolment in social security schemes is relegated to the domestic legal context; employers must pay equal amounts to the state budget, in case national legislation excludes irregular workers or specific activities.128 As an instrument of migration management, the Employer Sanctions Directive refrains from prescribing the legalisation of irregular stay, with the exception of temporary residence permits for purposes of criminal prosecution against the employer, mirroring similar guarantees for victims of trafficking.129 The idea of a grace period during which removal

118 See Sabine Voglrieder, ‘Die Sanktionsrichtlinie’ [2009] Zeitschrift für Ausländerrecht 168, 171. 119 See Employer Sanctions Directive 2009/52/EC, art 3(3). 120 ibid arts 4(2), 5(3), 7(2). 121 ibid art 4(1). 122 ibid arts 5, 7, 9–12; and Florian Schierle, ‘Employers Sanctions Directive 2009/52/EC’ in Thym and Hailbronner (n 28) Articles 9–12. 123 See Schierle (n 122) Article 14 MN 2–6. 124 See Commission, ‘Report on the application of Directive 2009/52/EC’ COM(2021) 592 final, 4–9, 15–19. 125 ibid 9–14; and FRA, Protecting Migrants in an Irregular Situation from Labour Exploitation (Publications Office 2021) ch 1. 126 See Employer Sanctions Directive 2009/52/EC, arts 6, 8, 13. 127 ibid art 6(1), (3); Schierle (n 122) Article 6 MN 11–14; and Andreas Inghammar, ‘The Employment Contract Revisited’ (2010) 12 EJML 193, 201–206. 128 Employer Sanctions Directive 2009/52/EC, art 6(1)(b). 129 ibid art 13(4).

524 Irregular Presence and Return would be postponed until back payments are recovered was abandoned during the negotiations.130 It depends on the specific national context whether regularisation schemes are available.

16.4 Bans on Entry and Stay Entry bans for ‘unwanted aliens’ have featured in the Schengen Information System (SIS) from the beginning. Chapter 9.2.1 explained that hundreds of thousands of such refusals of entry have been registered in the database at any point in time over the past thirty years. Conditions for entering alerts had been loosely regulated originally.131 As a result, domestic practices varied considerably. By way of example, Germany was the source of a significant share of the alerts, as unsuccessful asylum applicants were registered systematically.132 Adoption of the Return Directive 2008/115/EC was a decisive step towards harmonisation. Additional grounds are scheduled to become operational during 2022 under the new SIS framework (16.4.1), which will complement entry bans in accordance with Article 11 Return Directive (16.4.2).

16.4.1 Refusal for different reasons EU institutions rightly chose to speak of ‘alerts for refusal of entry and stay’133 when updating the SIS framework. Why? The definition of ‘entry ban’ in Article 3(6) Return Directive specifies that it is about ‘prohibiting entry into and stay on the territory of the Member States’. That definition is broader that the term ‘entry ban’ suggests, since it covers migrants staying unlawfully within the territory. That broader outlook becomes relevant when we consider the transnational component of refusal of entry or stay. An alert entered by one Schengen member prevents other countries from issuing a residence permit during the period of validity of the alert. That obligation is rendered effective by an obligation to consult the SIS during border checks; moreover, any domestic authorities may consult the database before issuing any type of residence permit.134 Inter-state consultation is supposed to overcome divergent views as to whether an entry ban shall be supplanted by a residence permit or not. Judgments on the legal consequence of inter-state communication failure were presented in Chapter 7.3 on the administrative dimension. Article 11(5) Return Directive reiterates the transnational consultation requirement, albeit without saying how the authorities that consider delivering a residence permit should learn about an entry ban issued elsewhere. The obvious answer would be ‘via the SIS’, but the legislature had not laid down an obligation to systematically enter alerts under the Return Directive into the database. This shortcoming was remedied in the new legislation.135 130 cf Proposal for an Employer Sanctions Directive, COM(2007) 249 final, art 7(4). 131 See CISA (n 62) art 96(3). 132 See (French) Council of State (Conseil d’Etat), No 190384 Mr and Mrs Forabosco (judgment of 9 June 1999). 133 SIS Border Checks Regulation (EU) 2018/1861, art 24. 134 ibid arts 27, 34(1)(d); Return Directive 2008/115/EC, art 11(4); Schengen Borders Code Regulation (EU) 2016/399, art 8(3)(a)(iv); and also SIS Return Regulation (EU) 2018/1860, art 9; on gradual extension see Steve Peers, ‘Key Legislative Developments on Migration in the European Union: SIS II’ (2008) 10 EJML 77, 87–93. 135 SIS Border Checks Regulation (EU) 2018/1861, arts 24(1)(b); and previously Commission Recommendation (EU) 2017/432 on making returns more effective [2017] OJ L66/15, No 24(c).

Bans on Entry and Stay 525 Judges confirmed that inter-state consultation establishes an obligation of result, not conduct.136 Member States are not, in other words, obliged to find a common ground. In case of disagreement, the residence permit shall prevail over entry and stay, under the former legislative framework at least. Migration-specific alerts can be issued in three situations: firstly, on the occasion of a return decision; secondly, after apprehension during border surveillance in the context of an irregular border crossing; thirdly, whenever illegal stay is discovered during an exit check, for instance when someone leaves after having overstayed his visa. While entry bans accompanying return decisions are governed by the Return Directive, the latter scenarios are new. They are construed as refusal of entry or stay on public policy grounds in accordance with the SIS Border Checks Regulation, which expressly covers irregular border crossings and may be interpreted to include other public policy threats than the ones mentioned by way of example.137 The Commission proposes disconnecting alerts on occasions of exit from the public policy criterion, thus making it easier to issue a re-entry ban.138 The bearing of public policy was discussed in Chapter 10.3; the right to be heard, giving reasons, and effective remedies follow general principles. A judgment on the previous legal framework can hardly be read to prohibit entry bans on public policy grounds at present, given that judges did not address the interaction of different pieces of legislation.139 The revised SIS Border Checks Regulation recognises explicitly that entry bans under the Return Directive are only ‘one of [several] conditions’ for entering alerts.140 Anyone studying entry bans should not, therefore, focus on the Return Directive primarily. The SIS Border Checks Regulation (EU) 2018/1861 is equally important.

16.4.2 Entry bans under the Return Directive Conditions for entry bans in the context of return are subject to a combination of strictness and opening clauses. Article 11(1) Return Directive requires Member States to issue entry bans mandatorily whenever no period for voluntary departure was granted and if the obligation to return is not complied with. The wording (‘shall’) leaves little room for a proportionality assessment,141 although the personal situation must be considered when determining the period of validity of the entry ban, as well as before any administrative decision not to give a period for voluntary departure.142 To be sure, these individualised assessments will usually be performed in the same administrative procedure as the one resulting in an entry ban. Such parallelism does not, however, turn the obligation to issue the entry ban 136 See ch 7.3.2; and Case C-240/17 E EU:C:2018:8, paras 53–55. 137 See SIS Border Checks Regulation (EU) 2018/1861, arts 24(1)(a), 2(c); building upon Recommendation (EU) 2017/432 (n 135) No 24(d). 138 See Proposal for a Recast of the Return Directive COM(2018) 634 final, art 13(2). 139 See Case C-546/19 Westerwaldkreis EU:C:2021:432, paras 40–48; the case concerned a peculiar German argument that an entry ban under the Return Directive had mutated into an alert on public policy grounds later on; see AG Priit Pikamäe, Opinion in ibid EU:C:2021:105, points 86–87. 140 See SIS Border Checks Regulation (EU) 2018/1861, art 24(1). 141 Contra Diego Acosta Arcarazo and Elspeth Guild, ‘The Returns Directive’ in Peers and others (n 116) 483, 498–99; and Peter Boeles, ‘Entry Bans and SIS Alerts’ in Karin Zwaan (ed), The Returns Directive (Wolf 2011) 39, 41–43. 142 See Return Directive 2008/115/EC, art 7(4); and Case C-554/13 Zh and O EU:C:2015:377, para 50.

526 Irregular Presence and Return into an option, subject to proportionality; administrative discretion concerns the length of validity, not the choice whether to issue an entry ban. Moreover, an individualised assessment must be performed whenever national authorities issue non-mandatory entry bans or refrain from issuing mandatory entry bans on humanitarian grounds.143 These optional provisions leave ample room to comply with human rights whenever Article 8 ECHR requires states not to issue an entry ban or to limit its duration.144 Additional guarantees are available for family members of Union citizens.145 Ongoing debates about the recast of the Return Directive do not concern entry bans, although the mandatory character of the criteria when periods for voluntary departure shall not be granted would result in more compulsory entry bans. Procedural guarantees in Articles 12 and 13 Return Directive must be applied in accordance with the case law and the general principles discussed in Chapter 7.2 on the administrative dimension. Whenever national authorities issue an entry ban, they must determine its length motu proprio, irrespective of whether the returnee had requested so.146 States can issue entry bans and return decisions in a joint administrative procedure.147 They may also disentangle both decisions, especially when a period for voluntary departure was granted148 (alternatively, they may issue a mandatory return decision subject to the condition precedent that the obligation to return is not being complied with). The clock for the calculation of the period of validity starts ticking when the addressee leaves Union territory, not at the moment of the administrative decision.149 Member States are free to impose custodial sentences for illegal stay once the return procedure has been followed unsuccessfully. They may not, however, punish non-compliance with an entry ban, as opposed to sanctioning unlawful stay, before someone has left Union territory.150 Entry bans must be withdrawn together with the return decision, or anytime on humanitarian grounds.151

16.5 Structure of the Return Directive Instead of harmonising the reasons for illegal stay, the Return Directive 2008/115/EC concentrates on procedural and substantive standards for return (16.5.1). Several exceptions limit the scope ratione materiae and loci, thus giving Member States leeway to organise the return process without much supranational intervention (16.5.2). Whenever the Return Directive applies, domestic authorities are required to follow a sequence of successive stages, starting with the adoption of a return decision (16.5.3), followed by a period for voluntary departure and the option of forced removal (16.5.4). By contrast, Union law scarcely addresses the status of ‘non-removable’ migrants (16.5.5). Entry bans, detention, regularisation, and return to other Member States are mentioned elsewhere. Negotiations 143 Return Directive 2008/115/EC, art 11(1)(2), (2), (3)(3). 144 See ch 15.2.2; by way of example Savran v Denmark App no 57467/15 (ECtHR [GC], 7 December 2021) § 182; and Majcher (n 43) ch 6. 145 See ch 14.3.1; and Case C-82/16 KA and others EU:C:2018:308, paras 85–97. 146 See Case C-297/12 Filev and Osmani EU:C:2013:569, paras 27–30. 147 See Return Directive 2008/115/EC, art 6(6). 148 See also Case C-225/16 Ouhrami EU:C:2017:590, para 50. 149 ibid paras 38–53. 150 ibid paras 54–56; and JZ (n 90) paras 30–40. 151 See Return Directive 2008/115/EC, art 11(3)(3); and Westerwaldkreis (n 139) paras 52–54.

Structure of the Return Directive 527 on a Recast Proposal, tabled in 2018, were ongoing at the time of writing; it was uncertain whether they would be completed before the European elections in 2024. Their contents will be mentioned whenever appropriate.

16.5.1 Reasons for illegal stay In a multilingual legal order, the same word can have different meanings across the Union. ‘Expulsion’ is one such false friend, referring to the administrative act terminating legal residence in some countries, while primarily addressing the process of removal elsewhere. Early legislation used the term expulsion nonetheless, often in an open-ended manner without clarifying whether it was about the criteria for terminating legal stay or departure from the territory.152 Similarly, the direct linguistic equivalent of ‘deportation’ is associated with illegal renditions during the Nazi dictatorship or Soviet rule in some countries. EU institutions evaded these linguistic pitfalls by introducing ‘return’ as an overarching category, with ‘removal’ as the enforcement component. Member States are not obliged to follow this supranational terminology when transposing the Return Directive, provided that domestic legislation respects its contents. A central choice was to limit the scope of the Return Directive to the return process, instead of dealing with the reasons for illegal stay. Any application of the Directive presupposes that a third country national is illegally staying because he does not fulfil, or no longer fulfils, the conditions for entry and stay.153 Reasons for legal entry and stay are governed by either Union law or domestic legislation.154 By way of example, a residence permit may be withdrawn on public policy grounds; some may fail the conditions for status renewal; others may be overstayers. The legal and factual reasons for unlawful stay are as diverse as the instruments discussed in this volume. A special case are asylum seekers. They are not covered for as long as they have a right to remain under Article 9(1) Asylum Procedures Directive 2013/32/EU, namely until the administrative authority has taken a decision. If the outcome is negative, the asylum applicant must be considered ‘staying illegally’ for the purposes of the Return Directive, irrespective of whether she may remain on the territory until a domestic court has decided on appeal in accordance with Article 46(5)–(10) Asylum Procedures Directive.155 This outcome may be counter-intuitive but it follows, amongst others, from the option, enshrined in Article 6(6) Return Directive, to adopt the return decision jointly with the asylum decision.156 In two respects, however, asylum legislation supplants the Return Directive: living standards and detention are governed by the Reception Conditions cited in Directive 2013/33/EU pending the outcome of any appeal.157

152

By way of example see Long-Term Residents Directive 2003/109/EC, arts 9, 12; and CISA (n 62) art 23(3). See Return Directive 2008/115/EC, arts 2(1), 3(2). 154 See also Achughbabian (n 83) paras 28–30. 155 See Case C-181/16 Gnandi EU:C:2018:465, paras 36–59. 156 ibid paras 49–50. 157 ibid paras 43–45, 63; building upon Case C-534/11 Arslan EU:C:2013:343, paras 40–49. 153

528 Irregular Presence and Return

16.5.2 Exceptions from the scope Member States may decide not to apply the Return Directive to third country nationals subject to a refusal of entry or apprehended ‘in connection with the irregular crossing’ under Article 2(2)(a)—an option most countries are using.158 That exception effectively excludes external border controls from the scope of the Directive. Geographically, any application of the exception requires ‘a direct temporal and spatial link with that crossing of the border’.159 Whether that is the case, depends on the circumstances of the individual case and may cover, for instance, boat arrivals, people who have climbed border fences, or those who are intercepted in forests.160 Judges held that temporary border controls within the Schengen area do not transform an internal border into an external border where states may decide not to apply the Directive.161 Another limitation ratione materiae concerns return as a consequence of a criminal conviction, such as drug trafficking, provided that national laws foresees such linkage.162 The exception does not authorise, as we have seen, the criminalisation of illegal stay. Whenever one of the exceptions applies, some guarantees apply under Article 4(4). This minimum threshold of protection would gain significance if the Proposal for an Instrumentalisation Regulation in response to ‘hybrid threats’ was adopted. The Commission envisages introducing, by means of a lex specialis, another exception from the Return Directive.163 ‘Emergency return management procedures’ would cover asylum applicants who are not currently covered by Article 2(2)(a) as a consequence of the right to remain until the outcome of the procedure.164 The Proposal is another example of asylum reform extending leeway by means of deference to domestic laws.

16.5.3 Return decision Return is not a matter of state discretion: both the adoption of the return decision and enforcement are an obligation under Union law, subject to several exceptions.165 The return decision presents itself as the quasi-automatic consequence of illegal stay: it ‘stat[es] or declar[es]’166 stay to be illegal. States are not required to check the legality of the previous administrative decision terminating lawful stay anew on the occasion of legal remedies against the return decision.167 The Commission proposes laying down an explicit obligation to issue a return decision immediately after any decision terminating legal stay (an obligation that is implicit in Article 6(1) at present); moreover, the option to combine a negative asylum decision with a return decision, under Article 6(6), shall become mandatory.168 The 158 See Majcher (n 19) 43–45. 159 See Affum (n 77) paras 41–42. 160 See Commission Recommendation (EU) 2017/2338 establishing a common ‘Return Handbook’ [2017] OJ L339/83, No 2.1. 161 ibid Nos 59–60; and Case C-444/17 Arib and others EU:C:2019:220, paras 44–67. 162 See Return Directive 2008/115/EC, art 2(2)(b); and Filev and Osmani (n 146) paras 50–56. 163 See ch 10.4; and Proposal for an Instrumentalisation Regulation, COM(2021) 890 final, art 4. 164 See Fabian Lutz, ‘Return Directive 2008/115/EC’ in Thym and Hailbronner (n 28) Article 2 MN 14. 165 See Return Directive 2008/115/EC, arts 6(1), 8(1). 166 ibid art 3(4). 167 See Case C-184/16 Petrea EU:C:2017:684, paras 57–65. 168 See Amended Proposal for an Asylum Procedures Regulation COM(2020) 611 final, art 35a; and Return Directive Recast Proposal (n 138) art 8(6)(1); see also Recommendation (EU) 2017/432 (n 133) No 5(c).

Structure of the Return Directive 529 underlying idea is swift return, notably in the context of future asylum border procedures for carrying out return, which would be subject to shorter time limits and, possibly, stricter rules on detention.169 Accelerated decision-making would put the spotlight on the relationship between the administrative decision that gives rise to illegal stay (qua domestic law) and the return procedure (qua Union law). Their interaction may work reasonably well whenever domestic procedures preceding the return decision ascertain in a reliable manner whether someone can be returned legally. On the contrary, the focus of the Return Directive on departure from the territory can cause frictions when the preceding domestic measures fail to ensure compliance with the prohibition of refoulement. National practices are diverse in this respect: some countries foresee a distinct non-refoulement screening during the return process; others assess human rights under the heading of complementary protection as an integral part of the asylum procedure (new risks can be asserted by means of a subsequent application); some countries, however, rely on administrative authorities implicitly checking the prohibition of refoulement when issuing a return decision, or national courts doing so in the context of legal remedies under Article 13(1) Return Directive.170 The example illustrates the pitfalls of a multi-level system whose effectiveness depends on the seamless interaction of domestic laws and supranational legislation. Introduction of an additional procedure qua Union law to prevent illegal refoulement would complicate things considerably for Member States providing for adequate safeguards. It is crystal clear that the Return Directive requires domestic authorities to ensure full compliance with human rights throughout the return process, as reaffirmed by Articles 5 and 9(1)(a), but the Return Directive leaves the instruments and methods how that is to be achieved to the domestic level. Judges give Member States leeway in this respect. A real risk of refoulement which arises after a return decision has been issued does not necessitate a withdrawal of the return decision. Rather, removal shall be postponed, until the risk disappears or domestic authorities issue a residence permit.171 By contrast, the Court concluded that Member States must refrain, in light of the general scheme of the Return Directive, from issuing a return decision whenever the real risk of refoulement exists before that date.172 That judgment effectively introduces another reason not to adopt a return decision, in addition to the list in Article 6(2)–(5) Return Directive. Enhanced protection is available, as we shall see, for children and parents. Articles 12 and 13 establishes a skeleton of procedural safeguards, which must be interpreted in accordance with the general principles and human rights discussed in Chapter 7.2 on the administrative dimension. Judges generally interpret these principles stricter in the context of return than in judgments on asylum, but they establish a basic safety net nonetheless: the right to be heard does not require separate hearings during the various stages of the return procedure; denial of a hearing need not entail the illegality of the return decision; reasons need not be detailed; suspensive effect of legal remedies is warranted whenever there is a real risk of refoulement; time limits may be short but not too short; and case law on legal assistance is complex. The only element where the Return Directive evidently falls

169

Amended APR Proposal (n 168) arts 41g–41i; and ch 13.4.4. See Majcher (n 19) 49–54. 171 See Return Directive 2008/115/EC, art 9(1)(a); and Westerwaldkreis (n 139) paras 56–59. 172 See Case C-69/21 Staatssecretaris van Justitie en Veiligheid EU:C:2022:913, paras 55–58. 170

530 Irregular Presence and Return foul of human rights is the option in Article 13(1) that an ‘administrative authority’ may decide legal remedies; the Charter requires domestic courts to exercise this function.173 Commission proposals to streamline the procedure, on the occasion of the ongoing recast negotiations, would have to comply with human rights.

16.5.4 Voluntary departure and removal Directive 2008/115 conceives of return as ‘the process of a third-country national going back—whether in voluntary compliance with an obligation to return, or enforced—to the country of origin’.174 Return is, in other words, an overarching term comprising both voluntary departure and forced removal. The general scheme of the Return Directive sets out the procedure and fixes the order in which its stages should take place; their intensity must be applied in accordance with the principle of proportionality.175 Voluntary departure takes priority over removal and less coercive measures shall be used as a last resort according to Article 8(4), which also requires Member States only to resort to ‘reasonable force’. Periods for voluntary departure of between seven and thirty days must be specified in the return decision in accordance with Article 7(1), although third country nationals may leave earlier. Member States may extend the period in due consideration of the individual case, for instance to facilitate legalization for purposes of family reunification under domestic laws.176 Judges confirmed that the exceptions where no period of voluntary departure is granted are enumerative.177 Interpretation of the public policy caveat was discussed in Chapter 10.3, and the meaning of ‘risk of absconding’ will be mentioned in the context of detention. The Commission wants to make the exceptions mandatory and render voluntary departure optional during border procedures.178 Doing so would speed up the return procedure and result in more mandatory entry bans. Voluntary departure is the preferred option in many respects: it is less invasive of personal liberty; much cheaper from a financial perspective; allows policy-makers to give return a humanitarian appearance; and can support return where forced removal is no realistic option, for instance if countries of origin do not accept return flights. Several Member States have introduced return counselling and dedicated ‘assisted voluntary return’ (AVR) packages. Migrants receive money to pay for transportation to the country of origin, which states are not generally obliged to subsidise; many countries also offer financial ‘starter kits’ as an additional sweetener. Further support may be available upon return for reintegration in the labour market.179 AVR originates in state practices towards ‘guest workers’ in the 1980s and beneficiaries of temporary protection during the wars in the former Yugoslavia.

173 See ch 3.1.3. 174 Return Directive 2008/115/EC, art 3(3). 175 Reaffirmed by El Dridi (n 82) paras 34, 38, 41. 176 See Case C-409/20 Subdelegación del Gobierno en Pontevedra EU:C:2022:148, paras 51–52, 56–58. 177 See El Dridi (n 82) para 37. 178 See Return Directive Recast Proposal (n 138) art 9(4); and Amended APR Proposal (n 168) art 41(4). 179 See Ulrike Brandl, ‘Voluntary Departure as a Priority’ in Madalina Moraru and others (eds), Law and Judicial Dialogue on the Return of Irregular Migrants from the European Union (Hart Publishing 2020) 83.

Structure of the Return Directive 531 At the time, substantial numbers returned home, mostly voluntarily, and support packages helped to tilt the decision towards repatriation.180 The Commission supports AVR as best practice, financial support has been made available under the migration funds, and Frontex will introduce support tools.181 Ongoing recast negotiations are likely to result in an obligation of the Member States to introduce programmes for assisted return and reintegration; their scope ratione personae is to be determined domestically, also to prevent financial support upon return from rendering irregular entry attractive.182 Use of the term ‘voluntary’ marks a linguistic distance away from forced removal; it does not usually indicate that going back is a matter of free choice, although that may be the case (think of repatriation when a war ends).183 ‘Voluntary’ return takes place in the shadow of the option of enforcement and the reality of prolonged limbo situations whenever removal is postponed. Practically speaking, migrants will usually return to the country of origin or third state to which they have a personal connection. From a legal perspective, Article 3(3) defines return as the process of going back to either the home state or ‘countries of transit’. That formulation may cover safe third countries and first countries of asylum under the Asylum Procedures Directive, or any other third state through which the migrant has passed. The notion of ‘transit’ implies physical presence, without the need for any additional qualitative criteria.184 Countries where someone had resided permanently, without having nationality, are probably covered as well. By contrast, third states where returnees had never been before are not a legitimate destination, unless the returnees agree to being sent there.185 The country of destination must be specified in the return decision, and any change of that country requires the adoption of a new return decision, subject to legal remedies.186 Expiration of the period for voluntary departure entails, as we have seen, an obligation to proceed with forced removal. Hardly any Member State seems to require a separate removal order under Article 8(3), which the Commission had originally proposed to be mandatory. Legal remedies are available against ‘decisions’ related to return under Article 13(1); this formulation covers the return decision, while it remains unclear whether legal remedies must also be available against operational action as a matter of Union law.187 Chapter 5.4.4 on human rights explained that it has to be ascertained on a case-by-case basis whether operational enforcement measures are covered by the Charter or not. Monitoring has become widely used, and Article 8(6) Return Directive leaves Member States much flexibility to organise the ‘monitoring system’, which need not foresee the use of monitors during every removal operation.188 Curricula for training and a pool of return monitors will be made available under the auspices of Frontex. 180 See Rosemarie Rogers, ‘Migration Return Policies and Countries of Origin’ in Kay Hailbronner and others (eds), Immigration Admissions (Berghahn Books 1997) 147, 152–65; and Decision 97/340/JHA on the exchange of information concerning assistance for the voluntary repatriation [1997] OJ L147/3. 181 See Commission, ‘Communication: The EU strategy on voluntary return and reintegration’ COM(2021) 120 final. 182 See Return Directive Recast Proposal (n 138) art 14(3). 183 See also NA v Finland App no 25244/18 (ECtHR, 14 November 2019) §§ 58–60. 184 Argumentum e contrario Asylum Procedures Directive 2013/32/EU, art 38(2)(a). 185 See Return Directive 2008/115/EC, art 3(3), third indent. 186 Országos Idegenrendészeti Főigazgatóság Dél-alföldi Regionális Igazgatóság (n 33) paras 114–23. 187 See ch 7.2.3.2; and, for a generous view, see Sergo Mananashvili and Madalina Moraru, ‘Return Directive 2008/115/EC’ in Thym and Hailbronner (n 28) Article 13 MN 10. 188 See Lutz (n 164) Article 8 MN 10–18.

532 Irregular Presence and Return

16.5.5 Limbo of ‘non-removable’ returnees Widespread failure of voluntary return and forced removal entails a phenomenon that is often discussed under the heading of ‘non-removable’ migrants in EU law circles, or under the generic notion of ‘limbo’. These scenarios were a bone of contention during the legislative process: the Council rejected the idea to apply the Reception Conditions Directive, thus requiring states to guarantee an adequate standard of living, including accommodation.189 Article 14 Return Directive was the compromise institutions finally agreed upon, and the Commission does not suggest reopening the debate during the ongoing recast negotiations. The provision establishes a basic safety net. Guarantees under Article 14 are available once the return decision has been issued, including during the period for voluntary departure and whenever legal remedies have suspensive effect.190 The Directive remains ambiguous as to whether there is room for a status of non-removability beyond the reach of Article 14, whenever removal has not been formally postponed in accordance with Article 9.191 That may be the case, in particular, whilst domestic authorities engage in practical preparations for enforcement, or explore the logistics to decide whether removal is feasible. Even if we concluded that Article 14 did not apply during this initial phase, the provision could be relied upon once it has become apparent that there is no reasonable prospect of removal. Article 14(2) reaffirms that de facto suspension of removal is covered by Article 14(1), irrespective of whether a formal administrative decision on postponement has been taken.192 By contrast, illegally staying migrants for which no return decision has been issued are covered by domestic law alone.193 In scenarios where the Return Directive does not apply, namely at the external borders, basic guarantees must be provided in accordance with Article 4(4). Human rights may prove more promising than the Return Directive for anyone trying to advance the position of migrants in an irregular situation. Previous comments, on the constitutional foundations, demonstrated that recent judgments may be interpreted to comply states to provide for basic needs, including in areas not mentioned in Article 14(1)(b)–(d), such as housing or food.194 Having said this, the Court has emphasised the primary responsibility of Member States to ensure compliance.195 Judges have hesitated to move further into the sensitive terrain of basic needs. A development of great practical and constitutional significance are a series of judgments which put flesh on the bones of the best interests of the child; they build on judgments on other segments of EU migration law, mentioned in Chapter 5.4.2 on human rights. While the Court did not outlaw the return of children, limbo situations were found to place children in a ‘situation of great uncertainty as to [their] legal status and [their] future’, in particular as regards schooling, links with a foster family, or the length of stay—a situation 189 cf Proposal for a Return Directive, COM(2005) 391 final, art 13. 190 Reaffirmed by Case C-562/13 Abdida EU:C:2014:2453, paras 55–59. 191 See Peers (n 89) 516. 192 See Case C-402/19 CPAS de Seraing EU:C:2020:759, paras 30, 50; Westerwaldkreis (n 139) para 57; and Lutz (n 164) Article 9 MN 5. 193 See Case C-146/14 Mahdi EU:C:2014:1320, para 87. 194 Abdida (n 190) para 60 maintained that emergency healthcare would be ‘meaningless’ without the provision of unspecified other ‘basic needs’. 195 ibid paras 54, 61; CPAS de Seraing (n 192) paras 51–45; and Case C-673/19 M and others EU:C:2021:127, paras 45–47.

Inter-state Cooperation 533 ‘contrary to’ their best interests.196 Member States must consider the prospect of removal and, in the case of unaccompanied minors, reception in accordance with Article 10(2) before issuing a return decision. If these conditions are not fulfilled, no return decision shall be issued. Parents benefit as well, by means of a legal reflex.197 Special protection for children does not extend to other adults: a return decision shall be issued where the prohibition of refoulement requires removal to be postponed.198 The effects of these judgments are less clear-cut than it might appear at first. Not issuing a return decision does not put an end to illegal stay. One might argue that the best interests of the child require states to legalise illegal stay a fortiori if there is no realistic prospect of removal. However, the Court has not said so and might not have the jurisdiction anyway, given that the status of irregularly staying migrants is not usually covered by Union law and, hence, the Charter whenever the Return Directive does not apply. One may build an argument for mandatory regularisation on the basis of the human rights case law of the ECtHR, which, as we have seen, emphasises the need to assess personal circumstances. Somewhat paradoxically, the guarantees in Article 14(1) Return Directive do not apply either in the absence of a return decision. Such doctrinal considerations will not, however, have much relevance in practice. The level of commitment under both Article 14(1) and human rights law is quite abstract and comparative surveys illustrate that domestic practices vary to a great extent.199 Experts in EU migration law may intuitively call for a common response, but this appears unlikely at this juncture, considering the principled resistance of the Member States to further supranational intervention.200 That outcome may be unsatisfactory from the perspective of policy coherence. For the time being, however, the choice of how to deal with limbo situations rests with the domestic authorities at the national, regional, and local levels, where contrasting policy dynamics can be observed in line with our introductory comments.

16.6 Inter-state Cooperation Removal is understood, in the EU context, as the enforcement of a return decision. Whereas the Return Directive regulates the administrative procedure and requires Member States to resort to coercive measures in accordance with the principle of proportionality, supplementary instruments deal with inter-state cooperation. Early legislation on mutual recognition of expulsion decisions (16.6.1) is superseded by enhanced activities under the auspices of Frontex nowadays (16.6.2). As a consequence of a provision in the Return Directive that is often overlooked, intra-European removals are a rare occurrence whenever the Dublin III Regulation does not apply (16.6.3). Readmission agreements with third states facilitate forced return. They play a prominent role in the public discourse, although they are no legal prerequisite for removal (16.6.4). 196 See Case C-441/19 Staatssecretaris van Justitie en Veiligheid EU:C:2021:9, paras 53–54; and Case C-484/22 Bundesrepublik Deutschland EU:C:2023:122, paras 23–27. 197 See Case C-112/20 État belge EU:C:2021:197, paras 31–42. 198 See Westerwaldkreis (n 139) paras 56–59, delivered a few weeks after the rulings on the best interests of children. 199 See EMN (n 53); and Paul Schoukens and Danny Pieters, ‘Irregular Labour Migrants and Access to Social Protection’ in Gijsbert Vonk (ed), Cross-Border Welfare State (Intersentia 2012) 91. 200 See Fabian Lutz, ‘Non-Removable Returnees under Union Law’ (2018) 20 EJML 28.

534 Irregular Presence and Return

16.6.1 Mutual recognition of expulsion decisions Directive 2001/40/EC on the mutual recognition of expulsion decisions201 is representative of the focus of the early debate on inter-state cooperation. The rationale behind the Directive was to accelerate return, with states enforcing an expulsion order (return decision) issued elsewhere.202 That is easier said than done. Withdrawing a residence permit can be complex, but the response to unlawful stay in another member of the Schengen area is not.203 However, it is the second scenario the Directive deals with: France shall enforce an expulsion order (return decision) the Spanish authorities had issued to remove a Moroccan national who had crossed the Pyrenees. Instead of activating Directive 2001/40/EC, France could simply adopt a return decision of its own, oftentimes without periods for voluntary departure.204 The added value of mutual recognition is limited, and the instrument seems not to be widely used. Nevertheless, the Directive still features in the EU migration law rulebook and cannot be ignored therefore. Adopting a second return decision has the advantage of evading the legal and practical pitfalls mutual recognition inevitably gives rise to. Does de facto suspension of return in the first state, in the sense of lapse of time without removal proceedings being pursued, suffice to prevent recognition by the second state? Difficult to say.205 Can the returnee claim before the authorities of the second state that the authorities of the first state had adopted expulsion decision erroneously? Probably not, since the Directive foresees legal remedies in the issuing country, except when removal would violate human rights.206 Indeed, both the issuing country and the enforcing state have to respect human rights and the mandatory postponement of removal Article 9(1)(a) Return Directive.207 In short, the legal and practical pitfalls of mutual recognition render recourse to the Directive unattractive. That is why it would have been preferable if the legislature had repealed Directive 2001/40/EC when adopting the Return Directive.208 Somewhat surprisingly, Directive 2001/40/EC may possibly experience a renaissance in years to come. EU institutions are discussing whether to introduce ‘return sponsorship’ as a means of ‘flexibly solidarity’ in the successor instrument to the Dublin III Regulation. By way of example, Ireland would assume responsibility for returning an asylum seeker whose application was rejected in a hotspot in Sicily.209 During the initial phase of the sponsorship, Ireland would support Italy, as an administrative assistant, in enforcing the return decision.210 However, Ireland would transfer the person to its territory whenever the person 201 [2001] OJ L149/34. 202 See Kees Groenendijk, ‘The Directive on Mutual Recognition of Expulsion Decisions’ in Philippe De Bruycker (ed), The Emergence of a European Immigration Policy (Bruylant 2003) 447, 455–56. 203 It is not evident that an ‘expulsion decision’ is the same, in part at least, as a ‘return decision’, since it may alternatively be construed as the preceding administrative act that terminates legal stay; if that was the case, the second country would have to adopt a return decision anyway; see Directive 2001/40/EC, art 3(1)(a); and Maren-Kathrin Diekmann, Menschenrechtliche Grenzen des Rückführungsverfahrens in Europa (Nomos 2016) 57–60. 204 Irregular movement from Spain to France will often be an indicator for a risk of absconding under Return Directive 2008/115/EC, art 7(4). 205 See Directive 2001/40/EC, art 3(1)(2). 206 ibid art 4; and Groenendijk, ‘Mutual Recognition’ (n 202) 452–53. 207 See Peers (n 89) 522. 208 cf Proposal for a Return Directive (n 189) arts 16, 20. 209 See ch 13.1.5; and Olivia Sundberg Diez, Florian Trauner, and Marie De Somer, ‘Return Sponsorships in the EU’s New Pact on Migration and Asylum’ (2021) 23 EJML 219, 226–36. 210 See Proposal for an Asylum and Migration Management Regulation, COM(2020) 610 final, art 55(1), (4).

Inter-state Cooperation 535 is not returned within 8 months.211 At that stage, the Irish authorities might be tempted to recognise the Italian return decision, in whose enforcement they had been involved already, under Directive 2001/40/EC, which applies whenever the returnee is present in the enforcing country. If Ireland does so, it might even ask Italy for financial compensation under an instrument most governments have probably forgotten, although doing so might run counter to the effet utile of the idea behind return sponsorship as a solidarity measure.212 At the time of writing, the fate of return sponsorship was wholly uncertain in the absence of any indication that the instrument would be adopted in the foreseeable future.

16.6.2 Cooperation between the Member States Coordination of return policies began as a spillover of the abolition of internal border controls in the Schengen area. As a result, movements between the Member States played a prominent part in the early years, before the focus of attention shifted towards improving return to third states generally. A telling example is Directive 2003/110/EC on transit for the purposes of removal by air, which builds upon earlier intergovernmental arrangements.213 Member States may request neighbouring countries to authorise the transit of returnees to an airport on their territory. By way of example, Luxembourg could ask Belgium, France, or Germany to use commercial or charter flights that take off in Brussels, Paris, or Frankfurt. While the Directive sets out grounds for refusal, silence on the part of the requested authorities is interpreted as consent.214 Practical implementation depends on bilateral arrangements: the requesting country may use its own escorts or rely, in part or in full, upon the assistance of the neighbouring country, whose personnel can support the return operation.215 Additional options for transnational cooperation may be available under bi-or plurilateral agreements on police cooperation, such as the Prüm Convention of 2005, which also covers return by land.216 Legal oversight shall be available in the requesting country, whose enforcement measure the transit country is supporting. Directive 2003/110/EC is about transnational assistance, not about the mutual recognition of return decisions.217 In that respect, the Directive was a precursor of the assistance of Frontex. Chapter 8.3.1 explained that the mandate of the agency has been reinforced considerably. Frontex may support states under pressure in the preparation, adoption, and enforcement of return decisions. The agency also serves as the locus of enhanced operational cooperation on return. Articles 48–53 Frontex Regulation (EU) 2019/1896 lay down multiple tasks ranging from the design of training materials and financial support over the provision 211 ibid art 55(2). 212 See ch 6.4.5; and Decision 2004/191/EC setting out the criteria and practical arrangements for the compensation of the financial imbalances resulting from the application of Directive 2001/40/EC [2004] OJ L60/55. 213 [2003] OJ L321/26; and Decision SCH/Com-ex (98)10 of the Executive Committee on cooperation in returning foreign nationals by air [2000] OJ L239/193. 214 Directive 2003/110/EC (n 213) arts 3–4. 215 ibid arts 5, 7. 216 Prüm Convention on the stepping up of cross-border cooperation, particularly in combating terrorism, cross-border crime, and illegal migration (adopted 27 May 2005, entered into force 23 November 2006) [2006] Federal Law Gazette (Bundesgesetzblatt) II/626, art 23, which has been ratified by many Member States and whose illegal migration component was not superseded by Decision 2008/615/JHA on the stepping up of cross-border cooperation [2008] OJ L210/1. 217 Contra Peers (n 89) 525.

536 Irregular Presence and Return of interpretation to a pool of return monitors to operational guidance, for instance on reference models for IT systems. Dedicated return operations, such as chartering aircrafts and organising joint return flights, are entrusted to Frontex, thus effectively supplanting bilateral cooperation of the Member States.218 The remarkable increase in supranational staff and the upsurge of financial muscle indicate that Frontex will become a critical facilitator of inter-state cooperation in years to come.

16.6.3 Return to other Member States Member States must generally refrain from issuing a return decision whenever someone holds a residence permit for another Member State. Return to that country takes precedence in accordance with Article 6(2) Return Directive 2008/115/EC.219 Administrative procedures for such intra-European expulsion are governed by domestic laws, not the Return Directive. While the Return Directive obliges Member States to permit third country nationals to return a country for which they hold a residence permit, the Directive prevents states from enforcing an obligation to do so. Instead, the country where the migrants is physically present shall respond to the failure of voluntary return to another Member State by means of adopting a return decision. Crucially, that return decision must designate a third state, not another Member State, as the country of destination.220 That outcome may be counter-intuitive, but the Court has confirmed that anyone holding a residence permit for another Member State who is not willing to go there voluntarily is illegally staying in the other state.221 This seemingly peculiar scenario has gained relevance as a result of irregular movements of beneficiaries of international or complementary protection. Chapter 13.3.7 explained that the take back procedure under the Dublin III Regulation (EU) No 604/2013 does not cover secondary movements of beneficiaries of international protection at present, although EU institutions are discussing whether to extend its scope. Against this background, the combined effect of the Dublin III Regulation and the Return Directive is that Member States have to prioritise the removal to countries of origin or transit, even if someone holds a valid residence permit for another Member State. Priority of return to a third state, instead of intra-European transfers, extends to situations where forced removal to a third state is not possible for reasons of law or fact.222 The end result may be a situation hardly anyone will find satisfactory (except for the Member State which had issued the residence permit). Think of an Afghan national who received subsidiary protection in Bulgaria, before moving on to Austria. Austrian authorities will reject an application for asylum as inadmissible under Article 33(2)(a) Asylum Procedures Directive 2013/32/ EU. At the same time, however, they cannot forcibly return that person to Bulgaria and are hindered, for reasons of non-refoulement, to proceed with deportation to the home state. A coherent supranational policy framework looks different.

218 They remain an option under Decision 2004/573/EC on the organisation of joint flights for removals from the territory of two or more Member States [2004] OJ L261/28. 219 See also E (n 136) paras 45–47; and CISA (n 62) art 23. 220 See Return Directive 2008/115/EC, art 3(3); and Affum (n 77) paras 82–87. 221 See M and others (n 195) paras 29–37. 222 ibid paras 38–42.

Inter-state Cooperation 537 By way of exception, states can continue applying bilateral readmission agreements, or other arrangements such as treaties on policy cooperation, under the condition that they had existed when the Return Directive entered into force on 13 January 2009.223 On that basis, third country nationals can be forcibly returned to other Member States, irrespective of whether they have a residence permit for that country or not. The procedure will be governed by domestic legislation and the bilateral agreement in question, since the Return Directive does not apply. France appears to be using that option extensively in relations with Spain, whereas the refusal of entry of migrants coming from Italy follows a different legal rationale.224 An initiative of the Commission to introduce another exception for transfers back to neighbouring countries as part of cross-border police cooperation was mentioned in Chapter 13.3.7 on secondary movements.

16.6.4 Readmission agreements with third states States are obliged to readmit nationals under customary international law, as reaffirmed by the Global Compact for Migration.225 Nevertheless, practical collaboration is often flawed and presents a major reason for the substantial mismatch between the number of return decisions and actual returns. Lack of cooperation relates to various steps of the return procedure: identity determination, acquisition of travel documents, acceptance of the transfer, and organisation of the removal operation.226 Readmission agreements address these practical deficits. Their provisions define procedures for cooperation, means of evidence regarding nationality, necessary travel documents, time limits, and modalities of transportation; supplementary implementing protocols, drawn up by Member States and the third state, designate the competent authorities and provide for practical arrangements.227 Common European standards for a travel document (laissez passer) are supposed to facilitate the return of people without passports or other identity documents of the home state.228 Such technicalities are the backbone of readmission agreements. Member States spearheaded the conclusion of bilateral agreements with countries in Central and Eastern Europe in the 1990s. A specimen bilateral treaty and implementing protocol contained all the practical components that define readmission agreements until today.229 That cooperation was reasonably successful, precisely because it was integrated into a broader external policy framework comprising economic cooperation and some 223 See Return Directive 2008/115/EC, art 6(3); and Lutz (n 164) Article 6 MN 16–19. 224 See ch 12.4.1.2; and Iker Barbero, ‘A Ubiquitous Border for Migrants in Transit and their Rights’ (2020) 22 EJML 366. 225 See ch 5.5.4; and Kay Hailbronner, ‘Readmission Agreements and the Obligation on States under Public International Law to Readmit their Own and Foreign Nationals’ (1997) 57 Heidelberg J Intl L 1. 226 See Court of Auditors (n 17) 38–46. 227 See Nils Coleman, European Readmission Policy (Martinus Nijhoff 2008) ch 5; and see eg Agreement between the European Community and the Islamic Republic of Pakistan on the readmission of persons residing without authorisation (adopted 26 October 2010, entered into force 1 December 2010) [2010] OJ L287/52, arts 2, 5–9, 16–17. 228 See Regulation (EU) 2016/1953 on the establishment of a European travel document for the return of illegally staying third-country nationals [2016] OJ L311/13; which replaced a Council Recommendation [1996] OJ C274/18. 229 See Council Recommendation concerning a specimen bilateral readmission agreement [1996] OJ C274/20; and Council Recommendation on the guiding principles to be followed in drawing up protocols [1996] OJ C274/ 25.

538 Irregular Presence and Return legal pathways.230 EU migration law entered the picture in the form of abstract readmission clauses in association agreements, namely the Cotonou Agreement with more than 70 former colonies, which will be reinforced somewhat in the successor agreement.231 Such abstract readmission clauses, without any practical arrangements, did not deliver tangible results. EU institutions changed course, therefore, and concentrated on negotiating readmission agreements. Years of debate over supranational competences finally subsided when Article 79(3) TFEU established an express legal base in the form of a shared competence, thus allowing Member States to conclude bilateral agreements whenever the EU remains inactive.232 By the end of 2009, the EU had signed thirteen readmission agreements, many of them with neighbouring European countries. Effective cooperation on return was made possible by the political context of good neighbourly relations and attractive ‘package deals’ in the form combined readmission and visa liberalisation agreements.233 Conversely, negotiations with important countries of origin or transit, such as Morocco, Turkey, or Nigeria, proved difficult. Chapter 18.1.4 will explain that negotiations often fail in the absence of a mutually beneficial give-and-take in the field of migration or beyond. Moreover, ratification did not always translate into effective implementation either, for instance in relations with Turkey.234 One stumbling block was removed when the Commission proposed not to insist on the inclusion of irregularly staying third country nationals any longer,235 but the challenge to design a mutually beneficial package remains.236 Chapter 18.3 will introduce readers to the EU’s relentless search for ‘carrots’ and ‘sticks’ to convince third states to cooperate. Widespread focus on non-arrival in recent years can be explained, in part at last, by the entrenched failure to develop a reasonably well-functioning return policy. Our comments on the external dimension will emphasise that legally binding readmission agreements have increasingly given way to non-binding ‘standard operation procedures’, or other informal arrangements. The benefits and pitfalls of informalisation are manifold. In the case of readmission, the objective of smooth practical cooperation on the basis of administrative arrangements could similarly be achieved through informal commitments.237 At the time of writing, there were eighteen readmission agreements and at least six informal cooperation frameworks,238 in addition to numerous bilateral agreements of the Member States.239 While informalisation evades institutional safeguards, returnees can rely on the Return Directive 2008/115/EC and human rights. 230 See ch 18.3.1; and Daphné Bouteillet-Paquet, ‘Passing the Buck’ (2003) 5 EJML 359, 360–68. 231 The Partnership Agreement between the members of the African, Caribbean and Pacific Group of States and the European Community and its Member States (adopted 23 June 2000, entered into force 1 April 2003) [2000] OJ L317/6, art 13(5)(c) will be replaced by the Partnership Agreement between the European Union, of the one part, and the members of the Organisation of African, Caribbean and Pacific States (OACPS), of the other part (not signed yet), art 74, Annex I, which emulates core features of many readmission agreements, albeit at a high level of abstraction; see Commission Proposal for a Council Decision on the signing, COM(2021) 312 final, Annex. 232 See ch 18.2.1; and Martin Schieffer, ‘Community Readmission Agreements with Third Countries’ (2003) 5 EJML 343. 233 See ch 11.3.2; and Coleman (n 227) ch 7. 234 See Çiğdem Akın Yavuz, ‘Analysis of the EU-Turkey Readmission Agreement’ (2019) 21 EJML 486. 235 See Commission, ‘Communication: Evaluation of EU readmission agreements’ COM(2011) 76 final, 5, 9. 236 See Letter from the European Commission, ‘Subject: EU Readmission Cooperation with Partner Countries— State of Play’ (Ares(2022)656813, 28 January 2022), which is not publicly available but can be found online. 237 See ch 18.2.2. 238 See Mark Provera, ‘External Dimension of the Return Directive (Part III): European Implementation Assessment’ (PE 642.840, June 2020) Annexes I–III. 239 See the useful ‘Inventory of the Bilateral Agreements Linked to Readmission‘, coordinated by Jean-Pierre Cassarino https://www.jeanpierrecassarino.com/datasets/ra (accessed 1 March 2023).

Detention 539 As a general lesson, we may conclude that readmission is in essence about administrative inter-state cooperation, whose effectiveness depends on a mutually beneficial package. If that is correct, cooperation with third states on readmission is bound to require ongoing coordination, on the part of the EU, between the national and supranational levels. While Member States organise returns, deliver operational cooperation, and design additional legal pathways, EU institutions provide money and the policy framework, together with the expanding external capacities of Frontex.240 The ensuing balance between national responsibilities, supranational autonomy, and multilevel coordination will have to be constantly adapted in order to create synergies instead of inconsistencies.241 The appointment of a ‘Return Coordinator’ within the Commission, supported by a high level network, pursues the aim of ‘facilitating a seamless and interlinked implementation’ of return processes.242 We shall learn whether the new position will deliver.

16.7 Detention Western democracies portray personal freedom as the normative foundation of constitutionalism. They are, as a result, sceptical when individuals are detained without having committed a crime. This pivot to liberty explains why the ‘administrative’ detention of migrants has turned controversial, both politically and legally.243 Corresponding provisions can be found in different pieces of legislation, whose scope overlaps to some extent (16.7.1). Grounds for detention include the ‘risk of absconding’, as well as other criteria (16.7.2). Crucially, the grounds are a necessary but insufficient justification for detention. Alternatives must be given priority, and the length must be limited in accordance with procedural safeguards (16.7.3). Legislation lays down specific prerequisites for humane conditions, with a focus on vulnerable persons (16.7.4).

16.7.1 Delimitation of EU legislation Administrative detention mostly occurs in the context of entry and return, including asylum procedures. Detention for other purposes is not covered by EU migration law, especially criminal proceedings and extradition.244 A special case are criminal sanctions for illegal stay, which can be incompatible, as we have seen, with the effet utile of the Return Directive. Previous comments on human rights explained that not any ‘restriction’ of liberty will amount to a ‘deprivation’, thus giving states some flexibility to restrict mobility during entry clearance below the threshold of detention. Controversial proposals for a ‘fiction of non-entry’ in the Commission’s asylum reform package, discussed in Chapter 13.8.4, build upon the distinction between restriction and deprivation. It similarly informs debates

240 See ch 18.3–4. 241 See Jean-Pierre Cassarino, ‘Readmission Policy in the European Union: Study for the European Parliament’ (PE 425.632, September 2018) 24–38; and Court of Auditors (n 17) 26–37. 242 See Commission, ‘Communication on a new Pact on Migration and Asylum’ COM(2020) 609 final 8; reviving an idea of ‘The Hague Programme’ [2005] OJ C53/1, Pt III.1.6.4. 243 See Daniel Wilsher, Immigration Detention (CUP 2012). 244 See also Case C-808/18 Commission v Hungary EU:C:2020:1029, para 169.

540 Irregular Presence and Return about whether the treatment of asylum applicants in the ‘hotspots’ on the Greek islands amounts to detention or not.245 The Court confirmed that the notion of ‘detention’ must be interpreted uniformly in EU migration law and coincides with the concept of ‘deprivation’,246 thus not covering less intense restrictions of liberty. Articles 8–11 Reception Conditions Directive 2013/33/EU apply whenever someone has expressed the wish to request asylum, not formal registration with the competent authorities.247 Detention of asylum applicants does not affect their right to remain on the territory pending the outcome of the procedure.248 Grounds for detention in Article 8 Reception Conditions Directive are supplanted by Article 28 Dublin III Regulation (EU) No 604/2013, once national authorities submit a take back or take charge request to another Member State, until the failure of the transfer.249 During that period, a ‘significant risk of absconding’ is the only ground for detention. When an application for asylum is rejected, the Reception Conditions Directive remains applicable, with regard to detention, until the domestic court has delivered a judgment,250 or until the applicant, exceptionally, loses the right to remain.251 Articles 15–18 Return Directive 2008/115/EC take over as the relevant legal benchmark thereafter, provided the situation is covered by the Directive ratione materiae. If one of the exceptions applies, states remain free to regulate the grounds for detention autonomously, in compliance with human rights law.252 The Commission proposes, as we have seen, extending the scope of the exceptions. Moreover, draft provisions on a border procedure for carrying out return can be read to authorise the automatic detention of asylum applicants after rejection.253 Presumably, the idea was that an existing detention order shall remain valid (instead of having to be adopted anew), subject to mandatory review to verify whether detention still complies with Article 9(5) Reception Conditions Directive.

16.7.2 Grounds for detention Grounds for the detention can be found in different pieces of legislation: Article 9 Reception Conditions Directive 2013/33/EU, Article 15 Return Directive 2008/115/EC, and Article 28 Dublin III Regulation (EU) No 604/2013. These provisions revolve around different grounds for detention, including the ‘risk of absconding’. Detention is not always legal if we conclude that one of the grounds for detention applies. Legality requires, in a second step, compliance with the principle of proportionality, discussed hereafter.

245 See ‘UNHCR Redefines Role in Greece as EU‐Turkey Deal Comes into Effect’ (UNHCR Briefing Note, 22 March 2016); and FRA, ‘Update of the 2016 Opinion [Hotspots Update]’ (Opinion 3/2019, 4 March 2019) 58–59. 246 See Országos Idegenrendészeti Főigazgatóság Dél-alföldi Regionális Igazgatóság (n 33) paras 216–25. 247 See ch 13.4.1; and Case C-36/20 PPU Ministerio Fiscal EU:C:2020:495, paras 96–99. 248 See ch 13.8.4; and N (n 37) para 74. 249 See Case C-60/16 Khir Amayry EU:C:2017:675, paras 23–26. 250 See n 159 and accompanying text. 251 See ch 13.8.4; and Case C-269/18 C and others EU:C:2018:544, paras 52–54. 252 See Return Directive 2008/115/EC, art 4(4)(a). 253 See Amended APR Proposal (n 168), recital 40(i), art 41a(5).

Detention 541

16.7.2.1 Risk of absconding A common feature of the different legislative provisions on detention is the ‘risk of absconding’, which defines administrative practice. Reference to ‘objective criteria defined by [national] law’254 entailed that the risk of absconding does not establish an autonomous concept of Union law, to be interpreted uniformly across Europe.255 Rather, the Court concluded that ‘rules of general application’ must exist at the national level, whose ‘wording sets out the limits of the flexibility of [administrative] authorities . . . in a manner that is binding and known in advance’.256 While the legal benchmark for assessing whether there is a risk of absconding will vary from country to country, the question of whether these criteria are objective and sufficiently precise can be assessed in light of Union law by the CJEU. Requiring domestic law to lay down predictable standards satisfies the need, under human rights law, that any interference with the right to liberty shall be foreseeable.257 Authors criticising the directives for being imprecise misunderstand the multilevel character of supranational harmonisation, as laid down in Article 288(3) TFEU. Practices on the ground are informed by domestic legislation, and it is these rules which must be predictable. Moreover, the CJEU’s insistence on ‘rules of general application’ whose ‘wording’ sets out objective standards arguably goes beyond human rights case law, which does not require a formal catalogue of objective criteria for the risk of absconding at the national level. When the Court required so qua Union law, this had a ripple effect on apex courts in several countries demanding legislative change.258 National criteria vary considerably, including non-cooperation in the identification process, opposition to forced return, violation of an entry ban, unauthorised secondary movements, and non-compliance with less coercive alternatives to detention.259 To support further unity, the Commission recommended establishing these criteria as a rebuttable presumption, whereas other circumstances, such as non-compliance with a period for voluntary departure, should be treated as an indicator that there is a risk of absconding.260 While this recommendation relates to the current legal framework, the Recast Proposal comprises a long and non-exclusive list of no fewer than sixteen objective criteria, including rebuttable presumptions.261 Earlier draft legislation on reception conditions during the asylum procedure was more abstract but would probably be aligned to the model set out in the final version of the Return Directive during negotiations. Legislation continues that the risk of absconding must be determined, with respect to the objective criteria, ‘on the basis of an individual assessment of each case’.262 Judges essentially reiterated this legislative prescription when concluding that the risk of absconding ‘must be based on an individual examination of that person’s case’.263 Some academics interpret 254 Directive 2008/115/EC, art 3(7); Dublin III Regulation (EU) No 604/2013, art 2(n); and, similarly, Reception Conditions Directive 2013/33/EU, art 8(3)(2). 255 See Case C-528/15 Al Chodor EU:C:2017:213, paras 27–40; and ch 6.4.4. 256 Al Chodor (n 255) para 44. 257 ibid paras 38–39, for the ECHR and the Charter. 258 See Madalina Moraru, ‘Judicial Dialogue in Action’ in Moraru and others (n 179) 125, 134–39. 259 See Majcher (n 19) 86–90. 260 See Recommendation (EU) 2017/432 (n 135) Nos 15–16. 261 See Return Directive Recast Proposal (n 138) art 6(1); and the far-reaching Proposal for a Crisis and Force Majeure Regulation, COM(2020) 613 final, art 5(1)(c). 262 Reception Conditions Directive 2013/33/EU, art 8(2); similarly, Return Directive 2008/115/EC, recital 6, art 3(7). 263 Mahdi (n 193) para 70; and also Al Chodor (n 255) para 34.

542 Irregular Presence and Return this statement to establish a high threshold, not allowing national law to lay down rebuttable presumptions.264 I am not convinced that the Court would follow that argument. Chapter 10.3 and 10.5 illustrated that abstract formula like ‘individual examination’ can mean different things in the judicial practice. After all, domestic authorities have to assess, in light of the situation of each person’s case, whether the factual requirements of the presumption have been met and whether there are grounds for rebuttal. Proportionality is the main corrective tool; it applies even if we conclude, in a first step, that a ground for detention can be relied upon in casu.

16.7.2.2 Other criteria for asylum seekers Article 8(3) Reception Conditions Directive 2013/ 33/ EU establishes five additional grounds for detention, in addition to the risk of absconding. This list is exhaustive,265 in contrast to the open-ended Article 7 of the former Reception Conditions Directive 2003/ 9/EC.266 To lay down prevents national authorities from invoking a justification that is not mentioned explicitly. By way of example, assignment to a detention facility cannot be justified by the unavailability of suitable accommodation, the absence of sufficient resources, or irregular presence that does not constitute a risk of absconding at the same time.267 Judges confirmed that the verification of identity is a ground for detention under Article 8(3)(a), in order to prevent illegal entry and secondary movements,268 while saying so will require a strict proportionality assessment.269 Article 8(3)(c) was interpreted, notwithstanding the open-ended wording, to cover only border procedures, as defined in the Asylum Procedures Directive, thus prohibiting detention during border procedures under domestic legislation.270 These judgments can be read not to require an individualised assessment, and possibly not even a proportionality test, during the maximum time limit of four week for border procedures.271 That question is of great significance in light of the ongoing debate about the expanded use of asylum border procedures. Remember that the ECtHR gives states leeway during entry clearance, whereas the position of the Charter remains subject to debate. Article 8(3)(d) prevents the release of returnees who, during detention for purposes of return, file a (subsequent) application for asylum.272 At the same time, however, judges concluded that the prolongation of detention is not automatic, since the motivation of delaying or frustrating removal has to be substantiated separately.273 With regard to Article 8(3)(e), our comments in Chapter 10.3 explained that the Court interprets ‘national security’ and 264 See, amongst others, Acosta Arcarazo and Guild (n 141) 195; and Mananashvili and Moraru (n 187) Article 15 MN 34. 265 This has been settled case law ever since N (n 37) para 59. 266 See Arslan (n 157) paras 51–56; and Wilsher (n 243) 185–89. 267 See Ministerio Fiscal (n 247) paras 106–13; Országos Idegenrendészeti Főigazgatóság Dél-alföldi Regionális Igazgatóság (n 33) paras 249–55; and Case C-72/22 PPU Valstybės sienos apsaugos tarnyba EU:C:2022:505, para 85. 268 See K (n 37) paras 36, 39. 269 ibid paras 37–38, 40–43. 270 See Commission v Hungary (n 244) paras 178–86; and Országos Idegenrendészeti Főigazgatóság Dél-alföldi Regionális Igazgatóság (n 33) paras 238–41. 271 Commission v Hungary (n 244) paras 176–77 passim; see also Daniel Thym, ‘Expert Opinion on the Reform of the Common European Asylum System’ (January 2017) 42–44 http://ssrn.com/abstract = 3163014 (accessed 1 March 2023). 272 See Evangelia (Lilian) Tsourdi, ‘Asylum Reception Conditions Directive 2013/ 33/ EU’ in Thym and Hailbronner (n 28) Article 8 MN 22–24. 273 See Case C-186/21 Republika Slovenija EU:C:2021:447, paras 35–39.

Detention 543 ‘public policy’ narrowly. Again, irregular presence does not justify, as such, recourse to this ground for detention, unless additional elements substantiate the conclusion that someone poses a public policy or national security threat.274 Having said this, judges found detention on security grounds to be compatible with the Charter275—a conclusion that may possibly deviate from ECtHR case law.276

16.7.2.3 Additional grounds for returnees Article 15(1) Return Directive lists two grounds for detention, without stating clearly (‘in particular’) whether the catalogue is exhaustive or not. The Commission suggests establishing an exclusive list on the occasion of a future amendment, while adding public policy and security.277 This debate about exclusivity is relevant, as the Court had interpreted the former Reception Conditions Directive not to be exhaustive, thus permitting the detention of asylum seekers for reasons of public policy (the criteria was added to today’s exclusive list)—a view it has recently extended to the Return Directive.278 As a consequence, national legislation can authorise detention on other grounds, including security. Before the latest ruling, the Kadzoev judgment had often been cited to have said otherwise, although it had concerned a different question, namely detention beyond the maximum time limit for detention.279 Optional recourse to additional grounds for detention not mentioned in the Return Directive does not give Member States a carte blanche, since they have to comply with the principle of proportionality. In line with the comments on the constitutional foundations, proportionality comprises an obligation to refrain from arbitrary detention, for instance by means of open-ended formula in national laws. Having said this, administrative detention of returnees for other purposes than return remains possible, in full compliance with human rights.280 Hampering or avoiding return is recognised as a justification for detention under Article 15(1)(b) Return Directive. In contrast to the risk of absconding, national legislation need not define objective criteria, and national practices are diverse. ‘Avoiding’ may include failure to appear before competent authorities, contrary to an explicit obligation to stay in touch; examples of ‘hampering’ include false or misleading statements and the refusal to board a plane.281 The Court has addressed the interpretation of hampering on one occasion, albeit indirectly. Prolongation of detention beyond the initial six-month time limit was found to be justified for ‘lack of cooperation’, which appears to overlap with the notion of hampering, whenever the returnee refuses to declare, before the consular staff of the home state, that return will be voluntary.282

274 See Valstybės sienos apsaugos tarnyba (n 267) paras 89–91. 275 See N (n 37) paras 47–53. 276 See n 38 and accompanying text. 277 See Return Directive Recast Proposal (n 138) art 18(1)(c). 278 See Arslan (n 157) paras 57–62; and Politsei-ja Piirivalveamet (n 38) paras 34–36. 279 See Case C-357/09 PPU Kadzoev EU:C:2009:741, paras 68–71; and Majcher (n 43) 383–88. 280 See Galina Cornelisse, ‘Annotation of Case C-357/09 PPU: Kadzoev’ (2011) 48 CML Rev 925, 938–40; and also Országos Idegenrendészeti Főigazgatóság Dél-alföldi Regionális Igazgatóság (n 33) para 255. 281 See Mananashvili and Moraru (n 187) Article 15 MN 35–38. 282 See Mahdi (n 193) paras 76–82, 85.

544 Irregular Presence and Return

16.7.3 Alternatives and length Detention is not automatically lawful whenever one of the grounds applies. Secondary legislation affords migrants a ‘high level of protection’ and ‘places significant limitations on’ state powers, especially in the form of the proportionality test.283 The principle of proportionality is broken down into specific guarantees, meaning that detention is ‘strictly regulated’.284 Nevertheless, the effectiveness of these safeguards ultimately depends on everyday practices. To decide about alternatives to detention or a reasonable prospect of removal is not so much a question of legal interpretation, on the part of the Court in Luxembourg, but of the assessment of the individual case at the national level. Comparative surveys have revealed substantial difference.285 Alternatives to detention take precedence in accordance with Article 8(2) and (4) Reception Conditions Directive and Article 15(1) Return Directive. While Union law was instrumental in the spread of ‘alternatives to detention’ (ATD), an administrative culture of alternatives appears not to have emerged in all the Member States, so far at least.286 Possible alternatives include deposits, reporting conditions, residence requirements, or semi-open reception centres.287 Detention shall generally be ‘for as short a period as possible’,288 reiterating a formula that had been coined by the Council of Europe.289 This prerequisite must be fulfilled at the point of initial detention and at any moment thereafter; if not, the migrant shall be released immediately.290 In the context of return, additional safeguards command release whenever removal arrangements are no longer in progress or whenever it becomes apparent that there is no reasonable prospect of removal, for example since the country of origin will not take back the person.291 Member States must introduce a mechanism of periodic review, ex officio or at the request of the detainee.292 The highly relevant procedural detail for these procedural safeguards has to be established at the national level. Commission proposals to introduce a minimum period of at least three months for the initial detention order would not release domestic authorities from assessing the reasonable prospect of removal during the first weeks of detention.293 Maximum periods were discussed controversially during the negotiations for the Return Directive, which laid down an initial limit of six months which can be extended by another 12 months in accordance with Article 15(6). Not all Member States use this option, and transposition reduced permissible time limits for detention.294 Periods of detention during 283 See Al Chodor (n 253) para 34; and K (n 37) para 44. 284 El Dridi (n 82) para 42. 285 See Moraru and others (n 179); and Madalina Moraru and Géraldine Renaudiere, ‘European Synthesis Report on the Judicial Implementation of Chapter IV of the Return Directive’ (REDIAL Research Report 2016/05, EUI 2016). 286 See Philippe De Bruycker and others, ‘Alternatives to Immigration and Asylum Detention in the EU’ (Final Report of the Made Real Project, January 2015). 287 See UNHCR (n 42) Annex A. 288 Return Directive 2008/115/EC, art 15(1)(2); Reception Conditions Directive 2013/33/EU, art 9(1); and Dublin III Regulation (EU) No 604/2013, art 28(3)(1). 289 See Committee of Ministers, ‘Twenty Guidelines on Forced Return’ (Council of Europe, 4 May 2005) No 8(1). 290 See Return Directive 2008/115/EC, art 15(4). 291 ibid art 15(1), (4), (5); and Kadzoev (n 279) paras 54–56. 292 Return Directive, art 15(3); and Reception Conditions Directive 2013/33/EU, art 9(5). 293 See Return Directive Recast Proposal (n 138) art 18(5). 294 Commission, ‘Communication on EU return policy’ COM(2014) 199 final, 17.

Detention 545 the asylum procedure must be calculated towards the eighteen-month limit under the Return Directive, which, moreover, cannot be extended, even if the other conditions for detention are still fulfilled.295 Ordering detention on the same ground, after release for a short period, would be incompatible with the effet utile of the Directive. No upper limit exists for asylum procedures, but the principle of proportionality must be respected.296 Detention in border procedures cannot, at present, last for more than four weeks.297 Other grounds for detention can be activated thereafter. Dublin transfers are subject to a distinct time limit of six weeks, starting with the acceptance of the take charge or take back request, provided legal remedies do not have suspensive effect;298 detention thereafter is not limited in time but must be proportionate.299 Most Member States empower administrative authorities to order detention, subject to legal remedies. Some countries go further and require judicial authorisation, mirroring guarantees for criminal proceedings. Judicial authorisation usually entails a higher degree of scrutiny.300 The obligation to give reasons extends to the initial detention order and its renewal, not, however, to periodic review;301 the reasoning ought to reflect on alternatives to detention.302 Domestic courts must be able to rule on all relevant matters of fact and of law; they shall be able to order release.303 These procedural guarantees must be interpreted in light of the general principles and fundamental rights described in Chapter 7.2 on the administrative dimension.

16.7.4 Conditions of detention Administrative detention during the asylum procedures and in preparation of return follows a different rationale than the deprivation of liberty in criminal matters. EU legislation obliges Member States to reflect this administrative, non-punitive character in the design, organisation, and management of detention facilities.304 These requirements complement human rights, in relation to which the ECtHR affords states considerable leeway when detaining adult migrants without specific vulnerabilities.305 Generic guarantees in EU legislation concern access to open-air spaces and contact with family, legal advisers, and NGO representatives.306 Additional safeguards exist for minors, families, and other vulnerable persons, again in addition to human rights.307 Much depends on how these abstract standards are being implemented in practice.308 295 See Kadzoev (n 279) paras 49–57, 68–71. 296 See Országos Idegenrendészeti Főigazgatóság Dél-alföldi Regionális Igazgatóság (n 33) paras 262–64. 297 ibid paras 242–46. 298 See Khir Amayry (n 249) paras 27–39; on the scope ratione temporis of Dublin III Regulation (EU) No 604/ 2013, art 28(3)(3)–(4). 299 Khir Amayry (n 249) paras 41–47. 300 See Moraru and Renaudiere (n 285) 27–32. 301 See Mahdi (n 193) paras 41–52. 302 See Országos Idegenrendészeti Főigazgatóság Dél-alföldi Regionális Igazgatóság (n 33) paras 257–59, 273–75. 303 ibid paras 293–96; and Mahdi (n 193) paras 53–64. 304 See Case C-519/20 Landkreis Gifhorn EU:C:2022:178, paras 33–46. 305 See MSS v Belgium and Greece App no 30696/09 (ECtHR [GC], 21 January 2011) §§ 216–22; and Ilias and Ahmed v Hungary (n 31) §§ 186–94. 306 See Reception Conditions Directive 2013/33/EU, art 10(2)–(5); and Return Directive 2008/115/EC, art 16(2), (4)–(5). 307 Reception Conditions Directive, arts 11, 16(3), 17; and MH and others v Croatia (n 41) §§ 183–86. 308 See Majcher (n 43) 507–23; and Diekmann (n 203) 266–304.

546 Irregular Presence and Return Recourse to specialised facilities, separate from ordinary prisons, underlines the administrative character. Article 16(1) Return Directive has given rise to four judgments, in response to preliminary references from Germany; their findings can be projected upon Article 11(1) Reception Conditions Directive.309 Judges confirmed that the availability of specialised detention facilities related to the state territory as a whole, not federal subunits, even if they are larger than many European countries.310 Consent of the returnee is no justification for using regular facilities,311 in contrast to public policy, which can be relied upon as an unwritten justification for persons who pose a security risk.312 An explicit exception for emergency situations under Article 18, whenever ‘an exceptionally large number’ of returnees ‘places an unforeseen heavy burden on the capacity of the detention facilities’, does not absolve domestic authorities from making specialised facilities available and must be applied diligently, taking into account various factors.313 Having said this, separate buildings within an overall complex may qualify as a specialised facility if conditions there fulfil the specific requirements for administrative detention.314

16.8 Regularisation Regularisation comes in diverse forms and is a frequent occurrence across Europe. It essentially designates a change of status from illegal to legal stay, notwithstanding grey areas of semi-legality in line with our introductory comments. However, even generous regularisation will not prevail over ‘illegal stay’ once and for all. If law is a counterfactually stabilised behavioural expectation, as Luhmann famously put it, new arrivals, changing circumstances, the temporary character of many regularisation measures, and statal control imperatives entail that irregularity is here to stay.315 EU migration law addresses regularisation indirectly, mainly by means of provisions in the Return Directive. Article 6(4) Return Directive recognises that domestic authorities ‘may’ issue a residence permit ‘at any moment’. As a consequence of regularisation, a return decision shall be withdrawn, or not be issued in the first place. At the same time, the Directive does not leave Member States with the binary choice of either return or regularisation. Return proceedings can be suspended de facto or de jure, thus giving rise to limbo situations discussed previously. The Court stated unambiguously that a state ‘cannot be obliged to issue an autonomous residence permit, or other authorisation conferring a right to stay’, if there is no reasonable prospect of removal, for reasons of fact or law; Member States have ‘a very broad discretion’ in this respect.316 Not issuing a return decision would be incompatible with the objective of effective return.317 However, return decisions shall not be adopted, as we have 309 All judgments concerned return, as German law does not authorise the detention of asylum seekers outside border procedures at airports. 310 See Joined Cases C-473/13 and C-514/13 Bero and Bouzalmate EU:C:2014:2095, paras 30–32. 311 See Case C-474/13 Pham EU:C:2014:2096, paras 17–22. 312 See Case C-18/19 Stadt Frankfurt am Main EU:C:2020:511, paras 39–46. 313 See Landkreis Gifhorn (n 304) paras 68–103. 314 ibid paras 51–56. 315 See Philippe De Bruycker, ‘Introduction générale’ in Philippe De Bruycker (ed), Les régularisations des étrangers illégaux dans l’Union européenne (Bruylant 2000) 1. 316 See Mahdi (n 193) paras 86–89; M and others (n 195) para 44; and Case C-825/21 Centre public d’action sociale de Liège EU:C:2022:810, para 44; contra Kevin Fredy Hinterberger, Regularisations of Irregularly Staying Migrants in the EU (Nomos 2023) 102–11. 317 See Westerwaldkreis (n 139) paras 56–60.

Summary 547 seen, whenever the best interests of the child command so. This may possibly prepare the ground for an obligation to regularise the illegal stay of ‘non-removable’ children, going beyond human rights law. Comparative analyses of regularisation practices should take into account not only measures that are officially labelled as such. Status change from illegal stay to regular presence can be based on legal migration instruments (family unity, work, studies), humanitarian residence permits (various forms of complementary protection, described in Chapter 13.6.4), or distinct programmes for those staying illegally (regularisation sensu stricto).318 Whereas legal migration instruments often involve an individual right to receive a residence permit, distinct regularisation schemes are mostly discretionary and often temporary.319 National practices are diverse and employ an abundance of criteria to determine whose status will be legalised, including lapse of time, humanitarian considerations, and indicators of meritocratic ‘earned regularisation’ (work, economic self-sufficiency, language skills, sponsorship by local communities).320 Attempts by the Commission to get a foot in the door were met with resistance by the Council in the early 2000s and, then again, in 2014.321 Having said this, one measure was adopted in response to disputes about mass regularisation in Spain, which some governments had criticised as an incentive for illegal entry.322 EU institutions could go much further, as we have seen, by adopting legislation on the regularisation and complementary protection, although an activation of these competences remains highly unlikely at this juncture.323 An earlier Recommendation by the Parliamentary Assembly of the Council of Europe has not been reiterated in recent years.324 Arguably, the diversity of the legal, social, political, and economic context across Europe renders it difficult to design a regularisation programme for the continent as a whole. To try to do so may not be a smart move anyway, keeping in mind the contrasting policy dynamics described at the outset. National governments may be more inclined to implement pragmatic solutions in response to local needs than officially to defy the Treaty policy objective ‘to combat illegal immigration’ in highly politicised inter-institutional debates at the supranational level.

16.9 Summary The ‘fight’ against illegal migration has been high on the EU’s agenda ever since the initial phase of intergovernmental cooperation. Low ‘return rates’ are often mentioned as a justification for restrictions, both internally and in cooperation with third states. In this overall context, this chapter concentrated on the internal dimension of illegal presence and return. Various factors influence the design and effectiveness of the legislation on return. Sociolegal 318 See Albert Kraler, ‘Fixing, Adjusting, Regulating, Protecting Human Rights’ (2011) 13 EJML 297. 319 See Joanna Apap, Philippe De Bruycker, and Catherine Schmitter, ‘Regularisation of Illegal Aliens in the European Union’ (2000) 2 EJML 263, 266–71; and Hinterberger (n 316) ch 3. 320 Apap, De Bruycker, and Schmitter (n 319) 272–82; EMN (n 53) 30–32; and Martin Baldwin-Edwards and Albert Kraler, REGINE: Regularisations in Europe (ICMPD 2009). 321 See Costello (n 8) 98–101; and Lutz (n 200) 46–50. 322 See Decision 2006/688/EC on the establishment of a mutual information mechanism concerning Member States’ measures [2006] OJ L283/40. 323 For a proposal see Kevin Fredy Hinterberger, ‘An EU Regularization Directive’ (2019) 26 MJECL 736. 324 See PACE, ‘Regularisation Programmes for Irregular migrants’ (Recommendation 1807, 1 October 2007).

548 Irregular Presence and Return research emphasises migrant agency, and social scientists highlight structural factors, such as administrative capacities or labour market needs. They help to explain why there can be a contrast between the political focus on restrictive ‘control signals’ and the pragmatism of the local response. Careful use of language, such as the term ‘(ir)regular’, serves as a semantic stumbling block inviting readers to reflect upon the multiple legal and social facets of irregular presence beyond binary description of (il)legality. EU migration law was instrumental in the spread of criminal sanctions and other measures against ‘smugglers’ and people facilitating illegal entry and stay, building upon the UN’s Palermo Protocol. Particularly controversial are criminal sanctions for humanitarian assistance and the intimidation of search and rescue by private vessels through meticulous security and safety inspections. With the exception of sanctions for ‘facilitators’, Member States remain free to decide whether to criminalise the illegal entry and stay of migrants, subject to an important limitation established and fine-tuned by numerous CJEU judgments: custodial sentences must not impede the primary objective of departure. It depends on the domestic context whether custodial sanctions for illegal entry and stay comply with this obligation. Moreover, criminal sanctions are compulsory for the employment of migrants in an irregular situation under the Employer Sanctions Directive 2009/52/EC, which contains protective elements as well. Of little practical relevance is the option to grant temporary residence permits to victims of trafficking for the distinct purpose of cooperation in criminal proceedings against perpetrators. These sectoral instruments complement the Return Directive 2008/115/EC as the focal point of the EU’s legislative response to illegal presence. The Directive does not harmonise the loss of residence permits, focusing on the consequences of irregularity instead. Member States are under an obligation to issue a return decision to any third country national staying unlawfully on their territory and to designate a third state to which she should return. Adoption of the return decision sets in motion a sequence of events, unless courts temporarily suspend enforcement. If migrants do not comply with a period for voluntary departure, domestic authorities shall proceed with removal mandatorily; in practice, however, many migrants do not leave voluntarily and are not forcibly removed either, resulting in the de facto or de jure suspension of return proceedings. EU migration law recognises this ubiquity of ‘non-removal’ by prescribing basic minimum standards, which are supplemented by further-reaching human rights guarantees. A more effective return policy—something the supranational institutions have repeatedly declared a priority—is mainly a matter of practical measures. The EU supports national endeavours through multiple forms of operational support and inter-state cooperation, which are increasingly brought together under the auspices of Frontex. These practical measures are much more relevant than the Directives on the mutual recognition of expulsion decisions and on transit for purposes of removal by air. Improving operational cooperation is also the main objective of readmission agreements with third states, concluded by either the EU as a whole or Member States bilaterally. Readmission agreements lay down practical procedures for identity determination, acquisition of travel documents, or removal operations—objectives that can similarly be achieved on the basis of non-binding arrangements. Negotiating readmission agreements and improving implementation has proven difficult, thus triggering the search for incentives to make third states cooperate. This trend towards the externalisation of migration control will take centre stage in Chapter 18.

Summary 549 Detention is a severe interference with personal liberty. It is politically controversial and the object of legal debates about the level of protection under the ECHR and the Charter. EU law authorises and constrains the administrative detention of migrants during the asylum procedure and in order to prepare return. Grounds for detention are a necessary but insufficient condition, with proportionality serving as an essential control standard. Member States must lay down objective criteria for the ‘risk of absconding’, which the Commission wants to harmonise, including the use of rebuttable presumptions. Assessing proportionality and the use of alternatives to detention ultimately hinges on what states do in practice, which differs considerably. Procedural safeguards help to render proportionality more effective, for instance through periodic review; detainees must be released whenever there is no reasonable prospect of removal. European migration law authorises the regularisation of illegal stay at any moment, although Member States are—with the possible exception of children—under no obligation to introduce regularisation schemes domestically.

17

Association Agreements with Neighbours To facilitate cross-border movements for economic purposes is an essential objective of the EU Treaties. At the same time, some countries do not want to join the Union, have left, or are in the waiting room for accession. Association agreements can narrow the gap between those inside and outside the single market by ‘creating special, privileged links’1 which allow non-members to partake in economic integration. Diplomatic negotiations determine to what extent the freedom-enhancing rationale of free movement is projected upon neighbours. Corresponding guarantees are of great practical significance. Turkish, British, Albanian, and Ukrainian nationals are among the biggest groups of third country nationals residing in the Member States. Most of them benefit from residence security as long-term residents, while others have entered only recently. Their legal status is governed by the legislation discussed in this volume, unless the association agreements provide for varying degrees of more favourable treatment. Cooperation with neighbouring states need not be confined to reciprocal rights for nationals. Agreements have been signed with Norway, Switzerland, and Iceland on their participation in the Schengen area and the Dublin system; the British government similarly aspires to deepen cooperation on asylum and return. In doing so, the management paradigm of justice and home affairs complements the facilitation of cross-border movements in the tradition of the single market. While cooperation with ‘best friends’ in Norway, Switzerland, and Iceland combines internal mobility with external closure, migration management has become the focal point of the bilateral relations with neighbours to the east and south. Turkey and Morocco are prime examples of how the European Union seeks to externalise migration control. These initiatives will be presented in Chapter 18. Comments hereafter will assess the legal status of citizens of neighbouring states residing in Union territory under the association agreements. The status of Turkish nationals, in particular, has given rise to numerous Court judgments interpreting association agreements with neighbours in light of the single market rulebook; they featured prominently in early analyses of EU migration law.2 While these rulings remain highly relevant for Turkish nationals, we may conclude, with the wisdom of hindsight, that the legislation and case law on the area of freedom, security, and justice follow an independent rationale nowadays. The rules and verdicts discussed in this chapter are specific to bilateral agreements, which subject to several constitutional specificities (17.1). Norway, Iceland, Switzerland, and Liechtenstein belong to the Schengen area and have subscribed to free movement, even though the small print unearths caveats (17.2). The UK, by contrast, preferred to have leeway with respect to future movements, thus limiting the benefits of the contractual relations to those living abroad already (17.3). Turkey is a 1 See Case 12/86 Demirel EU:C:1987:400, para 9. 2 See ch 1.3.3; Kay Hailbronner, Immigration and Asylum Law and Policy of the European Union (Kluwer 2000) 212–60; and Elspeth Guild, The Legal Elements of European Identity (Kluwer 2004) ch 8. European Migration Law. Daniel Thym, Oxford University Press. © Daniel Thym 2023. DOI: 10.1093/oso/9780192894274.003.0018

Constitutional Foundations 551 special case as a result of decisions made by the Association Council forty years ago (17.4). Nationals of other states waiting for membership benefit from limited guarantees (17.5). The same holds for neighbouring countries where the justice and home affairs rationale has overtaken the single market as the lodestar of association policies (17.6).

17.1 Constitutional Foundations The EU Treaties authorise the institutions to negotiate association agreements with reciprocal rights and obligations. Their content depends on the outcome of the diplomatic negotiations and may go as far as extending the free movement to nationals of neighbouring states (17.1.1). These agreements are an integral part of the supranational legal order, but their effects and rank differ from directives and regulations (17.1.2). Provisions on Turkish nationals features prominently in the case law. While the Court had traditionally emphasised parallel interpretation in line with Union citizenship, it has emphasised differences more recently (17.1.3).

17.1.1 Supranational competences From a legal perspective, not all treaties mentioned in this chapter are association agreements on the basis of Article 217 of the Treaty on the Functioning of the European Union (TFEU). EU institutions use that provision for broad agreements covering multiple subject matters. Prominent examples are the Trade and Cooperation Agreement (TCA) with the UK, the Agreement on the European Economic Area (EEA), and the Cotonou Agreement with former colonies. Other legal bases are used for specific questions, such as readmission and visa facilitation, participation in the Schengen area and the Dublin system, or free trade. For our purposes, this distinction is relevant with respect to majority requirements, the scope of Union competences, and Member States with an opt-out. Association agreements are adopted unanimously by the Council, after the consent of the European Parliament.3 An activation of Article 217 TFEU presupposes that the third state ‘must, at least to a certain extent, take part in the Community system’.4 The degree of participation can vary: free movement characterises the EEA, not, however, the TCA and the Cotonou Agreement. Crucially, Article 217 TFEU is more than the sum of other Treaty articles. Judges interpreted that provision early on to ‘empower the [Union] to guarantee commitments towards non-member countries in all the fields covered by the Treaty’.5 What may sound self-evident at first had tangible consequences, as the example of free movement demonstrates. EU institutions were unable to adopt legislation on the entry or stay of third country nationals prior to the Treaty of Amsterdam. Nevertheless, Article 217 TFEU covered the partial extension of the free movement of workers by means of association. If the Treaty competence for association had not been interpreted generously, the dynamic case law on Turkish nationals would not have occurred. The Court does not have the authority

3

See TFEU, art 218(6)(a)(i), (8)(2).

4 See Demirel (n 1) para 9. 5 ibid.

552 Association Agreements with Neighbours on ‘mixed agreements’, in which both the Union and the Member States participate, whenever a provision transcends the scope of supranational powers.6 Alternative legal bases must be relied upon for sectoral agreements on specific topics. For the entry and stay of third country nationals, Article 79(2) TFEU nowadays establishes a shared competence the EU may activate to conclude international agreements in line with the comments on ‘implied powers’ in Chapter 18.2.1 on the external dimension. That distinction is practically relevant for Ireland and Denmark, whose opt-outs cover any measure based upon Article 79(2) TFEU. Turkish nationals residing in Denmark cannot rely on the Long-Term Residents Directive 2003/109/EC, but they can invoke the association agreement with Turkey, since the opt-out does not apply. Similarly, British nationals living in Ireland benefit from the TCA, as an association agreement, and the Withdrawal Agreement, which was based on Article 50(2) TFEU.7 The scope of alternative legal bases, such as development or trade, will be presented in the last chapter. Whenever we conclude that the contents of an international treaty is covered comprehensively by either Article 217 TFEU or a collection of alternative legal bases, it should be adopted as an ‘EU only’ agreement without the parallel participation of the Member States as separate state parties under international law; shared powers, such as Article 79(2) TFEU, may be exercised at the time of adoption. Doing so accelerates ratification, which would almost have derailed the entry into force of the association agreement with Ukraine after a negative consultative referendum in the Netherlands. Given that Article 217 TFEU applied, it was not convincing that the Council adopted a separate decision for the justice and home affairs components, thus allowing Ireland and Denmark (as well as the UK, before Brexit) not to participate.8

17.1.2 Status in the EU legal order Judges in Luxembourg have traditionally emphasised the relative independence of the autonomous legal order they famously found the EU Treaties to have established. On that basis, the Court controlled the legal effects and the hierarchical rank of international law in the supranational legal order. It holds that international agreements ‘form an integral part of [Union] law’9 and can benefit, as a matter of principle, from the same effects as regular Union law in domestic legal orders. Provisions of international agreements concluded by the EU have a hierarchical status above secondary legislation but below that of the EU Treaties.10 An important difference concerns justiciability. Settled case law requires that a provision of an international treaty must be directly applicable to guide the resolution of individual cases; without direct effect, the Court and domestic courts will usually ignore an 6 ibid 9; in contrast to Joined Cases C-300/98 and C-392/98 Christian Dior EU:C:2000:688, para 48; see also Kay Hailbronner, ‘Die Freizügigkeit türkischer Staatsangehöriger nach dem Assoziationsabkommen EWG/Türkei’ [1984] Europarecht 54, 55–61. 7 See Case C-479/21 PPU Governor of Cloverhill Prison and others EU:C:2021:929, paras 57–58, 66–69; Case C-431/11 United Kingdom v Council EU:C:2013:589, paras 63–65. 8 See Decision 2014/668/EU on the signing and provisional application of the Association Agreement, as regards Title III [2014] OJ L278/1. 9 Case 181/73 Haegeman EU:C:1974:41, para 5. 10 See Piet Eeckhout, External Relations of the European Union (2nd edn, OUP 2011) ch 9.

Constitutional Foundations 553 agreement.11 In order to determine whether a provision has direct effect or not, the Court interprets the agreement. In doing so, judges have traditionally been strict towards agreements of global reach; they found world trade law not to be directly applicable (and, hence, irrelevant for determining the validity of secondary legislation). By contrast, they were often more generous with regard to association agreements. Direct effect extends to the implementing decisions of the association councils.12 Contemporary agreements often include explicit provisions on direct applicability. In the case of the UK, both the Withdrawal Agreement and the Protocol on Social Security Coordination state that they can be relied upon directly, unlike all the other provisions of the TCA.13 International agreements concluded by the Member States are not binding on EU institutions. While the Refugee Convention and international human rights instruments can be invoked indirectly, via Article 78(1) TFEU and similar guarantees in the Charter, other agreements cannot be invoked, even if they have been ratified by all Member States.14 When an agreement was concluded prior to accession (or the conferral of competences to the EU), the safeguard clause under Article 351 TFEU applies. Union law does not prevent Member States from fulfilling their obligations under pre-existing agreements. By way of example, Monaco and the Vatican indirectly participate in the Schengen area on the basis of older treaties.15 Deference to earlier agreements is static. States lose the capacity to amend them, unless the EU authorises them to do so.16 In practice, respect for existing obligations is often guaranteed through safeguard clauses in the directives on more favourable provisions, whose distinct layout was considered in Chapter 10.

17.1.3 Interpretative parallelism and its limits A starting point of many Court rulings is the conceptual autonomy of Union law; generic notions, such as ‘public policy’, and the terminology of secondary legislation must be interpreted independent of their meaning in domestic legislation and public international law.17 What at first sounds highly technical, has proven to have had profound effects on the status of millions of Turkish nationals, since judges interpreting the association acquis in light of the rules on free movement in the single market. Such an outcome was not a foregone conclusion, since the Court generally insists that treaties concluded by the EU are subject to the rules of interpretation of the international law of treaties.18 On that basis, it recognised early on that even an article with wording identical to a single market provision can have a

11 See Case C-308/06 Intertanko EU:C:2008:312, paras 42–45. 12 See Case C-192/89 Sevince EU:C:1990:322, paras 13–36. 13 See Agreement on the Withdrawal of the UK from the EU (adopted of 24 January 2020, entered into force 1 February 2020) [2020] OJ L29/7, art 4; and Protocol on Social Security Coordination [2021] OJ L149/2292, art SSC.67, which is attached to the Trade and Cooperation Agreement (adopted 30 December 2020, provisionally applied 1 January 2021, entered into force 1 May 2021) [2021] OJ L149/10. 14 See Intertanko (n 11) paras 48–52. 15 France signed an agreement with Monaco in 1863, which was upgraded in 1997; the Lateran Agreement between Italy and the Vatican foresees public accessibility of St Peter, which may be read to imply the absence of border controls. 16 See Case C-205/06 Commission v Austria EU:C:2009:118. 17 See ch 6.4.4. 18 See Opinion 1/91 EEA I EU:C:1991:490, para 14.

554 Association Agreements with Neighbours different meaning, if the purpose supports such divergence.19 Association agreements can be synchronised with Union law, but whether this is the case or not depends on the context of the rules in question. The operationalisation of this formula is bound to remain controversial. While the Court had often advanced the status of individuals up until the 2000s, later judgments have accentuated differentiation. Two examples taken from the rich case law on Turkish nationals illustrate the potential of parallel interpretation and its limits. On the one hand, the Ankara Agreement stated explicitly that the High Contracting Parties ‘agree to be guided by [Articles 45–47 TFEU] for the purpose of progressively securing freedom of movement for workers between them’.20 Building upon this proviso, judges advanced parallel interpretation ‘so far as is possible’.21 Doing so considerably extended the rights of Turkish nationals: the term ‘worker’ covers part-time employment,22 a wide reading of the notion ‘family member’ is warranted,23 and generous interpretation of ‘public policy’ considerably curtailed state powers to expel Turkish nationals.24 On the other hand, interpretative dynamics were reversed when the Court was asked whether Turkish nationals benefit from the higher threshold for the expulsion of Union citizens that have lived elsewhere for more than ten years, which the EU legislature had introduced.25 Judges answered that question to the negative, since the new rules transcended the predominantly economic rationale of the association agreement.26 The protection of Turkish ‘workers’ does not embrace the membership rationale of Union ‘citizenship’, since ‘the two legal schemes in question cannot be considered equivalent’.27 The potential for parallel interpretation ran out of steam at a time when it became apparent that accession negotiations were stagnating. When it comes to contemporary agreements, the room for interpretative dynamics is curtailed by cautious drafting. Stating explicitly what an agreement requires was, in part at least, a response to innovative Court rulings on Turkish nationals, which national governments perceived to be overtly dynamic.28 In the case of the UK, the European Union insisted on thorough prescriptions to prevent the British government from indirectly undermining the acquired rights of Union citizens residing there by means of restrictive interpretation. Many provisions on citizens’ rights in the Withdrawal Agreement refer explicitly to Free Movement Directive 2004/38/EC, and the provisions on social security coordination in the TCA are mostly detailed.29

19 See Case 270/80 Polydor EU:C:1982:43, paras 14–21; and Christa Tobler, ‘Context-Related Interpretation of Association Agreements’ in Daniel Thym and Margarite Zoeteweij-Turhan (eds), Rights of Third-Country Nationals under EU Association Agreements (Brill Nijhoff 2015) 101. 20 See [Ankara] Agreement establishing an Association between the European Economic Community and Turkey (adopted 12 September 1963, entered into force 1 December 1964) [1977] OJ L361/29, art 12. 21 Case C-434/93 Bozkurt EU:C:1995:168, para 20; and Daniel Thym, ‘Constitutional Foundations of the Judgments on the EEC-Turkey Association Agreement’ in Thym and Zoeteweij-Turhan (n 19) 13, 16–24. 22 See Case C-14/09 Genc EU:C:2010:57, paras 19–32. 23 See Case C-275/02 Ayaz EU:C:2004:570, para 45. 24 See Case C-340/97 Nazli EU:C:2000:77, para 56; and Case C-467/02 Cetinkaya EU:C:2004:708, paras 41–48. 25 See Free Movement Directive 2004/38/EC, art 28(3)(a). 26 See Case C-371/08 Ziebell EU:C:2011:809, paras 64, 68, 72; and Ankara Agreement (n 20) art 2(1). 27 Ziebell (n 26) para 74. 28 I learnt about this motivation when talking to people involved in the negotiations. 29 See Withdrawal Agreement (n 13) arts 9–38; and the technical rules in the Protocol on Social Security Coordination (n 13).

‘Best friends’: European Economic Area and Switzerland 555

17.2 ‘Best friends’: European Economic Area and Switzerland Some European countries prefer not to join the Union. The European Economic Area (EEA) allows them to partake in the single market; it comprises Norway, Iceland, and Liechtenstein nowadays. Switzerland has concluded two sets of bilateral agreements instead. While relations between the EU and the EEA are trouble-free, Switzerland can be a difficult partner. At the time of writing, the government had walked out of the negotiating room for an institutional framework agreement, which the EU considers to be a prerequisite for the conclusion of further bilateral agreements on economic cooperation. However, these difficulties do not have immediate repercussions for the existing treaties on cross-border movements. In assessing relations with the EEA countries and Switzerland, we should distinguish the distinct rationale and layout of the agreements on free movement (17.2.1) as well as Schengen and Dublin membership (17.2.2). A peculiarity are the status of micro states, such as Monaco, and Gibraltar (17.2.3).

17.2.1 Free movement of persons When the EU established the single market, neighbouring countries worried about economic isolation. The EEA Agreement overcame these concerns. Unlike previous free trade agreements, concluded during the 1970s, it embraces the free movement of persons, including social security coordination and the recognition of professional qualifications. Norway, Iceland, and Liechtenstein apply these guarantees unreservedly.30 Minor exceptions include the option of charging higher tuition fees for foreign students.31 New legislation is integrated into the EEA framework through decisions of the Joint Committee, for instance with regard to the Free Movement Directive 2004/38/EC.32 The European Free Trade Association (EFTA) Tribunal, which holds jurisdiction on preliminary references from EEA states, turned out to be a trustworthy companion of the Court of Justice of the European Union (CJEU). It repeatedly opted for a dynamic interpretation in line with later case law, even though that outcome was not officially required.33 In short, the EEA countries are de facto members of the single market. The Swiss situation follows the free movement agreement, which was signed in 1999 as an integral part of the first set of bilateral treaties. This treaty covers most aspects of free movement, including mutual recognition of professional qualifications and social security coordination.34 Having said this, contractual relations with Switzerland are static. Incorporating new legislation would require an official amendment subject to ratification.35 As a consequence, previous legislation applies instead of the Free Movement Directive 2004/38/EC, the Social Security Coordination Regulation (EC) No 883/2004, and the Recognition of 30 See Agreement on the European Economic Area (adopted 2 May 1992, entered into force 1 January 1994) [1994] OJ L1/3, arts 28–45, Annex V, Annex VIII. 31 See Protocol 29 on Vocational Training [1994] OJ L1/196. 32 See EEA Agreement (n 30) arts 93, 98–104; and Decision No 158/2007 of the EEA Joint Committee amending Annex V and Annex VIII [2008] OJ L124/205. 33 EEA Agreement (n 30) art 6. 34 See Agreement on the Free Movement of Persons (adopted 21 June 1999, entered into force 1 June 2002) [2002] OJ L114/6, arts 2–9, 16, Annexes I–III. 35 See Agreement on the Free Movement of Persons (n 34) arts 17–18.

556 Association Agreements with Neighbours Professional Qualifications Directive 2005/36/EC. Switzerland has agreed, however, to extend free movement after accession, and the country even pays a financial contribution to support regional development. As in the case of the EEA, Swiss courts generally follow the lead of the Court of Justice.36 Dynamic alignment with new legislation is one reason why the Commission insists on the conclusion of an institutional framework agreement. Negotiations failed, amongst others, since the Swiss government aims at stating explicitly that it is not obliged to accept the free movement of economically inactive persons under the Free Movement Directive; moreover, it wants assurances about domestic wage levels.37 These requests respond to domestic anxieties about free movement. More than 25 per cent of the population do not have Swiss nationality; many come from Italy, Portugal, Germany, and France. Bilateral relations were put under pressure when the population narrowly supported a referendum ‘against mass immigration’; another referendum, which would have required the government to undertake more than cosmetic steps to implement the referendum, was rejected by a strong majority in another referendum in 2020.38 Its adoption might have triggered withdrawal from the first set of bilateral agreements. These political fissures need not signal entrenched problems in the everyday application of the free movement acquis in Switzerland, Norway, Island, and Liechtenstein. It unfolds by and large trouble-free, notwithstanding legally tricky disputes over selected themes. In the Swiss case, the CJEU confirmed previous Swiss rulings that the bilateral agreement does not embrace passive service reception and a general prohibition of discrimination on grounds of nationality; tourists, for instance, cannot invoke equal treatment in service charges.39 By contrast, nationals of the EEA countries benefit from far-reaching equal treatment with Union citizens, including in matters of extradition.40 Nevertheless, it remains uncertain to what extent the outer limits of the CJEU’s citizenship case law beyond the confines of the Free Movement Directive can be applied in the EEA context, in particular for family members in situations of return and immobility. Icelandic and Norwegians courts have revolted against the dynamic interpretation of the EFTA Tribunal.41 In a symbolic move, the CJEU found the Charter not to be an integral part of a bilateral agreement with Switzerland, even though the ECHR establishes a comparable level of protection.42

36 See Narin Idriz and Christa Tobler, ‘“Citizenship of the Association’ in Dora Kostakopoulou and Daniel Thym (eds), Research Handbook on European Union Citizenship Law and Policy (Edward Elgar Publishing 2022) 320, 331–39; and Astrid Epiney, ‘Das Freizügigkeitsabkommen Schweiz-EU’ in Ferdinand Wollenschläger (ed), Enzyklopädie Europarecht. Band X (2nd edn, Nomos 2021) § 13. 37 See (Swiss) Federal Council, ‘Institutional agreement between Switzerland and the European Union’ (Letter to Ursula von der Leyen, 26 May 2021). 38 See https://en.wikipedia.org/wiki/2014_Swiss_immigration_initiative (accessed 1 March 2023); and Idriz and Tobler (n 36) 339–40. 39 See Case C-70/09 Hengartner and Gasser EU:C:2010:430, paras 36–42. 40 See Case C-897/19 PPU Ruska Federacija EU:C:2020:262, paras 55–68. 41 See ch 14.3.1; and Christian NK Franklin and Halvard H Fredriksen, ‘Differentiated Citizenship in the European Economic Area’ in Kostakopoulou and Thym (n 36) 297, 299–312. 42 See Case C-680/17 Vethanayagam EU:C:2019:627, paras 86–87.

‘Best friends’: European Economic Area and Switzerland 557

17.2.2 Membership in Schengen and Dublin When Sweden and Finland joined the Schengen area in 1996, one year after EU accession, the other Member States agreed to sign an association agreement with Norway and Iceland. Doing so allowed for the maintenance of the Nordic Passport Union. A revised agreement on the participation of Norway and Iceland was signed when the Schengen acquis was integrated in the EU framework by the Treaty of Amsterdam.43 Border controls were discontinued from March 2001 onwards. In the same year, Norway and Iceland were associated with the Dublin system.44 Switzerland similarly joined Schengen and Dublin in the context of the second set of bilateral agreements; border controls were discontinued in 2008, in the wake of an affirmative referendum.45 Liechtenstein followed its neighbour.46 The geographic location of the four countries facilitated swift membership, which has not caused major controversies or crises ever since. A common feature of the agreements on Schengen and Dublin is the dynamic incorporation of new legislation. To facilitate smooth cooperation, the EU even granted Norway, Iceland, and Switzerland an indirect say in decision-making via a ‘mixed committee’, meeting regularly in the margins of the Justice and Home Affairs Council.47 The associated Schengen and Dublin members may also intervene before the CJEU; even though domestic courts cannot send preliminary references. The agreements aim for uniform interpretation and application, while staying short of an official obligation to follow the supranational case law.48 Chapter 12.3.3 on border controls explained that it can be challenging to define the scope of the Schengen and Dublin acquis when new legislation does not build upon pre- existing provisions. By way of example, the associated countries participate in Frontex as equal members, including voting rights in the management board and contributions to the budget; similarly arrangements can be introduced for the Asylum Agency.49 Equal membership in the agencies is testimony to close cooperation. Switzerland was not given a strict deadline for the participation in measures building on the Schengen or Dublin acquis, in order to give the country time to organise a referendum. In total, 72 per cent of the population supported continued participation in Frontex in 2022, thus flatly rejecting the arguments of the opponents who had criticised

43 See Protocol [No 19] integrating the Schengen acquis into the framework of the European Union [2008] OJ C115/290, art 6; and Agreement concerning the association with the implementation, application and development of the Schengen acquis (adopted 18 May 1999, entered into force 26 June 2000) [1999] OJ L176/36. 44 See Agreement concerning the criteria and mechanisms for establishing the State responsible for examining a request for asylum lodged in a Member State or in Iceland or Norway (adopted 19 January 2001, entered into force 1 April 2001) [2001] OJ L93/40. 45 See Agreement on the Swiss Confederation’s association with the implementation, application and development of the Schengen acquis (adopted 26 October 2004, entered into force 1 March 2008) [2008] OJ L53/52; and Agreement concerning the criteria and mechanisms for establishing the State responsible for examining a request for asylum lodged in a Member State or in Switzerland (adopted of 26 October 2004, entered into force 1 March 2008) [2008] OJ L53/5. 46 ibid art 16; and art 15; as well as Protocol on the accession of the Principality of Liechtenstein [2011] OJ L160/ 3; and Protocol on the accession of the Principality of Liechtenstein [2011] OJ L160/39. 47 ibid arts 2–4; and arts 2–4 for Switzerland; as well as Agreement concerning the Schengen acquis (n 43) arts 2, 8; and Agreement on asylum jurisdiction (n 44) for Norway and Iceland, arts 2–4; see also Fabrice Filliez, ‘Schengen/Dublin’ in Bernd Martenczuk and Servaas van Thiel (eds), Justice, Liberty, Security (VUB Press 2008) 145. 48 Agreement on the Schengen acquis (n 45) arts 8–10 by way of example. 49 See Frontex Regulation (EU) 2019/1896, recitals 121–24, art 101(3); and EUAA Regulation (EU) 2021/2303, art 34.

558 Association Agreements with Neighbours pushbacks, externalisation, and the deaths of thousands of migrants at sea.50 Refusal to participate in new legislation would entail the automatic termination of Schengen or Dublin membership, unless the Join Committee decides otherwise.51 Remember that the Asylum Procedures Directive, the Qualification Directive, and the Reception Conditions Directive do not build upon the Schengen or Dublin acquis. References to the ‘Member States’ in these instruments cannot, therefore, be interpreted to cover the associated countries. As a result, Germany could not reject an asylum application as inadmissible in light of a previous decision in Norway, irrespective of the level of protection there.52

17.2.3 Micro states and Gibraltar A peculiarity are the micro states of Monaco, San Marino, Andorra, Liechtenstein, and the Vatican, as well as Gibraltar as a British Overseas Territory. Neither of them accepts unlimited inward migration. Rather, Andorra, Monaco, and San Marino have signed asymmetric free movement agreements with neighbours that allow their nationals to live in adjoining territory without the promise of reciprocity.53 Similar exceptions apply to Liechtenstein under the EEA Agreement, which authorises the principality to operate a quota system.54 These unusual arrangements are a pragmatic recognition of the special situation of micro states, whose social fabric and public infrastructure may not sustain the influx of a few thousand people to a destination many perceive to be attractive, including for tax purposes. Nationals of the micro states are covered, as third country nationals, by EU migration law, unless they hold dual nationality. Moreover, citizens of Andorra and San Marino may use the ‘EU lines’ at border crossing points and are exempted from electronic exit and entry stamps under the Entry/Exit System.55 Monaco and the Vatican indirectly participate in Schengen via international agreements with France and Italy,56 while the absence of border controls towards San Marino is not regulated. Andorra exercises some controls and unilaterally recognises Schengen visas to allow third country nationals to visit the principality.57 Brexit brought the status of Gibraltar back on the political agenda. A political agreement was reached on 31 December 2020, on the eve of Brexit, that the rock territory should join the Schengen area on the basis of a bilateral agreement, which had originally been supposed to foresee operational powers of Frontex.58 At the time of writing, the negotiations were ongoing.59 Remember that de facto 50 See https://frontex-referendum.ch (accessed 1 March 2023). 51 See Agreement on the Swiss association with the Schengen acquis (n 45) art 5(4). 52 See Case C-8/20 LR EU:C:2021:404. 53 See Jorri Duursma, Fragmentation and the International Relations of Micro-States (CUP 1996); and Francesco Maiani, ‘European Union and Andorra, Monaco and San Marino’ in Sieglinde Gstohl and David Phinnemore (eds), The Proliferation of Privileged Partnerships between the European Union and its Neighbours (Routledge 2019) 84. 54 See Decision No 158/2007 (n 32) recital 12; and Franklin and Fredriksen (n 41) 299–312. 55 See ‘Andorra and San Marino’ (Council doc 13020/1/04, 6 October 2004); and Schengen Borders Code Regulation (EU) 2016/399, art 11(2), 6a(3), as amended by Regulation (EU) 2017/2225 [2017] OJ L327/28. 56 See n 15 and accompanying text. 57 See Commission, ‘Communication: EU relations with the Principality of Andorra, the Principality of Monaco and the Republic of San Marino’ COM(2012) 680 final, 8–10. 58 See ‘Spain and UK Reach Draft Deal on Post-Brexit Status of Gibraltar’ The Guardian (31 December 2020). 59 See Recommendation for a Council Decision authorising the opening of negotiations for an agreement in respect of Gibraltar, COM(2021) 411 final.

United Kingdom: Reversed Dynamics 559 Schengen membership transcends the question of border controls. If Spain performs the controls, this will reassert its claim to territorial sovereignty over Gibraltar, although it may possibly seek the support of Frontex.

17.3 United Kingdom: Reversed Dynamics A period of forty-eight years of free movement between the UK and the continent came to an end on 31 December 2020 with the expiration of the transitional period after Brexit.60 Not to subscribe to further free movement had been an essential objective of the British government, building upon the political promise to ‘take back control’ during the campaign in the run-up to the Brexit referendum.61 London was willing to pay a price in the form of limited access to the single market for services, since the EU insisted that the four freedoms were intricately linked. As a result, we have to distinguish between the acquired rights of British nationals and Union citizens residing abroad at the end of 2020 (17.3.1) and movements thereafter (17.3.2).

17.3.1 Acquired rights of (former) Union citizens Almost four million Union citizens and several hundreds of thousands of UK nationals were living on the territory of the other Party at the end of 2020. Their legal status is protected under Articles 9–39 Withdrawal Agreement, which cover all categories of persons under the Free Movement Directive 2004/38/EC, as well as frontier workers.62 The will receive a right to permanent residence after five years, which offers life-long protection, unless beneficiaries move elsewhere for a period exceeding five consecutive years.63 Social security coordination is maintained, not, however, voting in municipal elections. EU institutions insisted on detailed prescriptions for application procedures to make sure that the UK does not undermine citizens’ rights through administrative practices and the ‘hostile environment’ policy towards migrants who are irregularly present.64 As mentioned previously, the citizens’ rights in the Withdrawal Agreement are directly applicable. They shall be interpreted with ‘due regard’ to later judgments, and the CJEU has jurisdiction for all preliminary references initiated by British courts during the transitional period.65 These procedural and substantive guarantees ideally support that the transition from free movement qua Union law to residence documents in accordance with the Withdrawal Agreement unfolds without major difficulties. Permanent residence will also be available for those who pass the five-year threshold after the end of 2020, provided they entered before that date.66 Some aspects are left to states to regulate, especially family reunification 60 See Withdrawal Agreement (n 13) art 126. 61 See John Curtice, ‘Why Leave Won the UK’s EU Referendum’ (2017) 55 JCMS Spec Ed 19. 62 For detailed comments see Thomas Liefländer, Manuel Kellerbauer, and Eugenia Dumitriu-Segnana (eds), The UK-EU Withdrawal Agreement. A Commentary (OUP 2021). 63 See Withdrawal Agreement (n 13) arts 15(3), 39. 64 ibid arts 18–19; and Eleanor Spaventa, ‘The Rights of Citizens under the Withdrawal Agreement’ (2020) 45 EL Rev 193. 65 Withdrawal Agreement (n 13) arts 4(5), 86(2), 127; and, by way of example, Case C-247/20 Commissioners for Her Majesty’s Revenue and Customs EU:C:2022:177. 66 Withdrawal Agreement (n 13) arts 16–17.

560 Association Agreements with Neighbours and expulsion on public policy grounds from 2021 onwards.67 This gives the British government some flexibility, whereas British nationals residing on the continent will often benefit from enhanced protection under the Family Reunification Directive and the Long-Term Residents Directive. Like in the case of the EEA Agreement, the outer limits of the CJEU’s citizenship case law are not covered by the Withdrawal Agreement, especially with regard to family members in scenarios of return and immobility.68

17.3.2 Sovereign control over future movements Domestic law has governed the entry and stay of UK national and Union citizens moving from 2021 onwards. The Trade and Cooperation Agreement (TCA) does not provide for free movement, and it did not even realise the original intention to consider rules for selected purposes, such as research or studies.69 On the continent, British nationals are, as third country nationals, covered by the EU migration law instruments, such as the Blue Card Directive, the Students and Researchers Directive, or the Single Permit Directive. By contrast, it remains the sovereign choice of the UK how to regulate the entry and stay of Union citizens in future. Far-reaching restrictions for unskilled workers were an immediate consequence of Brexit. Having said this, there are some corollary rules on migration. With regard to visas, the TCA notes that ‘short-term visits’ are not currently subject to visa requirements, although the parties do not commit to retaining the status quo indefinitely.70 British nationals will not be exempt from the future ETIAS requiring visitors to apply for electronic travel authorisation, as described in Chapter 9.3.4 on databases, and the UK plans to introduce a similar scheme. Some narrowly defined guarantees exist for the entry and temporary stay of business visitors, contractual service suppliers, intra-corporate transferees, and independent professionals.71 These entry channels mirror provisions on posted workers and transnational service provision in secondary legislation and selected free trade agreements, which were analysed in Chapter 14.5 on legal migration. They are not directly effective, neither in the UK nor in the EU. Once the UK or a Member State has authorised entry and stay, the Protocol on Social Security Coordination can be relied upon. It is directly applicable and covers all persons— not only nationals—‘legally residing’.72 By way of example, Polish nationals working in the UK, Moroccans moving to London after having previously worked in Paris, and British pensioners residing in Spain will benefit from the Protocol on Social Security Coordination. Its provisions are complex, thus enhancing legal certainty. They mirror guarantees in the Social Security Coordination Regulation (EC) No 883/2004, such as the export of pensions 67 ibid arts 10(1)(e)–(f), 20–21. 68 ibid arts 9–10; ch 14.3.1; and Michael Dougan, ‘The UK’s Withdrawal Package’ (2020) 57 CML Rev 631, 667–75. 69 cf (Revised) Political declaration setting out the framework for the future relationship [2019] OJ C384I/178, Nos 48–57. 70 See Trade and Cooperation Agreement (n 13) art 492; and ch 11.3.2. 71 ibid art 140; and Catherine Barnard and Emilija Leinarte, ‘Mobility of Persons’ in Federico Fabbrini (ed), The Law and Politics of Brexit, vol III (OUP 2021) 134. 72 Trade and Cooperation Agreement (n 13) arts 488–91; and the Protocol on Social Security Coordination (n 13).

Turkey: ‘Consolation Prize’ with Much Practical Effects 561 or health insurance in scenarios of transnational mobility. These technical rules are of great practical importance, and we can expect the EU and the UK to extend their validity beyond the initial fifteen-year period. Negotiations on the TCA were cumbersome and need not present the end point. Supplementary agreements can be incorporated in the overall governance structure at a later stage.73 Such sector-specific agreements may cover, for instance, asylum or the Erasmus +Programme.74 At the time of writing, there were no indications that the TCA might be upgraded in the near future. Rather, political tensions were rising on the Northern Ireland Protocol. One area where cooperation might be reinforced in the years to come is return. Chapter 2.4.2 explained that it is an irony of Brexit that the UK lost the option of flexibly opting in control instruments. The Dublin III Regulation (EU) No 604/2013 does not apply any longer. The British government had sought to negotiate bilateral or multilateral readmission agreements as an integral part of the post-Brexit package, only to be rebutted by the EU side.75 Public outcry over the death of migrants trying to cross the Channel put pressure on France to agree to some sort of enhanced cooperation. In the summer of 2022, no supplementary measures had been agreed upon. Rather, the UK had negotiated an external processing deal with Rwanda, mentioned in Chapter 13.4.7 on asylum. Moreover, bilateral French–British cooperation on juxtaposed border controls on both sides of the Channel continued, as described in Chapter 11.6.3 on liaison officers. Bilateral arrangements also exist between Ireland and the UK.

17.4 Turkey: ‘Consolation Prize’ with Much Practical Effects Turkey was the first country ever, together with Greece, to negotiate an association agreement with the former European Economic Community. Turkey insisted on being treated on par with its neighbour (and political rival). It was successful. The Ankara Agreement of 1963 comprises the objective of progressively realising the free movement of workers and other persons, while deferring concrete steps to unanimous decisions of the Association Council.76 This promise was realised to a limited extent, and judges interpreted these decisions in a highly dynamic manner. Workers and family members residing in the EU are the main beneficiaries (17.4.1), and ‘standstill’ provisions had unexpected consequences for first admission (17.4.2).

17.4.1 Turkish nationals residing in the Member States In 1970, Turkey and the former EEC agreed to establish the free movement of workers by the 1990s.77 That promise was overtaken by developments in the main destination countries when 73 Trade and Cooperation Agreement (n 13) art 2(1). 74 ibid art 710 requires a simple decision, no separate agreement, for participation in Union programmes. 75 See Joint Political Declaration on Asylum and Returns [2020] OJ L444/1477; and Ben Hulme, ‘Analysis of the United Kingdom’s Proposal for a UK-EU Readmission Agreement’ EULawAnalysis (20 August 2020). 76 See Ankara Agreement (n 20) arts 12–16, 22–23. 77 See Additional Protocol (adopted 23 November 1970, entered into force 1 January 1973) [1977] OJ L361/60, arts 36–40.

562 Association Agreements with Neighbours the economic recession following the ‘oil crisis’ of 1973 coincided with the abrupt termination of the guest worker programmes.78 As a sort of ‘consolation prize’, the Association Council decided to protect the rights of those Turkish nationals who were residing on the territory of the Member States already. Articles 6–9 Association Council Decision No 1/80, which was not published in the Official Journal, guaranteed residence security to former ‘guest workers’ and their family members. To do so was no trivial feature, given that permanent residence was not yet available everywhere at the time. Dynamic case law increased the level of protection.79 To start with, national residence permits are declaratory, like in the case of Union citizens.80 Unconditional labour market access under Article 6 after four years of regular employment was found to be accompanied with residence security, thus introducing a distinct category of de facto permanent residence.81 Generous interpretation of the notion of ‘worker’ in line with single market rules, mentioned previously, and the retention of the status during periods of unemployment and imprisonment reinforced residence security further.82 While these guarantees were designed for former ‘guest workers’, they remain valid for Turkish nationals entering nowadays. Students or refugees, for instance, are equally covered by Article 6, provided they are working legally for a certain number of years. The association rulebook supplants the EU migration law instruments as a more favourable provision of international law.83 Residence security is confined, however, to one state and does not authorise transnational mobility in the single market.84 Family members are covered by Article 7 Association Council Decision No 1/80. Again, judges supported a generous reading, thus developing an autonomous residence status for spouses and children. The concept of ‘family member’ was interpreted in line with single market rules, and, crucially, autonomy is almost absolute. Children born on the territory benefit unconditionally, notwithstanding the wording.85 It is irrelevant whether the ‘child’ is an adult or resides with the sponsor; death or naturalisation of the parent does not affect residence security either. Longer periods of absence and public policy grounds under Article 14 are the only reasons justifying the loss of status.86 Article 14 was aligned, as we have seen, with Union citizenship as a matter of principle, despite some exceptions. Grandchildren of former ‘guest workers’ will equally benefit from Article 7, provided their parents work legally in accordance with Article 6. Finally, Association Council Decision No 3/ 80 provides for equal treatment and transnational coordination in core areas of social security. The provision on the ‘export’ of benefits in the case of return was found to be directly applicable, especially regarding pensions.87 Similarly, equal treatment with nationals in the domain of family benefits has

78 See ch 14.1.1. 79 See generally Pieter Boeles and others, European Migration Law (2nd edn, Intersentia 2014) 97–116; and Stefan Lorenzmeier, ‘Assoziierungsfreizügigkeit (Türkei, Balkan, Ukraine, Mittelmeer) und EFTA/EWR’ in Wollenschläger (n 36) § 12 MN 17–28. 80 See ch 10.1.1; and Case C-351/95 Kadiman EU:C:1997:205, para 51. 81 See Sevince (n 12) paras 28–29. 82 See Case C-171/95 Tetik EU:C:1997:31; and Case C-383/03 Dogan EU:C:2005:436. 83 See Case C-294/06 Payir EU:C:2008:36, paras 47–48. 84 See Tetik (n 82) para 29; and Case C-171/13 Demirci and others EU:C:2015:8, paras 54–56. 85 See Case C-502/04 Torun EU:C:2006:112, paras 22–25. 86 See the summary by Case C-720/19 Stadt Duisburg EU:C:2020:847, paras 24–26. 87 See Decision No 3/80 of the Association Council on the application of the social security schemes to Turkish workers and members of their families [1983] OJ C110/60, art 6; and also Paul Minderhoud, ‘Decision No 3/80 of the EEC-Turkey Association Council’ in Thym and Zoeteweij-Turhan (n 19) 169.

Turkey: ‘Consolation Prize’ with Much Practical Effects 563 tangible benefits for Turkish nationals residing in the EU.88 The remainder of Decision No 3/80 is not practically relevant. Politically negotiations to replace it by the Social Security Coordination Regulation (EC) No 883/2004 appear not to have resulted in the adoption of a successor instrument.89

17.4.2 ‘Standstill’ as dynamism for first admission Residence security on the basis of the Association Council Decision No 1/80 presupposes that Turkish nationals have been authorised to enter or stay. First admission is not covered. The association acquis ‘does not encroach upon the competence retained by the Member States to regulate both the entry into their territories . . . and the conditions under which they may take up their first employment’.90 For a while, it seemed as if this bastion might fall as well, thus effectively realising the political promise, in the Ankara Agreement, of free movement by means of judicial fiat. To do so the Court relied on the ‘standstill clauses’ preventing Member States from introducing new restrictions on the first entry workers and cross-border service provision.91 Judges interpreted the standstill clauses to be directly applicable. Practical effects ultimately depend on national legislation, since standstill provisions operate as a ‘quasi- procedural rule’.92 They oblige Member States not to introduce new restrictions that had not existed when Article 41 Additional Protocol entered into force in 1973 (or the date of EU accession if that was later). With regard to Article 13 Association Council Decision No 1/80, the legal framework at the end of 1980 is decisive. More lenient national legislation introduced thereafter cannot be can cancelled either, even if it does not fall behind the status quo ante of the date when the standstill clause entered into force.93 Interior ministries across Europe were upset when the Court outlawed visa requirements for lorry drivers that had not existed in the 1970s; the judgment effectively set aside visa requirements under the former Visa List Regulation.94 Creative lawyers used the dynamism of ‘standstill’ to challenge language skills as a precondition for family reunification, which had been introduced in the new millennium. Judges confirmed this reading as a matter of principle, while highlighting, at the same time, that the notion of ‘restriction’ embraced unwritten justification requirements; interpretative dynamism supporting convergence with settled single market case law backfired.95 The Court also distinguished, like in the Swiss case, passive service reception from active service provision, thus maintaining visa requirements for tourists.96 These judgments curtailed the 88 ibid art 3; ch 15.3.3; and, by way of example, Case C-262/96 Sürül EU:C:1999:228. 89 cf Council Decision 2012/776/EU on the position to be taken within the Association Council with regard to the adoption of provisions on the coordination of social security systems [2012] OJ L340/19. 90 Case C-237/91 Kus EU:C:1992:527, para 25; see also Demirel (n 1) paras 23–25. 91 See Additional Protocol (n 77) art 41; and Association Council Decision No 1/80, art 13. 92 Case C-16/05 Tum and Dari EU:C:2007:530, para 55. 93 See Joined Cases C-300/09 and C-301/09 Toprak and Oguz EU:C:2010:756, paras 51–56; and Case C-123/17 Yön EU:C:2018:632, paras 40–56. 94 See Case C-228/06 Soysal and Savatli EU:C:2009:101, paras 50–61; and Steve Peers, ‘EC Immigration Law and EC Association Agreements’ (2009) 34 EL Rev 628. 95 See Case C-138/13 Dogan EU:C:2014:2066, paras 37–38; and Narin Tezcan Idriz, ‘The Puzzle Posed by Demir for the Free Movement of Turkish Workers’ in Thym and Zoeteweij-Turhan (n 19) 223. 96 See Case C-221/11 Demirkan EU:C:2013:583, paras 42–59.

564 Association Agreements with Neighbours practical effects of the standstill provisions, which, nevertheless, remain a source of interpretative revelations. A series of more recent rulings entrusted domestic courts with assessing the proportionality of integration requirements for spouses and children.97 We saw in Chapter 15.4 on integration that this conclusion mirrors the Court’s position on language requirements under the Family Reunification Directive.

17.5 Western Balkans: Pre-accession Guarantees Countries on the path to membership habitually receive association agreements as a first step towards accession. ‘Europe Agreements’ were concluded with countries in Central and Eastern Europe in the 1990s, and ‘Stabilisation and Association Agreement’ are in force with the countries in the Western Balkans. They support the gradual integration of future members in the single market, justice and home affairs, and other Union policies. In contrast to the agreement with Turkey, contemporary pre-accession treaties do not embrace free movement (17.5.1). Instead, they contain limited guarantees with clearly delineated rights and obligations, which prevent the Court from interpreting them in an overtly dynamic fashion (17.5.2).

17.5.1 Former agreements and transitional periods The contents of ‘Europe Agreements’ with states in Central and Eastern Europe, which joined in 2004/07, remained meagre when it comes to free movement and migration. They were essentially limited to the equal treatment of workers residing legally, social security coordination, and the admission of the self-employed, which gained some relevance in practice.98 Additional legal pathways were available domestically, such as German quotas for seasonal and posted workers, which will be mentioned in the next chapter. Concerns about the economic and political fallout of free movement in light of stark income disparities explain that outcome. Governments of the ‘old’ Member States worried about the entry of literally millions of workers and that the economic development of home states might be slowed down as a result of ‘brain drain’.99 Even after accession, free movement was phased in to allow the economies of the ‘new’ members to catch up. Treaties of Accession laid down transnational periods of seven to ten years for Greece, Spain, Portugal, and, more recently, Central and Eastern Europe.100 Member States decide autonomously whether to activate the transitional period; they are subject to fairly loose constraints when taking that decision.101 The UK and Ireland decided not to use this option; they experienced 97 See ch 3.3.3.3; Case C-561/14 Genc EU:C:2016:247, paras 37–38; Yön (n 93) paras 78–88; Case C-89/18 A EU:C:2019:580, paras 36–42; and Case C-379/20 Udlændingenævnet EU:C:2021:660, paras 31–35. 98 By way of example see Europe Agreement establishing an association between the European Communities and their Member States, of the one part, and the Republic of Poland, of the other part (adopted 16 December 1991, entered into force 1 February 1994) [1993] OJ L348/2, arts 37–54; and Norbert Reich, ‘The Constitutional Relevance of Citizenship and Free Movement in an Enlarged Union ’ (2005) 11 ELJ 675, 684–87. 99 See Martin Kahanec and Klaus F Zimmermann (eds), EU Labor Markets after Post-Enlargement Migration (Springer 2010). 100 See eg Act concerning the conditions of accession of the Republic of Croatia [2012] OJ L112/21, Annex V, art 18, No 2. 101 See Adelina Adinolfi, ‘Free Movement and Access to Work of Citizens of the New Member States’ (2005) 42 CML Rev 469, 485–96.

Western Balkans: Pre-accession Guarantees 565 net inward migration of about one million citizens before Brexit. Germany become a major destination country when transitional periods expired; net inward migration amounted to more than two million during the 2010s.

17.5.2 Limited impact of contemporary agreements Stabilisation and Association Agreements (SAA) with countries in the Western Balkans essentially reiterate guarantees under the former Europe Agreements. They do not bring about individual rights to be admitted for work purposes, except for narrowly defined categories of self-employed. Similarly, companies with subsidiaries in the Union may sponsor the entry of ‘key personnel’ as intra-corporate transferees.102 These guarantees reiterate our previous conclusion, in Chapter 14.5 on legal migration, that free trade may embrace cautiously circumscribed entry channels. While these provisions are directly effective,103 they do not turn national migration laws upside down. States can require visas for the self-employed.104 One step further, workers from these countries benefit from equal treatment with regard to working conditions, remuneration, and dismissal once they have been admitted to the labour market.105 Agreements also foresee base guarantees of social security coordination, with additional protection being available under domestic legislation or bilateral treaties.106 Equal treatment may be used, for instance, to challenge fixed term contracts for foreign language teachers.107 Having said this, the practical relevance is limited nowadays, as the Single Permit Directive 2011/98/EU establishes a higher level of protection for those with labour market access. The most recent SAA with Kosovo, concluded in 2015, does not foresee guarantees for workers and the self-employed anymore. Instead, it simply refers to individual rights under the EU migration law, which are indeed wider than provisions in previous agreements (with the exception of the self-employed).108 Instead of legal migration, justice and home affairs are the focus of attention. Kosovo is promised capacity building on border controls, the country agrees to readmission, and shall cooperate on irregular movements.109 Agreements with neighbouring countries contain similar provisions, which are buttressed by bilateral agreements on specific subject matters, such as visa facilitation and readmission, as mentioned in Chapters 11.3.2 and 16.6.4. Reasonably generous financial and operational support is available under the Instrument of Pre-Accession, and Frontex has started deploying 102 By way of example see Stabilisation and Association Agreement between the European Communities and their Member States of the one part, and the Republic of Serbia, of the other part (adopted 29 April 2008, entered into force 1 September 2013) [2013] OJ L278/16, arts 52(d), 54–57, 58(2); on the other agreements Lorenzmeier (n 79) MN 38–44. 103 See Case C-257/99 Barkoci and Malik EU:C:2001:491; and Case C-268/99 Jany EU:C:2001:616 on the Europe Agreements. 104 See Case C-327/02 Panayotova and others EU:C:2004:718; and Katharina Eisele, The External Dimension of the EU’s Migration Policy (Brill/Nijhoff 2014) 390–401. 105 By way of example see SAA with Serbia (n 102) art 49(1). 106 ibid art 51; and ch 15.3.3. 107 See Case C-162/00 Pokrzeptowiez-Meyer EU:C:2002:57; and also Eisele (n 104) 316–35. 108 See Stabilisation and Association Agreement between the European Union and the European Atomic Energy Community, of the one part, and Kosovo, of the other part (adopted 27 October 2015, entered into force 1 April 2016) [2016] OJ L71/3, art 86(2). 109 ibid arts 85–89.

566 Association Agreements with Neighbours police missions in the Western Balkans. This change of emphasis reflects a general trend: the migration management objective gradually replaced the single market paradigm as the lodestar for rules on migration in association agreements.

17.6 Neighbourhood: Novel Focus on Migration Control For decades, EU institutions had been successful in influencing policy change in their surroundings. The prospect of accession served as a powerful incentive. The European Neighbourhood Policy (ENP) was introduced with much fanfare as a substitute to support transformation in the near abroad at a time when enlargement became a distant prospect, or was unavailable altogether. We may only conclude that the EU has failed to prevent civil wars, slippage into autocratic rule, and economic stagnation. Think of Syria, Ukraine, Libya, Georgia, and other neighbours to the east and south. Inspection of the Eastern Partnership (17.6.1) and the Union for the Mediterranean (17.6.2) reaffirms that justice and home affairs have dethroned the single market as the primary point of reference. Legal and political commitments can be found in a conglomerate of association agreements, unilateral financial support, and informal action plans.110

17.6.1 Eastern partnership: pragmatic cooperation Relations with Eastern neighbours in the field of migration have proven comparatively unproblematic. Bilateral agreements with countries in the near abroad, which have a common border with a Member State, have traditionally contained rudimentary provisions on the equal treatment of workers residing legally and the transfer of key personnel.111 These guarantees were removed from more recent agreements in light of the guarantees enshrined in the Single Permit Directive 2011/98/EU and the ICT Directive 2014/66/EU. Instead, more recent agreements provide for the temporary presence of independent professionals for transnational service provision and the posting of workers with specific qualifications, which are not covered by supranational legislation on economic migration.112 Social security coordination had always been relegated to bilateral agreements, and, nowadays, the equal treatment clause in Article 12 Single Permit Directive. Relations with Russia are not technically part of the neighbourhood policy, but cooperation followed the same pattern before it was suspended after the armed attack on Ukraine.113 Ukraine had vehemently opposed being treated as a ‘neighbour’, when it signed an association agreement in 2014 without a firm promise of membership. Migration was an ancillary 110 See generally Peter Van Elsuwege, ‘The EU and Its Neighbours’ in Ramses A Wessel and Joris Larik (eds), EU External Relations Law (Hart Publishing 2020) 437; and Paul James Cardwell, EU External Relations and Systems of Governance (Routledge 2009) ch 5. 111 By way of example see Partnership and Cooperation Agreement between the European Communities and their Member States, of the one part, and the Russian Federation, of the other part (adopted 24 June 1996, entered into force 1 December 1997) [1997] OJ L327/1, arts 23(1), 32; the first provision was found to be directly applicable in Case C-265/03 Simutenkov EU:C:2005:213. 112 See eg Association Agreement between the European Union and its Member States, of the one part, and Ukraine, of the other part (adopted 27 June 2014, entered into force 1 September 2017) [2014] OJ L161/3, arts 17, 97–102, Annex XXVII-C. 113 See Oleg Korneev, ‘European Union-Russia Cooperation on Migration Management’ (2012) 17 EFA Rev 605.

Neighbourhood: Novel Focus on Migration Control 567 aspect at the time, even though additional entry channels were made available under domestic law, especially in Poland, as mentioned in Chapter 14.5.5 on legal migration. Lack of substantial political, economic, and military backing meant that the association agreement did little to prevent the Russian aggression. Thereafter, the first ever activation of the Temporary Protection Directive 2001/55/EC, analysed in Chapter 13.7 on asylum, could be read, from the perspective of association policies, as an exercise in de facto free movement prior to accession. Developments will show whether the country will proceed quickly along the path to membership. Besides rudimentary rules on legal migration in relations with Eastern neighbours, cooperation on irregular movements, asylum, and border controls has gathered momentum.114 The Association Agreements are testimony to a sequence of ever more sophisticated articles indicating a gradual increase of ambition.115 These abstract provisions serve as a point of reference for project-based cooperation, which can similarly unfold without an anchor in the bilateral agreements. Many Eastern neighbours agreed to readmission agreements in return for visa facilitation, as described in Chapters 11.3.2 and 16.6.4. These agreements tend to work well in practice precisely because they reflect the mutual interest of both partners. By contrast, informal mobility partnerships with Eastern neighbours such as Moldova, Georgia, and Armenia have not delivered much in practice. They will feature in Chapter 18.

17.6.2 Union for the Mediterranean: hotbed of control practices Relations with countries in the Mediterranean basin have become a veritable hotbed of control-oriented cooperation—albeit under reversed circumstances than originally foreseen. The Barcelona Declaration of 1995, which kick-started closer cooperation, focused on peace-building, free trade, and economic prosperity, mentioning migration as a corollary issue alongside the media, health, or cultural dialogue.116 Accordingly, the first generation of ‘Euromed’ Agreements with Maghreb countries did not pay much attention to migration besides equal treatment as regards working conditions and social security of workers who are residing legally.117 These guarantees were practically relevant for states like France, Italy, or the Netherlands, where many nationals from the Maghreb countries were living; they gave rise to a number of preliminary references.118 From today’s perspective, they have mostly been supplanted by equal treatment guarantees in secondary legislation, such as the

114 See Peter Van Elsuwege and Zvezda Vankova, ‘Migration and Mobility in the EU’s Eastern Neighbourhood’ in Ippolito Francesca and others (eds), Bilateral Relations in the Mediterranean (Edward Elgar Publishing 2020) 160. 115 Compare Association Agreement with Ukraine (n 112) arts 16–19 to the previous Partnership and Cooperation Agreement between the European Communities and their Member States, of the one part, and the Republic of Azerbaijan, of the other part (adopted 22 April 1996, entered into force 1 July 1999) [1999] OJ L246/3, art 75. 116 See the ‘Barcelona Declaration’ Euro-Mediterranean Conference (27–28 November 1995) https://eeas.eur opa.eu/archives/docs/euromed/docs/bd_en.pdf (accessed 1 March 2023). 117 By way of example see Euro-Mediterranean Agreement establishing an association between the European Communities and their Member States, of the one part, and the Republic of Tunisia, of the other part (adopted 17 July 1995, entered into force 1 March 1998) [1998] OJ L97/2, arts 64–67. 118 See eg Case C-336/05 Eckhouik EU:C:2006:394; Eisele (n 104) 243–47, 367–79; and Anja Wiesbrock, ‘Social Security Rights of Third Country Nationals under the Euro-Mediterranean Association Agreements’ in Elspeth Guild and others (eds), Social Benefits and Migration (CEPS 2013) 45.

568 Association Agreements with Neighbours Single Permit Directive. EU legislation also applies to nationals from beyond the Maghreb, for which association agreements had always been limited to abstract declarations of intent concerning fair treatment of workers residing legally.119 These declarations of intent were not put into practice (and are arguably superfluous in light of the secondary legislation adopted in the meantime). Instead, Morocco, Tunisia, and, controversially so, Libya have become leading partners in the externalisation of migration control practices, to be discussed in the chapter that follows.120 Cooperation on migration management is mentioned in the association agreements at a high level of abstraction,121 but this has not prevented cooperation from evolving dynamically in practice. Constitutionally, such dynamism is unproblematic given that operational cooperation does not require a legal basis in an association agreement. It has to comply, of course, with international law and supranational legislation. Migration has become a central pillar of the bilateral relations, and occasionally dominates the agenda, as explained in Chapter 13.1.6 on geopolitics. Notwithstanding the practical dominance of migration control, the EU institutions highlight the desire for legal pathways in the abstract.122 In practice, however, they are rarely delivered, as discussed at the very end of this volume.

17.7 Summary The geographic region of ‘Europe’ extends beyond the external borders of the EU, which, moreover has neighbours to the East and South. Close cooperation with these countries is neither surprising nor new. Article 217 TFEU establishes a broad competence allowing the institutions to project single market rules upon neighbouring states. The degree and scope of association agreements is defined during diplomatic negotiations. In contrast to multilateral treaties, the Court has generally found bilateral agreements to be directly applicable. Member States and the EU institutions must respect their contents. One step further, judges promoted parallel interpretation of important provisions in light of the single market rulebook, especially for Turkish nationals. The protective credentials of this settled case law on parallel interpretation remain intact, even if we recognise that judges have emphasised differences recently. Cautious drafting of contemporary association agreements limits the room for further judicial innovation. Norway, Iceland, Switzerland, and tiny Liechtenstein accept free movement and are members of the Schengen area and the Dublin system, even though some corollary aspects of free movement are not covered. While the EEA Agreement allows for the dynamic incorporation of new legislation, the bilateral agreement on free movement with Switzerland is static; the legal framework of twenty years ago continues to apply. Conversely, the treaties on the participation of the ‘best friends’ in the Schengen area and the Dublin system embrace new legislation automatically. That is practically relevant, since the Schengen acquis,

119 See eg Euro-Mediterranean Agreement establishing an Association between the European Communities and their Member States, of the one part, and the Arab Republic of Egypt, of the other part (adopted 25 June 2001, entered into force 1 June 2004) [2004] OJ L304/39, art 62. 120 For the change of direction see Cardwell (n 110) ch 7. 121 For ‘dialogue in social matters’ see Euro-Mediterranean Agreement with Tunisia (n 117) art 69(3); and Euro- Mediterranean Agreement with Egypt (n 119) arts 63, 70. 122 See Commission and High Representative, ‘A New Agenda for the Mediterranean’ JOIN(2021) 2 final, 16–18.

Summary 569 in particular, has evolved dynamically. Cooperation unfolds without major problems. A peculiarity are micro states, such as Andorra or Monaco, and the British territory of Gibraltar. They partake in some measures. Brexit brought to an end a period of almost fifty years of free movement between the UK and the continent. ‘To take back control’ over migration was a central objective of the British government. That is why the guarantees for future movements in the Trade and Cooperation Agreement are confined to narrowly defined categories of contractual service suppliers, intra-corporate transferees, and independent professionals; they are not directly effective. By contrast, the Withdrawal Agreement contains far-reaching and directly applicable provisions on the acquired rights of British nationals and Union citizens residing on the territory of the other party at the end of 2020. They benefit from life-long protection as permanent residents, subject to few limitations. Moreover, the EU and the UK have agreed on complex provisions on social security coordination, which will apply whenever someone stays legally in the EU or the UK in future. Turkish nationals equally benefit from enhanced guarantees once they have been permitted to enter and stay in the Member States. Judges interpreted the Association Council Decision No 1/80 dynamically and established a de facto permanent residence for former ‘guest workers’ and their families. Decision 1/80 was found to replicate single market rules, thus bringing about an almost unconditional right to remain. Turkish nationals entering nowadays equally benefit from these guarantees. Member States decide autonomously about first admission, in accordance with EU migration legislation. The original idea of progressively realising the free movement of workers was not realised. ‘Standstill provisions’ can be relied upon to challenge new restrictions, for instance visa requirements for cross-border service provision, even though judges have reined in the integrationist potential of the standstill provisions. Contemporary association agreements with countries in the Western Balkans, which have been put on the path to membership, and other neighbours to the South and East do not contain far-reaching provisions on legal migration anymore. Ancillary previous rules on equal treatment have been supplanted by more generous guarantees under the Single Permit Directive and other instruments discussed in this volume. Entry channels for some independent professionals and service suppliers are the exception to the rule. The migration management objective has replaced the single market paradigm as the lodestar for the latest generation of agreements with neighbours to the East and South. These provisions serve as an abstract anchor for operational activities and sectoral treaties, for instance on readmission and visa liberalisation. These highly dynamic practices will be discussed in the last chapter on international cooperation.

18

International Cooperation with Third States Migration law has traditionally belonged to the realm of ‘home affairs’. Movements of people across borders was subject to few international obligations besides the Refugee Convention and international human rights law, as traced in Chapter 5. The legal dimension of globalisation remains Janus-faced: while goods can circulate largely unhindered, persons often face legal and practical hurdles. The European Union continues this tradition in many respects and has been instrumental in promoting international cooperation nevertheless. On the one hand, it advanced the free movement of Union citizens and integrated ever more countries into an area without internal border controls. On the other hand, EU institutions have reinforced cooperation with third states on control practices ever since they started dealing with migratory matters. In the Tampere Conclusions, the European Council stressed ‘the need for more efficient management of migration flows at all their stages’, in close cooperation with countries of origin or transit.1 The institutions have taken that maxim seriously. The external dimension is one of the most dynamic—and controversial—topics dealt with in this book. At the same time, it can be difficult to grasp from a legal perspective. Crucial developments are not subject to legal prescriptions, in either secondary legislation or international agreements, and informal practices and policy frameworks have gained significance in recent years. We would lose sight of crucial changes if we concentrated on legally binding output. Informalisation has been accompanied by a greater emphasis on migration control. A leitmotiv of our comments will be a conceptual tension between the promotion of mobility across borders and the management rationale of home affairs. The focus of attention has shifted from association agreements, discussed in the previous chapter, towards the externalisation of migration control by means of cooperation with third states. A short history of the policy framework demonstrates the steady rise of the management paradigm and introduces legal experts to distinctively European acronyms like ‘GAMM’ and policy documents like the ‘New Partnership Framework’ (18.1). The external dimension has an overarching character, comprising measures adopted on the basis of Articles 77–80 of the Treaty on the Functioning of the European Union (TFEU) and other policy fields, such as development or trade; to define the scope of Union competences can be challenging (18.2). Comments on the policy substance will be structured along overarching themes, meaning that relations with specific countries, such as Turkey, may feature in different sections. The management paradigm gathered momentum after the policy crisis of 2015/16 and is widely discussed under the disapproving heading of ‘externalisation’ nowadays (18.3). The EU often has a long-standing development partnership with countries further afield, which increasingly have a migration component, for instance with regard to the prevention of ‘root causes’ and legal pathways (18.4).

1

See European Council, ‘Presidency Conclusions of the Meeting in Tampere’ (15–16 October 1999) No 22.

European Migration Law. Daniel Thym, Oxford University Press. © Daniel Thym 2023. DOI: 10.1093/oso/9780192894274.003.0019

Theory and Policy Design 571

18.1 Theory and Policy Design EU institutions are well-known for designing ‘agendas’ or ‘strategies’, which lay down ambitious objectives which often contrasts with a lacklustre performance in terms of policy output—on migration and many other foreign policy matters. A gap emerges between the semblance of strategic vision and the reality of ad hoc measures and continuous ‘muddling through’. Conflicts of interests among the Member States and between the EU institutions often reinforce the sense of disorder. While trade policy and financial support are reasonably successful, other domains are often less influential. Nevertheless, perseverance and the investment of political and financial capital meant that the EU institutions have gradually built a remarkable—although controversial—baseline of external migration policies. Their inspection benefits from an awareness that commitment to ‘partnership’ and ‘management’ at Treaty level and in policy papers (18.1.1) coincided with a flurry of declarations of intent (18.1.2), whose implementation was defined by a succession of trial and error (18.1.3). International relations theory indicates why it can be difficult to find common ground with countries of origin or transit (18.1.4).

18.1.1 ‘Migration management’ as an overarching narrative Many experts in migration law will perceive the notion of ‘migration management’ to stand for restrictions in light of the contemporary focus on externalisation. The original usage of the term, however, was more positive. Commitment to ‘management’2 and ‘partnership and cooperation with third countries for the purpose of managing inflows of people applying for asylum’3 at Treaty level coincided with similar pledges in the Tampere Conclusions, The Hague Programme, and the Stockholm Programme. Doing so reinforced by the EU’s self- image as a ‘normative power’4 and a force for good, whose ‘action on the international scene shall be guided by the principles which have inspired its own creation’.5 Such an idealised vision of international actorness may appear naïve in light of today’s interest-based external migration policy, but it served as the conceptual starting point for international EU activities nonetheless. The focus on partnership, management, and cooperation in the EU Treaties and strategy documents coincided with the reorientation of the international debate in the early 2000s. Academic experts and international actors started emphasising positive feedback loops between migration and development, which contrasted with the sceptical views on migration as an expression of development failure earlier. States in the Global South might benefit from financial remittances and the transfer of knowledge as a result of migration.6 At the same time, governments in the Global North realised that domestic migration controls can be insufficient and that their economies may benefit from regular migration. Events during 2005/6 were testimony to a staggering dynamic. The report of the Global Commission on 2 TFEU, art 79(1). 3 Ibid, art 78(2)(g). 4 See Ian Manners, ‘Normative Power Europe’ (2002) 40 JCMS 235. 5 TEU, art 21(1) on the general objectives for the EU’s external action. 6 See Hein de Haas, ‘Migration and Development’ (2010) 44 IM Rev 227; and Oliver Bakewell, ‘Keeping Them in Their Place’ (IMI Working Paper 8, Oxford International Migration Institute 2007).

572 International Cooperation with Third States International Migration (GCIM) prominently emphasised the objective of ‘maximizing positive outcomes’ for sending and receiving states.7 This was taken up by the EU’s Global Approach, mentioned below, and, more recently, the non-binding Global Compact for Safe, Orderly, and Regular Migration, adopted in 2018. In this overall context, the term ‘management’ was favoured by those emphasising the element of state control. Nevertheless, it was much more than an euphemism for securitarian externalisation; the management paradigm acknowledged both the need for transnational cooperation and the benefits of safe and orderly movements.8 Migration management was widely understood, in the academic and policy debate, to signal a reorientation of migration policies, within Europe and beyond, from domestic measures towards inter-state cooperation, including the involvement of private actors and a novel focus on migrant agency.9 This paradigm shift accepted that public authorities will never be able fully to control real life events, while striving for regulatory leverage nonetheless. The management paradigm has been criticised for depoliticising the discourse through the use of technical language.10 Moreover, seemingly flawless ideas such as ‘triple-win’ or ‘circular migration’ can be deconstructed to hide dilemmas and conflicts of interests any external action on migration will inevitably be confronted with.11 Until today, a major problem of abstract strategy papers and international documents is the predominance of diplomatic compromise formulae, concealing deeper tensions between interests and values. The contents of the Global Compacts, adopted in 2018, proved a case in point, as shown in Chapter 5.5.4 on human rights. The EU’s external migration policy can be read as a succession of initiatives which gradually abandoned idealist accounts of ‘triple-win’ to the benefit of a realist emphasis on the interests under the banner of ‘a Europe that protects’.12 Commission President von der Leyen promised to lead ‘a geopolitical Commission’13—a description that suits the external dimension of migration policy particularly well. Europe wants to assert itself in a world, which it increasingly perceives to be threating and potentially even hostile. Even so, the positive narrative of well-managed migration remains on the table. The ‘New Pact on Migration and Asylum’ generally put much emphasis on control elements but also highlighted the need for legal pathways and the desire to maximise opportunities for ‘partner countries’.14

7 See GCIM, ‘Migration in an Interconnected World. New Directions for Action’ (October 2005) 1; and Vincent Chetail, ‘Paradigm and Paradox of the Migration-Development Nexus’ (2008) 52 German Yearb Intl L 183, 187– 92, 200–205. 8 An early reference point was the ‘International Agenda for Migration Management’ (IAMM), presented in 2004 by the ‘Berne Initiative’; see https://www.iom.int/berne-initiative (accessed 1 March 2023). 9 See Thomas Spijkerboer, ‘Changing Paradigms in Migration Law Research’ in Carolus Grütter and others (eds), Migration on the Move (Brill 2017) 13, 15–24; and Savitri Taylor, ‘From Border Control to Migration Management’ (2005) 39 Soc Poly & Admin 563. 10 See Jürgen Bast, Aufenthaltsrecht und Migrationssteuerung (Mohr Siebeck 2011) 6–12. 11 See Martin Geiger and Antoine Pécoud, ‘The Politics of International Migration Management’ in Martin Geiger and Antoine Pécoud (eds), The Politics of International Migration Management (Palgrave Macmillan 2010) 1, 11–16. 12 See ch 1.2.6. 13 See ‘The von der Leyen Commission’ (Press Release IP/19/5542, 10 September 2019). 14 See Commission, ‘Communication on a new Pact on Migration and Asylum’ COM(2020) 609 final, 17–24.

Theory and Policy Design 573

18.1.2 A short history of the EU’s strategic vision The historic background to early intergovernmental cooperation within the Schengen framework and under the Treaty of Maastricht was what might be called Europe’s first asylum policy crisis. Countries such as Germany, Denmark, or the Netherlands witnessed an increase of asylum applications governments deemed problematic. One of the first informal measures were the London Resolutions which promoted restrictive domestic asylum legislation on, amongst others, safe third countries.15 These internal measures were complemented by cooperation with countries of origin or transit, including many states in Central and Eastern Europe.16 Germany was particularly active on that front, concluding bilateral readmission agreements, promoting operational cooperation, and advancing legal pathways for seasonal workers and specific forms of posted workers.17 National capitals dominated external policy initiatives at the time, but also the Commission recognised the need for international cooperation early on.18 When heads of state or government supported close cooperation with countries of origin or transit at their meeting in Tampere, the rise in the number of asylum applications and the wars in the former Yugoslavia were still fresh in everyone’s mind. Nevertheless, EU institutions concentrated on the internal dimension at first. That changed during 2005 when the international debate paved the way for the European Council introducing the ‘Global Approach to Migration’ (GAM) as an overarching frame.19 From the beginning, the Global Approach was more than a contribution to the global debate. Heads of state or government emphasised that legal pathways for economic purposes and refugee protection in regions of origin should go hand in hand with the prevention of irregular arrivals and effective return. British proposal for external processing of asylum applications in third states, mentioned in Chapter 13.4.7 on asylum, triggered the search for alternative mechanisms of external cooperation. The essential commitment ‘to ensure that migration works to the benefit of all countries’20 included a noticeable dose of self-interest. Rereading the policy documents, one tends to focus on Commission Communications, which are generally more accessible and present a coherent message. We should be careful, however, not to overlook the dynamism—and the occasional tensions—underlying interinstitutional relations during a period which saw a veritable ‘ping-pong’ of reasonably progressive Commission documents emphasising the need for genuine cooperation and statements by the (European) Council stressing the significance of migration control.21 Such interaction reaffirms our earlier findings, in Chapter 2.2.1, that the Commission combined agenda-setting with pragmatic alignment to governmental preferences. The external dimension emerged on the basis of the usual arm-wrestling between the EU institutions. 15 See ch 13.1.2. 16 See Sandra Lavenex, Safe Third Countries: Extending the EU Asylum and Immigration Policies to Central and Eastern Europe (CEU Press 1999) 78–89. 17 Between 150,000 and 300,000 seasonal workers and 30,000 to 90,000 posted workers per annum were allowed to enter from Central and Eastern Europe during the 1990s; see Federal Ministry of the Interior (Bundesministerium des Innern), ‘Migrationsbericht im Auftrag der Bundesregierung’ (November 2004) 42–47 https://ub01.uni-tuebin gen.de/xmlui/handle/10900/63358 (accessed 1 March 2023). 18 See Commission, ‘Communication on immigration’ SEC(91) 1855 final, 6–8. 19 See Annex I to European Council, ‘Presidency Conclusions’ (Council doc 15914/05, 17 December 2005); and Commission, ‘Communication: Migration and Development’ COM(2005) 390 final. 20 ibid 9. 21 See Sandra Lavenex, ‘Shifting Up and Out’ (2006) 29 West Eur Politics 329.

574 International Cooperation with Third States In addition, bilateral relations of the Member States played a critical role at a time when the EU’s position was still in the making. External measures were often reactive and ad hoc, in response to tragic events on the ground. In September 2005, hundreds of migrants stormed the border fence of the Spanish exclave of Ceuta; one year later, the Canary Islands witnessed the arrival of several thousand boat people from West Africa; moreover, Italy signed a deal with the former Libyan dictator Muammar Ghaddafi, further developing a model for externalised migration control it had tested with Albania during the 1990s.22 Italy, Spain, and France employed a strategy spearheaded by Germany towards Central and Eastern Europe a decade earlier: domestic measures were combined with operational cooperation, capacity building, financial support, and (modest) legal pathways.23 Specific examples will be mentioned in our comments on externalisation. While relations with Southern neighbours have played a central role in political debates after the millennium change,24 cooperation with Eastern Europe was equally important in practice. Central and Northern European Member States were active there.25 In all these initiatives, national capitals were at the helm but received increasing political, financial, and operational support from the EU institutions. Thus, the institutions gradually became engulfed in control-oriented migration management. The next phase of international cooperation was set off by migratory movements from Tunisia to Italy in the aftermath of the Arab Spring, which resulted in fierce disputes about internal border controls and pushback practices between France and Italy, mentioned in Chapter 12.3.2 on the Schengen area. The Commission tried to seize the initiative by upgrading what became to be known as the ‘Global Approach to Migration and Mobility’ (GAMM)—with the novel emphasis on ‘mobility’ standing for visas and other travel for shorter periods.26 The contents was pretty much a remake of earlier initiatives, with control- oriented projects gradually gaining the upper hand over development aspects, which retained a prominent place nevertheless.27 Essential features of the GAMM were reiterated by the ‘European Agenda on Migration’, proclaimed by the Commission after a series of devastating shipwrecks in the Central Mediterranean in the spring of 2015; like earlier policy documents, the European Agenda combined border controls with support for neighbours, socioeconomic development, and legal pathways.28 Cooperation was raised to a new level with the EU–Turkey Statement of 18 March 2016 and the ‘New Partnership Framework’, presented shortly thereafter, which responded to almost one million asylum seekers entering via Greece and continuous boat arrivals in Italy and Malta. The change of name, from GAMM to ‘New Partnership

22 See David Scott FitzGerald, Refuge beyond Reach (OUP 2019) 183–99. 23 See Francesca Ippolito, Gianluca Borzoni, and Federico Casolari (eds), Bilateral Relations in the Mediterranean (Edward Elgar Publishing 2020); and Francesca Ippolito and Seline Trevisanut (eds), Migration in the Mediterranean (CUP 2015). 24 See María-Teresa Gil-Bazo, ‘The Practice of Mediterranean States in the Context of the European Union’s Justice and Home Affairs External Dimension’ (2006) 18 IJRL 571, 584–92; and Chetail (n 7) 192–99. 25 See Rosemary Byrne, Gregor Noll, and Jens Vedsted-Hansen, New Asylum Countries? (Kluwer 2002) chs 2–4. 26 See Commission, ‘Communication: The Global Approach to Migration and Mobility’ COM(2011) 743 final. 27 See Georgia Papagianni, ‘Forging an External EU Migration Policy’ (2013) 15 EJML 283; and Tineke Strik, ‘The Global Approach to Migration and Mobility’ (2017) 5 Groningen J Intl L 311, 316–22. 28 See Commission, ‘Communication: A European Agenda on Migration’ COM(2015) 240 final; and Anne Millet-Devalle and Nadia Hammami, ‘Le renforcement des capacités des Etats tiers par l’Union européenne’ in Josiane Auvret-Finck and Anne-Sophie Millet-Devalle (eds), Crise des réfugiés, crise de l’Union européenne? (Pedone 2017) 307.

Theory and Policy Design 575 Framework’ was more than an exercise in rebranding.29 Strategy papers openly embraced the prevention of irregular movements and effective return as the epicentre of the external dimension.30 Development assistance and legal pathways remained on the table as a bargaining chip, but the shift of attention was apparent. Control-oriented externalisation continued unabated under the ‘New Pact on Migration and Asylum’ of 2020. The Pact is mostly known for domestic reform proposals, but it similarly reiterated the basic principles of external migration policies in previous years.31 It replaced the ‘New Partnership Framework’, which had been portrayed, by the Commission, as an integral part of the European Agenda on Migration. This constant change of name may confuse external observers, but it can be rationalised by the desire of each new Commission to present the proverbial old wine in new bottles by means of re-branding. A new phenomenon was the ‘instrumentalisation’ of migration by neighbouring states, mentioned in Chapter 10.7 on general features and Chapter 13.1.6 on asylum. Controversies about instrumentalisation remind us that third states are actors in their own right which pursue strategic objectives and will not simply implement what the EU tells them to do.

18.1.3 Trial and error of external migration policy Declarations of intent regarding ‘management’ and ‘partnership’ in the EU Treaties do not translate into policy substance easily. The output has often been more modest than the official fanfare of strategy papers and the invention of new acronyms like ‘GAMM’ or ‘talent partnership’ might suggest. At an intermediate level, we may describe the evolution of the external dimension as a series of trial and error subscribing to a common trend: while cooperation in the field of legal migration remained patchy and support for refugees was typically ad hoc and project-based, cooperation on control elements was pursued in a more structured manner and had a greater impact. The external dimension remains lop-sided and tilts towards migration control, irrespective of what the EU institutions proclaim in their strategy papers. We can identify a toolbox of instruments.32 First, semi-structural political dialogue at regional and bilateral levels provides a framework for the regular exchange of views. Secondly, treaties and non-binding arrangements may deliver tangible results in terms of rules and opportunities for cross-border movements. Whereas combined readmission agreements and visa facilitation treaties have been reasonably successful, as mentioned in Chapter 11.3.2 on visas and Chapter 16.6.4 on return, the original idea of ‘mobility partnerships’ did not deliver; projects on legal migration were ineffective, and it remains to be seen whether ‘talent partnerships’ will suffer the same fate. Thirdly, operational cooperation was pursued with determination; it comes in diverse forms, ranging from capacity building 29 See also Melissa Mouthaan, ‘Old Wine in New Bottles?’ (2021) 59 JCMS 1177, 1181. 30 See Commission, ‘Communication on establishing a new Partnership Framework with third countries’ COM(2016) 385 final; and the regular progress reports up to the ‘Fifth Progress Report on the Partnership Framework with third countries’ COM(2017) 471 final; updates thereafter can be found in general reports up to ‘Progress Report on the implementation of the European Agenda on Migration’ COM(2019) 481 final. 31 See the Commission (n 14) 17–24. 32 See Commission (n 30) 3–13; Commission, ‘The Global Approach’ (n 26) 5–21; European Council (n 19), Annex I; and Paula García Andrade, Iván Martín, and Sergo Mananashvili, EU Cooperation with Third Countries in the Field of Migration: Study for the European Parliament’ (PE 536.469, October 2015).

576 International Cooperation with Third States over liaison officers to working arrangements with Frontex. Fourthly, refugee protection in regions of origin and cooperation on development to eradicate ‘root causes’ are less controversial but financially important nonetheless. A recurring feature has been the lack of coordination and occasional turf battles between and within the Brussels-based institutions, as well as reticence of national governments to follow their lead. External migration policy spans across policy fields which have traditionally been assigned to different actors, both within EU institutions and at the national level.33 One factor that explains the failure of legal pathways was the unwillingness of home affairs actors to subscribe to the development agenda. States ferociously resisted attempts by the Commission to nudge them into substantive commitments for legal pathways. Conversely, foreign affairs and development cooperation specialists dismissed the use of ‘their’ policies for migration control purposes, at least until the asylum policy crisis of 2015/16 resulted in a reorientation of political priorities. Intra-and inter-institutional antagonism can be as damaging as lack of cooperation with countries of origin or transit, which proved cumbersome as well. EU institutions have introduced mechanisms to improve coordination in crises and with regard to operational cooperation and the European Council has identified ‘increased external action’ as a policy priority.34

18.1.4 Give and take in the mutual interest To be successful, cooperation with third states requires ‘tailor-made comprehensive and balanced’35 solutions, which take into account the specific needs and preferences of the partner countries. International relations theory explains why initiatives are bound to fail if they do not live up to this promise. While invocations of common norms and values are important, cooperation will rarely succeed if it conflicts with what the relevant actors perceive to be essential interests.36 Readmission is a case in point. International law requires states to take back nationals. Nevertheless, they are often unwilling to do so: public opinion dislikes it, economic difficulties might increase, and financial remittances can be a substantial part of gross domestic product (GDP). Effective cooperation requires the EU to make concessions elsewhere; without reciprocal give-and-take, cooperation risks being ineffective. Experience shows that third states agree to conclude and implement readmission agreements when the EU offers something meaningful in return. An attractive reward is visa- free access to the Schengen zone—a valuable asset in the eyes of citizens from the Western Balkans, Ukraine, or Georgia, as mentioned in Chapter 11.3.2. By contrast, many other third states refused to sign readmission agreements, or were reluctant to implement them in practice, without a political sweetener. To give EU institutions additional leverage, a 33 See García Andrade, Martín, and Mananashvili (n 32) 88–94; and Tine Van Criekinge, ‘Integration of Migration Issues in EPAs’ in Gerrit Faber and Jan Orbie (eds), Beyond Market Access for Economic Development (Routledge 2009) 173. 34 See European Council, ‘Conclusions’ (EUCO 1/23, 9 February 2023) No 20–25; and Implementing Decision (EU) 2018/1993 on the EU Integrated Political Crisis Response Arrangements [2018] OJ L320/28, which shall serve as the broader framework for an ‘Operational Coordination Mechanism for the External Dimension of Migration’ (Council doc 15299/21, 30 December 2021). 35 Commission (n 14) 24. 36 See Randall Hansen, ‘Making Cooperation Work’ in Randall Hansen and others (eds), Migration, Nation States, and International Cooperation (Routledge 2011) 14; and Alexander Betts, ‘Introduction’ in Alexander Betts (ed), Global Migration Governance (OUP 2011) 1, 11–23.

Constitutional Foundations 577 revision of the Visa Code Regulation introduced, as we shall see, the option of punitive measures. Other incentives include political prestige, for instance in the form of high profile visits to small countries;37 targeted cooperation with the security sector through capacity building; financial support—or the denial thereof—for migration-related projects; and legal pathways for economic purposes. Of course, negotiations are often defined by pronounced power asymmetries, which are particularly problematic in light of the history of colonial domination and exploitation. From a purely analytical perspective, however, power asymmetries do not hamper reciprocal trade-offs. The outcome may be ‘unfair’, but it can present a realistic package nonetheless, provided the EU institutions engage in real negotiations instead of simply issuing instructions. Such ‘orders’ will often not work in practice, as the example of readmission agreements illustrates. EU institutions should take the views and preferences of partner countries seriously—something they have not always done in the past.38 One option to exploit power asymmetries is to resort to bilateral negotiations instead of multilateral formats, although doing so is not a guarantee of success. Failure to ‘convince’ third states to set up disembarkation platforms for people rescued in the Mediterranean, mentioned in Chapter 13.4.7 on asylum, demonstrates that there are limits to what the EU can achieve on the basis of attractive financial offers and political pressure alone. Package deals can be based on reciprocal trade-offs on migration, such as readmission in return for visa liberalisation, or they may concern cross-sectoral issue linkage. To do so is standard practice in international relations, including among the Member States in the EU context.39 The EU–Turkey cooperation is a telling example: President Erdoğan received a promise to discuss an upgrade of the customs union and progress in the accession talks as a quid pro quo for cooperation on migration, and there might have been an unspoken anticipation that the EU side would refrain from being overtly critical of his authoritarian rule.40 Similarly, EU institutions were crystal clear in negotiations with the UK after Brexit that full access to the single market for services would not be available without the free movement of people.

18.2 Constitutional Foundations Experts in migration law who are not familiar with the idiosyncrasies of Union law may be confused by debates about ‘implied powers’ and the consequences they have for the delineation of spheres of influence of the EU institutions and Member States. Yet, these questions are important precisely because they can have practical consequences (18.2.1). Judges will not usually pronounce themselves on the legal effects—or the absence thereof—of informal cooperation frameworks. They have always existed in the field of migration, while 37 See Commission, ‘First Progress Report on the Partnership Framework with third countries’ COM(2016) 700 final, 2–3 and Annex II. 38 See Sandra Lavenex and Rachel Nellen-Stucky, ‘“Partnering” for Migration in EU External Relations’ in Rahel Kunz and others (eds), Multilayered Migration Governance (Routledge 2011) 116, 129–32; and José Pina-Delgado, ‘The Current International Legal Framework of Economic Migration Management’ in Guy S Goodwin-Gill and Philippe Weckel (eds), Migration & Refugee Protection in the 21st Century (Martinus Nijhoff 2015) 127. 39 See generally Hansen (n 36) 17–18. 40 See ‘EU–Turkey Statement’ (Press Release 144/16, 18 March 2016); and ‘Statement by President von der Leyen following the meeting with Turkish President Erdoğan’ (Statement/21/1603, 8 April 2021).

578 International Cooperation with Third States spreading significantly recently; they are problematic, as they are non-transparent and can evade accountability mechanisms (18.2.2).

18.2.1 Supranational competences Competences for cooperation with third states and international organisations are not limited to scenarios of express authorisation, such as Article 79(3) TFEU for readmission agreements, since the EU benefits from an implied external power in all areas covered by domestic competences.41 As a shared competence, the activation of implied powers is a matter of political choice, to be made by the EU institutions.42 Shared powers turn into an exclusive external competence for international treaty-making after the adoption of secondary legislation insofar as an international treaty may affect internal rules or alter their scope. Member States are prohibited from acting autonomously in such scenarios.43 It is irrelevant whether a treaty openly conflicts with Union law, as thematic overlap is sufficient to bring about external exclusivity.44 In short, the thematic scope of secondary legislation defines the EU’s exclusive external competence. Similarly, minor gaps in the regulatory reach of secondary legislation do not prevent the emergence of exclusive external powers whenever Union law has ‘largely covered’ a field.45 By way of example, a revision of the Refugee Convention would probably be covered by an exclusive Union competence nowadays, since its contents largely coincides with the Qualification Directive. Likewise, the EU institutions may coordinate the position that national governments represent in international fora to which the EU institutions do not have access.46 Decision-making procedures for international agreements were mentioned in Chapter 2.3.5, in particular with regard to factual constraints on the powers of the European Parliament. The integrationist thrust of the case law on competences primarily concerns legislative harmonisation. By contrast, adoption of the Return Directive does not bring about an exclusive external competence for any external action on return. We have to distinguish, rather, between administrative procedures pending removal, which define the contents of the Return Directive, and practical arrangements with third states on actual transfers in readmission agreements or informal cooperation frameworks. Legislative harmonisation within the Return Directive does not give rise to an exclusive competence for practical cooperation with third states.47 Similarly, Member States may pursue operational cooperation with neighbouring countries and states further afield on border controls and migration management. Protocol No 23 limits the pre-emptive effects of the Schengen Borders Code Regulation for such bilateral agreements.48 These are exceptions to the rule that secondary 41 The term ‘implied power’ reflects the historic origin in case law, before the Treaty of Lisbon codified the idea in TFEU, arts 3(2), 216(1). 42 See TFEU, art 2(2); and Case C-600/14 Germany v Council EU:C:2017:935, paras 45–52. 43 See TFEU, arts 3(2), 2(1). 44 See Opinion 1/13 Accession of Third States to the Hague Convention EU:C:2014:2303, paras 69–74. 45 See Opinion 1/03 Lugano Convention EU:C:2006:81, para 126. 46 See TFEU, art 218(9); and ch 5.6.1 for the Refugee Convention. 47 See Paula García Andrade, ‘EU External Competences in the Field of Migration’ (2018) 55 CML Rev 157, 169–73; and Carole Billet, ‘EC Readmission Agreements’ (2010) 12 EJML 45, 61–63. 48 ibid 168; on Protocol [No 23] on external relations of the Member States with regard to the crossing of external borders [2008] OJ C115/304.

Constitutional Foundations 579 legalisation brings about an exclusive competence for international treaties on the same matter. International agreements are sometimes migration-specific, for instance with regard to readmission or visa facilitation. On other occasions, a treaty on a different topic may touch upon migration as a corollary, such as physical presence of natural persons for service provision under trade agreements mentioned in Chapter 14.5.2 on legal migration. In such scenarios, the Court applies a ‘centre of gravity’ test to identify the correct legal basis in light of the aims and contents of the measure; incidental aspects are covered by the prevailing theme.49 As a result, corollary rules on migration in free trade agreements are covered by the Common Commercial Policy,50 while the three migration-specific articles in the Istanbul Convention on violence against women were found to justify recourse to Article 78(2) TFEU as a separate legal basis.51 Such additional legal bases are warranted whenever an agreement simultaneously has several centres of gravity without one being incidental in relation to the other.52 Overlap with the Common Security and Defence Policy (CSDP) will feature in our comments on externalisation that follow. In the same vein as trade agreements cover service provision, corollary rules on illegal migration in contractual relations on development are covered by Articles 206–10 TFEU.53 Judges similarly recognised that development funding can support border management projects to counter terrorism as a precondition for security, without which economic development rarely succeeds.54 It is less clear, however, whether the same applies to control- oriented externalisation projects, which the EU promotes to advance its own interests.55 Such policies may even have negative consequences for development if they buttress authoritarian regimes or advance the position of the security sector in partner countries.56 If development cooperation does not apply, EU institutions can activate justice and home affairs as an alternative or complementary legal basis.57 Article 78(2)(g) TFEU and Article 79(2)(c) TFEU, in particular, establish shared powers in this respect.

18.2.2 Spread of informal cooperation frameworks Foreign affairs differ from domestic policies insofar as legal rules are generally less relevant. Of course, states and other actors must respect international custom and treaty obligations, including human rights, but their activities are much less ‘legalised’ than in domestic affairs. Foreign policy headline news often concern developments that do not translate into legally binding commitments. Think of the Global Compacts on refugees and for migration of 2018, donor conferences to raise humanitarian aid for refugees in the Global South, 49 See Case C-91/05 Commission v Council EU:C:2008:288, para 73. 50 See Opinion 2/15 EU-Singapore FTA EU:C:2017:376, paras 40ff, 217. 51 See Opinion 1/19 Istanbul Convention EU:C:2021:832, paras 302–304. 52 See Case C-244/17 Commission v Council EU:C:2018:662, para 37. 53 On readmission clauses see Case C-377/12 Commission v Council EU:C:2014:1903, paras 42–44, 57–59; and also García Andrade (n 47) 178–82. 54 Case C-403/05 Parliament v Commission EU:C:2007:624, paras 56–60. 55 See Mauro Gatti, ‘The Gendarmes of Europe’ in Ippolito and others (n 23) 140, 149–50. 56 See Strik (n 27) 323–26; and Jean-Pierre Cassarino, ‘Channelled Policy Transfers’ (2014) 16 EJML 97. 57 See Daniel Thym, ‘Legal Framework for EU Immigration Policy’ in Daniel Thym and Kay Hailbronner (eds), EU Immigration and Asylum Law: Article-by-Article Commentary (3rd edn, CH Beck/Hart Publishing/Nomos 2022) MN 20–23.

580 International Cooperation with Third States or peace talks to end the devastating civil wars in Syria or Libya. Successful foreign policy requires the identification of strategic goals and the constant adjustment of methods for their realisation, in line with the diplomatic tradition.58 Law can be a useful instrument in foreign affairs, but it is neither the only nor necessarily the best tool available. Against that background, it is not unusual per se that the external dimension of migration law covers non-binding measures.59 That being said, recent years have seen a proliferation of informal cooperation frameworks, including in the area of readmission where treaties had been agreed upon previously. This left observers with the impression that EU institutions and third states deliberately evaded legal commitments and corresponding institutional safeguards.60 Several episodes support the informalisation hypothesis. First, the substitution of legally binding readmission agreements by informal arrangements, dealt with in Chapter 16.6.4 on return. Secondly, non-binding mobility partnerships and cooperation frameworks, which will be traced in the development context. Thirdly, widespread focus on operational cooperation and financial support for capacity building, discussed below. Fourthly, many legal experts were frustrated by a decision of the Court, mentioned in Chapter 3.3.1 on the judicial output, nonchalantly declaring inadmissible an action against the EU–Turkey Statement.61 The sense of disappointment was reinforced by ‘pushback’ allegations. Generally speaking, we may identify several driving forces behind informalisation. Important benefits are speed, without lengthy ratification, and flexibility, by means of easy adaptation to changing circumstances. In that respect, informal ‘gentlemen’s agreements’ have the same advantages as soft law, which has gained relevance for the internal dimension.62 Moreover, absence of legal obligations may facilitate a political compromise which partner countries would not have accepted in the form of hard legal obligations. The Global Compacts and the, albeit meagre, commitments to legal pathways in mobility partnerships are a case in point. The alternative to soft law would have been the failure of the negotiations. EU institutions generally insist on handing out financial support to third states on the basis of internal legislation, so as to preserve flexibility to alter or stop the payment schedule. President Erdoğan, in return, might not have been willing to commit legally to the statement: ‘All new irregular migrants crossing from Turkey into Greek islands as from 20 March 2016 will be returned to Turkey.’63 In short, informalisation may support the emergence of a political consensus precisely because it is not legally binding. The opposite scenario might also occur. One reason for the proliferation of informal readmission arrangements was the practical failure of legally binding agreements. The EU had invested much political capital in the signature of formal readmission agreements, only to learn that third states often failed to live up to them in practice.64 Governments prefer 58 See generally Daniel Thym, ‘Foreign Affairs’ in Armin von Bogdandy and Jürgen Bast (eds), Principles of European Constitutional Law (2nd edn, Hart Publishing 2009) 310–16, 333–34. 59 On continuity over time see Natasja Reslow, ‘Transformation or Continuity?’ in Sergio Carrera and others (eds), Constitutionalising the External Dimension of EU Migration Policies in Times of Crisis (Edward Elgar Publishing 2019) 95. 60 See Eva Kassoti and Narin Idriz (eds), The Informalisation of the EU’s External Action in the Field of Migration and Asylum (Springer 2022). 61 See also Luc Leboeuf, ‘La Cour de justice face aux dimensions externes de la politique commune de l’asile et de l’immigration’ [2019] Revue trimestrielle de droit européen 55. 62 See ch 2.3.4. 63 ‘EU–Turkey Statement’ (n 40) No 1. 64 See Commission, ‘Communication: Evaluation of EU readmission agreements’ COM(2011) 76 final, 3–5.

Externalisation of Control Practices 581 non-binding arrangements that work to largely dysfunctional formal agreements. Such an outcome need not be to the disadvantage of returnees who can invoke guarantees in the Return Directive and human rights, nonetheless. Informalisation is, however, detrimental to parliamentary prerogatives, as explained in Chapter 2.3.5 on the institutions. Moreover, informal cooperation is notoriously intransparent, thus impeding public scrutiny.65 That argument runs both ways. Third states have occasionally asked for an informal measure instead of a legally binding readmission agreement, whose visibility might be met with scepticism by public opinion in countries of origin and transit.66 This evasion of political and institutional accountability mechanisms renders the informalisation of the external dimension problematic, notwithstanding legitimate benefits.

18.3 Externalisation of Control Practices ‘Externalisation’ has become widely used as a generic term to describe cooperation with third states. It highlights the motivation, on the part of the European Union, to control migration by means of international cooperation. Third states do a job, Member States and EU institutions are unwilling or unable to perform. ‘Externalisation’ is not a neutral description, which the EU institutions use as well. Speaking of ‘externalisation’ involves a criticism of the predominance of control measures over human rights and economic development, echoing the interdisciplinary concept of ‘securitisation’ described in Chapter 4.3.4. The underlying idea is not new and, yet, recent years have witnessed an unprecedented rise of control-oriented externalisation (18.3.1). To gain further leverage, EU institutions have devised various forms of conditionality to ‘convince’ partner countries of the merits of cooperation by means of positive incentives or negative sanctions (18.3.2). Capacity building and operational support is an important component, which often relies on funding from the EU budget (18.3.3). Moreover, Frontex has started deploying operational assets, thus enhancing the profile of the EU’s external action (18.3.4).

18.3.1 A recurring theme in the policy discourse Migration is an international phenomenon by its own definition. It does not come as a surprise, therefore, that governments try to increase leverage by means of collaboration. This tactic was pioneered across the Atlantic in the early days of sovereignty-based migration laws. Successive US administrations cooperated with states of origin or transit to prevent the arrival of migrants deemed undesirable for moral, health, racial, or other reasons67— around the same time as the famous poem of ‘huddled masses yearning to breathe free’ was cast onto a bronze plaque at the statue of liberty.68 An early academic observer called the external cooperation agenda ‘prophetic’,69 and we can only agree with this prediction 65 See Paul James Cardwell, ‘Rethinking Law and New Governance in the European Union’ (2016) 41 EL Rev 362, 373–78. 66 See Jean-Pierre Cassarino, ‘Informalizing EU Readmission Policy’ in Ariadna Ripoll Servent and Florian Trauner (eds), The Routledge Handbook of Justice and Home Affairs Research (Routledge 2018) 83, 90–93. 67 See Scott FitzGerald (n 22) ch 4. 68 See https://en.wikipedia.org/wiki/The_New_Colossus (accessed 1 March 2023). 69 Paul S Peirce, ‘The Control of Immigration as an Administrative Problem’ (1910) 4 Am Pol Sci Rev 374, 389.

582 International Cooperation with Third States from an analytical perspective. EU institutions have invested considerable political, financial, and operational resources in rendering externalisation effective. They even replicate the American model of giving it a semblance of humanitarian legitimacy by highlighting the prevention of deaths on the high seas, the fight against human trafficking, or the eradication of root causes as a positive narrative alongside control imperatives.70 Asylum policies of countries like Germany or Denmark are a direct precursor of contemporary externalisation in response to an increase in asylum applications during the late 1980s and early 1990s. They were among the first to promote visa requirements and carrier sanctions, presented in Chapter 11. Veteran German politician Wolfgang Schäuble served as head of the chancellery when the government negotiated deals with communist East Germany. In return for a sizeable credit line to prevent bankruptcy, the communist regime required potential asylum seekers to present a visa for Western Germany before boarding a plane bound for East Berlin.71 Sweden and Denmark reached similar arrangements, at a time when more than half of the asylum applicants entered via East Berlin.72 A few years later, Schäuble was instrumental, this time as German interior minister, in devising cooperation with Eastern neighbours in the form of readmission agreements and quotas with legal pathways for seasonal and posted workers.73 In 2006, he presented, again as interior minister and together with his French counterpart, the blueprint for mobility partnerships as a ‘balanced partnership’, including control measures.74 Finally, Schäuble is reported to have played a prominent role, as finance minister, in designing the EU–Turkey Statement whose basic outline was agreed upon during a visit of the German Chancellor Angela Merkel to Ankara at the height of the asylum policy crisis at the end of October 2015.75 These anecdotal references are indicative of a general lesson. Externalisation may have reached unprecedented levels in recent years, to an extent that it dominates the foreign policy agenda. Yet, it is entrenched in the history of EU migration law, which, generally, was described as ‘venue shopping’ early on to designate the promotion of securitarian approaches by interior ministries via the European level, as traced in Chapter 2.1.1. From the perspective of the classic destination countries, Schengen cooperation and the initial Dublin Convention had been instruments of externalisation. Contemporary externalisation towards third states replicates this model.76 Third states serve as a ‘buffer zone’ or ‘filter’, in line with the basic commitment in Article 78(2)(g) TFEU that international partnership serves ‘the purpose of managing inflows of people applying for asylum’. After the millennium change, the Mediterranean became the hotbed of externalisation. Italy and Spain, in particular, started cooperating with neighbours—in addition to ongoing

70 See Violeta Moreno-Lax, ‘The EU Humanitarian Border and the Securitization of Human Rights’ (2018) 56 JCMS 119, 121–22; and Tjaša Učakar, ‘The Rhetoric of European Migration Policy and Its Role in Criminalization of Migration’ in Neža Kogovšek Šalamon (ed), Causes and Consequences of Migrant Criminalization (Springer 2020) 91, 99–102. 71 See ‘Interne Sache’ Der Spiegel (7 July 1985) https://www.spiegel.de/politik/interne-sache-a-e59f43ac-0002- 0001-0000-000013514331 (accessed 1 March 2023). 72 See Heiner Busch, Grenzenlose Polizei? (Westfälisches Dampfboot 1995) 10–12. 73 See n 17 and accompanying text. 74 See Zvezda Vankova, Circular Migration and the Rights of Migrant Workers in Central and Eastern Europe (Springer 2020) 54–55; and Andreas Ette, Migration and Refugee Policies in Germany (Barbara Budrich 2017) 203–204. 75 See Robin Alexander, Die Getriebenen (Siedler 2017) 139–43. 76 See Virginie Guiraudon, ‘Before the EU Border: Remote Control of the “Huddled Masses”’ in Kees Groenendijk and others (eds), In Search of Europe’s Borders (Kluwer 2003) 191, 192–96.

Externalisation of Control Practices 583 cooperation between the EU and several Member States with countries in the Western Balkans, Ukraine, or Georgia. While Spanish practices will be mentioned on the pages that follow as an example of capacity building and operational cooperation, the Italian case is representative of formalised arrangements agreed upon at the highest political level. It was spearheaded in relations with Albania when cargo ships carried thousands of Albanians to Italy after the fall of the Iron Curtain, resulting in widely televised pictures.77 Italian practices gave rise to a European Court of Human Rights (ECtHR) judgment, which found no human rights violation.78 Albania agreed to return in the context of broader political and economic advantages it aspired to reap from cooperation at a time of profound economic upheaval. Libya first became the centre of attention during the 2000s when Italy negotiated informal arrangements and an official Treaty of Friendship and Cooperation during the reign of the former dictator Muammar Ghaddafi. The Treaty comprised a symbolically important apology for crimes during Italian colonialism, compensation amounting to US$5 billion, and partnership on diverse matters including migration. The rest of the story is well-known as a result of the ECtHR’s Hirsi judgment, which declared Italian pushback practices to be illegal.79 While Italy was the decisive player, EU institutions had supported cooperation with Libya from the beginning.80 Cooperation with Libya resumed when the chaos after the demise of the dictatorship turned the country into a major source for migrants and refugees trying to reach Europe. Debates during the 2010s encompassed mourning for the deaths of more than 10,000 people at sea, fierce political disputes over search and rescue by state ships and private actors, Frontex missions, the military operation ‘Irini’, support for the UNHCR and IOM, as well as capacity building for the Libyan coastguard.81 According to rumours, the noticeable decrease in the number of boat arrivals from Libya during the summer of 2017 was the result of informal ‘deals’ the Italian government had struck with local warlords, possibly with the help of the Italian gas supplier ENI. Restrictions for private non-governmental organisations operating in the Mediterranean were mentioned in Chapter 16.3.1 on return. Cooperation with Libya has been maintained by successive governments of varying political background. It served as a model for EU–Turkey cooperation mentioned in Chapter 13.4.5 on asylum procedures and in the section on development that will follow.

18.3.2 Conditionality: ‘carrots and sticks’ for cooperation Cooperation usually requires, as we have seen, a give and take in the mutual interest. Third states will not usually cooperate, unless doing so has tangible benefits. While combined visa 77 See https://en.wikipedia.org/wiki/Vlora_(ship) (accessed 1 March 2023); and Alessia Di Pascale, ‘Migration Control at Sea. The Italian Case’ in Bernard Ryan and Valsamis Mitsilegas (eds), Extraterritorial Immigration Control (Martinus Nijhoff 2010) 281, 293–96. 78 See Xhavara and others v Italy and Albania App no 39473/98 (ECtHR, 11 January 2001), which is only available in French. 79 See also Di Pascale (n 77) 296–300. 80 See Sara Hamood, ‘EU-Libya Cooperation on Migration’ (2008) 21 J Refugee Stud 19. 81 See Luigi Francesco Gatta, ‘The European and International Response to the Human Rights Situation of Migrants in Libya’ in Markus Kotzur and others (eds), The External Dimension of EU Migration and Asylum Policies (Nomos 2020) 163.

584 International Cooperation with Third States liberalisation and readmission agreements are paradigmatic for migration-specific trade- offs, money and economic or political advantages can be the foundation for cross-sectoral package deals. In some scenarios, however, the Union asks for cooperation without being able to offer much in return, since visa-free access is not considered an option and because third states benefit from privileged trade relations and sizeable development aid already. To increase the leverage in such scenarios, interior ministries occasionally suggest a simple formula: if third states do not cooperate, we will stop doing so as well. Such proposals are usually called ‘conditionality’ by supporters, while critics speak of ‘sanctions’ or ‘less for less’. From the point of view of economic rational choice theory, they are conceptually sound, since individual decisions are influenced by means of either positive incentives or negative sanctions—the proverbial ‘carrots and sticks’. In the world of diplomacy, however, the threat of sanctions sits uncomfortably with the idea of ‘partnership’ upon which effective cooperation tends to be built; the threat of sanctions is usually reserved for rivals or opponents. For this reason, foreign policy experts prefer the positive language of ‘incentives’ and ‘more for more’, for which mobility partnerships or visa and readmission packages are prime examples.82 Having said this, the binary juxtaposition of sanctions and incentives can hide nuances. Visa requirements, for instance, were introduced at some point and could be lifted at any time. EU institutions thus create the need for positive incentives by means of visa facilitation agreements; along similar lines, they may subtly impose ‘sanctions’ in terms of rearranging development aid or shifting foreign policy priorities in the medium run, without resorting to formal conditionality. Partners are aware of these possibilities. An instructive example was the ‘Joint Way Forward’ with Afghanistan on return, which was agreed upon without a direct linkage in the broader context of substantial financial, military, and political support for peace- building in the war-torn country during the years before the Taliban retook power.83 In addition to these subtle and indirect forms of political pressure, EU institutions have devised at least four formal conditionality instruments. First, visa policy has been deployed as a ‘stick’ for decades. Visa requirements can be re-imposed on third states benefiting from visa free travel in case of a substantial increase in the number of irregular overstayers or unfounded asylum applications, or when cooperation on readmission decreases markedly.84 The suspension procedure has not been activated a single time with the indirect exception of Ecuador and Bolivia, for which visa requirements had been reintroduced in response to illegal migration before the introduction of the suspension mechanism.85 Moreover, the option of suspension can have pre-emptive effects in terms of reminding countries in the Western Balkans, in particular, that visa free travel is contingent upon effective cooperation; the Commission regularly assesses compliance on the basis of concrete indicators and issues policy recommendations.86 Secondly, visa-free access and the threat of suspension cannot be used as a ‘carrot’ and ‘stick’ when the EU is unwilling to waive visa requirements. Nevertheless, the institutions 82 See Cassarino (n 66) 85–89; and Mouthaan (n 29) 1186–91. 83 See Marieke van Houte, ‘The Evolution of Governance and Financing of Migration and Development Policy and Politics in Europe’ in Agnieszka Weinar and others (eds), The Routledge Handbook of the Politics of Migration in Europe (Routledge 2019) 420, 427. 84 See Visa List Regulation (EU) 2018/1806, arts 8–11; introduced first by Regulation (EU) No 1289/2013 [2013] OJ L347/74; and reformed by Regulation (EU) 2017/371 [2017] OJ L61/1). 85 See ch 11.3.1. 86 See ch 11.3.2.

Externalisation of Control Practices 585 agreed to use visa policy as a leverage during the 2019 revision of the Visa Code Regulation. A new mechanism was designed to sanction uncooperative third states, while rewarding what the EU perceives as good behaviour: the regular visa fee of €80 shall be either increased to €120 or 160 in case of problems and may be lowered to €60 as a reward for cooperation; processing times can be extended or shortened; and benefits for bona fide travellers and multi-entry visas can be suspended or extended.87 Sanctions were designed primarily to affect politicians, business people, and other members of the elite who might be eligible for Schengen visas. Based on a Commission Report assessing the performance of third states, Gambia became the test case: the fee waiver for diplomats and advantages for bona fide travellers were suspended.88 Heads of state or government are determined to ‘make full use’ of the instrument in the coming years; an activation towards Iraq and Bangladesh was being discussed at the time of writing, although cooperation with the latter seems to have improved markedly, once the use visa leverage had become a realistic option.89 Thirdly, development aid has been discussed as a bargaining chip ever since the European Council threatened to withhold money from third states ‘which do not cooperate in combating illegal immigration’.90 The idea resurfaced regularly thereafter, not least during the negotiations on the Multiannual Financial Framework 2021–27. The Council and the Parliament finally agreed on an ‘incentive-based approach’, together with highly abstract language about the potential reallocation of funding which stayed short of the 10 per cent quota the Commission had proposed to earmark for performance-based spending.91 It remains to be seen how and whether the new mechanism will be used in practice. Even if it turns out to remain dead letter, the option of subtle reprioritisation remains. Projects discussed throughout this chapter demonstrate that EU institutions have constantly rearranged spending without recourse to formal conditionality. Fourthly, the Commission’s New Pact of 2020 proposed a generic mechanism ‘to facilitate return and readmission’ through the identification of ‘any measure’ deemed appropriate. Political agreement on this proposal, which was wholly uncertain at the time of writing, would effectively introduce a procedure to link migration policy and other external action.92 Decisions would have to be taken in accordance with the legislation applicable, ‘taking into account the Union’s overall relations with the third country’. Suspension of development aid, for instance, would have to follow budgetary rules; economic sanctions or ‘targeted sanctions’ against individuals would have to be based on the Common Foreign and Security Policy and Article 215 TFEU. Thus, the new mechanism would primarily serve as a platform for debate within the institutions, raising the visibility of the migration 87 See Visa Code Regulation (EC) No 810/2009, art 25a(5), (8), as amended by Regulation (EU) 2019/1155 [2019] OJ L188/25); and Salvatore Fabio Nicolosi, ‘Refashioning the EU Visa Policy’ (2020) 22 EJML 467, 483–89. 88 See Council Implementing Decision (EU) 2021/ 1781 on the suspension of certain provisions of Regulation (EC) No 810/2009 with respect to The Gambia [2021] OJ L360/124; on the basis of Commission, ‘Communication: Enhancing cooperation on return and readmission’ COM(2021) 56 final; and the accompanying ‘Report’ COM(2021) 55 final, which is confidential. 89 See European Council (n 34) No 22; and the Commission Proposals, COM(2021) 412 final; and COM(2021) 414 final. 90 European Council, ‘Presidency Conclusions on the Meeting in Thessaloniki’ (22 June 2002) Nos 35–36, which were carefully drafted to honour ‘contractual commitments’ (not usually covering the amount of financial support) and development ‘objectives’ (not: spending targets). 91 See NDICI Regulation (EU) 2021/947, art 20; Proposal for a NDICI Regulation, COM(2018) 460 final, art 16; and Vince Chadwick, ‘EU Breaks Impasse on Aid Budget’ DEVEX (22 December 2020) https://www.devex.com/ news/eu-breaks-impasse-on-aid-budget-98828 (accessed 1 March 2023). 92 See Proposal for an Asylum and Migration Management Regulation, COM(2020) 610 final, art 7.

586 International Cooperation with Third States portfolio. It would not provide for a simplified decision-making. As a result, the idea behind the mechanism can be employed irrespective of formal adoption. By way of example, individuals and companies were threatened with harsh sanctions, in the autumn of 2021, for ‘contributing to activities of the [Belarusian] regime that facilitate . . . the illegal crossing of the external borders’, by means of flying ‘tourists’ from the Near East to Belarus for purposes of onward movement to Poland and Lithuania; pressure on airlines with an economic interest in access to the single market contributed, as one element amongst others, to the significant reduction in the number of arrivals.93 Another controversial move would be to link preferential access to the internal market by least developed countries via the Generalised Scheme of Tariff Preferences (GSP) to cooperation on readmission.94 EU institutions are learning the language of interest- based realpolitik, notwithstanding the classic description of the European project as a champion of values.

18.3.3 Capacity building and operational cooperation Spain was a forerunner of externalisation by means of capacity building in the Southern neighbourhood. As mentioned previously, hundreds of migrants stormed the border fence of the Spanish exclave of Ceuta in September 2005; one year later, the Canary Islands saw the arrival of several thousand boat people from West Africa. The Spanish reaction was swift and decisive. Border fences were fortified with military-style barbed wire, and an integrated surveillance system was set up, which served as a model for today’s Eurosur run by Frontex.95 Information gathered through that system was shared with Morocco, which was given substantial financial and operational support from Spain and the EU. During 2013 and 2020, Morocco received more than €300 million for border management, capacity building and other migration-related projects.96 The facts of the ND and NT judgment illustrates the ‘success’ of operational support: people with a sub-Saharan appearance climbed border fences, sine the Moroccan police prevented them from reaching Spanish border crossing points, where they could have applied for asylum.97 On the Atlantic, Spain spearheaded the prevention of seaborne arrivals to the Canary Islands. It supported the Mauritanian and Senegalese coast guard through the provision of patrol boats and engaged in joint operations, in accordance with the ‘shiprider model’, which the Committee against Torture found to amount to the exercise of Spanish jurisdiction.98 Frontex operations ‘Hera I/II’ were among the first to be deployed on the high seas, and Spanish boats were involved in pushback operations, which, at the time, were still 93 See Regulation (EC) No 765/2006 concerning restrictive measures in respect of Belarus [2006] OJ L134/1, art 2(6)(a)(i), (b), as amended by Regulation (EU) 2021/1985 [2021] OJ L405/1; and https://en.wikipedia.org/ wiki/2021_Belarus%E2%80%93European_Union_border_crisis (accessed 1 March 2023). 94 See Commission, ‘Proposal for a Regulation on applying a generalised scheme of tariff preferences’ COM(2021) 579 final, art 19(1)c. 95 See ch 8.3.4; Jørgen Carling, ‘Migration Control and Migrant Fatalities at the Spanish-African Borders’ (2007) 41 IM Rev 316; and Paula García Andrade, ‘Extraterritorial Strategies to Tackle Irregular Immigration by Sea. A Spanish Perspective’ in Ryan and Mitsilegas, Immigration Control (n 77) 311, 316–18. 96 See ‘EU Support on Migration in Morocco’ (Commission Factsheet, October 2020). 97 See ND and NT v Spain App nos 8675/15 and 8697/15 (ECtHR [GC], 13 February 2020) §§ 212–18. 98 See JHA v Spain (Marine I) Communication no 323/2007 (CAT, 21 November 2008); and García Andrade (n 95) 318–25.

Externalisation of Control Practices 587 considered to be permissible; the Spanish government initiated the ‘Seahorse Network’ to serve as a framework for operational cooperation and capacity building, with generous financial support from the EU budget.99 This network had originally been designed for the Atlantic region and was replicated in the Mediterranean later. Today’s support to the Libyan coast guard is partly delivered through this network with an innocent name; additional money comes from the Italian government.100 Operational cooperation elsewhere comes in the form of multiple projects, both smaller ones and bigger investments. Space precludes further comments. Whereas the near abroad plays a critical role, including the Western Balkans and the Eastern neighbourhood, countries further afield can equally be in the spotlight. Niger became the focus of attention after the events of 2015/16, since crucial transit routes through the Saharan dessert pass through the provincial town of Agadez.101 Chapter 7.4 explained that spending on the external dimension can be difficult to track down. Nevertheless, it is worthwhile to explore the richness of external operational cooperation, which promotes migration control in the European interest.

18.3.4 EU involvement via Frontex and CSDP missions The European Union has started contributing operational assets of its own. The substantial increase in the capacities and statutory muscle of Frontex, presented in Chapter 8.3.3, means that the border agency will increasingly partake in capacity building and operational cooperation with third states through the deployment of supranational personnel and equipment. The agency may send liaison officers, support return, coordinate technical assistance, foster information exchange, and deploy operations. Agency staff may even be authorised, as we have seen, to exercise executive powers, with the agreement of the third state in question. Frontex operations will have to cooperate with civilian and military missions under the Common Security and Defence Policy (CSDP) which are active in the same theatre, as some of these missions embrace a border management component.102 Several ‘Border Assistance Missions’ (BAM) have been set up. While some of these border missions pursue classic security objectives not related to migration, for instance in Moldova and between Egypt and the Gaza strip, others may have a direct effect on migration. By way of example, EUBAM Libya and operations in Sahel region on security sector reform reinforce border management and security apparatuses generally, thus allowing them to also deal with migration.103 The sheer weight of the operational and financial muscle of the border agency might entail

99 See Inmaculada González García, ‘Immigration in Spain’ in Ippolito and others (n 23) 2, 6–11. 100 See Violeta Moreno-Lax and others, ‘The EU Approach on Migration in the Mediterranean: Study for the European Parliament’ (PE 694.413, June 2021) 129–33, 151–52. 101 See Commission (n 30) 3–4; and Moreno-Lax and others (n 100) 104–107. 102 See https://www.eeas.europa.eu/eeas/eu-csdp-missions-and-operations_en (accessed 1 November 2022); and Panos Koutrakos, The EU Common Security and Defence Policy (OUP 2013) chs 5–6. 103 See Decision 2013/233/CFSP on the European Union Integrated Border Management Assistance Mission in Libya (EUBAM Libya) [2013] OJ L138/15, with later amendments; and Panos Koutrakos, ‘The Nexus between CFSP/CSDP and the Area of Freedom, Security and Justice’ in Steven Blockmans and Panos Koutrakos (eds), Research Handbook on the EU’s Common Foreign and Security Policy (Edward Elgar Publishing 2018) 296.

588 International Cooperation with Third States that border-related activities are gradually taken over by Frontex, as long as they do not require defence-specific equipment or training. The most prominent examples of migration-related CSDP operations have been naval missions in the Central Mediterranean. They officially pursued security and defence objectives but contributed to migration management nonetheless. Such operational overlap is unproblematic from a constitutional perspective whenever the use of security and defence personnel or equipment justifies, in light of the centre of gravity test, recourse to the legal bases for security and defence.104 EUNAVFOR MED operated from 2015–20 in parallel to the Frontex missions ‘Triton’ and ‘Themis’. EUNAVFOR MED was rebranded operation ‘Sophia’ after the name of a girl born on a German frigate in September 2015. The mission actively contributed to search and rescue and saved more than 40,000 people, despite the focus of the mandate being on fighting human trafficking and supervising of an arms embargo.105 After fierce political disputes on the involvement of state ships in search and rescue, the geographic focal point of the successor operation ‘Irini’ was relocated to reduce involvement in search and rescue.106 Frontex missions ‘Triton’ and ‘Themis’ similarly became less active. This gradual reorientation of the operational practices serves as a powerful symbol of the general drive towards externalisation.

18.4 Development Cooperation and Legal Pathways Policy initiatives overlap, and so does the drive towards the ‘selfish’ externalisation of migration control with more ‘altruistic’ projects on legal pathways as instruments of development. The Global Compact for Migration, adopted in 2018, reminded us of the underlying rationale: ‘safe, orderly, and regular migration works for all when it takes place in a well- informed, planned and consensual manner’.107 As mentioned at the outset, cooperation is meant to accommodate the interests of home states and destination countries. Projects described below are not about an idealised vision of a borderless world but about pragmatic and sustainable endeavours in the mutual interest. That is easier said than done. Contextual factors help to explain why it can be difficult to design projects that work in practice (18.4.1). Widely used are funding instruments, which focus on migration as a component of development cooperation (18.4.2). Refugee protection in regions of origin and transit also receives much attention (18.4.3), whereas the promise of legal pathways for economic purposes resembles the proverbial dead letter (18.4.4).

104 See TEU, arts 24, 31, 36, 42–43. 105 See Decision (CFSP) 2015/ 778 on a European Union military operation in the Southern Central Mediterranean (EUNAVFOR MED) [2015] OJ L122/31, with later extensions and amendments, including through Decision (CFSP) 2016/993 [2016] OJ L162/18; and Sophie Dura, ‘The EU in the Central Mediterranean’ (2018) 20 EJML 205. 106 See Decision (CFSP) 2020/472 on a European Union military operation in the Mediterranean (EUNAVFOR MED IRINI) [2020] OJ L101/4; and ECRE, ‘Weekly Bulletin’ (2 April 2021) https://mailchi.mp/ecre/ecre-weekly- bulletin-02042021 (accessed 1 March 2023). 107 UNGA, ‘Global Compact for Safe, Orderly and Regular Migration’ (Resolution No 73/195, 19 December 2018) No 13.

Development Cooperation and Legal Pathways 589

18.4.1 Contextual complexity of the migration-development nexus EU institutions recognise the nexus between migration and development, which the international debate placed on the agenda in the 2000s. The ‘New European Consensus on Development’ stresses the linkage,108 as does the UN’s 2030 Agenda for Sustainable Development.109 The Court found, as we have seen, that the Treaty competence for development can cover ancillary projects on migration. At the same time, it is one thing to have a competence and another to design a viable external migration policy. The abstract recognition of a positive nexus of migration and development does not easily translate into policy projects, since various contextual factors determine the impact migration will have on home states and host societies.110 It is difficult, in other words, to identify ideas that might work in practice and to deliver the much-touted ‘triple win’ situation benefiting home states, host societies, and individual migrants. Positive feedback loops are straightforward for the entry and stay of skilled migrants, which featured prominently in the initial debate. Skilled migrants fill vacancies in the Global North, support families abroad via remittances, and promote entrepreneurship and human capital in home states in case of return or circular migration.111 In the case of lesser skilled migration, however, the economic effects on home and host states are less obvious.112 Among the multiple factors, financial remittances are the most obvious reason why migration can support economic and social development.113 Remittances rose sharply during the 2000s, before dropping temporarily during the Covid-19 pandemic. They are higher than official development aid and can amount to substantial parts of the gross domestic product: more than 10 per cent for Egypt, 15 per cent for Gambia, and 6 per cent for Nigeria.114 The EU is an important source, although the footprint of financial flows does not necessarily coincide with politically salient migration corridors. Financial transfers within the EU and towards Eastern neighbours, for instance, are statistically far more relevant than those to the African continent. Moreover, remittances are often an echo of earlier migration from decades ago, from Algeria to France and from Surinam to the Netherlands for instance. Moreover, North and East African countries often receive substantial transfers from ‘guest workers’ in the Gulf States. Much attention has been paid to facilitate transfers and lower service costs. Generally speaking, the migration-development nexus is not primarily about legal pathways to Europe. Libya, Morocco, and Turkey, for example, are not only ‘transit states’ but destination countries in their own right. Oil-rich Libya attracted millions of workers from neighbouring Arab countries and sub-Saharan states before the civil war (and could do so again if safety returned); Morocco actively employs migration policy as an instrument to 108 See ‘The New European Consensus on Development: “Our World, Our Dignity, Our Future” ’ (Joint Statement, 7 June 2017) paras 39–42; building on the previous ‘The European Consensus’ [2006] OJ C46/1. 109 See UNGA, ‘Transforming our World. The 2030 Agenda for Sustainable Development’ (Resolution 70/1, 25 September 2015) No 29 and Goal 10.7. 110 See de Haas (n 6) 227–64. 111 See UN Secretary-General, ‘Report on International Migration and Development’ (UN doc A/60/871, 18 May 2006). 112 See ch 14.1.6; and UNDP, ‘Overcoming Barriers. Human Mobility and Development’ (Human Development Report 2009). 113 See World Bank, ‘Global Economic Prospects. Economic Implications of Remittances and Migration’ (2006). 114 See World Bank, ‘Migration and Development Brief No 33’ (November 2021); and later versions for up-to- date data.

590 International Cooperation with Third States present itself as a leader of West Africa; and Turkey is the home to more than three million refugees and economic migrants from the neighbourhood.115 Some basic figures illustrate underlying push and pull factors: GDP per capita in USD is about 1,000 for Niger, almost 4,000 for Senegal, 9,000 for Morocco, and 32,000 for Turkey—in-between 29,000 for Greece and 40,000 for Italy.116 Migration on the African continent is first and foremost regional, often along established migratory corridors which were interrupted by artificial colonial borders.117 Similarly, forced migration mainly takes places within regions of origin. A prominent motto of the external dimension is the eradication of ‘root causes’, by means of opportunities at home as an alternative to emigration. The popularity of the objective contrasts with the complexity of the economic rationale. Economists recognise what they instructively call the ‘migration hump’.118 It is not usually the very poor who migrate, since doing so requires financial and organisational capital they often do not have. Instead, emigration levels tend to increase with economic development, before they level out and start dropping once the gross domestic product has reached certain levels. Applying these theories to the EU’s neighbours, economists conclude that countries in Northern Africa and the Middle East may be close to the ‘hump’, while economic development South of the Sahara is likely to result in more—not less—migration. Of course, these models are generalisations at a high level of abstraction, since the drivers of migration cannot be reduced to simple economic push and pull factors, as examined in Chapter 4.1. Nevertheless, they illustrate why the idea of eradicating root causes is a simplification.119 It may ideally serve as a rallying cry to create political momentum in favour of a positive approach to international cooperation beyond the externalisation paradigm.

18.4.2 Financial support for multiple purposes Funding for the external dimension can be difficult to track. Chapter 7.4 explained that it is spread across several instruments and trust funds. Building upon these general comments, three examples illustrate the diversity of projects pursued in the name of migration and development. First, the Facility for Refugees in Turkey comprised €6 billion for the 2016–20 period, with the EU promising to pay another €3 billion until the end of 2023. The linkage of these funds to restrictive entry and return policies, described in Chapter 13.4.5 on asylum procedures, was politically controversial, but the projects as such are classic examples of humanitarian assistance and development aid. The Facility financed basic daily needs and supported socio-economic development through heavy investment in schooling and the health sector; only a small segment of the funds was delivered for migration management.120 More than €2 billion have been made available for similar projects in other countries under the EU Regional Trust Fund in Response to the Syrian Crisis,121 and the Council indicated its 115 See generally Kelsey P Norman, ‘Inclusion, Exclusion or Indifference?’ (2019) 45 JEMS 42; and Philippe Fargues, ‘Work, Refuge, Transit’ (2009) 43 IM Rev 544. 116 Data for 2021 in purchasing power parity. 117 See Marie- Laurence Flahaux, ‘Demystifying African Migration’ in Giovanni Carbone (ed), Out of Africa: Why People Migrate (ISPI 2017) 31. 118 See Phil Martin and Edward J Taylor, ‘The Anatomy of the Migration Hump’ in Edward J Taylor (ed), Development Strategy, Employment and Migration (OECD 1996) 43. 119 See Hein de Haas, ‘Turning the Tide?’ (2007) 38 Dev & Change 819. 120 See Commission, ‘Sixth Annual Report on the Facility for Refugees in Turkey’ COM(2021) 255 final. 121 See https://trustfund-syria-region.ec.europa.eu (accessed 1 March 2023).

Development Cooperation and Legal Pathways 591 willingness to do the same for Afghanistan when the Taliban took over in 2021. Funds from the EU budget and special facilities are handed out for projects administered by actors like UNHCR, UNICEF, other international or non-governmental agents, as well as the development agencies of the Member States or host countries. Secondly, projects financed under the EU Emergency Trust Fund (EUTF) for Africa had a more ambivalent character. Migration management was openly embraced as a central objective, and substantial funds have been spent on projects concerning, by way of example, reintegration upon return and information campaigns about the risks of irregular movements.122 Note, however, that the financial envelope of the EUTF was much smaller than supranational and national development funding. Moreover, projects on migration need not be confined to movements towards Europe. African states are, as we have seen, destination countries in their own right. It is a classic example of the migration-development nexus to use the EUTF money to build social services for migrants in Morocco, to combat racism against sub-Saharan migrants, or to develop labour migration policies.123 Capacity building need not be an expression of control-oriented externalisation. Thirdly, the Neighbourhood, Development, and International Cooperation Instrument (NDICI) includes a 10 per cent target for migration-related projects. Most of the money will be allocated under the European Fund for Sustainable Development Plus (EFSD+) with an overall financial envelope of €68 billion during 2020–27; it lists ‘addressing specific socioeconomic root causes of irregular migration’ as one priority amongst others.124 Country-specific multi-annual programmes were adopted during 2021, whose migration components remain to be assessed in detail.125 Experience shows that many initiatives are classic development projects in the form of job creation, education, or better health. Such ‘rebranding’ is not unusual in the development sector; it happened in a similar fashion when climate change became a new priority.126 There is nothing wrong with such dual objectives. They reiterate, rather, our more general conclusion that the migration-development nexus has a multitude of faces.

18.4.3 Refugee protection in the Global South Overall numbers of asylum seekers reaching Europe are small in comparison to the figure of around 80 million refugees worldwide. While roughly half of them are internally displaced, for instance within the Democratic Republic of the Congo, many others stay in regions of origin. For example, Syrians primarily went to Turkey, Lebanon, and Jordan, while Somalis often sought refuge in Kenia, Ethiopia, or tiny Djibouti. European states have supported the UNHCR and other actors in improving the living conditions of these refugees for decades. The EU and Member States are the biggest donor to the UNHCR budget behind the United 122 See https://ec.europa.eu/trustfundforafrica (accessed 1 March 2023); and Oxfam International, ‘An Emergency for Whom?’ (Briefing Note, November 2017). 123 See Factsheet (n 96); and Fanny Tittel-Mosser, ‘The Unintended Legal and Policy Relevance of EU Mobility Partnerships’ (2018) 20 EJML 314. 124 See ch 7.4; and NDICI Regulation (EU) 2021/947, recital 64, arts 31(2), 41(7), Annex V. 125 See https://ec.europa.eu/international-partnerships/global-europe-programming_en (accessed 1 March 2023). 126 See Ida Marie Vammen and Birgitte Mossin Brønden, ‘Donor‐Country Responses to the Migration– Development Buzz’ (2012) 50 Intl Migration 26; and Bakewell (n 6) 18–26.

592 International Cooperation with Third States States, contributing more than €1 billion per year.127 Additional funds are distributed to public and private actors. To do so responds to the preamble of the Refugee Convention that ‘a satisfactory solution . . . cannot . . . be achieved without international co-operation’. More recently, the Global Compact on Refugees emphasised the need for regional solutions and support for countries of first destination.128 In strategy papers at the EU level, the idea of Regional Development and Protection Programmes (RDPPs) emerged from 2005 onwards.129 On the one hand, they appeared as a new slogan for established practices to support refugees in regions of origin; pilot projects in the Great Lakes Region and at the Horn of Africa followed this humanitarian reasoning.130 On the other hand, the political context reveals nuances. RDPPs came about in the slipstream of proposals for external processing, mentioned in Chapter 13.4.5 on asylum. To be sure, the Commission tried to decouple both ideas, which have never been formally linked in the practice of the EU institutions.131 Nevertheless, the potential of overlap with externalisation shows that even seemingly altruistic initiatives can be part of a hidden agenda, among some actors at least. The linkage was put into practice by the EU–Turkey Statement. The Facility for Refugees in Turkey, mentioned previously, served humanitarian purposes and control objectives at the same time. Support was supposed to eliminate potential push factors for onward movements, served as a sweetener to convince President Erdoğan to cooperate with the EU, and prepared the ground for the—controversial—conclusion of Greek courts that the situation in Turkey fulfilled the legislative conditions for safe third countries. Member States openly embraced the linkage between assistance to neighbouring states and the objective ‘to prevent the recurrence of uncontrolled large-scale movements’ when the Taliban took control of Afghanistan.132 The linkage between refugee protection elsewhere and arrivals in Europe are less evident when the EU calls upon partner countries under the European Neighbourhood Policy to ratify important human rights documents. Doing so serves the primary objective of advancing democracy and the rule of law. Similarly, ratification of the Refugee Convention, the introduction of effective status determination procedures, and the improvement of the socioeconomic situation of migrants and refugees are not directly embedded in externalisation strategies.133 Nevertheless, a certain ambiguity remains inherent to these endeavours as well, which some promote as best practice. Instead of spending substantial resources on few refugees reaching Europe, economists and political scientists maintain that rich countries should use the money to improve the lot of literally tens of millions living in often desperate conditions in the Global South.134 127 See ‘Update on budgets and funding (2021 and 2022)’ (UN doc EC/73/SC/CRP.7, 17 February 2022). 128 See UNGA, ‘Global Compact on Refugees’ (Resolution 73/151, 17 December 2018) Nos 17–32. 129 See Commission, ‘Communication on Regional Protection Programmes’ COM(2005) 388 final; the development component was added to the former ‘Regional Protection Programmes’ (RPPs) in the early 2010s. 130 See Anneliese Baldaccini, ‘The External Dimension of the EU’s Asylum and Immigration Policies’ in Helen Toner and others (eds), Whose Freedom, Security and Justice? (Hart Publishing 2007) 277, 284–92. 131 See Karin de Vries, ‘An Assessment of “Protection in Regions of Origin” in Relation to European Asylum Law’ (2007) 9 EJML 83, 84–87; and Violeta Moreno-Lax, ‘External Dimension’ in Steve Peers and others (eds), EU Immigration and Asylum Law (Text and Commentary), Vol 3 (2nd edn, Brill Nijhoff 2015) 617, 633–44. 132 See Council, ‘Statement on the situation in Afghanistan’ (Press Release, 31 August 2021) Nos 4, 6. 133 See Commission, ‘Communication: Taking stock of the European Neighbourhood Policy’ COM(2010) 207 final, 3–4. 134 See Alexander Betts and Paul Collier, Refuge (Penguin 2017).

Development Cooperation and Legal Pathways 593 The best-known example of an innovative project combining refugee protection with economic development is the EU–Jordan Compact. It followed a more general reorientation of refugee support from humanitarian assistance to economic and social empowerment.135 Cooperation with Jordan concerned the creation of special economic zones, which were located close to refugee camps to empower refugees and nationals (to prevent resentment) to support themselves through work. EU institutions backed the idea financially and with the help of trade concessions.136 Similar strategies were implemented at a micro-level in Uganda, mainly for refugees from South Sudan.137 Again, projects on the ground received support from the EU and the Member States. While we cannot exclude that some refugees would have tried to reach Europe in the absence of opportunities in regions of origin, many others would not have done so. Support for refugees in the Global South need not be related to control objectives.

18.4.4 From ‘mobility’ to ‘talent’ partnerships Channels of entry for economic migrants have been an integral part of the migration- development nexus from the beginning. They were originally meant to promote the economic development of home countries by means of both remittances and voluntary return. ‘Circular migration’ was a buzzword in the early debate, which concentrated on skilled labour and foresaw the transfer of knowledge and entrepreneurship.138 Rudimentary rules to facilitate circular migration were adopted at the EU level, through the retention of residence permits during extended periods of absence in the Blue Card Directive and the option of recurring seasonal work.139 At the time, the harmonisation of national rules on the admission of students, researchers, and economic migrants were considered to be instruments of legal pathways precisely because the early debate focused on skilled migration, with seasonal work as the exception to the rule. In order to breathe life into the Global Approach, the Commission promoted ‘mobility partnerships’. They were supposed to combine legal pathways with migration management projects, thus accommodating the interests of states of origin and destination.140 The Commission had grounds for believing that the project might be successful. France and Spain, in particular, were experimenting with comparable packages involving legal migration with African states,141 mirroring the earlier German practice in relations with Central and Eastern Europe after the fall of the iron curtain mentioned previously. Indeed, mobility

135 ibid ch 6; and Global Compact on Refugees (n 128) Nos 31–46. 136 See Decision No 1/2016 EU-Jordan Association Committee [2016] OJ L233/6; and Sara Poli, ‘The Integration of Migration Concerns into EU External Policies’ (2020) 5 European Papers 71, 80–89. 137 See Alexander Betts and others, ‘Refugee Economies in Uganda’ (Report, Refugee Studies Centre 2019). 138 See Commission (n 19) 7–8; and UN Secretary-General (n 111) 56–69. 139 See today’s Blue Card Directive (EU) 2021/1883 recital 52, art 18(3)–(4); Seasonal Workers Directive 2014/ 36/EC, art 16; and generally Purdey Devisscher, ‘Legal Migration in the Relationship between the European Union and ACP Countries’ (2011) 13 EJML 53, 65–78. 140 See Commission, ‘Communication on circular migration and mobility partnerships’ COM(2007) 248 final, 8–13. 141 See Paula García Andrade, Alessia Di Pascale, and Asma Sarraj, ‘L’importance des relations bilatérales entre états membres et pays tiers en matière d’immigration’ in Marleen Maes and others (eds), External Dimensions of EU Migration and Asylum Law and Policy (Bruylant 2011) 251; and Natasha Ward, ‘Facilitating the Temporary Movement of Natural Persons’ in Kunz and others (n 38) 143, 148–58.

594 International Cooperation with Third States partnerships had originally been a Franco-German initiative.142 Negotiations took place on a bilateral basis with selected third states. Doing so reacted to the practical failure of multilateral debates about migration in the context of the Cotonou Convention with countries in Africa, the Caribbean, and the Pacific.143 First mobility partnerships were signed with Moldova, Cape Verde, and Georgia in 2008/09, and a second round of accords with Armenia, Morocco, Azerbaijan, Tunisia, and Jordan followed after the Arab Spring during 2011–14.144 A last mobility partnership was agreed upon with Belarus during 2016, while negotiations with other states like Senegal did not come to a conclusion. Mobility partnerships are prime examples of soft law instruments that ‘are not designed to create legal rights or obligations under international law.’145 While they served as stepping stones for legally binding visa facilitation and readmission agreements in some cases, mobility partnerships failed to live up to the great expectations.146 In particular, the circular migration concept hardly delivered new legal pathways, neither at the point of signature nor thereafter. Mobility partnerships had been meant to serve as frameworks for ongoing cooperation, but this did not happen. Only few specific commitments were made, for instance regarding students and information about entry options under national laws.147 In short, there was little mobility in the mobility partnerships. That may partly be the result of economic developments, since countries in Southern Europe, which would have been natural partners in circular migration, were hit hard by the financial crisis. However, not even France and Germany contributed much to the success of an idea they had originally launched. Most projects realised in the framework of mobility partnerships concentrated on migration management and capacity building, thus effectively involving third states in indirectly controlling movements towards Europe.148 Political scientists emphasise that practical failure reiterates basic lessons for the external dimension, mentioned at the outset. Negotiations were dominated by the European side instead of taking the interests of third states seriously and by making concessions they would find attractive, especially with respect to legal migration. A crucial hurdle was the reluctance of the Member States to commit meaningful channels of entry which the EU institutions could not create out of thin air without the consent of the Member States.149 Similar problems haunt contemporary attempts at creating legal pathways for workers with a lesser degree of qualification, presented in Chapter 14.5.5, although ongoing pilot projects indicate a certain willingness to breathe new life into an idea that had appeared obsolete. More recent initiatives are no longer being pursued under the heading of ‘mobility partnerships’. Similarly, the idea of ‘Common Agendas on Migration and Mobility’ (CAMM) 142 See Vankova (n 74) 54–55. 143 See Katharina Eisele, The External Dimension of the EU’s Migration Policy (Brill/Nijhoff 2014) 251–64; and Devisscher (n 139) 81–93. 144 See García Andrade, Martín, and Mananashvili (n 32) 30–34. 145 Joint Declaration on a Mobility Partnership between the EU and the Republic of Cape Verde, No 17 (Council doc 9460/08 ADD 2, 21 May 2008). 146 See Commission, ‘Mobility Partnerships as a tool of the Global Approach to Migration’ SEC(2009) 1240 final. 147 See eg Annex to the Mobility Partnership with Cape Verde (n 145) No 2; and Joint Declaration on the Mobility Partnership between Tunisia and the EU and its participating Member State, Nos 1–8 (Council doc 16371/1/13, 2 December 2013). 148 See Sergio Carrera and Raül Hernández i Sagrera, ‘Mobility Partnerships’ in Kunz and others (n 38) 97, 101– 106; and Natasja Reslow, ‘Making and Implementing Multi-Actor EU External Migration Policy’ in Sergio Carrera and others (eds), EU External Migration Policies in an Era of Global Mobilities (Brill 2019) 277, 285–91. 149 See ch 14.2.1; Lavenex and Nellen-Stucky (n 38) 129–33; Commission (n 146) 8; and Natasja Reslow and Maarten Vink, ‘Three-Level Games in EU External Migration Policy’ (2015) 53 JCMS 857.

Summary 595 has been effectively abandoned; they would have replicated the idea behind mobility partnerships, albeit without the option of visa-free travel.150 Two CAMMs were signed with Ethiopia in November 2015 and India in March 2016, but their contents remained general and did not cause much debate.151 The Commission wants to move beyond these ineffective predecessors by promoting ‘talent partnership’ as the new label. A first round of talent partnerships was scheduled to be negotiated with Morocco, Tunisia, and Egypt by the end of 2022, together with a feasibility study on future cooperation with Pakistan, Bangladesh, Senegal, and Nigeria.152 Irrespective of whether the talent partnerships will flourish, comments throughout this chapter indicate how important legal pathways for economic purposes can be for the external dimension.

18.5 Summary While most readers will perceive of Article 78(2)(g) TFEU on ‘partnership and cooperation with third countries’ as a euphemism for externalisation nowadays, the external dimension has traditionally pursued other objectives as well. It gained momentum in the new millennium when the international debate started recognising positive feedback loops between migration and development. The title of the Global Compact for Safe, Orderly, and Regular Migration, adopted in 2018, hints at the underlying idea: legal pathways for economic purposes are to be combined with measures preventing irregular movements, thus accommodating the interests of countries of origin and destination. To do so ideally requires reciprocal give and take in the form of package deals, which might involve political concessions on other subject matters than migration. An inspection of the institutional practice displayed multiple manifestations of international cooperation, which our comments organised into two thematic groups: externalisation of control practices and the migration-development nexus. Over the years, the centre of gravity shifted from association agreements, presented in the previous chapter, towards externalisation. This reorientation was accompanied by the rise of informal cooperation below the threshold of legally binding agreements. Nevertheless, primary law remains relevant. Case law on the delimitation of competences entails that other legal bases than the ones for justice and home affairs will often support cooperation with third states. Upon the adoption of internal legislation, the Union acquires an exclusive competence for the conclusion of treaties on the same subject. Such pre-emptive effect will not usually occur, however, with respect to operational cooperation. Comments throughout this chapter stressed how important the activities of the Member States besides EU action. Occasional tensions and disputes, for instance about national commitments under the mobility partnerships, coexist with constructive cooperation.

150 See Commission (n 26) 10–12; and ‘Draft Joint Declaration on a Common Agenda’ (Council doc 7600/13, 22 March 2013). 151 See ‘European Union and Ethiopia sign Common Agenda on Migration and Mobility’ (Press Release IP/15/ 6050, 11 November 2015); and ‘Joint Declaration on a Common Agenda on Migration and Mobility between India and the European Union and its Member States’ (29 March 2016) https://www.ilo.org/newdelhi/whatwedo/publi cations/WCMS_723582/lang--en/index.htm (accessed 1 March 2023). 152 See Commission, ‘Communication: Attracting skills and talent to the EU’ COM(2022) 657 final, 10–14.

596 International Cooperation with Third States Externalisation has become the pivot of the external dimension. Control-oriented cooperation was raised to a new level after the policy crisis of 2015/16, but it goes back much further. Asylum destination countries experimented with various forms of externalisation from the 1980s onwards. The EU’s ‘Global Approach to Migration and Mobility’ (GAMM) took up the idea at a time when Spain and Italy were experiencing migratory pressure. Most of the externalisation policy instruments were developed early on, such as capacity building to enhance the control capacities of neighbouring states and broader political ‘deals’ as a framework for strengthening bilateral relations. A controversial new element are sanctions against third states deemed uncooperative, in particular the suspension of visa-free travel, stricter visa procedures, and linkages between development aid and migration. EU institutions invest ever higher proportions of the external relations budget into projects on migration. Moreover, Frontex will play a greater role in the future, thus complementing the operational activities of the Member States. Strategy papers of the EU institutions continue emphasising the migration-development nexus, including the option of legal pathways. The idea remains conceptually relevant, even though the output in this domain is dwarfed by the dominance of control-oriented externalisation. Contextual factors help to explain why it can be difficult to achieve practical results. Widespread calls for the eradication of ‘root causes’ conceal that economic development can result in more—not less—migration. Similarly, the prevalent focus of the early debate on skilled workers ignored the need for legal pathways for the lesser skilled. That being said, some results have been delivered. Money is an essential tool to improve the situation of migrants and refugees who do not aim to travel to Europe; EU institutions and Member States are major donors. By contrast, mobility partnerships yielded little tangible output. In particular, they did not realise much legal pathways for economic purposes. It remains to be seen whether ‘talent partnership’ will be more successful in the years to come. They might herald a new phase of international cooperation beyond the externalisation paradigm with countries in the near abroad and further afield.

Bibliography Abella, Manolo, ‘Global Competition for Skilled Workers and Consequences’ in Christiane Kuptsch and Eng Fong Pang (eds), Competing for Global Talent (ILO 2006) 11–32. Abraham, David, ‘Immigrant Integration and Social Solidarity in a Time of Crisis’ (2014) 1 Crit Hist Stud 215. Achiume, E Tendayi, ‘Race, Refugees, and International Law’ in Cathryn Costello, Michelle Foster, and Jane McAdam (eds), The Oxford Handbook of International Refugee Law (OUP 2021) 43–59. Ackers, Doede, ‘The Negotiations on the Asylum Procedures Directive’ (2005) 7 EJML 1. Acosta Arcarazo, Diego, ‘The Good, the Bad and the Ugly in EU Migration Law: Is the European Parliament Becoming Bad and Ugly?’ (2009) 11 EJML 19. —— The Long-Term Residence Status as a Subsidiary Form of EU Citizenship (Martinus Nijhoff 2011). ——‘Civic Citizenship Reintroduced?’ (2015) 21 ELJ 200. ——‘Court of Justice of the European Union (Ninth Chamber), WT v Subdelegación del Gobierno en Guadalajara’ (2020) 22 EJML 457. ——and Geddes, Andrew, ‘The Development, Application and Implications of an EU Rule of Law in the Area of Migration Policy’ (2013) 51 JCMS 179. ——and Guild, Elspeth, ‘The Returns Directive’ in Steve Peers and others (eds), EU Immigration and Asylum Law (Text and Commentary), vol 2 (2nd edn, Martinus Nijhoff 2012) 483–524. ——and Martire, Jacopo, ‘Trapped in the Lobby: Europe’s Revolving Doors and the Other as Xenos’ (2014) 39 EL Rev 362. Adam, Ilke and Caponio, Tiziana, ‘Research on the Multi-Level Governance of Migration and Migrant Integration’ in Agnieszka Weinar, Saskia Bonjour, and Lyubov Zhyznomirska (eds), The Routledge Handbook of the Politics of Migration in Europe (Routledge 2019) 26–37. ——and Thym, Daniel, ‘Integration’ in Philippe De Bruycker, Marie De Somer, and Jean-Louis De Brouwer (eds), From Tampere 20 to Tampere 2.0 (EPC 2019) 73–86. Adamczyk, Marek, ‘Balancing Openness with Control’ in Joanna Apap (ed), Justice and Home Affairs Law in the EU (Edward Elgar Publishing 2004) 281–86. Adinolfi, Adelina, ‘Free Movement and Access to Work of Citizens of the New Member States’ (2005) 42 CML Rev 469. Agamben, Giorgio, Homo Sacer: Sovereign Power and Bare Life (Meridian 1998). Ager, Alastair and Strang, Alison, ‘Understanding Integration: A Conceptual Framework’ (2008) 21 J Refugee Stud 166. Aho, C Michael, ‘Fortress Europe: Will the EU Isolate itself from North America and Asia?’ (1994) 29 Colum J World Bus 32. Alba, Richard and Nee, Victor, Remaking the American Mainstream: Assimilation and Contemporary Immigration (Harvard UP 2003). Aleinikoff, Alexander, ‘State-Centered Refugee Law: From Resettlement to Containment’ (1992) 14 Mich J Intl L 120. ——‘The Unfinished Work of the Global Compact on Refugees’ (2018) 30 IJRL 611. Alexander, Robin, Die Getriebenen (Siedler 2017). Alston, Philip and Weiler, Joseph HH, ‘An “Ever Closer Union” in Need of a Human Rights Policy’ in Philip Alston and Joseph HH Weiler (eds), The EU and Human Rights (OUP 1999) 3–66. Alter, Karen J, ‘The European Court’s Political Power’ (1996) 19 Western Eur Politics 458. Anderson, Bridget, Us and Them? The Dangerous Politics of Immigration Control (OUP 2013). Andersson, Ruben, Illegality, Inc: Clandestine Migration and the Business of Bordering Europe (UC Press 2014). Ansems de Vries, Leonie and Guild, Elspeth, ‘Seeking Refuge in Europe’ (2019) 45 JEMS 2156.

598 Bibliography Apap, Joanna, De Bruycker, Philippe, and Schmitter, Catherine, ‘Regularisation of Illegal Aliens in the European Union: Summary Report of a Comparative Study’ (2000) 2 EJML 263. Arendt, Hannah, The Origins of Totalitarianism (Harcourt, Brace & Company 1951). Armstrong, Kenneth, ‘New Governance and the European Union’ in Gráinne de Búrca, Claire Kilpatrick, and Joanne Scott (eds), Critical Perspectives on Global Governance: Liber Amicorum David M Trubek (Hart Publishing 2013) 249–72. Aus, Jonathan P, ‘The Mechanisms of Consensus: Coming to Agreement on Community Asylum Policy’ in Daniel Naurin and Helen Wallace (eds), Games Governments Play in Brussels (Palgrave Macmillan 2008) 99–118. Azoulai, Loïc, ‘The Court of Justice and the Social Market Economy’ (2008) 45 CML Rev 1335. ——‘“Integration Through Law” and Us’ (2016) 14 ICON 449. ——‘Le droit européen de l’immigration, une analyse existentielle’ [2018] Revue trimestrielle de droit européen 519. ——and Coutts, Stephen, ‘Restricting Union Citizens’ Residence Rights on Grounds of Public Security’ (2013) 50 CML Rev 553. Bačić Selanec, Nika and Petrić, Davor, ‘Migrating with Dignity’ (2021) 17 EuConst 498. Bakewell, Oliver, ‘Research beyond the Categories’ (2008) 21 J Refugee Stud 432. Baldaccini, Anneliese, ‘The External Dimension of the EU’s Asylum and Immigration Policies’ in Helen Toner, Elspeth Guild, and Anneliese Baldaccini (eds), Whose Freedom, Security and Justice? (Hart Publishing 2007) 277–300. ——‘Counter-Terrorism and the EU Strategy for Border Security’ (2008) 10 EJML 31. Balibar, Etienne, We, the People of Europe? Reflections on Transnational Citizenship (Princeton UP 2003). Balzacq, Thierry, ‘The Policy Tools of Securitization’ (2008) 46 JCMS 75. Bank, Roland, ‘The Potential and Limitations of the Court of Justice of the European Union in Shaping International Refugee Law’ (2015) 27 IJRL 213. Barbero, Iker, ‘A Ubiquitous Border for Migrants in Transit and their Rights’ (2020) 22 EJML 366. Barbou des Places, Ségolène, ‘EU Asylum Policy and Regulatory Competition’ (2004) 24 J Public Policy 75. ——‘L’empreinte des nationalismes sur le droit de l’immigration de l’Union européenne’ [2018] Revue trimestrielle de droit européen 725. ——‘Droit de l’asile et de l’immigration’ [2019] Revue trimestrielle de droit européen 175. ——‘Taking the Language of “European Sovereignty” Seriously’ (2020) 5 European Papers 287. Barnard, Catherine and Leinarte, Emilija, ‘Mobility of Persons’ in Federico Fabbrini (ed), The Law and Politics of Brexit, vol 3 (OUP 2021) 134–56. Bast, Jürgen, ‘Transnationale Migrationsverwaltung des europäischen Migrationsraums’ (2007) 46 Der Staat 1. ——‘Of General Principles and Trojan Horses: Procedural Due Process in Immigration Proceedings under EU Law’ (2010) 11 German LJ 1006. —— Aufenthaltsrecht und Migrationssteuerung (Mohr Siebeck 2011). ——‘Deepening Supranational Integration: Interstate Solidarity in EU Migration Law’ (2016) 22 EPL 289. — —‘Völker-und unionsrechtliche Anstöße zur Entterritorialisierung des Rechts’ (2017) 76 Veröffentlichungen der Vereinigung Deutscher Staatsrechtslehrer 277. — —‘Vom subsidiären Schutz zum europäischen Flüchtlingsbegriff ’ [2018] Zeitschrift für Ausländerrecht 41. Battjes, Hemme, European Asylum Law and International Law (Martinus Nijhoff 2006). ——‘Piecemeal Engineering’ in Vincent Chetail, Philippe de Bruycker, and Francesco Maiani (eds), Reforming the Common European Asylum System (Brill 2016) 197–239. Bauböck, Rainer (ed), Transnational Citizenship (Edward Elgar Publishing 1994). ——‘Why European Citizenship? Normative Approaches to Supranational Union’ (2007) 8 Theo Inq L 453. —— (ed), Debating Transformations of National Citizenship (Springer 2018).

Bibliography 599 Bauder, Harald, ‘Why We Should Use the Term “Illegalized” Refugee or Immigrant: A Commentary’ (2014) 26 IJRL 327. Bauman, Zygmunt, Strangers at Our Door (Polity Press 2016). —— Globalization: The Human Consequences (Columbia UP 1998). Baumann, Mechthild, Der deutsche Fingerabdruck (Nomos 2006). Baumgärtel, Moritz, Demanding Rights: Europe’s Supranational Courts and the Dilemma of Migrant Vulnerability (CUP 2019). Bay Larsen, Lars, ‘Quelques remarques sur les relations entre le droit international et le régime d’asile européen commun’ in Antonio Tizzano and others (eds), La Cour de justice de l’Union européenne sous la présidence de Vassilios Skouris (Bruylant 2015) 73–88. Beck, Gunnar, The Legal Reasoning of the Court of Justice of the EU (Hart Publishing 2012). Behrman, Simon, ‘Legal Subjectivity and the Refugee’ (2014) 26 IJRL 1. —— (ed), Law and Asylum: Space, Subject, Resistance (Routledge 2018). Beljin, Saša, ‘Rights in EU Law’ in Sacha Prechal and Bert van Roermund (eds), The Coherence of EU Law: The Search for Unity in Divergent Concepts (OUP 2008) 91–122. Bellamy, Richard, Citizenship: A Very Short Introduction (OUP 2008). Benhabib, Seyla, The Rights of Others: Aliens, Residents, and Citizens (CUP 2004). ——‘Claiming Rights across Borders’ (2009) 103 Am Pol Sci Rev 691–704. Bergström, Carl Fredrik, ‘Shaping the New System for Delegation of Powers to EU Agencies’ (2015) 52 CML Rev 219. Berlemann, Michael and Steinhardt, Max Friedrich, ‘Climate Change, Natural Disasters, and Migration: A Survey of the Empirical Evidence’ (2017) 63 CESifo Econ Stud 353. Berlinghoff, Marcel, ‘The Failure of the Commission’s Attempt to Concert a Common European Immigration Policy’ in Centre international de formation européenne (ed), L’Europe en Formation (CAIRN 2009) 183–95. —— Das Ende der „Gastarbeit“. Europäische Anwerbestopps 1970–1974 (Schöningh 2013). Berneri, Chiara, Family Reunification in the EU. The Movement and Residence Rights of Third Country National Family Members of EU Citizens (Hart Publishing 2017). Bernhard, Laurent and Kaufmann, David, ‘Coping with the Asylum Challenge’ (2018) 44 JEMS 2506. Bernsdorff, Norbert, ‘Artikel 18’ in Jürgen Meyer and Sven Hölscheidt (eds), Die Charta der Grundrechte der Europäischen Union (5th edn, Nomos 2019) 382–96. Berthelet, Pierre, ‘Les conséquences des crises migratoires de 2011 et de 2015’ (2018) 54 Cahiers de droit européen 395–436. Bertrand, Brunessen, ‘De quelques paradoxes de la directive procédures: Aspects de droit européen’ [2016] Revue trimestrielle de droit européen 21. Besters, Michiel and Brom, Frans WA, ‘Greedy Information Technology: The Digitalization of the European Migration Policy’ (2010) 12 EJML 455. Betts, Alexander, ‘Survival Migration: A New Protection Framework’ (2010) 16 Glob Gov 361. ——‘Introduction’ in Alexander Betts (ed), Global Migration Governance (OUP 2011) 1–33. ——and Collier, Paul, Refuge: Transforming a Broken Refugee System (Penguin 2017). ——Loescher, Gil, and Milner, James, UNHCR: The Politics and Practice of Refugee Protection (2nd edn, Routledge 2012). Bhagwati, Jagdish and Partington, Martin (eds), Taxing the Brain Drain (North Holland Publishing 1977). Bhandari, Romit, Human Rights and the Revision of Refugee Law (Routledge 2020). Bianchini, Katia, Protecting Stateless Persons (Brill/Nijhoff 2018). Bigo, Didier, ‘Criminalisation of “Migrants”’ in Barbara Bogusz and others (eds), Irregular Migration and Human Rights (Martinus Nijhoff 2004) 61–91. ——Carrera, Sergio, and Guild, Elspeth (eds), Foreigners, Refugees or Minorities? (Ashgate 2013). Billet, Carole, ‘EC Readmission Agreements: A Prime Instrument of the External Dimension of the EU’s Fight against Irregular Immigration’ (2010) 12 EJML 45. Bivand Erdal, Marta and Oeppen, Ceri, ‘Forced to Leave? The Discursive and Analytical Significance of Describing Migration as Forced and Voluntary’ (2018) 44 JEMS 981. Blake, Sir Nicholas, ‘Luxembourg, Strasbourg and the National Court’ (2013) 25 IJRL 349.

600 Bibliography Block, Laura and Bonjour, Saskia, ‘Fortress Europe or Europe of Rights? The Europeanisation of Family Migration Policies in France, Germany and the Netherlands’ (2013) 15 EJML 203. Bloemraad, Irene, ‘Theorising the Power of Citizenship as Claims-Making’ (2018) 44 JEMS 4. Böcker, Anita and Strik, Tineke, ‘Language and Knowledge Tests for Permanent Residence Rights’ (2011) 13 EJML 157. Boelaert-Suominen, Sonja, ‘Non-EU Nationals and Council Directive 2003/109/EC on the Status of Third-Country Nationals Who are Long-Term Residents’ (2005) 42 CML Rev 1011. Boeles, Peter, ‘What Rights Have Migrating Third Country Nationals?’ in Jaap de Zwaan and Flora Goudappel (eds), Freedom, Security and Justice in the European Union (TMC Asser Press 2006) 151–63. ——‘Entry Bans and SIS Alerts’ in Karin Zwaan (ed), The Returns Directive (Wolf 2011) 39–45. ——den Heijer, Maarten, Lodder, Gerrie, and Wouters, Kees, European Migration Law (2nd edn, Intersentia 2014). Bomhoff, Jacco A, ‘Perfectionism in EU Law’ (2014) 12 Camb Yearb Eur Leg Stud 75. Bommes, Michael, ‘Migration Research in Germany’ in Dietrich Thränhardt and Michael Bommes (eds), National Paradigms of Migration Research (Vandenhoeck & Ruprecht 2010) 127–64. Bonjour, Saskia, ‘The Power and Morals of Policy Makers: Reassessing the Control Gap Debate’ (2011) 45 IM Rev 89. ——‘Speaking of Rights’ (2016) 38 Law & Policy 328. — —Ripoll Servent, Ariadna, and Thielemann, Eiko, ‘Beyond Venue Shopping and Liberal Constraint: A New Research Agenda for EU Migration Policies and Politics’ (2017) 25 JEPP 409. Borges, Isabel M, Environmental Change, Forced Displacement and International Law (Routledge 2019). Borjas, George Jesus, ‘The Economics of Immigration’ (1994) 32 Journal of Economic Literature 1667–1717. ——‘Immigration and Welfare Magnets’ (1999) 17 J Lab Econ 607. Bornemann, Jonas, ‘The Guises of and Guidance to Administrative Discretion in the European Court of Justice’s Interpretation of EU Immigration Law’ (2019) 12 Rev E Admin L 97. ——‘The Role of Member State Governments in Migration Litigation before the ECJ’ (2020) 22 EJML 541. —— The Transformation of Discretionary Power in European Migration Law (PhD Thesis, University of Konstanz 2022). Bosniak, Linda, ‘Being Here: Ethical Territoriality and the Rights of Immigrants’ (2007) 8 Theo Inq L 389. Bossuyt, Marc, Strasbourg et les demandeurs d’asile: des juges sur un terrain glissant (Bruylant 2010). Boswell, Christina, ‘Theorizing Migration Policy: Is There a Third Way?’ (2007) 41 IM Rev 75. —— The Political Uses of Expert Knowledge: Immigration Policy and Social Research (CUP 2009). ——‘The “Epistemic Turn” in Immigration Policy Analysis’ in Gary P Freeman and Nikola Mirilovic (eds), Handbook on Migration and Social Policy (Edward Elgar Publishing 2016) 11–27. Boucher, Anna Katherine, ‘How “Skill” Definition Affects the Diversity of Skilled Immigration Policies’ (2020) 46 JEMS 2533. Bouteillet- Paquet, Daphné, ‘Passing the Buck: A Critical Analysis of the Readmission Policy Implemented by the European Union and Its Member States’ (2003) 5 EJML 359. Bouthinon-Dumas, Hughes and Masson, Antoine, ‘Quelles sont les revues juridiques qui comptent à la Cour de justice de l’Union européenne?’ [2013] Revue trimestrielle de droit européen 781. Bovens, Mark, Curtin, Deirdre, and ’t Hart Publishing, Paul, ‘Framework and Design’ in Mark Bovens, Deirdre Curtin, and Paul ’t Hart (eds), The Real World of EU Accountability (OUP 2010) 31–62. Brady, Hugo, ‘Openness versus Helplessness: Europe’s 2015–2017 Border Crisis’ in Tómas JoensenIan Taylor (ed), Small States and the European Migrant Crisis (Springer 2021) 67–85. Brandl, Ulrike, ‘Voluntary Departure as a Priority’ in Madalina Moraru, Galina Cornelisse, and Philippe De Bruycker (eds), Law and Judicial Dialogue on the Return of Irregular Migrants from the European Union (Hart Publishing 2020) 83–104. Braudel, Fernand, ‘La longue durée’ (1958) 13 Annales: Économies, Sociétés, Civilisations 725.

Bibliography 601 Bredimas, Antonis, ‘Le régime des migrants dans le cadre de la Charte Sociale Européenne et de la pratique du Comité européens de droits sociaux’ in Julia Iliopoulos-Strangas and others (eds), Migration –Migration –Migrations (Nomos/Stämpfli 2017) 191–208. Breitenmoser, Stephan and Weyeneth, Robert, ‘Artikel 72 AEUV’ in Hans von der Groeben, Jürgen Schwarze, and Armin Hatje (eds), Europäisches Unionsrecht: EUV, AEUV, GRC. Band 2 (7th edn, Nomos 2015). Breuer, Marten (ed), Principled Resistance to ECtHR Judgments (Springer 2019). Brinkmann, Gisbert, ‘Opinion of Germany on the Single Permit Proposal’ (2012) 14 EJML 351. Brito Bastos, Filipe, ‘An Administrative Crack in the EU’s Rule of Law’ (2020) 16 EuConst 63. Broeders, Denis, Breaking Down Anonymity: Digital Surveillance of Illegal Migrants in Germany and the Netherlands (Amsterdam UP 2009). Brouwer, Evelien, Digital Borders and Real Rights (Martinus Nijhoff 2008). ——and de Vries, Karin, ‘Third-Country Nationals and Discrimination on the Ground of Nationality’ in Marjolein van den Brink, Susanne Burri, and Jenny Goldschmidt (eds), Equality and Human Rights (SIM 2015) 123–46. Brown, Wendy, Walled States, Waning Sovereignty (Zone Books 2010). Buckler, Julius, ‘Auf dem Weg zu einer digitalen europäischen Grenzkontrollarchitektur?’ [2018] Bayerische Verwaltungsblätter 73. ——‘Schengen-Grenzmanagement 2022’ [2020] Die Öffentliche Verwaltung 749. Burson, Bruce, ‘Refugee Status Determination’ in Cathryn Costello, Michelle Foster, and Jane McAdam (eds), The Oxford Handbook of International Refugee Law (OUP 2021) 569–87. Busch, Heiner, Grenzenlose Polizei? Neue Grenzen und polizeiliche Zusammenarbeit in Europa (Westfälisches Dampfboot 1995). Buzan, Barry, Wæver, Ole, and de Wilde, Jaap, Security: A New Framework for Analysis (Rienner 1998). Byrne, Rosemary and Gammeltoft-Hansen, Thomas, ‘International Refugee Law between Scholarship and Practice’ (2020) 32 IJRL 181. ——Noll, Gregor, and Vedsted-Hansen, Jens, ‘Understanding the Crisis of Refugee Law’ (2020) 33 Leiden J Intl L 871–92. ——Noll, Gregor, and Vedsted-Hansen, Jens, New Asylum Countries? Migration Control and Refugee Protection in an Enlarged European Union (Kluwer 2002). Caicedo Camacho, Natalia, ‘Social Rights and Migrants before the European Court of Human Rights’ in David Moya and Georgios Milios (eds), Aliens before the European Court of Human Rights (Brill 2021) 191–210. Caiola, Antonio, ‘Une base juridique pour la solidarité’ (2018) 54 Cahiers de droit européen 437. Çalı, Başak, Costello, Cathryn, and Cunningham, Stewart, ‘Hard Protection through Soft Courts? Non-Refoulement before the United Nations Treaty Bodies’ (2020) 21 German LJ 355. Canor, Iris, ‘My Brother’s Keeper? Horizontal Solange’ (2013) 50 CML Rev 383. Cardwell, Paul James, EU External Relations and Systems of Governance: The CFSP, Euro- Mediterranean Partnership and Migration (Routledge 2009). Carens, Joseph H, ‘Aliens and Citizens: The Case for Open Borders’ (1987) 49 Rev Pol 251. ——‘Live-in Domestics, Seasonal Workers, Foreign Students, and Others Hard to Locate on the Map of Democracy’ (2008) 16 J Political Philos 419. —— The Ethics of Immigration (OUP 2013). Carlier, Jean-Yves, ‘Droit d’asile et des réfugiés: de la protection aux droits’ (2007) 332 Recueil des cours 9–354. ——and Lebœuf, Luc, ‘Droit européen des migrations’ [2015] Journal de droit européen 111. ——and Lebœuf, Luc, ‘Droit européen des migrations’ [2020] Journal de droit européen 132. Carling, Jørgen, ‘Migration Control and Migrant Fatalities at the Spanish-African Borders’ (2007) 41 IM Rev 316. Carrera, Sergio, In Search of the Perfect Citizen? The Intersection between Integration, Immigration and Nationality in the EU (Martinus Nijhoff 2009). ——‘The Impact of the Treaty of Lisbon over EU Policies on Migration, Asylum and Borders’ in Elspeth Guild and Paul Minderhoud (eds), The First Decade of EU Migration and Asylum Law (Martinus Nijhoff 2012) 229–54.

602 Bibliography ——den Hertog, Leonhard, and Parkin, Joanna, ‘The Peculiar Nature of EU Home Affairs Agencies in Migration Control’ (2013) 15 EJML 337. ——and Hernández i Sagrera, Raül, ‘Mobility Partnerships’ in Rahel Kunz, Sandra Lavenex, and Marion Panizzon (eds), Multilayered Migration Governance (Routledge 2011) 97–115. ——Mitsilegas, Valsamis, Allsopp, Jennifer, and Vosyliute, Lina, Policing Humanitarianism: EU Policies Against Human Smuggling and their Impact on Civil Society (Hart Publishing 2019). ——and Petkova, Bilyana, ‘The Potential of Civil Society and Human Rights Organizations through Third-Party Interventions before the European Courts’ in Mark Dawson, Bruno de Witte, and Elise Muir (eds), Judicial Activism at the European Court of Justice (Edward Elgar Publishing 2013) 233–63. Carrubba, Clifford and Gabel, Matthew, International Courts and the Performance of International Agreements (CUP 2014). Casella Colombeau, Sara, ‘Crisis of Schengen?’ (2020) 46 JEMS 2258. Casolari, Federico, ‘The Unbearable “Lightness” of Soft Law’ in Francesca Ippolito, Gianluca Borzoni, and Federico Casolari (eds), Bilateral Relations in the Mediterranean (Edward Elgar Publishing 2020) 215–28. Cassarino, Jean-Pierre, ‘Channelled Policy Transfers: EU-Tunisia Interactions on Migration Matters’ (2014) 16 EJML 97. ——‘Informalizing EU Readmission Policy’ in Ariadna Ripoll Servent and Florian Trauner (eds), The Routledge Handbook of Justice and Home Affairs Research (Routledge 2018) 83–98. Castles, Stephen, ‘The Factors that Make and Unmake Migration Policies’ (2004) 38 IM Rev 852. ——‘Understanding Global Migration: A Social Transformation Perspective’ (2010) 36 JEMS 1565. ——and Kosack, Godula, Immigrant Workers and Class Structure in Western Europe (OUP 1973). Catanzariti, Mariavittoria and Türk, Alexander H, ‘EU Agencies and the Rise of a Mixed Administration in the EU Multi- Jurisdictional Setting’ in Miroslava Scholten and Alex Brenninkmeijer (eds), Controlling EU Agencies (Edward Elgar Publishing 2020) 18–38. Ceyhan, Ayse and Tsoukala, Anastassia, ‘The Securitization of Migration in Western Societies’ (2002) 27 Alternatives: Global, Local, Political 21. Chalmers, Damian, Davies, Gareth, and Monti, Giorgio, European Union Law: Text and Materials (3rd edn, CUP 2019). Chamon, Merjin, EU Agencies (OUP 2016). ——‘A Constitutional Twilight Zone: EU Decentralized Agencies’ External Relations’ (2019) 56 CML Rev 1509. Chatty, Dawn, ‘Anthropology and Forced Migration’ in Elena Fiddian-Qasmiyeh and others (eds), The Oxford Handbook of Refugee & Forced Migration Studies (OUP 2014) 74–85. Chauvin, Sébastien and Garcés-Mascareñas, Blanca, ‘Contradictions in the Moral Economy of Migrant Irregularity’ in Sarah Spencer and Anna Triandafyllidou (eds), Migrants with Irregular Status in Europe (Springer 2020) 33–49. Chetail, Vincent, ‘Paradigm and Paradox of the Migration-Development Nexus’ (2008) 52 German YB Intl L 183. ——‘Are Refugee Rights Human Rights?’ in Ruth Rubio-Marín (ed), Human Rights and Immigration (OUP 2014) 19–72. ——‘The Common European Asylum System: Bric-à-brac of System?’ in Vincent Chetail, Philippe De Bruycker, and Francesco Maiani (eds), Reforming the Common European Asylum System (Brill 2016) 3–38. ——‘Sovereignty and Migration in the Doctrine of the Law of Nations’ (2016) 27 EJIL 901. —— International Migration Law (OUP 2019). Chimni, BS, ‘The Geopolitics of Refugee Studies: A View from the South’ (1998) 11 J Refugee Stud 350. Chiti, Edoardo, ‘Decentralisation and Integration into the Community Administrations: A New Perspective on European Agencies’ (2004) 10 ELJ 402. Chlebny, Jalek, ‘Public Order, National Security and the Rights of Third-Country Nationals in Immigration Cases’ (2018) 20 EJML 115. Cholewinski, Ryszard, Migrant Workers in International Human Rights Law (Clarendon 1997).

Bibliography 603 ——‘Family Reunification as a Constitutional Right?’ in Joanna Apap (ed), Justice and Home Affairs Law in the EU (Edward Elgar Publishing 2004) 259–74. ——‘International Labour Migration’ in Brian Opeskin, Richard Perruchoud, and Jillyanne Redpath- Cross (eds), Foundations of International Migration Law (CUP 2012) 283–311. Clemens, Michael A, ‘Global Skill Partnerships’ (2015) 4 IZA J Lab Poly 1–18. Cohen, Antonin, ‘Ten Majestic Figures in Long Amaranth Robes’ in Antoine Vauchez and Bruno de Witte (eds), Lawyering Europe (Hart Publishing 2013) 21–43. Coleman, Nils, European Readmission Policy (Martinus Nijhoff 2008). Coles, GJL, ‘Approaching the Refugee Problem Today’ in Gil Loescher and Laila Monahan (eds), Refugees and International Relations (OUP 1989) 373–410. Collier, Paul, Exodus: How Migration is Changing Our World (OUP 2013). Colliot-Thélène, Catherine, ‘What Europe Does to Citizenship’ in Damian Chalmers, Markus Jachtenfuchs, and Christian Joerges (eds), The End of the Eurocrats’ Dream (CUP 2016) 127–46. Coman-Kund, Florin, European Union Agencies as Global Actors (Routledge 2018). Cooper, Frederick, Citizenship between Empire and Nation (Princeton UP 2014). Coppens, Jasmine, ‘Interception of Migrant Boats at Sea’ in Violeta Moreno-Lax and Efthymios Papastavridis (eds), ‘Boat Refugees’ and Migrants at Sea (Brill Nijhoff 2016) 199–221. Cornelisse, Galina, Immigration Detention and Human Rights (Martinus Nijhoff 2010). ——‘What’s Wrong with Schengen?’ (2014) 51 CML Rev 741. ——‘The Scope of the Return Directive’ in Madalina Moraru, Galina Cornelisse, and Philippe De Bruycker (eds), Law and Judicial Dialogue on the Return of Irregular Migrants from the European Union (Hart Publishing 2020) 41–62. ——‘Annotation of Case C-357/09 PPU: Kadzoev’ (2011) 48 CML Rev 925. ——and Moraru, Madalina, ‘Introduction’ in Madalina Moraru, Galina Cornelisse, and Philippe De Bruycker (eds), Law and Judicial Dialogue on the Return of Irregular Migrants from the European Union (Hart Publishing 2020) 17–38. Cornelius, Wayne, Martin, Philip, and Hollifield, James F (eds), Controlling Immigration: A Global Perspective (1st edn, Stanford UP 1994). Costello, Cathryn, The Human Rights of Migrants and Refugees in European Law (OUP 2016). ——‘Administrative Governance and the Europeanisation of Asylum and Immigration Policy’ in Herwig CH Hofmann and Alexander H Türk (eds), EU Administrative Governance (Edward Elgar Publishing 2006) 287–340. ——‘The Asylum Procedures Directive in Legal Context’ in Helen Toner, Elspeth Guild, and Anneliese Baldaccini (eds), Whose Freedom, Security and Justice? (Hart Publishing 2007) 151–94. ——and Hancox, Emily, ‘The Recast Asylum Procedures Directive 2013/32/EU’ in Vincent Chetail, Philippe De Bruycker, and Francesco Maiani (eds), Reforming the Common European Asylum System (Brill 2016) 377–445. ——and Mann, Itamar, ‘Border Justice: Migration and Accountability for Human Rights Violations’ (2020) 21 German LJ 311. ——and Moreno-Lax, Violeta, ‘The Extraterritorial Application of the EU Charter of Fundamental Rights’ in Steve Peers and others (eds), The EU Charter of Fundamental Rights: A Commentary (1st edn, CH Beck/Hart Publishing/Nomos 2014) 1657–84. ——and Mouzourakis, Minos, ‘The Common European Asylum System: Where Did it All Go Wrong?’ in Maria Fletcher, Ester Herlin-Karnell, and Claudio Matera (eds), The European Union as an Area of Freedom, Security and Justice (Routledge 2018) 263–300. Countouris, Nicola and Engblom, Samuel, ‘“Protection or Protectionism”. A Legal Deconstruction of the Emerging False Dilemma in European Integration’ (2015) 6 Eur Labour LJ 19. Cousins, Mel, ‘The European Convention on Human Rights and Residence Requirement for the Purposes of Social Assistance Benefits’ (2019) 21 EJML 541. Coutts, Stephen, ‘Citizenship, Territory and COVID-19’ in Dora Kostakopoulou and Daniel Thym (eds), Research Handbook on European Union Citizenship Law and Policy (Edward Elgar Publishing 2022) 116–30. Cover, Robert R, ‘Foreword: Nomos and Narrative’ (1983) 97 Harv L Rev 4. Craig, Paul, EU Administrative Law (3rd edn, OUP 2018).

604 Bibliography Crawley, Heaven and Hagen-Zanker, Jessica, ‘Deciding Where to Go’ (2019) 57 Intl Migration 20. ——and Skleparis, Dimitris, ‘Refugees, Migrants, Neither, Both: Categorical Fetishism and the Politics of Bounding in Europe’s “Migration Crisis”’ (2018) 44 JEMS 48. Crepaz, Markus ML, ‘Rumors that Diversity is the Death of the Welfare State are Greatly Exaggerated’ in Gary P Freeman and Nikola Mirilovic (eds), Handbook on Migration and Social Policy (Edward Elgar Publishing 2016) 315–29. Crépeau, Francois and Hastie, Bethany, ‘The Case for “Firewall” Protections for Irregular Migrants’ (2015) 17 EJML 157. Croon-Gestefeld, Johanna, Reconceptualising European Equality Law (Hart Publishing 2017). Cruz, António, Shifting Responsibility: Carrier Liability in the Member States of the European Union and North America (Trentham Books 1995). Cryer, Robert, Hervey, Tamara, and Sokhi-Bulley, Bal, Research Methodologies in EU and International Law (Hart Publishing 2011). Curtice, John, Executive Power of the European Union (OUP 2009). ——‘Why Leave Won the UK’s EU Referendum’ (2017) 55 JCMS Special Edition 19–37. Czaika, Mathias and Trauner, Florian, ‘EU Visa Policy’ in Ariadna Ripoll Servent and Florian Trauner (eds), The Routledge Handbook of Justice and Home Affairs Research (Routledge 2018) 110–23. Dahinden, Janine, ‘A Plea for the “De-migranticization” of Research on Migration and Integration’ (2016) 39 Ethnic and Racial Stud 2207. Dahlberg, Kenneth A, ‘The EEC Commission and the Politics of the Free Movement of Labour’ (1968) 6 JCMS 310. Dani, Marco, ‘Rehabilitating Social Conflicts in European Public Law’ (2012) 18 ELJ 621. Dann, Philipp, ‘The Political Institutions’ in Armin von Bogdandy and Jürgen Bast (eds), Principles of European Constitutional Law (2nd edn, Hart Publishing 2009) 233–70. Dauvergne, Catherine, Making People Illegal (CUP 2008). —— The New Politics of Immigration and the End of Settler Societies (CUP 2016). Davies, Gareth, ‘Posted Workers’ (1997) 34 CML Rev 571. ——‘Bureaucracy and Free Movement’ [2003] Nederlands Tijdschrift voor Europees Recht 81. Davies, Sara E, ‘Redundant or Essential? How Politics Shaped the Outcome of the 1967 Protocol’ (2007) 19 IJRL 703–28. Dawson, Mark, ‘Three Waves of New Governance in the European Union’ (2011) 37 EL Rev 208. ——‘Better Regulation and the Future of EU Regulatory Law and Politics’ (2016) 53 CML Rev 1209. —— The Governance of EU Fundamental Rights (CUP 2017). De Baere, Geert, ‘The Basics of EU External Relations Law’ in Marleen Maes and others (eds), External Dimensions of EU Migration and Asylum Law and Policy (Bruylant 2011) 121–74. de Boer, Tom and Zieck, Marjoleine, ‘The Legal Abyss of Discretion in the Resettlement of Refugees’ (2020) 32 IJRL 63–71. De Bruycker, Philippe, ‘Introduction générale’ in Philippe De Bruycker (ed), Les régularisations des étrangers illégaux dans l’Union européenne (Bruylant 2000) 1–11. ——‘L’émergence d’une politique européenne d’immigration’ in Philippe De Bruycker (ed), The Emergence of a European Immigration Policy (Bruylant 2003) 1–96. ——‘Legislative Harmonization in European Immigration Policy’ in Ryszard Cholewinski, Richard Perruchoud, and Euan MacDonald (eds), International Migration Law (TMC Asser Press 2007) 329–48. ——‘Chronique de jurisprudence consacrée à l’espace de liberté, de sécurité et de justice’ (2010) 46 Cahiers de droit européen 137. ——‘The European Border and Coast Guard: A New Model Built on an Old Logic’ (2016) 1 European Papers 559. ——and Labayle, Henri, ‘Chronique de jurisprudence consacrée à l’Espace de liberté, de sécurité et de justice’ (2014) 49 Cahiers de droit européen 719. ——and Mananashvili, Sergo, ‘Audi alteram partem in Immigration Detention Procedures between the ECJ, the ECtHR and Member States: G & R’ (2015) 52 CML Rev 569. ——and Tsourdi, Evangelia (Lilian), ‘Building the Common European Asylum System beyond Legislative Harmonisation: Practical Cooperation, Solidarity and External Dimension’ in Vincent

Bibliography 605 Chetail, Philippe De Bruycker, and Francesco Maiani (eds), Reforming the Common European Asylum System (Brill 2016) 473–538. De Coninck, Joyce, ‘Rétention de demandeurs d’asile dans l’Union européenne et instruments parallèles de protection des droits fondamentaux’ (2017) 52 Cahiers de droit européen 83. De Genova, Nicholas, ‘The Legal Production of Mexican/Migrant “Illegality”’ (2004) 2 Latino Studies 160. ——‘Introduction: The Borders of “Europe” and the European Question’ in Nicholas De Genova (ed), The Borders of ‘Europe’ (Duke UP 2017) 1–36. de Groot, Gerard-René and Luk, Ngo Chun, ‘Twenty Years of CJEU Jurisprudence on Citizenship’ (2014) 15 German LJ 821. de Haas, Hein, ‘Turning the Tide? Why Development Will Not Stop Migration’ (2007) 38 Development and Change 819. ——‘Migration and Development: A Theoretical Perspective’ (2010) 44 IM Rev 227. ——‘The Internal Dynamics of Migration Processes: A Theoretical Inquiry’ (2010) 36 JEMS 1587. ——Castles, Stephen, and Miller, Mark J, The Age of Migration (6th edn, Palgrave 2020). de Hart, Betty, ‘The Europeanization of Love: The Marriage of Convenience in European Migration Law’ (2017) 19 EJML 281. de Jong, Petra W and de Valk, Helga AG, ‘Intra-European Migration Decisions and Welfare Systems’ (2020) 46 JEMS 1773. de Lange, Tesseltje, ‘A Low Level of Trust in EU Labour Migration Regulation’ in Carolus Grütters and Tineke Strik (eds), The Blue Card Directive (Wolf 2013) 17–26. ——‘A “Guildian” Analysis of the Equivocal Trusted Sponsorship under EU Labour Migration Law’ in Paul Minderhoud, Sandra Mantu, and Karin Zwaan (eds), Caught in Between Borders (Wolf 2019) 209–16. —— Welcoming Talent (Radboud University 2021). de Lourdes Peroni, Maria, ‘Challenging Culturally Dominant Conceptions in Human Rights Law’ (2010) 4 Hum Rts Int Leg Disc 241. De Somer, Marie, Precedents and Judicial Politics in EU Immigration Law (Routledge 2018). ——‘Schengen: Quo Vadis?’ (2020) 22 EJML 178. de Vries, Karin, ‘An Assessment of “Protection in Regions of Origin” in Relation to European Asylum Law’ (2007) 9 EJML 83. —— Integration at the Border (Hart Publishing 2013). —— ‘Rewriting Abdulaziz’ (2016) 18 EJML 467. ——‘The Integration Exception’ in Daniel Thym (ed), Questioning EU Citizenship (Bloomsbury/Hart Publishing 2017) 267–86. ——‘The Non-National as “The Other”’ in Moritz Jesse (ed), European Societies, Migration, and the Law (CUP 2020) 192–209. de Witte, Bruno, ‘European Union Law: A Unified Academic Discipline?’ in Antoine Vauchez and Bruno de Witte (eds), Lawyering Europe (Hart Publishing 2013) 101–16. ——‘Institutional Variation Across Policy Fields in the Evolution of European Union Law’ (2013) 20 MJECL 465. ——‘Direct Effect, Primacy and the Nature of the Legal Order’ in Paul Craig and Gráinne de Búrca (eds), The Evolution of EU Law (3rd edn, OUP 2021) 187–227. ——and Tsourdi, Evangelia (Lilian), ‘The Court of Justice Endorses the Emergency Scheme for Compulsory Relocation of Asylum Seekers within the European Union’ (2018) 55 CML Rev 1457. de Witte, Floris, Justice in the EU (OUP 2015). Delcour, Laure, ‘The EU. Shaping Migration Patterns in its Neighbourhood and Beyond’ in Dimitry Kochenov and Fabian Amtenbrink (eds), The European Union’s Shaping of the International Legal Order (CUP 2014) 261–81. della Cananea, Giacinto, ‘The European Union’s Mixed Administrative Proceedings’ (2004) 68 L & Contemp Probs 197. ——and Bussani, Mauro (eds), Judicial Review of Administration in Europe (OUP 2021). Dembour, Marie-Bénédicte, When Humans Become Migrants (OUP 2015).

606 Bibliography den Boer, Monica, ‘Justice and Home Affairs Cooperation in the Treaty on European Union’ (1997) 4 MJECL 310. den Heijer, Maarten, Europe and Extraterritorial Asylum (Hart Publishing 2012). ——‘Reflections on Refoulement and Collective Expulsion in the Hirsi Case’ (2013) 25 IJRL 265. ——‘Frontex and the Shifting Approaches to Boat Migration in the European Union’ in Ruben Zaiotti (ed), Externalizing Migration Management (Routledge 2016) 53–71. ——‘Remedies in the Dublin Regulation: Ghezelbash and Karim’ (2017) 54 CML Rev 859. ——‘Transferring a Refugee to Homelessness in Another Member State: Jawo and Ibrahim’ (2020) 57 CML Rev 539. ——‘The Pitfalls of Border Procedures’ (2022) 59 CML Rev 641. ——‘Article 18’ in Steve Peers and others (eds), The EU Charter of Fundamental Rights: A Commentary (2nd edn, CH Beck/Hart Publishing/Nomos 2021) 551–72. ——Rijpma, Jorrit, and Spijkerboer, Thomas, ‘Coercion, Prohibition, and Great Expectations: The Continuing Failure of the Common European Asylum System’ (2016) 53 CML Rev 607. Denza, Eileen, The Intergovernmental Pillars of the European Union (OUP 2002). Derrida, Jacques, De l’hospitalité (Calmann-Lévy 1997). Detjen, Stephan and Steinbeis, Maximilian, Die Zauberlehrlinge (Klett-Cotta 2019). Devisscher, Purdey, ‘Legal Migration in the Relationship between the European Union and ACP Countries’ (2011) 13 EJML 53. Di Pascale, Alessia, ‘Migration Control at Sea: The Italian Case’ in Bernard Ryan and Valsamis Mitsilegas (eds), Extraterritorial Immigration Control (Martinus Nijhoff 2010) 281–310. Diekmann, Maren-Kathrin, Menschenrechtliche Grenzen des Rückführungsverfahrens in Europa (Nomos 2016). Dörig, Harald, ‘Asylum Qualification Directive 2011/95/EU’ in Daniel Thym and Kay Hailbronner (eds), EU Immigration and Asylum Law: Article-by-Article Commentary (3rd edn, CH Beck/Hart Publishing/Nomos 2022) 1229–1314. ——and Langenfeld, Christine, ‘Vollharmonisierung des Flüchtlingsrechts in Europa’ [2016] Neue Juristische Wochenschrift 1. Dougan, Michael, ‘So Long, Farewell, auf Wiedersehen, Goodbye: The UK’s Withdrawal Package’ (2020) 57 CML Rev 631. Drescher, Seymour and Finkelman, Paul, ‘Slavery’ in Bardo Fassbender and Anne Peters (eds), The Oxford Handbook of the History of International Law (OUP 2012) 890–916. Dreyer-Plum, Domenica, ‘Dublin Revisited: Geographic Asymmetry and Corrective Allocation Mechanisms’ (2020) 22 EJML 386. Drywood, Eleanor, ‘Who’s in and Who’s Out? The Court’s Emerging Case Law on the Definition of a Refugee’ (2014) 51 CML Rev 1115. Dubout, Édouard, ‘Les enjeux constitutionnels du pouvoir de substitution de l’agence Frontex’ [2017] Revue trimestrielle de droit européen 457. Dura, Sophie, ‘The EU in the Central Mediterranean’ (2018) 20 EJML 205. Duursma, Jorri, Fragmentation and the International Relations of Micro-States: Self-determination and Statehood (CUP 1996). Düvell, Franck, ‘Paths into Irregularity’ (2011) 13 EJML 275. ——‘The “Great Migration” of Summer 2015. Analysing the Assemblage of Key Drivers in Turkey’ (2019) 45 JEMS 2227. Eckes, Christina, ‘External Relations Law’ in Diego Acosta Arcarazo and Cian C Murphy (eds), EU Security and Justice Law (Hart Publishing 2014) 186–206. Edwards, Alice, ‘Human Rights, Refugees, and the Right “to Enjoy” Asylum’ (2005) 17 IJRL 293. Eeckhout, Piet, External Relations of the European Union (2nd edn, OUP 2011). Eichenhofer, Johannes, ‘Das Recht auf Daueraufenthalt’ in Ferdinand Wollenschläger (ed), Enzyklopädie Europarecht, Band X (2nd edn, Nomos 2021) § 25. Eicke, Tim, ‘Paradise Lost? Exclusion and Expulsion from the EU’ in Kees Groenendijk, Elspeth Guild, and Paul Minderhoud (eds), In Search of Europe’s Borders (Kluwer 2003) 147–69. Eilmansberger, Thomas, ‘The Relationship between Rights and Remedies in EC Law’ (2004) 41 CML Rev 1199.

Bibliography 607 Eisele, Katharina, ‘Public Security and Admission to the EU of Foreign Students: Fahimian’ (2018) 55 CML Rev 279. —— The External Dimension of the EU’s Migration Policy (Brill/Nijhoff 2014). El-Enany, Nadine, Bordering Britain: Law, Race and Empire (Manchester UP 2020). Ellermann, Antje, States against Migrants: Deportation in Germany and the United States (CUP 2009). ——‘Human-Capital Citizenship and the Changing Logic of Immigrant Admissions’ (2020) 46 JEMS 2515. Epiney, Astrid, ‘Das Freizügigkeitsabkommen Schweiz- EU’ in Ferdinand Wollenschläger (ed), Enzyklopädie Europarecht, Band X (2nd edn, Nomos 2021) § 13. Errera, Roger, ‘The CJEU and Subsidiary Protection’ (2011) 23 IJRL 93. Esping-Andersen, Gosta, The Three Worlds of Welfare Capitalism (Princeton UP 1990). Ette, Andreas, Migration and Refugee Policies in Germany (Barbara Budrich 2017). Eule, Tobias G, Borrelli, Lisa Marie, Lindberg, Annika, and Wyss, Anna, Migrants before the Law (Palgrave 2019). Evans, Andrew, ‘European Citizenship’ (1982) 45 MLR 497. Fabbrini, Federico, Fundamental Rights in Europe (OUP 2014). Fahey, Elaine, ‘Swimming in a Sea of Law: Reflections on Water Borders, Irish(-British)-Euro Relations and Opting-out and Opting-in after the Treaty of Lisbon’ (2010) 47 CML Rev 673. Fahrmeir, Andreas, Citizenship: The Rise and Fall of a Modern Concept (Yale UP 2007). ——‘Staatliche Abgrenzungen durch Passwesen und Visumzwang’ in Jochen Oltmer (ed), Handbuch Staat und Migration in Deutschland seit dem 17. Jahrhundert (De Gruyter 2015) 221–44. Faist, Thomas, Fauser, Margit, and Reisenauer, Eveline, Transnational Migration (Polity 2013). Farahat, Anuscheh, Progressive Inklusion (Springer 2014). Farcy, Jean-Baptiste, ‘Labour Immigration Policy in the European Union’ (2020) 22 EJML 198. —— L’Union européenne et l’immigration économique (Anthemis 2021). Fargues, Philippe, ‘Work, Refuge, Transit’ (2009) 43 IM Rev 544. Farris, Sara R, In the Name of Women′s Rights: The Rise of Femonationalism (Duke UP 2017). Favell, Adrian, ‘Integration Nations: The Nation-State and the Research on Immigrants in Western Europe’ in Grete Brochmann (ed), Multicultural Challenge (Emerald 2003) 13–42. ——‘Rebooting Migration Theory’ in Caroline B Brettell and James F Hollifield (eds), Migration Theory (2nd edn, Routledge 2008) 259–78. ——‘Migration Theory Rebooted?’ in Caroline B Brettell and James F Hollifield (eds), Migration Theory (3rd edn, Routledge 2015) 318–28. Fehling, Michael, ‘Eigenwert des Verfahrens im Verwaltungsrecht’ (2011) 70 Veröffentlichungen der Vereinigung Deutscher Staatsrechtslehrer 278. Feldman, David, ‘Was the Nineteenth Century a Golden Age for Immigrants?’ in Andreas Fahrmeir and others (eds), Migration Control in the North Atlantic World (Berghahn 2003) 167–77. Fernand, Braudel, La Méditerranée: L’espace et l’histoire (Arts et Métiers Graphiques 1977). Fernández Rojo, David, EU Migration Agencies: The Operation and Cooperation of Frontex, EASO and Europol (Edward Elgar Publishing 2020). Fernández, Christian, ‘The Unbearable Lightness of Being Swedish?’ (2019) 19 Ethnicities 674. Ferwerda, Jeremy and Gest, Justin, ‘Pull Factors and Migration Preferences’ (2021) 55 IM Rev 431. Fichera, Massimo, ‘Sketches of a Theory of Europe as an Area of Freedom, Security and Justice’ in Maria Fletcher, Ester Herlin-Karnell, and Claudio Matera (eds), The European Union as an Area of Freedom, Security and Justice (Routledge 2018) 34–56. Filliez, Fabrice, ‘Schengen/Dublin’ in Bernd Martenczuk and Servaas van Thiel (eds), Justice, Liberty, Security: New Challenges for EU External Relations (VUB Press 2008) 145–86. Fine, Sarah and Ypi, Lea (eds), Migration in Political Theory (OUP 2016). Fink, Melanie, Frontex and Human Rights (OUP 2018) 35–79. Finotelli, Claudia and Ponzo, Irene, ‘Migrant Inclusion in Southern European Societies’ (2018) 44 JEMS 2303. FitzGerald, David Scott, Refuge beyond Reach: How Rich Democracies Repel Asylum Seekers (OUP 2019).

608 Bibliography —— and Cook-Martin, David, Culling the Masses: The Democratic Origins of Racist Immigration Policy in the Americas (Harvard UP 2014). ——Cook-Martín, David, García, Angela S, and Arar, Rawan, ‘Can You Become One of Us?’ (2018) 44 JEMS 27. Fitzpatrick, Joan, ‘Temporary Protection of Refugees’ (2000) 94 Am J Intl L 279. Fletcher, Maria, ‘Schengen, the European Court of Justice and Flexibility under the Lisbon Treaty’ (2009) 5 EuConst 71. Fontanelli, Filippo, ‘Implementation of EU Law through Domestic Measures after Fransson’ (2014) 39 EL Rev 682. Fornalé, Elisa, ‘A l’envers: Setting the Stage for a Protective Environment to Deal with “Climate Refugees” in Europe’ (2020) 22 EJML 518. Foster, Michelle, International Refugee Law and Socio-Economic Rights (CUP 2007). Foucault, Michel, L’ordre du discours (Gallimard 1971). —— Surveiller et punir: Naissance de la prison (Gallimard 1975). Franklin, Christian NK and Fredriksen, Halvard H, ‘Differentiated Citizenship in the European Economic Area’ in Dora Kostakopoulou and Daniel Thym (eds), Research Handbook on European Union Citizenship Law and Policy (Edward Elgar Publishing 2022) 297–319. Frasca, Eleonora and Gatta, Francesco Luigi: ‘Ebbs and Flows of EU Migration Law and Governance’ (2022) 24 EJML 58. Fredman, Sandra, ‘Substantive Equality Revisited’ (2016) 14 ICON 712. Freeman, Gary P, ‘Modes of Immigration Politics in the Liberal Democratic State’ (1995) 29 IM Rev 861. ——‘Immigrant Incorporation in Western Democracies’ (2004) 38 IM Rev 945. ——‘National Models, Policy Types, and the Politics of Immigration in Liberal Democracies’ (2006) 29 W Eur Pol 227. — —and Mirilovic, Nikola (eds), Handbook on Migration and Social Policy (Edward Elgar Publishing 2016). Friðriksdóttir, Bjarney, What Happened to Equality? (Brill/Nijhoff 2017). Fröhlich, Daniel, Das Asylrecht im Rahmen des Unionsrechts (Mohr Siebeck 2011). Fromont, Michel, ‘A Typology of Administrative Law in Europe’ in Armin von Bogdandy, Sabino Cassese, and Peter M Huber (eds), The Max Planck Handbooks in European Public Law, vol 1 (OUP 2017) 579–600. Fuchs, Maximilian and Cornelissen, Roberts (eds), EU Social Security Law: A Commentary on EU Regulations 883/2004 and 987/2009 (CH Beck/Nomos/Hart Publishing 2015). Fudge, Judy and Herzfeld Olsson, Petra, ‘The EU Seasonal Workers Directive’ (2014) 16 EJML 439. Fukuyama, Francis, The End of History and the Last Man (Free Press 1992). Fullerton, Maryellen, ‘Restricting the Flow of Asylum-Seekers in Belgium, Denmark, the Federal Republic of Germany, and the Netherlands’ (1988/89) 29 Va J Intl L 33. Funke, Andreas, ‘Primärrechtliche Grundlagen’ in Ferdinand Wollenschläger (ed), Enzyklopädie Europarecht, Band X (2nd edn, Nomos 2021) § 16. Gammeltoft-Hansen, Thomas, Access to Asylum (CUP 2011). ——and Hathaway, James C, ‘Non-Refoulement in a World of Cooperative Deterrence’ (2015) 53 Colum J Transnatl L 235. Ganty, Sarah, L’intégration des citoyens européens et des ressortissants de pays tiers en droit de l’Union européenne (Bruylant 2021). García Andrade, Paula, ‘Extraterritorial Strategies to Tackle Irregular Immigration by Sea: A Spanish Perspective’ in Bernard Ryan and Valsamis Mitsilegas (eds), Extraterritorial Immigration Control (Martinus Nijhoff 2010) 311–47. ——‘EU External Competences in the Field of Migration’ (2018) 55 CML Rev 157. ——‘The Duty of Cooperation in the External Dimension of the EU Migration Policy’ in Sergio Carrera and others (eds), EU External Migration Policies in an Era of Global Mobilities (Brill 2019) 299–325.

Bibliography 609 ——Di Pascale, Alessia, and Sarraj, Asma, ‘L’importance des relations bilatérales entre états membres et pays tiers en matière d’immigration’ in Marleen Maes and others (eds), External Dimensions of EU Migration and Asylum Law and Policy (Bruylant 2011) 251–339. Garlick, Madeline, ‘Reception of Asylum-Seekers in the EU in Times of “Crisis”’ in Paul Minderhoud and Karin Zwaan (eds), The Recast Reception Conditions Directive (Wolf 2016) 141–50. Gatta, Luigi Francesco, ‘The European and International Response to the Human Rights Situation of Migrants in Libya’ in Markus Kotzur and others (eds), The External Dimension of EU Migration and Asylum Policies (Nomos 2020) 163–94. Gatti, Mauro, ‘The Gendarmes of Europe’ in Francesca Ippolito, Gianluca Borzoni, and Federico Casolari (eds), Bilateral Relations in the Mediterranean (Edward Elgar Publishing 2020) 140–59. ——and Ott, Andrea, ‘The EU-Turkey Statement’ in Sergio Carrera, Juan Santos Vara, and Tineke Strik (eds), Constitutionalising the External Dimension of EU Migration Policies in Times of Crisis (Edward Elgar Publishing 2019) 175–200. Geddes, Andrew, Immigration and European Integration (Manchester UP 2000). —— The Politics of Migration and Immigration in Europe (1st edn, Sage 2003). ——and Scholten, Peter, The Politics of Migration and Immigration in Europe (2nd edn, Sage 2016). Geiger, Martin and Pécoud, Antoine, ‘The Politics of International Migration Management’ in Martin Geiger and Antoine Pécoud (eds), The Politics of International Migration Management (Palgrave Macmillan 2010) 1–20. Gerards, Janneke, ‘Pluralism, Deference and the Margin of Appreciation Doctrine’ (2011) 17 ELJ 80. —— General Principles of the European Convention on Human Rights (CUP 2019). Ghezelbash, Daniel, Refuge Lost (CUP 2018). ——‘Hyper-Legalism and Obfuscation: How States Evade their International Obligations Towards Refugees’ (2020) 68 Am J Comp L 479. Gibney, Matthew J, ‘Between Control and Humanitarianism: Temporary Protection in Contemporary Europe’ (1999) 14 Georgetown Imm LJ 689. Gil-Bazo, María-Teresa, ‘The Practice of Mediterranean States in the Context of the European Union’s Justice and Home Affairs External Dimension’ (2006) 18 IJRL 571. ——‘The Charter of Fundamental Rights of the European Union and the Right to be Granted Asylum in the Union’s Law’ (2008) 27 Refugee Surv Q 33. Gilbert, Geoff, ‘Is Europe Living Up to its Obligations to Refugees?’ (2004) 15 EJIL 963. Givens, Rhonda and Luedtke, Adam, ‘The Politics of European Union Immigration Policy’ (2004) 32 Poly Stud J 145. Gkliati, Mariana, ‘The EU Returns Agency’ (2022) 24 EJML 545–69. Glick Schiller, Nina and Salazar, Noel B, ‘Regimes of Mobility Across the Globe’ (2013) 39 JEMS 183. Gloninger, Gemma Marolda, ‘From Humanitarian Rescue to Border Security: Managing Migration in the Central Mediterranean’ (2019) 21 EJML 459. Goedings, Simone A, Labor Migration in an Integrating Europe (SDU Uitgevers 2005). Goldner Lang, Iris, ‘No Solidarity without Loyalty: Why Do Member States Violate EU Migration and Asylum Law and What Can Be Done?’ (2020) 22 EJML 39. Golub, Jonathan, ‘In the Shadow of the Vote?’ (1999) 53 IO 733. González García, Inmaculada, ‘Immigration in Spain’ in Francesca Ippolito, Gianluca Borzoni, and Federico Casolari (eds), Bilateral Relations in the Mediterranean (Edward Elgar Publishing 2020) 2–23. Goodman Wallace, Sara, Immigration and Membership Politics in Western Europe (CUP 2014). Goodwin-Gill, Guy S, ‘The Politics of Refugee Protection’ (2008) 27 Refugee Surv Q 8. ——and Lambert, Hélène (eds), The Limits of Transnational Law: Refugee Law, Policy Harmonization and Judicial Dialogue in the European Union (CUP 2010). ——and McAdam, Jane, The Refugee in International Law (4th edn, OUP 2021). Gordon, Jennifer, ‘People are not Bananas: How Immigration Differs from Trade’ (2010) 104 NW U L Rev 1109. Gordon, Milton M, Assimilation in American Life (OUP 1964). Grahl-Madsen, Atle, The Status of Refugees in International Law, vol 2 (AW Sijthoff 1972).

610 Bibliography Gray, Emily and Statham, Paul, ‘Becoming European? The Transformation of the British Pro-migrant NGO Sector in Response to Europeanization’ (2005) 43 JCMS 877. Greer, Steven, The European Convention on Human Rights (CUP 2006). Grimm, Dieter, ‘The Democratic Costs of Constitutionalisation: The European Case’ (2015) 21 ELJ 460. Groenendijk, Kees, ‘Long-Term Immigrants and the Council of Europe’ (1999) 1 EJML 275. ——‘New Borders behind Old Ones’ in Kees Groenendijk, Elspeth Guild, and Paul Minderhoud (eds), In Search of Europe’s Borders (Kluwer 2003) 131–46. ——‘The Directive on Mutual Recognition of Expulsion Decisions’ in Philippe De Bruycker (ed), The Emergence of a European Immigration Policy (Bruylant 2003) 447–63. ——‘Legal Concepts of Integration in EU Migration Law’ (2004) 6 EJML 111. ——‘Reinstatement of Controls at the Internal Borders of Europe: Why and Against Whom?’ (2004) 10 ELJ 150. ——‘Family Reunification as a Right under Community Law’ (2006) 8 EJML 215. ——‘Pre-Departure Integration Strategies in the European Union: Integration or Immigration Policy?’ (2011) 13 EJML 1. ——‘Categorizing Human Beings in EU Migration Law’ in Saskia Bonjour, Andrea Rea, and Dirk Jacobs (eds), The Others in Europe (Institut d’études européennes 2011) 21–36. ——‘Long-Term Residents’ in Steve Peers and others (eds), EU Immigration and Asylum Law (Text and Commentary), vol 2 (2nd edn, Martinus Nijhoff 2012) 287–354. ——‘Recent Developments in EU Law on Migration: The Legislative Patchwork and the Court’s Approach’ (2014) 16 EJML 313. ——‘The Court of Justice and the Development of EEC-Turkey Association Law’ in Daniel Thym and Margarite Zoeteweij-Turhan (eds), Rights of Third-Country Nationals under EU Association Agreements (Brill Nijhoff 2015) 39–61. ——and Guild, Elspeth, ‘Converging Criteria: Creating an Area of Security of Residence for Europe’s Third Country Nationals’ (2001) 3 EJML 37. ——Rondhuis, Merle, and Strik, Tineke, ‘Klagen bij de Europese Commissie’ (2015) 6 Asiel & Migrantenrecht 24. Groß, Thomas, ‘Die Kooperation zwischen europäischen Agenturen und nationalen Behörden’ (2005) 40 Europarecht 54. — —‘Defizite des Grundrechtsschutzes bei FRONTEX- Einsätzen’ [2020] Zeitschrift für Ausländerrecht 51. Grütters, Carolus, Guild, Elspeth, and de Groot, Sebastiaan (eds), Assessment of Credibility by Judges in Asylum Cases in the EU (Wolf 2013). Guild, Elspeth, Immigration Law in the European Community (Kluwer 2001). ——‘The Border Abroad: Visas and Border Controls’ in Kees Groenendijk, Elspeth Guild, and Paul Minderhoud (eds), In Search of Europe’s Borders (Kluwer 2003) 87–104. ——‘Mechanisms of Exclusion: Labour Migration in the European Union’ in Joanna Apap (ed), Justice and Home Affairs Law in the EU (Edward Elgar Publishing 2004) 211–24. —— The Legal Elements of European Identity (Kluwer 2004). ——‘Sanctions for Employers of Irregular Migrants’ in Steve Peers and others (eds), EU Immigration and Asylum Law (Text and Commentary), vol 2 (2nd edn, Martinus Nijhoff 2012) 431–60. ——‘Does the EU Need a European Migration and Protection Agency?’ (2016) 28 IJRL 585. ——and Bigo, Didier, ‘The Transformation of European Border Controls’ in Bernard Ryan and Valsamis Mitsilegas (eds), Extraterritorial Immigration Control (Martinus Nijhoff 2010) 257–79. ——Groenendijk, Kees, and Carrera, Sergio (eds), Illiberal Liberal States: Immigration, Citizenship and Integration in the EU (Ashgate 2009). ——and Peers, Steve, ‘Out of the Ghetto? The Personal Scope of EU Law’ in Steve Peers and Nicola Rogers (eds), EU Immigration and Asylum Law: Text and Commentary (Martinus Nijhoff 2006) 81–114. ——Peers, Steve, and Tomkins, Jonathan, The EU Citizenship Directive: A Commentary (2nd edn, OUP 2019).

Bibliography 611 Guiraudon, Virginie, ‘European Integration and Migration Policy: Vertical Policy-Making as Venue Shopping’ (2000) 38 JCMS 251. — —‘The Marshallian Triptych Reordered’ in Michael Bommes and Andrew Geddes (eds), Immigration and Welfare (Routledge 2000) 72–90. ——‘Before the EU Border’ in Kees Groenendijk, Elspeth Guild, and Paul Minderhoud (eds), In Search of Europe’s Borders (Kluwer 2003) 191–214. Gündoğdu, Ayten, Rightlessness in an Age of Rights: Hannah Arendt and the Contemporary Struggles of Migrants (OUP 2015). Gunnarsdóttir, Kristrún and Rommetveit, Kjetil, ‘The Biometric Imaginary’ (2017) 26 Public Understanding of Science 195. Habbe, Heiko, ‘Übergang der Verantwortung für anerkannte GFK- Flüchtlinge’ [2021] Anwaltsnachrichten Ausländer-und Asylrecht (ANA-ZAR) 49. Haddad, Emma, The Refugee in International Society: Between Sovereigns (CUP 2008). Hahamovitch, Cindy, ‘Creating Perfect Immigrants: Guestworkers of the World in Historical Perspective’ (2003) 44 Lab Hist 69. Hailbronner, Kay, ‘Die Freizügigkeit türkischer Staatsangehöriger nach dem Assoziationsabkommen EWG/Türkei’ [1984] Europarecht 54–83. —— Möglichkeiten und Grenzen einer europäischen Koordinierung des Einreise-und Asylrechts (Nomos 1989). ——‘Readmission Agreements and the Obligation on States under Public International Law’ (1997) 57 Heidelberg J Intl L 1. ——‘Artikel 3 EMRK: Ein neues europäisches Konzept der Schutzgewährung?’ [1999] Die Öffentliche Verwaltung 617. —— Immigration and Asylum Law and Policy of the European Union (Kluwer 2000). ——‘Die Speicherung personenbezogener Daten im AZR’ [2009] Zeitschrift für Ausländerrecht 178. —and Higgins, Imelda, ‘General Rapporteur Report’ in Imelda Higgins (ed), Migration and Asylum Law and Policy in the European Union: FIDE 2004 National Reports (CUP 2004) 455–76. ——and Thiery, Claus, ‘Schengen II and Dublin’ (1997) 34 CML Rev 957–89. ——and Thym, Daniel, ‘Case C-34/09, Gerardo Ruiz Zambrano v Office national de l’emploi’ (2011) 48 CML Rev 1253. ——and Thym, Daniel, ‘Grenzenloses Asylrecht?’ [2016] Juristenzeitung 753. Halfmann, Jost, ‘World Society and Migrations’ in Michael Bommes and Ewa Morawska (eds), International Migration Research (Ashgate 2005) 129–51. Hall, Peter A and Taylor, Rosemary CR, ‘Political Science and the Three New Institutionalisms’ (1995) 44 Pol Stud 936–57. Hammar, Tomas, ‘Comparative Analysis’ in Tomas Hammar (ed), European Immigration Policy (CUP 1985) 239–304. —— Democracy and the Nation-State: Aliens, Denizens and Citizens in a World of International Migration (Avebury 1990). Hamood, Sara, ‘EU-Libya Cooperation on Migration’ (2008) 21 J Refugee Stud 19. Hampshire, James, The Politics of Immigration: Contradictions of the Liberal State (Polity Press 2013). Hancox, Emily, ‘Judicial Approaches to Norm Overlaps in EU Law: A Case Study on the Free Movement of Workers’ (2021) 58 CML Rev 1057. Hansen, Peo and Jonsson, Stefan, Eurafrica: The Untold History of European Integration and Colonialism (Bloomsbury 2014). Hansen, Randall, ‘Globalization, Embedded Realism, and Immigrants to Europe’ (2000) 35 Comp Pol Stud 259–83. ——‘Making Cooperation Work’ in Randall Hansen, Jobst Koehler, and Jeannette Money (eds), Migration, Nation States, and International Cooperation (Routledge 2011) 14–30. Harlow, Carol, Accountability in the European Union (OUP 2002). ——and Rawlings, Richard, Law and Administration (3rd edn, CUP 2009). Harzig, Christiane and Hoerder, Dirk, What is Migration History? (Polity 2009) 66–72. Hathaway, James C, ‘The Evolution of International Refugee Law: 1920–1950’ (1984) 33 ICLQ 348.

612 Bibliography ——‘A Reconsideration of the Underlying Premise of Refugee Law’ (1990) 31 Harvard International Law Journal 129. ——The Law of Refugee Status (1st edn, Butterworths 1991). ——‘The Emerging Politics of Non-entrée’ (1992) 9 Refugees 40. ——‘What’s in a Label?’ (2003) 5 EJML 1. ——‘The Global Cop-Out on Refugees’ (2018) 30 IJRL 591. —— The Rights of Refugees under International Law (2nd edn, CUP 2021). ——and Foster, Michelle, The Law of Refugee Status (2nd edn, CUP 2014). Hatzopoulos, Vassilis, ‘Actively Talking to Each Other’ in Mark Dawson, Bruno de Witte, and Elise Muir (eds), Judicial Activism at the European Court of Justice (Edward Elgar Publishing 2013) 102–41. Hayes-Renshaw, Fiona, ‘The Council of Ministers’ in Dermot Hodson and John Peterson (eds), Institutions of the European Union (4th edn, OUP 2017) ch 4. Heath Wellman, Christopher and Cole, Phillip, Debating the Ethics of Immigration: Is There a Right to Exclude? (OUP 2011). Heath, Anthony F and Richards, Lindsay, ‘Contested Boundaries: Consensus and Dissensus in European Attitudes to Immigration’ (2020) 46 JEMS 489. Hecker, Jan, ‘Zur Europäisierung des Ausländerrechts’ [2011] Zeitschrift für Ausländerrecht 46. Hein, Jeremy, ‘Refugees, Immigrants and the State’ (1993) 19 Ann Rev Sociology 43. Herlin-Karnell, Ester, The Constitutional Structure of Europe’s Area of ‘Freedom, Security and Justice’ and the Right to Justification (Hart Publishing 2019). Hernández i Sagrera, Raül, ‘Reframing the EU Visa Cooperation with Third Countries’ in Didier Bigo, Sergio Carrera, and Elspeth Guild (eds), Foreigners, Refugees or Minorities? (Ashgate 2013) 189–203. Heussner, Kristina, Informationssysteme im Europäischen Verwaltungsverbund (Mohr Siebeck 2007). Hilbrink, Eva, Adjudicating the Public Interest: A Systematic Content Analysis of Strasbourg and Luxembourg Case Law on Legal Restrictions to Immigration and Free Movement (PhD Thesis, Vrije Universiteit Amsterdam 2017). Hinterberger, Kevin Fredy, ‘An EU Regularization Directive’ (2019) 26 MJECL 736. —— Regularisations of Irregularly Staying Migrants in the EU (Nomos 2023). Hoevenaars, Jos, A People’s Court: A Bottom-Up Approach to Litigation Before the European Court of Justice (Eleven 2018). Hofman, Herwig CH, Rowe, Gerard C, and Türk, Alexander H, Administrative Law and Policy of the European Union (OUP 2011) 405–10. ——and Türk, Alexander H, ‘The Development of Integrated Administration in the EU and its Consequences’ (2007) 13 ELJ 253. Hollifield, James F, Immigrants, Markets and States (Harvard UP 1992). ——‘Migration and the “New” International Order: The Missing Regime’ in Bimal Ghosh (ed), Managing Migration (OUP 2000) 75–109. ——Martin, Philip, and Orrenius, Pia M, ‘The Dilemmas of Immigration Control’ in James F Hollifield, Philip Martin, and Pia M Orrenius (eds), Controlling Immigration (3rd edn, Stanford UP 2014) 3–34. ——and Wong, Tom K, ‘The Politics of International Migration’ in Caroline B Brettell and James F Hollifield (eds), Migration Theory (2nd edn, Routledge 2008) 227–88. Honig, Bonnie, ‘Proximity and Paradox’ in Hans Lindahl (ed), A Right to Inclusion and Exclusion? (Hart Publishing 2009) 209–40. Hooghe, Liesbet and Marks, Gary, ‘A Postfunctionalist Theory of European Integration’ (2009) 39 Br J Pol Sci 1. Howard, Marc, The Politics of Citizenship in Europe (CUP 2009). Howarth, David, ‘The Compromise on Denmark and the Treaty on European Union’ (1994) 31 CML Rev 765. Hruschka, Constantin, ‘Anerkennung und Aufnahme als Flüchtling (insb. RL 2011/95/EU und RL 2013/33/EU)’ in Ferdinand Wollenschläger (ed), Enzyklopädie Europarecht, Band X (2nd edn, Nomos 2021) § 21.

Bibliography 613 ——and Maiani, Francesco, ‘Dublin III Regulation (EU) No 604/2013’ in Daniel Thym and Kay Hailbronner (eds), EU Immigration and Asylum Law: Article-by-Article Commentary (3rd edn, CH Beck/Hart Publishing/Nomos 2022) 1639–763. Hublet, Chloé, ‘The Scope of Article 12 of the Treaty of the European Communities vis-à-vis Third- Country Nationals’ (2009) 15 ELJ 757. Hurwitz, Agnes, ‘The 1990 Dublin Convention: A Comprehensive Assessment’ (1999) 11 IJRL 646. —— The Collective Responsibility of States to Protect Refugees (OUP 2009). Idriz, Narin and Tobler, Christa, ‘“Citizenship of the Association”. The Examples of Turkey and Switzerland’ in Dora Kostakopoulou and Daniel Thym (eds), Research Handbook on European Union Citizenship Law and Policy (Edward Elgar Publishing 2022) 320–42. Iglesias Sánchez, Sara, ‘Free Movement of Third Country Nationals in the European Union?’ (2009) 15 ELJ 791. ——‘Fundamental Rights Protection for Third Country Nationals and Citizens of the Union’ (2013) 15 EJML 137. ——‘Nationality’ in Elspeth Guild, Cristina Gortázar Rotaeche, and Dora Kostakopoulou (eds), The Reconceptualization of European Union Citizenship (Brill/Nijhoff 2014) 68–87. ——‘Irregular Migrants in Europe’ in Francesca Ippolito and Sara Iglesias Sánchez (eds), Protecting Vulnerable Groups (Hart Publishing 2015) 429–51. ——‘Constitutional Identity and Integration: EU Citizenship and the Emergence of a Supranational Alienage Law’ (2017) 18 German LJ 1797. — —‘Single Permit Directive 2011/ 98/ EU’ in Daniel Thym and Kay Hailbronner (eds), EU Immigration and Asylum Law: Article-by-Article Commentary (3rd edn, CH Beck/Hart Publishing/ Nomos 2022) 910–58. Ilareva, Valeria, ‘The Right to be Heard’ in Madalina Moraru, Galina Cornelisse, and Philippe De Bruycker (eds), Law and Judicial Dialogue on the Return of Irregular Migrants from the European Union (Hart Publishing 2020) 351–68. Iliopolou-Penot, Anastasia (ed), Directive 2004/338 relative au droit de séjour des citoyens de l’Union européenne et des membres de leur famille: Commentaire article par article (Bruylant 2020). İneli Ciğer, Meltem, ‘Time to Activate the Temporary Protection Directive’ (2016) 18 EJML 1. —— Temporary Protection in Law and Practice (Brill/Nijhoff 2018). ——‘Resettlement Framework Regulation (Proposal)’ in Daniel Thym and Kay Hailbronner (eds), EU Immigration and Asylum Law: Article-by-Article Commentary (3rd edn, CH Beck/Hart Publishing/ Nomos 2022) 1764–806. ——‘Immediate Protection in the New Pact on Migration and Asylum’ in Daniel Thym and Odysseus Academic Network (eds), Reforming the Common European Asylum System (Nomos 2022) 149–60. Infantino, Federica, ‘How Does Policy Change at the Street Level?’ (2021) 47 JEMS 1028. Inghammar, Andreas, ‘The Employment Contract Revisited’ (2010) 12 EJML 193. Ippolito, Francesca, ‘Le migrant confronté aux discriminations selon l’origine’ in Guy S Goodwin-Gill and Philippe Weckel (eds), Migration & Refugee Protection in the 21st Century (Martinus Nijhoff 2015) 245–63. ——Borzoni, Gianluca, and Casolari, Federico (eds), Bilateral Relations in the Mediterranean (Edward Elgar Publishing 2020). ——and Trevisanut, Seline (eds), Migration in the Mediterranean (CUP 2015). Ipsen, Hans-Peter, Europäisches Gemeinschaftsrecht (Mohr Siebeck 1972). Itzcovich, Giulio, ‘Legal Order, Legal Pluralism, Fundamental Principles’ (2012) 18 ELJ 358. Jääskinen, Niilo, ‘Fundamental Social Rights in the Charter’ in Steve Peers and others (eds), The EU Charter of Fundamental Rights: A Commentary (2nd edn, CH Beck/Hart Publishing/Nomos 2021) 1855–73. Jacob, Marc A, Precedents and Case-Based Reasoning in the European Court of Justice (CUP 2014). Jakuleviciene, Lyra, ‘Pre-Screening at the Border in Asylum and Migration Pact’ in Daniel Thym and Odysseus Academic Network (eds), Reforming the Common European Asylum System (Nomos 2022) 81–98. Janda, Constanze, Migranten im Sozialstaat (Mohr Siebeck 2012).

614 Bibliography Jaroszyński, Tomasz, ‘National Parliaments’ Scrutiny of the Principle of Subsidiarity’ (2020) 16 EuConst 91. Jeandesboz, Julien, ‘Ceci n’est pas un contrôle: PNR Data Processing and the Reshaping of Borderless Travel in the Schengen Area’ (2021) 23 EJML 431. Jesse, Moritz, The Civic Citizens of Europe (Brill/Nijhoff 2017). Joerges, Christian, ‘Europe a Großraum?’ in Christian Joerges and Navraj Singh Ghaleigh (eds), Darker Legacies of Law in Europe (Hart Publishing 2003) 167–91. Johnson, Kevin R, ‘Race Matters: Immigration Law and Policy Scholarship, Law in the Ivory Tower, and the Legal Indifference of the Race Critique’ [2000] U Ill L Rev 525. Joly, Daniele, ‘The Porous Dam: European Harmonization on Asylum in the Nineties’ (1994) 6 IJRL 159. Jönsson, Christer, Tägil, Sven, and Törnqvist, Gunnar, Organizing European Space (Sage 2000). Joppke, Christian, ‘Why Liberal States Accept Unwanted Immigration’ (1998) 50 World Pol 26. —— Immigration and the Nation State (OUP 1999). ——‘Comparative Citizenship: A Restrictive Turn in Europe?’ (2008) 2 JL & Ethics Hum Rights 1. —— Citizenship and Immigration (Polity Press 2010). —— Is Multiculturalism Dead? (Polity 2016). —— Neoliberal Nationalism: Immigration and the Rise of the Populist Right (CUP 2021). Jørgensen, Carsten Willemoes and Sørensen, Karsten Engsig, ‘Internal Border Controls in the European Union’ (2012) 37 EL Rev 249. Judt, Tony, Postwar: A History of Europe since 1945 (Penguin 2005). Kabera Karanja, Stephen, The Schengen Information System and Border Control Co- operation (Martinus Nijhoff 2008). Kahanec, Martin and Zimmermann, Klaus F (eds), EU Labor Markets after Post-Enlargement Migration (Springer 2010). Kahl, Wolfgang, ‘Rechts-und Sachkontrolle in grenzüberschreitenden Sachverhalten’ in Josef Isensee and Paul Kirchhof (eds), Handbuch des Staatsrechts, Band XI (3rd edn, CF Müller 2013) § 253. Kalpouzos, Ioannis, ‘International Criminal Law and the Violence against Migrants’ (2020) 21 German LJ 571. Kanetake, Machiko, ‘UN Human Rights Treaty Bodies Monitoring before Domestic Courts’ (2018) 67 ICLQ 201. Kassim, Hussein and others, The European Commission of the Twenty-First Century (OUP 2013). Kassoti, Eva and Idriz, Narin (eds), The Informalisation of the EU’s External Action in the Field of Migration and Asylum (Springer 2022). Kau, Marcel, ‘Human Trafficking Directive 2004/81/EC’ in Daniel Thym and Kay Hailbronner (eds), EU Immigration and Asylum Law: Article-by-Article Commentary (3rd edn, CH Beck/Hart Publishing/Nomos 2022) 642–90. Kaunert, Christian, Léonard, Sarah, and Hoffmann, Ulrike, ‘Venue-Shopping and the Role of Non- Governmental Organisations in the Development of the European Union Asylum Policy’ (2013) 1 Comp Migration Stud 179. ——and Yakubov, Ikrom, ‘Securitization’ in Ariadna Ripoll Servent and Florian Trauner (eds), The Routledge Handbook of Justice and Home Affairs Research (Routledge 2018) 30–40. Kawar, Leila, Contesting Immigration Policy in Court (CUP 2015). Kelemen, R Daniel, ‘The Political Foundations of Judicial Independence in the European Union’ (2012) 19 J Eur Pub Pol 43. Kennedy, David, The Dark Sides of Virtue (Princeton UP 2004). Kerber, Karoline, ‘The Temporary Protection Directive’ (2002) 4 EJML 193. Kessler, Alan E and Freeman, Gary P, ‘Public Opinion in the EU on Immigration from Outside the Community’ (2005) 43 JCMS 825. Khoo, Siew-Ean, Hugo, Graeme, and McDonald, Peter, ‘Which Skilled Temporary Migrants Become Permanent Residents and Why?’ (2008) 42 IM Rev 193. Kienast, Julia, Feith Tan, Nikolas and Vedsted- Hansen, Jens, ‘Preferential, Differential or Discriminatory?’ in Sergio Carrera and Meltem İneli Ciğer (eds), EU Responses to the Large-Scale Refugee Displacement from Ukraine (Migration Policy Centre 2023) 383–99.

Bibliography 615 King, Russel, Lazaridis, Gabriella, and Tsardanidis, Charalambos (eds), Eldorado or Fortress? Migration in Southern Europe (Palgrave Macmillan 2010). Klaassen, Mark and Rodrigues, Peter, ‘The Best Interests of the Child in EU Family Reunification Law’ (2017) 19 EJML 191. Klarmann, Tobias, Illegalisierte Migration (Nomos 2021). Klatt, Matthias and Meister, Moritz, The Constitutional Structure of Proportionality (OUP 2012). Klein Solomon, Michele, ‘GATS Mode 4 and the Mobility of Labour’ in Ryszard Cholewinski, Richard Perruchoud, and Euan MacDonald (eds), International Migration Law (TMC Asser Press 2007) 107–28. Kleizen, Bjorn and Verhoest, Koen, ‘Opportunities and Threats of Agency Autonomy in EU Governance’ in Miroslava Scholten and Alex Brenninkmeijer (eds), Controlling EU Agencies (Edward Elgar Publishing 2020) 39–59. Kneebone, Susan and Macklin, Audrey, ‘Resettlement’ in Cathryn Costello, Michelle Foster, and Jane McAdam (eds), The Oxford Handbook of International Refugee Law (OUP 2021) 1080–98. Kocharov, Anna, Republican Europe (Hart Publishing 2017). Kochenov, Dimitry, ‘The Essence of EU Citizenship Emerging from the Last Ten Years of Academic Debate’ (2013) 62 ICLQ 97. Kogovšek Šalamon, Neža (ed), Causes and Consequences of Migrant Criminalization (Springer 2020). Koka, Enkelejda and Veshi, Denard, ‘Irregular Migration by Sea’ (2019) 21 EJML 26. Konsta, Anna-Maria, ‘Is There a Right to Human Dignity? The Example of the Right to Education of Refugees’ (2019) 21 EJML 261. Koo, John, ‘Mass Influxes and Protection in Europe’ (2018) 20 EJML 157. Koopmans, Ruud, Michalowski, Ines, and Waibel, S, ‘Citizenship Rights for Immigrants’ (2012) 117 Am J Soc 1202. Korneev, Oleg, ‘Deeper and Wider than a Common Space: European Union-Russia Cooperation on Migration Management’ (2012) 17 EFA Rev 605. ——and Kluczewska, Olga, ‘The Globalised Third Sector in the Migration Policy Field’ in Anna Triandafyllidou (ed), Handbook of Migration and Globalisation (Edward Elgar Publishing 2018) 54–68. Kornezov, Alexander, ‘The New Format of the Acte Clair Doctrine and Its Consequences’ (2016) 53 CML Rev 1317. Kosińska, Anna Magdalena and Mikołajczyk, Barbara, ‘Does the Right to Migration Security Already Exist?’ (2019) 21 EJML 83. Koskenniemi, Martti, From Apology to Utopia: The Structure of International Legal Argument (2nd edn, CUP 2005). ——‘Constitutionalism as Mindset’ (2007) 8 Theo Inq L 9. Koslowski, Rey, ‘International Travel’ in Alexander Betts (ed), Global Migration Governance (OUP 2011) 109–32. Kostakopoulou, Dora, ‘European Union Citizenship: Writing the Future’ (2007) 13 ELJ 623. ——‘The Area of Freedom, Security and Justice and the European Union’s Constitutional Dialogue’ in Catherine Barnard (ed), The Fundamentals of EU Law Revisited (OUP 2007) 153–92. ——Acosta Arcarazo, Diego, and Munk, Tine, ‘EU Migration Law: The Opportunities and Challenges Ahead’ in Diego Acosta Arcarazo and Cian C Murphy (eds), EU Security and Justice Law (Hart Publishing 2014) 129–45. Koutrakos, Panos, The EU Common Security and Defence Policy (OUP 2013). ——‘The Nexus between CFSP/CSDP and the Area of Freedom, Security and Justice’ in Steven Blockmans and Panos Koutrakos (eds), Research Handbook on the EU’s Common Foreign and Security Policy (Edward Elgar Publishing 2018) 296–311. Kraft, Ingo, ‘Asylum Qualification Directive 2011/95/EU’ in Daniel Thym and Kay Hailbronner (eds), EU Immigration and Asylum Law: Article-by-Article Commentary (3rd edn, CH Beck/Hart Publishing/Nomos 2022) 1314–61. Kraler, Albert, ‘Fixing, Adjusting, Regulating, Protecting Human Rights: Shifting Uses of Regularisations in the European Union’ (2011) 13 EJML 297. Krieg, Sarah, Multilevel Regulation against Trafficking in Human Beings (Nomos 2014).

616 Bibliography Krisch, Nico, Beyond Constitutionalism (OUP 2010). Krommendijk, Jasper, ‘The Preliminary Reference Dance between the CJEU and Dutch Courts in the Field of Migration’ (2018) 10 Eur J Leg Stud 101. —— National Courts and Preliminary References to the Court of Justice (Edward Elgar Publishing 2021). Kubal, Agnieszka, ‘Conceptualizing Semi-Legality in Migration Research’ (2013) 47 L & Soc Rev 555. Küçük, Esin, ‘Solidarity in EU Law: An Elusive Political Statement or a Legal Principle with Substance?’ (2016) 23 MJECL 965. ——‘The Principle of Solidarity and Fairness’ (2016) 22 ELJ 448. Kuijper, Pieter Jan, ‘Some Legal Problems Associated with the Communitarization of Policy on Visas, Asylum and Immigration’ (2000) 37 CML Rev 345. ——‘The Evolution of the Third Pillar from Maastricht to the European Constitution’ (2004) 41 CML Rev 609. Kuptsch, Christiane and Martin, Philip, ‘Low-Skilled Labour Migration’ in Alexander Betts (ed), Global Migration Governance (OUP 2011). Kymlicka, Will, Multicultural Odysseys (OUP 2007). Laas, Matthias, Die Entstehung eines europäischen Migrationsverwaltungsraumes (Nomos 2008). Labayle, Henri, ‘L’espace de liberté, sécurité et justice dans la Constitution pour l’Europe’ [2005] Revue trimestrielle de droit européen 437. ——‘La suppression des contrôles aux frontières intérieures de l’Union’ in Claude Blumann (ed), Les frontières de l’Union européenne (Bruylant 2013) 19–54. ——‘Droit d’asile et confiance mutuelle’ (2014) 49 Cahiers de droit européen 501. ——and Bergé, Jean-Sylvestre, ‘Les principes de l’Espace de liberté, de sécurité et de justice’ [2016] Revue trimestrielle de droit européen 589. Ladenburger, Clemens, ‘Police and Criminal Law in the Treaty of Lisbon’ (2008) 4 EuConst 20. ——‘Session on Protection of Fundamental Rights Post-Lisbon’ in Julia Laffranque (ed), Reports of the XXV FIDE Congress, vol 1 (Tartu UP 2012) 165–67. ——and Verwilghen, Stephane, ‘Policies Relating to the Area of Freedom, Security and Justice’ in Giuliano Amato, Hervé Bribosia, and Bruno de Witte (eds), Genèse et destinée de la Constitution européenne (Bruylant 2007) 743–72. Laffan, Brigid, ‘From Policy Entrepreneur to Policy Manager: The Challenge Facing the European Commission’ (1997) 4 JEPP 422. Lahav, Gallya, Immigration and Politics in the New Europe (CUP 2004). Lambert, Hélène, Seeking Asylum: Comparative Law and Practice in Selected European Countries (Martinus Nijhoff 1995). ——‘Transnational Law and Refugee Identity’ in Susan Kneebone, Dallal Stevens, and Loretta Baldassar (eds), Refugee Protection and the Role of Law (Routledge 2014) 203–14. Lang, Alessandra, ‘Social Integration’ (2018) 3 European Papers 663. Lansbergen, Anja and Shaw, Jo, ‘National Membership Models in a Multilevel Europe’ (2010) 8 ICON 50. Larin, Stephen J, ‘Is It Really about Values? Civic Nationalism and Migrant Integration’ (2020) 46 JEMS 127. Larsson, Olof and Naurin, Daniel, ‘Judicial Independence and Political Uncertainty’ (2016) 70 IO 1. Lauterpacht, Hersch, Oppenheim’s International Law, vol 1 (8th edn, Longham 1955). Lavenex, Sandra, Safe Third Countries: Extending the EU Asylum and Immigration Policies to Central and Eastern Europe (CEU Press 1999) 78–89. —— The Europeanisation of Refugee Politics (Ashgate 2002). ——‘Shifting Up and Out: The Foreign Policy of European Immigration Control’ (2006) 29 W Eur Pol 329. ——‘“Failing Forward” Towards Which Europe? Organized Hypocrisy in the Common European Asylum System’ (2018) 56 JCMS 1195. ——and Nellen-Stucky, Rachel, ‘“Partnering” for Migration in EU External Relations’ in Rahel Kunz, Sandra Lavenex, and Marion Panizzon (eds), Multilayered Migration Governance (Routledge 2011) 116–42.

Bibliography 617 Law, Stephanie, ‘Humanitarian Admission and the Charter of Fundamental Rights’ in Marie-Claire Foblets and Luc Leboeuf (eds), Humanitarian Admission to Europe (Nomos/Hart Publishing 2020) 77–114. Leboeuf, Luc, ‘La Cour de justice face aux dimensions externes de la politique commune de l’asile et de l’immigration’ [2019] Revue trimestrielle de droit européen 55. Lebon, Lydia, La territorialité et l’Union européenne (Bruylant 2015). Leerkes, Arjen and Kulu-Glasgow, Isik, ‘Playing Hard(er) to Get: The State, International Couples, and the Income Requirement’ (2011) 13 EJML 95. Legomsky, Stephen H, ‘Secondary Refugee Movements and the Return of Asylum Seekers to Third Countries’ (2003) 15 IJRL 567. Lehmann, Julian, ‘Protection’ in European Union Asylum Law (Brill Nijhoff 2020). Lehner, Roman, ‘Rechtliche Möglichkeiten zur Schaffung einer EU-Asylbehörde’ in Roman Lehner and Friederike Wapler (eds), Die herausgeforderte Rechtsordnung (BWV 2018) 183–221. ——‘The EU-Turkey-“Deal”. Legal Challenges and Pitfalls’ (2019) 57 Intl Migration 176. ——‘Gemeinsame Visapolitik’ in Ferdinand Wollenschläger (ed), Enzyklopädie Europarecht, Band X (2nd edn Nomos 2021) § 28. ——‘Gesetzgebung: Vor-und Nachteile der Reformschritte in der letzten Dekade’ [2022] Zeitschrift für Ausländerrecht 144. Lehnert, Matthias, Frontex und operative Maßnahmen an den europäischen Außengrenzen (Nomos 2014). Lenaerts, Koen, ‘The Court’s Outer and Inner Selves’ in Maurice Adams and others (eds), Judging Europe’s Judges (Hart Publishing 2013) 13–60. ——‘La vie après l’avis: Exploring the Principle of Mutual (yet not Blind) Trust’ (2017) 54 CML Rev 805. ——‘Europarecht und Zuwanderung’ in Julia Iliopoulos-Strangas and others (eds), Migration – Migration –Migrations (Nomos/Stämpfli 2017) 233–42. ——and Gutiérrez-Fons, José A, ‘The Place of the Charter in the EU Constitutional Ediﬁce’ in Steve Peers and others (eds), The EU Charter of Fundamental Rights: A Commentary (2nd edn, CH Beck/ Hart Publishing/Nomos 2021) 1711–34. ——Maselis, Ignace, and Gutman, Kathleen, EU Procedural Law (OUP 2014). Levi-Faur, David, The Oxford Handbook of Governance (OUP 2012). Levy, Carl, ‘Refugees, Europe, Camps/State of Exception: “Into the Zone”’ (2010) 29 Refugee Surv Q 92. Lewis, Corinne, UNHCR and International Refugee Law (Routledge 2012). Lewis, Jeffrey, ‘The Janus Face of Brussels’ (2005) 59 IO 937. Liefländer, Thomas, Kellerbauer, Manuel, and Dumitriu- Segnana, Eugenia (eds), The UK-EU Withdrawal Agreement: A Commentary (OUP 2021). Lindahl, Hans, ‘Finding a Place for Freedom, Security and Justice’ (2004) 29 EL Rev 461. Lippert, Randy and Rehaag, Sean (eds), Sanctuary Practices in International Perspectives (Routledge 2012). Lipsky, Michael, Street-Level Bureaucrats (Russell Sage 2010). Lisi, Gaia and Eliantonio, Mariolina, ‘The Gaps in Judicial Accountability of EASO in the Processing of Asylum Requests in Hotspots’ (2019) 4 European Papers 589. Long, Katy, ‘When Refugees Stopped Being Migrants’ (2013) 1 Migration Studies 4. Lopatin, Esther, ‘The Changing Position of the European Parliament on Irregular Migration and Asylum under Co-decision’ (2013) 51 JCMS 740. Lorenzmeier, Stefan, ‘Assoziierungsfreizügigkeit (Türkei, Balkan, Ukraine, Mittelmeer) und EFTA/ EWR’ in Ferdinand Wollenschläger (ed), Enzyklopädie Europarecht Band X (2nd edn, Nomos 2021) § 12. Lott, Gaia, ‘The Dublin Convention and the Introduction of the “First Entry Rule”’ (2022/23) 31/32 Contemporary European History, forthcoming. Lübbe, Anna, ‘“Systemic Flaws” and Dublin Transfers: Incompatible Tests before the CJEU and the ECtHR?’ (2015) 27 IRJL 135. Lucassen, Jan, Migrant Labour in Europe, 1600–1900 (Croom Helm 1987).

618 Bibliography ——and Lucassen, Leo, ‘The Mobility Transition Revisited, 1500–1900’ (2009) 4 J Glob Hist 347. Luedtke, Adam, ‘Migration Governance in Europe: A Historical Perspective’ in Agnieszka Weinar, Saskia Bonjour, and Lyubov Zhyznomirska (eds), The Routledge Handbook of the Politics of Migration in Europe (Routledge 2019) 15–25. Lukas, Karin, The Revised European Social Charter: An Article by Article Commentary (Edward Elgar Publishing 2021). Lülf, Charlotte, Conflict Displacement and Legal Protection (Routledge 2019). Luthra, Renee, Platt, Lucinda, and Salamońska, Justyna, ‘Types of Migration’ (2018) 52 IM Rev 368. Lutz, Fabian, The Negotiations on the Return Directive (Wolf 2011). ——‘Transposition of the ICT Directive 2014/66/EU’ in Paul Minderhoud and Tesseltje de Lange (eds), The Intra Corporate Transferee Directive (Wolf 2018) 19–34. ——‘Non-Removable Returnees under Union Law’ (2018) 20 EJML 28. ——‘Return Directive 2008/115/EC’ in Daniel Thym and Kay Hailbronner (eds), EU Immigration and Asylum Law: Article-by-Article Commentary (3rd edn, CH Beck/Hart Publishing/Nomos 2022) 692–750. Lynskey, Orla, ‘A Legal Analysis of the Emerging Extraterritorial Elements of EU Refugee Protection Policy’ (2006) 31 EL Rev 230. Maas, Willem, Creating European Citizens (Rowman & Littlefield 2007). MacDonald, John S and MacDonald, Leatrice D, ‘Chain Migration, Ethnic Neighborhood Formation and Social Networks’ (1964) 42 Milbank Memorial Fund Quarterly 82. Mahoney, James, ‘Path Dependence in Historical Sociology’ (2000) 29 Theory and Society 507. Maiani, Francesco, ‘The Reform of the Dublin System and the Dystopia of “Sharing People”’ (2017) 24 MJECL 622. ——‘European Union and Andorra, Monaco and San Marino’ in Sieglinde Gstohl and David Phinnemore (eds), The Proliferation of Privileged Partnerships between the European Union and its Neighbours (Routledge 2019) 84–101. ——‘Into the Loop: The Doomed Reform of Dublin and Solidarity in the New Pact’ in Daniel Thym and Odysseus Academic Network (eds), Reforming the Common European Asylum System (Nomos 2022) 43–60. ——and Mogliorini, Sara, ‘One Principle to Rule Them All! Anatomy of Mutual Trust in the Law of the Area of Freedom, Security and Justice’ (2020) 57 CML Rev 7. Majcher, Izabella, The European Union Returns Directive and its Compatibility with International Human Rights Law (Brill Nijhoff 2020). Majone, Giandomenico, Dilemmas of European Integration (OUP 2005). —— Europe as the Would-be World Power (CUP 2009). —— Rethinking the Union of Europe Post-Crisis (CUP 2014). Mananashvili, Sergo and Moraru, Madalina, ‘Return Directive 2008/115/EC’ in Daniel Thym and Kay Hailbronner (eds), EU Immigration and Asylum Law: Article-by-Article Commentary (3rd edn, CH Beck/Hart Publishing/Nomos 2022) 750–804. Mancini, Frederico and Keeling, David, ‘Language, Culture and Politics in the Life of the European Court of Justice’ (1995) 1 Colum J Eur L 397. Mann, Itamar, Humanity at Sea: Maritime Migration and the Foundations of International Law (CUP 2016) 6–10. ——‘Maritime Legal Black Holes’ (2018) 29 EJIL 347. Manners, Ian, ‘Normative Power Europe: A Contradiction in Terms?’ (2002) 40 JCMS 235. Marin, Luisa, ‘Waiting (and Paying) for Godot: Analyzing the Systemic Consequences of the Solidarity Crisis in EU Asylum Law’ (2020) 22 EJML 60. ——and Krajčíková, Kamila, ‘Deploying Drones in Policing Southern European Borders’ in Aleš Završnik (ed), Drones and Unmanned Aerial Systems (Springer 2015) 101–27. Markard, Nora, Kriegsflüchtlinge (Mohr Siebeck 2012). ——‘The Right to Leave by Sea’ (2016) 27 EJIL 591. Martin, Phil and Taylor, Edward J, ‘The Anatomy of the Migration Hump’ in Edward J Taylor (ed), Development Strategy, Employment and Migration (OECD 1996) 43–62.

Bibliography 619 ——Abella, Manolo, and Kuptsch, Christiane, Managing Labour Migration in the Twenty-First Century (Yale UP 2006). Martin, Susan and Callaway, Amber, ‘Human Trafficking und Smuggling’ in Alexander Betts (ed), Global Migration Governance (OUP 2011) 224–41. Marx, Reinhard, ‘The Criteria of Applying the “Internal Flight Alternative” Test’ (2002) 14 IJRL 179. ——‘Sprachnachweis und Ehegattennachzug’ [2011] Zeitschrift für Ausländerrecht 15. Massey, Douglas S and others, ‘Theories of International Migration: A Review and Appraisal’ (1993) 19 Popul Dev Rev 431. Masterman, Roger, ‘The United Kingdom’ in Patricia Popelier, Sarah Lambrecht, and Koen Lemmens (eds), Criticism of the European Court of Human Rights (Intersentia 2016) 447–78. Mathew, Penelope and Harley, Tristan, Refugees, Regionalism and Responsibility (Edward Elgar Publishing 2016). Mau, Steffen, Brabandt, Heike, Laube, Lena, and Roos, Christof, Liberal States and the Freedom of Movement (Palgrave Macmillan 2012). Mayblin, Lucy and Turner, Joe, Migration Studies and Colonialism (Polity 2021). Mayer, Matthias M, Governmental Preferences on Liberalising Economic Migration Policies at the EU Level (PhD thesis, LSE 2011). Mazzotti, Paolo and Eliantonio, Mariolina, ‘Transnational Judicial Review in Horizontal Composite Procedures’ (2020) 5 European Papers 41. McAdam, Jane, ‘The European Union Qualification Directive: The Creation of a Subsidiary Protection Regime’ (2005) 17 IJRL 461. —— Complementary Protection in International Refugee Law (OUP 2007). ——‘Swimming Against the Tide: Why a Climate Change Displacement Treaty is Not the Answer’ (2011) 23 IJRL 2. —— Climate Change, Forced Migration, and International Law (OUP 2012). McColgan, Aileen, Discrimination, Equality and the Law (Hart Publishing 2016). McCormack- George, Dáire, ‘Equal Treatment of Third- Country Nationals in the European Union: Why Not?’ (2019) 21 EJML 53. McMahon, Richard, ‘Maastricht’s Third Pillar’ (1995) 22 Leg Issues Eur Integration 51. McNamara, Frank, ‘Member State Responsibility for Migration Control within Third States: Externalisation Revisited’ (2013) 15 EJML 319. Medland, Lydia, ‘Misconceiving “Seasons” in the Global Food Systems: The Case of the EU Seasonal Workers Directive’ (2017) 23 ELJ 157. Meloni, Annalisa, Visa Policy within the European Union Structure (Springer 2005). ——‘The Development of a Common Visa Policy under the Treaty of Amsterdam’ (2005) 42 CML Rev 1357. ——‘The Community Code on Visas’ (2009) 34 EL Rev 671. ——‘EU Visa Policy: What Kind of Solidarity?’ (2017) 24 MJECL 646. ——‘Visa Code Regulation (EC) No 810/2009’ in Daniel Thym and Kay Hailbronner (eds), EU Immigration and Asylum Law: Article-by-Article Commentary (3rd edn, CH Beck/Hart Publishing/ Nomos 2022) 84–178. Mendos Kuşkonmaz, Elif, Privacy and Border Controls in the Fight against Terrorism (Brill Nijhoff 2021). Menéndez, Agustín, ‘The Refugee Crisis’ (2016) 22 ELJ 388. ——and Olsen, Espen DH, Challenging European Citizenship (Palgrave 2020). Menz, Georg, The Political Economy of Managed Migration (OUP 2008). ——‘Stopping, Shaping and Moulding Europe’ (2011) 49 JCMS 437. Meyer, Frank, ‘Der Grundsatz des gegenseitigen Vertrauens’ [2017] Europarecht 163. Mezzadra, Sandro and Neilson, Brett, Border as Method, or, the Multiplication of Labor (Duke UP 2013). Michalowski, Ines, ‘Required to Assimilate? The Content of Citizenship Tests in Five Countries’ (2011) 15 Citizenship Stud 749. Milios, Georgios, ‘Family Reunification for Third-Country Nationals’ (2015) 17 EJML 127. Miller, David, Strangers in Our Midst (Harvard UP 2016).

620 Bibliography Millet-Devalle, Anne and Hammami, Nadia, ‘Le renforcement des capacités des Etats tiers par l’Union européenne’ in Josiane Auvret-Finck and Anne-Sophie Millet-Devalle (eds), Crise des réfugiés, crise de l’Union européenne? (Pedone 2017) 307–58. Milward, Alan S, The European Rescue of the Nation State (2nd edn, Routledge 2000). Minderhoud, Paul, ‘Social Security Rights of Third Country Nationals’ (2010) 17 J Soc Sec L 227. ——‘Decision No 3/80 of the EEC-Turkey Association Council’ in Daniel Thym and Margarite Zoeteweij-Turhan (eds), Rights of Third-Country Nationals under EU Association Agreements (Brill Nijhoff 2015) 169–85. Mitsilegas, Valsamis, The Criminalisation of Migration in Europe (Springer 2015). Möllers, Christoph, ‘European Governance: Meaning and Value of a Concept’ (2006) 43 CML Rev 313. Molnár, Tamás, The Interplay between the EU’s Return Acquis and International Law (Edward Elgar Publishing 2021). Monar, Jörg, ‘Die Vertragsreformen von Lissabon in den Bereichen Inneres und Justiz’ [2008] Integration 379, 385. ——‘The Area of Freedom, Security and Justice’ in Armin von Bogdandy and Jürgen Bast (eds), Principles of European Constitutional Law (2nd edn, Hart Publishing 2009) 543–76. ——and Morgan, Roger (eds), The Third Pillar of the European Union (Peter Lang 1995). Money, Jeannette, Fences and Neighbours (Cornell UP 1999). Monina, Giuliana, ‘“Judging” the Grounds for Detention of Asylum Seekers’ in Andrea Crescenzi, Rosita Forastiero, and Giuseppe Palmisano (eds), Asylum and the EU Charter of Fundamental Rights (Editoriale Scientifica 2018) 151–74. Moore, Margaret, ‘Cosmopolitanism and Political Communities’ (2006) 32 Soc Theory & Prac 627. Moraru, Madalina, ‘Judicial Dialogue in Action’ in Madalina Moraru, Galina Cornelisse, and Philippe De Bruycker (eds), Law and Judicial Dialogue on the Return of Irregular Migrants from the European Union (Hart Publishing 2020) 125–48. Moravcsik, Andrew, The Choice for Europe (Ithaka Press 1998). ——‘In Defence of the “Democratic Deficit”’ (2002) 40 JCMS 603. Moreno-Lax, Violeta, ‘Seeking Asylum in the Mediterranean’ (2011) 23 IJRL 174–220. ——‘Life after Lisbon: EU Asylum Policy as a Factor of Migration Control’ in Diego Acosta Arcarazo and Cian C Murphy (eds), EU Security and Justice Law (Hart Publishing 2014) 146–67. —— Accessing Asylum in Europe (OUP 2017). ——‘External Dimension’ in Steve Peers and others (eds), EU Immigration and Asylum Law (Text and Commentary), vol 3 (2nd edn, Brill Nijhoff 2015) 617–73. ——‘The EU Humanitarian Border and the Securitization of Human Rights’ (2018) 56 JCMS 119. ——‘The Architecture of Functional Jurisdiction’ (2020) 21 German LJ 385. ——‘Towards a Thousand Little Morias’ in Daniel Thym and Odysseus Academic Network (eds), Reforming the Common European Asylum System (Nomos 2022) 161–68. ——and Garlick, Madeline, ‘Qualification: Refugee Status and Subsidiary Protection’ in Steve Peers and others (eds), EU Immigration and Asylum Law (Text and Commentary), vol 3 (2nd edn, Brill Nijhoff 2015) 65–210. ——and Guild, Elspeth, ‘Temporary Protection’ in Steve Peers and others (eds), EU Immigration and Asylum Law (Text and Commentary), vol 3 (2nd edn, Brill Nijhoff 2015) 571–615. ——and Lemberg-Pedersen, Martin, ‘Border-Induced Displacement’ (2019) 56 Q Intl L 5. Morgades-Gil, Siliva, ‘The Right to Benefit from an Effective Remedy against Decisions Implying the Return of Asylum Seekers to European Safe Countries’ (2017) 19 EJML 255. Morgenstern, Felice, ‘The Right of Asylum’ (1949) 26 Br Yearb Intl L 327. Morijn, John, ‘Personal Conviction and Strategic Litigation in Wijsenbeek’ in Fernanda Nicola and Bill Davies (eds), EU Law Stories (CUP 2017) 178–200. Möschel, Mathias, Law, Lawyers and Race: Critical Race Theory from the United States to Europe (Routledge 2014). Motomura, Hiroshi, Americans in Waiting (OUP 2006). Mourão Permoser, Julia, ‘Civic Integration as Symbolic Politics’ (2012) 14 EJML 173–98. ——‘Redefining Membership’ (2017) 43 JEMS 2536.

Bibliography 621 Mouthaan, Melissa, ‘Old Wine in New Bottles? The European Union’s Organizational Response to Reforming EU-African Migration Cooperation’ (2021) 59 JCMS 1177. Mouzourakis, Minos, ‘More Laws, Less Law’ (2020) 26 ELJ 171. Moyn, Samuel, The Last Utopia: Human Rights in History (Harvard UP 2010). Muir, Elise, EU Equality Law (OUP 2018). Mungianu, Roberta, Frontex and Non-Refoulement (CUP 2016). Murphy, Cian C and Acosta Arcarazo, Diego, ‘Rethinking Europe’s Freedom, Security, and Justice’ in Diego Acosta Arcarazo and Cian C Murphy (eds), EU Security and Justice Law (Hart Publishing 2014) 1–16. Murphy, Clíodhna, Immigration, Integration and the Law (Ashgate 2013). Muzak, Gerhard, ‘Artikel 78 AEUV’ in Thomas Jaeger and Karl Stöger (eds), Kommentar zu EUV und AEUV unter Berücksichtigung der österreichischen Judikatur und Literatur (141th update, Manz 2012). Nail, Thomas, Theory of the Border (OUP 2016). Nascimbene, Bruno and Di Pascale, Alessia, ‘The “Arab Spring” and the Extraordinary Influx of People who Arrived in Italy from North Africa’ (2011) 13 EJML 341. Neumayer, Eric, ‘Asylum Destination Choice’ (2004) 5 EU Pol 155. Nic Shuibhne, Niamh, ‘Free Movement of Persons and the Wholly Internal Rule’ (2002) 39 CML Rev 731. ——‘The “Territory of the Union” in EU Citizenship Law’ (2019) 38 Yearb Eur L 267. Nicol, Andrew, ‘Between Immigration and Policing’ (2004) 10 ELJ 171. Nicolosi, Salvatore Fabio, ‘Going Unnoticed? Diagnosing the Right to Asylum in the Charter of Fundamental Rights’ (2017) 23 ELJ 94. ——‘Refashioning the EU Visa Policy’ (2020) 22 EJML 467. Niemann, Arne, Explaining Decisions in the European Union (CUP 2006). Niewöhner, Jörg, ‘Infrastructures of Society, Anthropology of ’ in James D Wright (ed), International Encyclopedia of the Social & Behavioral Sciences, vol 12 (2nd edn, Elsevier 2015) 119–25. Noll, Gregor, ‘Article 31’ in Andreas Zimmermann (ed), The 1951 Convention Relating to the Status of Refugees and its 1967 Protocol: A Commentary (OUP 2011) 1244–76. —— Negotiating Asylum (Martinus Nijhoff 2000). ——‘Empiricism vs Formalism: Some Reflections on the Dublin Convention on the Occasion of Recent European Case Law’ (2001) 70 Nordic J Intl L 161. ——‘Visions of the Exceptional: Legal and Theoretical Issues Raised by Transit Processing Centres and Protection Zones’ (2003) 5 EJML 303. Norman, Kelsey P, ‘Inclusion, Exclusion or Indifference?’ (2019) 45 JEMS 42. Nussberger, Angelika, The European Court of Human Rights (OUP 2020). O’Keeffe, David, ‘The Schengen Convention: A Suitable Model for European Integration’ (1991) 11 YEL 185. ——‘The New Draft External Frontiers Convention and the Draft Visa Regulation’ in Jörg Monar and Roger Morgan (eds), The Third Pillar of the European Union (Peter Lang 1995) 135–49. O’Leary, Síofra and Iglesias Sánchez, Sara, ‘Free Movement of Persons and Services’ in Paul Craig and Gráinne de Búrca (eds), The Evolution of EU Law (3rd edn, OUP 2021) 506–45. ——‘L’ELSJ à la Cour européenne des droits de l’homme’ in Véronique Beaugrand, David Mas, and Maud Vieux (eds), Sa justice: Liber Amicorum en hommage de Yves Bot (Bruylant 2022) 575–604. Oberloskamp, Eva, Codename TREVI (De Gruyter 2016). Oellers-Frahm, Karin, ‘The Contribution of the Council of Europe to the Legal Position of Aliens’ in Jochen Abr Frowein and Torsten Stein (eds), Die Rechtsstellung von Ausländern nach staatlichem Recht und Völkerrecht, vol 2 (Springer 1987) 1725–73. Oelgemöller, Christina, The Evolution of Migration Management in the Global North (Routledge 2017). Oepen-Mathey, Meta, Internationaler Schutz im Falle extremer Armut (Nomos 2021). Oosterom-Staples, Helen, ‘Effective Rights for Third-Country Nationals?’ in Hans Lindahl (ed), A Right to Inclusion and Exclusion? (Hart Publishing 2009) 65–93. Orgad, Liav, The Cultural Defense of Nations (OUP 2015).

622 Bibliography Ostrand, Nicole and Statham, Paul, ‘“Street-level” Agents Operating Beyond “Remote Control”’ (2021) 47 JEMS 25. Ott, Andrea, ‘EU-Turkey Cooperation in Migration Matters’ in Francesca Ippolito, Gianluca Borzoni, and Federico Casolari (eds), Bilateral Relations in the Mediterranean (Edward Elgar Publishing 2020) 184–214. Özden, Çağlar, ‘Educated Migrants: Is there Brain Waste?’ in Çağlar Özden and Maurice Schiff (eds), International Migration, Remittances, and the Brain Drain (World Bank 2006) 227–44. Pagenstecher, Cord, ‘Die „Illusion“ der Rückkehr’ (1996) 47 Soziale Welt 149. Palumbo, Letizia, ‘Exploitation in the Agri-Food Sector in Europe’ (2022) 24 EJML 287. Paoli, Simone, ‘The Schengen Agreements and their Impact on Euro-Mediterranean Relations’ (2015) 21 J Eur Integr Hist 125–45. ——‘France and the Origins of Schengen’ in Elena Calandri, Simone Paoli, and Antonio Varsori (eds), Peoples and Borders (Nomos 2017) 255–80. Papadopoulos, Dimitris, Stephenson, Niamh, and Tsianos, Vassilis, Escape Routes: Control and Subversion in the Twenty-First Century (Pluto Press 2008). Papagianni, Georgia, Institutional and Policy Dynamics of EU Migration Law (Martinus Nijhoff 2006). ——‘Forging an External EU Migration Policy’ (2013) 15 EJML 283. Papapanagiotou-Leza, Angeliki and Kofinis, Stergios, ‘The Case of Greece’ in Madalina Moraru, Galina Cornelisse, and Philippe De Bruycker (eds), Law and Judicial Dialogue on the Return of Irregular Migrants from the European Union (Hart Publishing 2020) 281–300. Pascouau, Yves, ‘Quel avenir pour Schengen?’ in Josiane Auvret-Finck and Anne-Sophie Millet- Devalle (eds), Crise des Réfugiés, Crise de l’Union Européenne? (Pedone 2017) 261–80. Passalacqua, Virginia, ‘Legal Mobilization via Preliminary Reference: Insights from the Case of Migrant Rights’ (2021) 58 CML Rev 751. Pastore, Ferruccio, ‘Visas, Borders, Immigration’ in Neil Walker (ed), Europe’s Area of Freedom, Security and Justice (OUP 2004) 89–142. ——‘The Governance of Migrant Labour Supply in Europe’ (2014) 2 Comp Migr Stud 385. Peers, Steve, ‘Undercutting Integration’ (1997) 22 EL Rev 76. ——‘Building Fortress Europe’ (1998) 35 CML Rev 1235. —— EU Justice and Home Affairs Law (1st edn, Longman 2000). ——‘Key Legislative Developments on Migration in the European Union: SIS II’ (2008) 10 EJML 77. ——‘EC Immigration Law and EC Association Agreements’ (2009) 34 EL Rev 628. ——‘Legislative Update: EC Immigration and Asylum Law Attracting and Deterring Labour Migration’ (2009) 11 EJML 387. ——‘Mission Accomplished? EU Justice and Home Affairs Law after the Treaty of Lisbon’ (2011) 48 CML Rev 661. ——‘Admission of Students and Others’ in Steve Peers and others (eds), EU Immigration and Asylum Law (Text and Commentary), vol 2 (2nd edn, Martinus Nijhoff 2012) 195–208. ——‘An EU Immigration Code’ (2012) 14 EJML 33. ——‘Family Reunion’ in Steve Peers and others (eds), EU Immigration and Asylum Law (Text and Commentary), vol 2 (2nd edn, Martinus Nijhoff 2012) 250–53. ——‘Transfer of International Protection and European Union Law’ (2012) 24 IJRL 527. ——‘The Court of Justice Lays the Foundations for the Long-Term Residents Directive’ (2013) 50 CML Rev 529. ——‘Institutional Framework’ in Steve Peers and others (eds), EU Immigration and Asylum Law (Text and Commentary), vol 3 (2nd edn, Brill Nijhoff 2015) 7–31. —— EU Justice and Home Affairs Law, vol I (4th edn, OUP 2016). ——Moreno-Lax, Violeta, Garlick, Madeline, and Guild, Elspeth, EU Immigration and Asylum Law (Text and Commentary), vol 3 (2nd edn, Brill Nijhoff 2015). Peirce, Paul S, ‘The Control of Immigration as an Administrative Problem’ (1910) 4 Am Pol Sci Rev 374–89. Penasa, Simone and Romeo, Graziella, ‘Sovereignty-based Arguments and the European Asylum System’ (2020) 22 EJML 11.

Bibliography 623 Perkowski, Nina and Squire, Vicki, ‘The Anti-Policy of European Anti-Smuggling as a Site of Contestation in the Mediterranean Migration “Crisis”’ (2019) 45 JEMS 2167. Pernice, Ingolf, ‘Multilevel Constitutionalism and the Treaty of Amsterdam’ (1999) 36 CML Rev 703. Peukert, Alexander and others, ‘Einreisen lassen oder zurückweisen?’ [2016] Zeitschrift für Ausländerrecht 131. Pijnenburg, Annick, At the Frontiers of State Responsibility: Socio-Economic Rights and Cooperation on Migration (Intersentia 2021). Pina- Delgado, José, ‘The Current International Legal Framework of Economic Migration Management’ in Guy S Goodwin-Gill and Philippe Weckel (eds), Migration & Refugee Protection in the 21st Century (Martinus Nijhoff 2015) 127–65. Pirjatanniemi, Elina, ‘Victims of Trafficking in the Migration Discourse’ in Rita Haverkamp, Ester Herlin-Karnell, and Claes Lernestedt (eds), What is Wrong with Human Trafficking? (Hart Publishing 2019) 77–98. Pirker, Benedikt, ‘Mapping the Scope of Application of EU Fundamental Rights’ (2018) 3 European Papers 133. Pistoia, Emanuela, ‘Social Integration of Refugees and Asylum Seekers through the Exercise of Socio- economic Rights’ (2018) 3 European Papers 781. ——‘The Right of Refugees and Beneficiaries of Subsidiary Protection to National Treatment in Social Assistance’ (2020) 45 EL Rev 396. Plappert, Liesa, ‘Der unionsrechtliche Verhältnismäßigkeitsvorbehalt’ [2020] Europarecht 364. Poiares Maduro, Miguel, ‘Interpreting European Law’ (2007) 2/1 Eur J Leg Stud 4. Polanyi, Karl, The Great Transformation (1944) (2nd edn, Beacon Press 2001). Poli, Simone, ‘Migration in European Integration’ (2016) 22 Journal of European Integration Policy 279. Poli, Sara, ‘The Integration of Migration Concerns into EU External Policies’ (2020) 5 European Papers 71. Pollack, Mark A, ‘Realist, Intergovernmentalist, and Institutionalist Approaches’ in Erik Jones, Anand Menon, and Stephen Weatherill (eds), The Oxford Handbook of the European Union (OUP 2012) 3–17. Portes, Alejandro and Rumbaut, Rubén G, Legacies: The Story of the Immigrant Second Generation (UC Press 2001). Priest, Dana and Arkin, William, Top Secret America (Little, Brown and Company 2011). Pusterla, Francesca, ‘Legal Perspectives on Solidarity Crime in Italy’ (2021) 59 Intl Migration 79. Raffaelli, Rosa, ‘Criminalizing Irregular Immigration and the Returns Directive’ (2011) 13 EJML 467. Ragazzi, Simona, ‘New Experiences in Investigating and Prosecuting Migrant Smuggling’ in Valsamis Mitsilegas, Violeta Moreno-Lax, and Niovi Vavoula (eds), Securitising Asylum Flows (Brill Nijhoff 2020) 9–33. Ramji-Nogales, Jaya, Schoenholtz, Andrew I, and Schrag, Philip G, Refugee Roulette: Disparities in Asylum Adjudication and Proposals for Reform (NYU Press 2009). Rass, Christopher, Institutionalisierunsprozesse auf einem internationalen Arbeitsmarkt (Schöningh 2010). Rawls, John, The Law of Peoples (Harvard UP 2001). Redpath, Jillyanne, ‘Biometrics and International Migration’ in Ryszard Cholewinski, Richard Perruchoud, and Euan MacDonald (eds), International Migration Law (TMC Asser Press 2007) 427–446. Reenen van, Pieter, ‘Impartiality in the EU Asylum Procedure’ (2018) 20 EJML 338. Reich, Norbert, ‘The Constitutional Relevance of Citizenship and Free Movement in an Enlarged Union ’ (2005) 11 ELJ 675. Relaño Pastor, Eugenia, ‘EU Initiatives on a European Humanitarian Visa’ in Marie-Claire Foblets and Luc Leboeuf (eds), Humanitarian Admission to Europe (Nomos/Hart Publishing 2020) 341–66. Reneman, Marcelle, EU Asylum Procedures and the Right to an Effective Remedy (Hart Publishing 2014). ——‘Asylum and Article 47 of the Charter’ in Andrea Crescenzi, Rosita Forastiero, and Giuseppe Palmisano (eds), Asylum and the EU Charter of Fundamental Rights (Editoriale Scientifica 2018) 59–78.

624 Bibliography ——‘Speedy Asylum Procedures in the EU’ (2013) 25 IJRL 717. Reslow, Natasja, ‘Making and Implementing Multi-Actor EU External Migration Policy’ in Sergio Carrera and others (eds), EU External Migration Policies in an Era of Global Mobilities (Brill 2019) 277–98. ——‘Transformation or Continuity?’ in Sergio Carrera, Juan Santos Vara, and Tineke Strik (eds), Constitutionalising the External Dimension of EU Migration Policies in Times of Crisis (Edward Elgar Publishing 2019) 95–115. ——and Vink, Maarten, ‘Three-Level Games in EU External Migration Policy’ (2015) 53 JCMS 857. Ricci, Carola, ‘The Necessity for Alternative Legal Pathways’ (2020) 21 German LJ 265. Richmond, Anthony H, ‘Sociological Theories of International Migration’ (1988) 36 Current Sociology 7–25. Rijpma, Jorrit J, ‘Hybrid Agencification in the Area of Freedom, Security and Justice and its Inherent Tensions: The Case of Frontex’ in Madalina Busuioc, Martijn Groenleer, and Jarle Trondal (eds), The Agency Phenomenon in the European Union (Manchester UP 2012) 84–102. ——‘Frontex and the European System of Border Guards’ in Maria Fletcher, Ester Herlin-Karnell, and Claudio Matera (eds), The European Union as an Area of Freedom, Security and Justice (Routledge 2017) 217–45. Ripoll Servent, Ariadna, ‘The European Parliament in Justice and Home Affairs’ in Ariadna Ripoll Servent and Florian Trauner (eds), The Routledge Handbook of Justice and Home Affairs Research (Routledge 2018) 385–95. Rizcallah, Cecilia, ‘Facing the Refugee Challenge in Europe’ (2019) 21 EJML 238. Rochel, Johan, ‘Working in Tandem: Proportionality and Procedural Guarantees in EU Immigration Law’ (2019) 20 German LJ 89. Rodekamp, Meike, Their Members’ Voice: Civil Society Organisations in the European Union (Springer 2014). Rodin, Siniša and Perišin, Tamara (eds), The Transformation or Reconstitution of Europe (Bloomsbury/ Hart Publishing 2018) 3–16. Rogers, Nicola, ‘Minimum Standards for Reception’ (2002) 4 EJML 215. Rogers, Rosemarie, ‘Migration Return Policies and Countries of Origin’ in Kay Hailbronner, David A Martin, and Hiroshi Motomura (eds), Immigration Admissions (Berghahn Books 1997) 147–204. Rokkan, Stein, State Formation, Nation-Building, and Mass Politics in Europe (OUP 1999). Roos, Christof, ‘The Council and European Council in EU Justice and Home Affairs Politics’ in Ariadna Ripoll Servent and Florian Trauner (eds), The Routledge Handbook of Justice and Home Affairs Research (Routledge 2018) 421–33. Rosenberg, Gerald N, The Hollow Hope: Can Courts Bring about Social Change? (2nd edn, University of Chicago Press 2008). Rosenfeldt, Herbert, Frontex im Zentrum der Europäischen Grenz-und Küstenwache (Mohr Siebeck 2021). Rubio-Marín, Ruth, Immigration as a Democratic Challenge (CUP 2000). Ruhs, Martin, The Price of Rights: Regulating International Labor Migration (Princeton UP 2013). Ryan, Bernard, ‘The European Union and Labour Migration’ in Helen Toner, Elspeth Guild, and Anneliese Baldaccini (eds), Whose Freedom, Security and Justice? (Hart Publishing 2007) 489–516. ——‘What Role for Legal Guarantees?’ in Bernard Ryan and Valsamis Mitsilegas (eds), Extraterritorial Immigration Control (Martinus Nijhoff 2010) 3–38. ——‘Sea Borders Regulation (EU) No 656/2014’ Policies’ in Daniel Thym and Kay Hailbronner (eds), EU Immigration and Asylum Law: Article-by-Article Commentary (3rd edn, CH Beck/Hart Publishing/Nomos 2022) 179–215. Sadowski, Piotr, ‘A Safe Harbour or a Sinking Ship? On the Protection of Fundamental Rights of Asylum Seekers in Recent CJEU Judgments’ (2019) 11 EJLS 29. Sainsbury, Diane, Welfare States and Immigrant Rights (OUP 2012). Salomon, Stefan, ‘Constructing Equality in EU Asylum Law’ (2022) 34 IJRL 608. ——and Rijpma, Jorrit J, ‘A Europe Without Internal Frontiers’ (2023) 24 German LJ section B (forthcoming).

Bibliography 625 Salt, John, ‘A Comparative Overview of International Trends and Types, 1950–80’ (1989) 23 IM Rev 431. Salter, Mark B, Rights of Passage: The Passport in International Relations (Lynne Rienner 2003). Samuelson, William and Zeckhauser, Richard, ‘Status Quo Bias in Decision Making’ (1988) 1 J Risk & Uncertainty 7. Sánchez-Barrueco, María-Luisa, ‘Business as Usual? Mapping Outsourcing Practices in Schengen Visa Processing’ (2018) 44 JEMS 382. Sangiovanni, Andrea, ‘Solidarity in the European Union’ (2013) 33 OJLS 213. Santos Vara, Juan and Pascual Matellán, Laura, ‘The Global Compact on Migration’ in Elaine Fahey (ed), Framing Convergence with the Global Legal Order (Hart Publishing 2020) 163–76. Sarat, Austin and Scheingold, Stuart, ‘The Dynamics of Cause Lawyering’ in Austin Sarat and Stuart Scheingold (eds), The Worlds Cause Lawyers Make (Stanford UP 2005) 1–34. Sarmiento, Daniel, ‘Half a Case at a Time’ in Catherine van de Heyning and Maartje de Visser (eds), Constitutional Conversations in Europe (Intersentia 2012) 13–40. Sarolea, Sylvie, Droits de l’homme et migrations (Bruylant 2006). ——‘Is Access to Asylum the Same as Access to Justice?’ in Marie-Claire Foblets and Luc Leboeuf (eds), Humanitarian Admission to Europe (Nomos/Hart Publishing 2020) 115–54. Sassen, Saskia, A Sociology of Globalisation (WW Norton 2007). —— Losing Control? (Columbia UP 1996). Savino, Mario, ‘Irregular Migration at the Crossroads’ (2016) 53 CML Rev 1419. Schapendonk, Joris, ‘Turbulent Trajectories: African Migrants on Their Way to the European Union’ (2012) 2 Societies 27. Scharpf, Fritz, ‘Community and Autonomy’ (1994) 1 JEPP 219. —— Governing in Europe (OUP 1999). ——‘The European Social Model’ (2002) 40 JCMS 645. Scheffler, Samuel, ‘Immigration and the Significance of Culture’ (2007) 35 Philosophy & Public Affairs 93. Scheingold, Stuart and Lindberg, Leon N, Europe’s Would-Be Polity (Prentice-Hall 1970). Schelling, Thomas, The Strategy of Conflict (Harvard UP 1980). Schepel, Harm and Wesseling, Rein, ‘The Legal Community’ (1997) 3 ELJ 165. Schewel, Kerilyn, ‘Understanding Immobility’ (2020) 54 IM Rev 328. Schieber, Julia, Komplementärer Schutz (Nomos 2013). Schieffer, Martin, ‘Community Readmission Agreements with Third Countries’ (2003) 5 EJML 343. Schierle, Florian, ‘Employers Sanctions Directive 2009/52/EC’ in Daniel Thym and Kay Hailbronner (eds), EU Immigration and Asylum Law: Article-by-Article Commentary (3rd edn, CH Beck/Hart Publishing/Nomos 2022) 869–909. Schinkel, Willem, Imagined Societies: A Critique of Immigrant Integration in Western Europe (CUP 2017). Schmalz, Dana, Refugees, Democracy and the Law (Routledge 2020). Schmidt, Susanne K, ‘Who Cares about Nationality? The Path-Dependent Case Law of the ECJ from Goods to Citizens’ (2012) 19 J Eur Pub Pol 8. —— The European Court of Justice and the Policy Process (OUP 2018). Schmidt-Aßmann, Eberhard, ‘Verfassungsprinzipien für den europäischen Verwaltungsverbund’ in Wolfgang Hoffmann- Riem, Eberhard Schmidt- Aßmann, and Andreas Voßkuhle (eds), Grundlagen des Verwaltungsrechts, Band I (2nd edn, CH Beck 2012) § 5. Schmitt, Carl, Politische Theologie (Duncker & Humblot 1922). Schmitter, Catherine, Julien-Laferriere, François, and Carlier, Jean-Yves, ‘Vers quelle politique communautaire des visas de court séjour?’ in Philippe De Bruycker (ed), The Emergence of a European Immigration Policy (Bruylant 2003) 97–170. Schmitz-Berning, Cornelia, ‘Festung Europa’ in Cornelia Schmitz-Berning (ed), Vokabular des Nationalsozialismus (De Gruyter 2007) 232–33. Schneider, Gerald, Segadlo, Nadine, and Leue, Miriam, ‘Forty-Eight Shades of Germany’ (2020) 29 German Pol 546.

626 Bibliography Schneider, Jens-Peter, ‘Basic Structures of Information Management in the European Administrative Union’ (2014) 20 EPL 89. Schon, Justin, ‘Focus on the Forest, not the Trees: A Changepoint Model of Forced Displacement’ (2015) 28 J Refugee Stud 437. Schoukens, Paul and Pieters, Danny, ‘Irregular Labour Migrants and Access to Social Protection’ in Gijsbert Vonk (ed), Cross-Border Welfare State (Intersentia 2012) 91–112. Schultz, Jessica, The Internal Protection Alternative in Refugee Law (Brill Nijhoff 2019). Schulze- Engler, Frank, ‘Irritating Europe’ in Graham Huggan (ed), The Oxford Handbook of Postcolonial Studies (OUP 2013) 669–91. Schulze-Fielitz, Helmuth, ‘Grundmodi der Aufgabenwahrnehmung’ in Wolfgang Hoffmann-Riem, Eberhard Schmidt-Aßmann, and Andreas Voßkuhle (eds), Grundlagen des Verwaltungsrechts, Band I (2nd edn, CH Beck 2012) § 12. Schütze, Robert, From Dual to Cooperative Federalism (OUP 2009). Schwarz, Michael, Grundlinien der Anerkennung im Raums der Freiheit, der Sicherheit und des Rechts (Mohr Siebeck 2016). Scicluna, Nicole, European Union Constitutionalism in Crisis (Routledge 2015). Scott, Matthew, ‘Natural Disasters, Climate Change and Non-Refoulement’ (2014) 26 IJRL 404. —— Climate Change, Disasters, and the Refugee Convention (CUP 2020). Shachar, Ayelet, ‘The Race for Talent: Highly Skilled Migrants and Competitive Immigration Regimes’ (2006) 81 NYU L Rev 148. ——‘The Shifting Border: Legal Cartographies of Migration and Mobility’ in Ayelet Shachar (ed), Shifting Border (Manchester UP 2020) 3–96. Sharpston, Eleanor, ‘Transparency and Clear Legal Language in the European Union’ (2009/10) 12 Camb Yearb Eur Leg Stud 409. Shaw, Jo, ‘European Union Legal Studies in Crisis?’ (1996) 16 OJLS 231. ——‘Citizenship of the Union’ in European University Institute (ed), Collected Courses of the Academy of European Law 1995, vol 6-1 (Kluwer 1998) 237–347. —— The People in Question (Bristol UP 2020). Sheller, Mimi and Urry, John, ‘The New Mobilities Paradigm’ (2006) 38 Environment & Planning A: Economy and Space 207. Siebold, Angela, ZwischenGrenzen: Die Geschichte des Schengen-Raums aus deutschen, französischen und polnischen Perspektiven (Ferdinand Schöningh 2013). Simon, Julian L, The Economic Consequences of Immigration (2nd edn, University of Michigan Press 1999). Simoncini, Marta, Administrative Regulation beyond the Non-Delegation Doctrine: A Study on EU Agencies (Bloomsbury/Hart Publishing 2018). Skordas, Achilles, ‘Immigration and the Market: The Long Term Residence Directive’ (2006) 13 Colum J Eur L 201. ——‘The Rise of the Neo-Hobbesian Age’ (2019) 79 Heidelberg J Intl L 469. ——‘Temporary Protection Directive 2001/55/EC’ in Daniel Thym and Kay Hailbronner (eds), EU Immigration and Asylum Law: Article-by-Article Commentary (3rd edn, CH Beck/Hart Publishing/ Nomos 2022) 1177–228. Slingenberg, Lieneke, The Reception of Asylum Seekers under International Law: Between Sovereignty and Equality (Hart Publishing 2014). ——‘The Right Not to be Dominated. The Case Law of the European Court of Human Rights on Migrants’ Destitution’ (2019) 19 Hum Rts L Rev 291. ——‘Political Agreement on a Recast Asylum Reception Conditions Directive’ in Daniel Thym and Odysseus Academic Network (eds), Reforming the Common European Asylum System (Nomos 2022) 257–76. Smyth, Ciara, ‘The Best Interests of the Child in the Expulsion and First-Entry Jurisprudence of the European Court of Human Rights’ (2015) 17 EJML 70. ——‘The Human Rights Approach to “Persecution” and its Child Rights Discontents’ (2021) 33 IJRL 238. Snyder, Francis, ‘New Directions in European Community Law’ (1987) 14 JL & Soc 167.

Bibliography 627 Soguk, Nevzat, States and Stranger: Refugees and Displacements of Statecraft (Minnesota UP 1999). Solomon, Solon, ‘Migrant Boats on the High Seas and their Interception through Psychologically Coercive Measures’ (2019) 37 Neth Q Hum Rights 36. Somek, Alexander, Engineering Equality Law (OUP 2011). —— The Cosmopolitan Constitution (OUP 2014). ——‘The Darling Dogma of Bourgeois Europeanists’ (2014) 20 ELJ 688. Song, Sarah, ‘Three Models of Civic Solidarity’ in Rogers M Smith (ed), Citizenship, Borders, and Human Needs (University of Pennsylvania Press 2011) 192–207. Soysal, Yasemin, Limits of Citizenship: Migrants and Postnational Membership in Europe (University of Chicago Press 1994). Spaventa, Eleanor, ‘The Rights of Citizens under the Withdrawal Agreement: A Critical Analysis’ (2020) 45 EL Rev 193. Spijkerboer, Thomas, ‘The Global Mobility Infrastructure: Reconceptualising the Externalisation of Migration Control’ (2018) 20 EJML 452. ——‘Subsidiarity in Asylum Law’ in Daphné Bouteillet-Paquet (ed), Subsidiary Protection of Refugees in the European Union (Bruylant 2002) 19–42. ——‘A Distributive Approach to Migration Law’ in Roland Pierik and Wouter Werner (eds), Cosmopolitanism in Context (CUP 2010) 249–76. ——‘Stretching the Limits’ in Marleen Maes and others (eds), External Dimensions of EU Migration and Asylum Law and Policy (Bruylant 2011) 387–413. ——‘Analysing European Case Law on Migration: Options for Critical Lawyers’ in Loïc Azoulai and Karin De Vries (eds), EU Migration Law: Legal Complexities and Political Rationales (OUP 2014) 188–218. ——‘Changing Paradigms in Migration Law Research’ in Carolus Grütter, Sandra Mantu, and Paul Minderhoud (eds), Migration on the Move (Brill 2017) 13–26. ——and Steyger, Elies, ‘European External Migration Funds and Public Procurement Law’ (2019) 4 European Papers 493. Squire, Vicki, ‘Unauthorised Migration Beyond Structure/Agency? Acts, Interventions, Effects’ (2017) 37 Politics 254. Staffans, Ida, Evidence in European Asylum Procedures (Martinus Nijhoff 2012). Staples, Helen, ‘Adjudicating the External Schengen Border’ in Kees Groenendijk, Elspeth Guild, and Paul Minderhoud (eds), In Search of Europe’s Borders (Kluwer 2003) 215–50. Staver, Anne, ‘Free Movement and the Fragmentation of Family Reunification Rights’ (2013) 15 EJML 69. Stein, Eric, ‘Lawyers, Judges, and the Making of a Transnational Constitution’ (1981) 75 Am J Intl L 1. Steinfeld, Robert J, Coercion, Contract and Free Labor in the Nineteenth Century (CUP 2001). Stephan, Walter G, Ybarra, Oscar, and Rios, Kimberly, ‘Intergroup Threat Theory’ in Todd D Nelson (ed), Handbook of Prejudice, Stereotyping, and Discrimination (2nd edn, Psychology Press 2016) 255–78. Stetter, Stephan, ‘Regulating Migration’ (2000) 7 J Eur Pub Pol 80. Storey, Hugo, ‘Asylum Qualification Directive 2011/95/EU’ in Daniel Thym and Kay Hailbronner (eds), EU Immigration and Asylum Law: Article-by-Article Commentary (3rd edn, CH Beck/Hart Publishing/Nomos 2022) 1362–86. ——‘EU Refugee Qualification Directive: A Brave New World?’ (2008) 20 IJRL 1. ——‘Briefing Note for Participants’ (2013) 25 IJRL 328. Stoyanova, Vladislava, ‘How Exceptional Must “Very Exceptional” Be? Non-Refoulement, Socio- Economic Deprivation, and Paposhvili v Belgium’ (2017) 29 IJRL 580. ——‘Human Trafficking and Refugee Law’ in Satvinder Singh Juss (ed), Research Handbook on International Refugee Law (Edward Elgar Publishing 2019) 324–42. ——‘The Right to Leave Any Country and the Interplay between Jurisdiction and Proportionality in Human Rights Law’ (2020) 32 IJRL 403. ——and Smet, Stijn (eds), Migrants’ Rights, Populism and Legal Resilience in Europe (CUP 2022). Strik, Tineke, ‘The Global Approach to Migration and Mobility’ (2017) 5 Groningen J Intl L 311.

628 Bibliography ——‘The European Parliament’ in Paul Minderhoud, Sandra Mantu, and Karin Zwaan (eds), Caught In Between Borders (Wolf 2019) 279–90. Strumia, Francesca, ‘European Citizenship and EU Immigration’ (2016) 22 ELJ 417. Stumpf, Juliet, ‘The Crimmigration Crisis’ (2006) 56 Am U L Rev 367. Sundberg Diez, Olivia, Trauner, Florian, and De Somer, Marie, ‘Return Sponsorships in the EU’s New Pact on Migration and Asylum’ (2021) 23 EJML 219. Sunstein, Cass R, Legal Reasoning and Political Conflict (OUP 1996). Tacea, Angela, ‘From Legal to Political Reasoning’ (2021) 59 JCMS 1573. Tans, Simon, Service Provision and Migration (Brill/Nijhoff 2017). Taylor, Savitri, ‘From Border Control to Migration Management’ (2005) 39 Soc Poly & Admin 563. Teitgen-Colly, Catherine, ‘The European Union and Asylum: An Illusion of Protection’ (2006) 43 CML Rev 1503. ten Teije, Irene, Coenders, Marcel, and Verkuyten, Maykel, ‘The Paradox of Integration’ (2013) 44 Soc Psych 278. ter Steeg, Marcus, Das Einwanderungskonzept der EU (Nomos 2006). Tezcan Idriz, Narin, ‘The Puzzle Posed by Demir for the Free Movement of Turkish Workers’ in Daniel Thym and Margarite Zoeteweij-Turhan (eds), Rights of Third-Country Nationals under EU Association Agreements (Brill Nijhoff 2015) 223–47. Thielemann, Eiko R, ‘Why Asylum Policy Harmonisation Undermines Refugee Burden-Sharing’ (2004) 6 EJML 47. ——and Zaun, Natascha, ‘Escaping Populism –Safeguarding Minority Rights: Non-Majoritarian Dynamics in European Policy-Making’ (2018) 56 JCMS 906. Thomas, Rebekah, ‘Biometrics, International Migrants and Human Rights’ (2005) 7 EJML 377. Thorburn, Joanne, ‘Transcending Boundaries: Temporary Protection and Burden-Sharing in Europe’ (1995) 7 IJRL 459. Thouez, Colleen, ‘Strengthening Migration Governance: The UN as “Wingman”’ (2019) 45 JEMS 1242. Thym, Daniel, ‘The Schengen Law: A Challenge for Legal Accountability in the European Union’ (2002) 8 ELJ 218. ——‘Respect for Private and Family Life under Article 8 ECHR in Immigration Cases’ (2008) 57 ICLQ 87. —— Ungleichzeitigkeit und Europäisches Verfassungsrecht (Nomos 2004). ——‘Foreign Affairs’ in Armin von Bogdandy and Jürgen Bast (eds), Principles of European Constitutional Law (2nd edn, Hart Publishing 2009) 309–43. —— Migrationsverwaltungsrecht (Mohr Siebeck 2010). —— ‘Residence as de facto Citizenship?’ in Ruth Rubio Marín (ed), Human Rights and Immigration (OUP 2014) 106–44. ——‘Constitutional Foundations of the Judgments on the EEC-Turkey Association Agreement’ in Daniel Thym and Margarite Zoeteweij-Turhan (eds), Rights of Third-Country Nationals under EU Association Agreements (Brill Nijhoff 2015) 13–38. ——‘The Elusive Limits of Solidarity: Residence Rights of and Social Benefits for Economically Inactive Union Citizens’ (2015) 52 CML Rev 17. ——‘Citizens and Foreigners in EU Law: Migration Law and Its Cosmopolitan Outlook’ (2016) 22 ELJ 296. ——‘The Constitutional Dimension of Public Policy Justifications’ in Panos Koutrakos, Niamh Nic Shuibhne, and Phil Syrpis (eds), Exceptions from EU Free Movement Law: Derogation, Justification and Proportionality (Hart Publishing 2016) 170–89. ——‘The “Refugee Crisis” as a Challenge of Legal Design and Institutional Legitimacy’ (2016) 53 CML Rev 1545. ——‘Towards a Contextual Conception of Social Integration in EU Immigration Law: Comments on P & S and K & A’ (2016) 18 EJML 89. ——‘Frontiers of EU Citizenship’ in Dimitry Kochenov (ed), EU Citizenship and Federalism: The Role of Rights (CUP 2017) 705–30. — —‘Migrationsfolgenrecht’ (2017) 76 Veröffentlichungen der Vereinigung Deutscher Staatsrechtslehrer 169.

Bibliography 629 ——‘The Evolution of Citizens’ Rights in Light of the EU’s Constitutional Development’ in Daniel Thym (ed), Questioning EU Citizenship (Bloomsbury/Hart Publishing 2017) 111–34. —— (ed), Questioning EU Citizenship (Bloomsbury/Hart Publishing 2017). ——‘Supranational Differentiation and Enhanced Cooperation’ in Takis Tridimas and Robert Schütze (eds), The Oxford Principles of European Union Law, vol 1 (OUP 2018) 848–55. ——‘Vom “Fremdenrecht” über die “Denizenship” zur “Bürgerschaft”’ (2018) 57 Der Staat 77. ——‘A Bird’s Eye View on ECJ Judgments on Immigration, Asylum and Border Control Cases’ (2019) 21 EJML 166. ——‘Between “Administrative Mindset” and “Constitutional Imagination”. The Role of the Court of Justice in Immigration, Asylum and Border Control Policy’ (2019) 44 EL Rev 138. ——‘Friendly Takeover, or: the Power of the “First Word”’ (2020) 16 EuConst 187. ——‘The End of Human Rights Dynamism? Judgments of the ECtHR on “Hot Returns” and Humanitarian Visas as a Focal Point of Contemporary European Asylum Law and Policy’ (2020) 32 IJRL 569. ——‘Supranational Courts in Europe: A Moderately Communitarian Turn in the Case Law on Immigration and Citizenship’ (2021) 47 JEMS 4534. ——‘Komplexität als Chance’ [2022] Zeitschrift für Ausländerrecht 139. — —‘Legal Framework for Asylum Policy’ in Daniel Thym and Kay Hailbronner (eds), EU Immigration and Asylum Law: Article-by-Article Commentary (3rd edn, CH Beck/Hart Publishing/ Nomos 2022) 1129–76. ——‘Legal Framework for Entry and Border Control Policies’ in Daniel Thym and Kay Hailbronner (eds), EU Immigration and Asylum Law: Article-by-Article Commentary (3rd edn, CH Beck/Hart Publishing/Nomos 2022) 47–83. ——‘Legal Framework for EU Immigration Policy’ in Daniel Thym and Kay Hailbronner (eds), EU Immigration and Asylum Law: Article-by-Article Commentary (3rd edn, CH Beck/Hart Publishing/ Nomos 2022) 383–431. ——‘Long-Term Residents Directive 2003/109/EC’ in Daniel Thym and Kay Hailbronner (eds), EU Immigration and Asylum Law: Article-by-Article Commentary (3rd edn, CH Beck/Hart Publishing/ Nomos 2022) 540–641. ——‘Never-Ending Story? Political Dynamics, Legislative Uncertainties, and Practical Drawbacks of the “New” Pact on Migration and Asylum’ in Daniel Thym and Odysseus Academic Network (eds), Reforming the Common European Asylum System (Nomos 2022) 11–32. ——‘Secondary Movements’ in Daniel Thym and Odysseus Academic Network (eds), Reforming the Common European Asylum System (Nomos 2022) 129–47. ——and Bornemann, Jonas, ‘Schengen and Free Movement Law During the First Phase of the Covid- 19 Pandemic’ (2020) 5 European Papers 1143. ——and Hailbronner, Kay (eds), EU Immigration and Asylum Law: Article-by-Article Commentary (3rd edn, CH Beck/Hart Publishing/Nomos 2022). ——and Odysseus Academic Network (eds), Reforming the Common European Asylum System (Nomos 2022). ——and Tsourdi, Evangelina (Lilian), ‘Searching for Solidarity in the EU Asylum and Border Policies’ (2017) 24 MJECL 605. Tittel-Mosser, Fanny, ‘The Unintended Legal and Policy Relevance of EU Mobility Partnerships’ (2018) 20 EJML 314. Tobler, Christa, ‘Context-Related Interpretation of Association Agreements’ in Daniel Thym and Margarite Zoeteweij- Turhan (eds), Rights of Third- Country Nationals under EU Association Agreements (Brill Nijhoff 2015) 101–26. Tomuschat, Christian, Human Rights (3rd edn, OUP 2014). Tonkiss, Katherine and Bloom, Tendayi, ‘Theorising Noncitizenship: Concepts, Debates and Challenges’ (2015) 19 Citizenship Stud 837. Torpey, John, The Invention of the Passport (2nd edn, CUP 2018). Torres Pérez, Aida, Conflicts of Rights in the European Union (OUP 2009). Trachtman, Joel P, The International Law of Economic Migration (WE Upjohn Institute for Employment Research 2009).

630 Bibliography ——‘Economic Migration and Mode 4 of GATS’ in Vincent Chetail and Céline Bauloz (eds), Research Handbook on International Law and Migration (Edward Elgar Publishing 2014) 346–67. Trauner, Florian and Kruse, Imke, ‘EC Visa Facilitation and Readmission Agreements’ (2008) 10 EJML 411–38. ——and Manigrassia, Emanuele, ‘When Visa-Free Travel Becomes Difficult to Achieve and Easy to Lose’ (2014) 16 EJML 125. ——and Ripoll Servent, Ariadna, ‘The Analytical Framework’ in Florian Trauner and Ariadna Ripoll Servent (eds), Policy Change in the Area of Freedom, Security and Justice (Routledge 2015) 11–32. ——and Ripoll Servent, Ariadna, ‘The Communitarization of the Area of Freedom, Security and Justice’ (2016) 54 JCMS 1417. Triandafyllidou, Anna and Bartolini, Laura, ‘Understanding Irregularity’ in Sarah Spencer and Anna Triandafyllidou (eds), Migrants with Irregular Status in Europe (Springer 2020) 11–31. Tridimas, Takis, The General Principles of EC Law (2nd edn, OUP 2007). Tryfonidou, Alina, Reverse Discrimination in EC Law (Kluwer 2009). Tsagourias, Nicholas, ‘EU Peacekeeping Operations’ in Martin Trybus and Nigel White (eds), European Security Law (OUP 2007) 102–33. Tsianos, Vassilis and Karakayalı, Serhat, ‘Transnational Migration and the Emergence of the European Border Regime’ (2010) 13 Eur J Soc Theory 373. Tsourdi, Evangelia (Lilian), ‘Bottom- up Salvation? From Practical Cooperation Towards Joint Implementation through the European Asylum Support Office’ (2016) 1 European Papers 997. ——‘EU Reception Conditions’ in Vincent Chetail, Philippe De Bruycker, and Francesco Maiani (eds), Reforming the Common European Asylum System (Brill 2016) 271–316. ——‘Holding the European Asylum Support Office Accountable for its Role in Asylum Decision- Making’ (2020) 21 German LJ 506. ——‘The Emerging Architecture of EU Asylum Policy’ in Francesca Bignami (ed), EU Law in Populist Times: Crises and Prospects (CUP 2020) 191–226. ——‘Relocation Blues: Refugee Protection Backsliding, Division of Competences, and the Purpose of Infringement Proceedings’ (2021) 58 CML Rev 1819. ——‘Asylum Reception Conditions Directive 2013/33/EU’ in Daniel Thym and Kay Hailbronner (eds), EU Immigration and Asylum Law: Article-by-Article Commentary (3rd edn, CH Beck/Hart Publishing/Nomos 2022) 1540–638. ——and Costello, Cathryn, ‘The Evolution of EU Law on Refugees and Asylum’ in Paul Craig and Gráinne de Búrca (eds), The Evolution of EU Law (3rd edn, OUP 2021) 793–823. Turack, Daniel C, ‘Freedom of Movement and the Travel Document in Benelux’ (1968) 17 ICLQ 191. ——‘The Scandinavian Passport Union’ (1968) 38 Nordic J Intl L 171. Turenne, Sophie, ‘Institutional Constraints and Collegiality at the Court of Justice of the European Union’ (2017) 24 MJECL 565. Türk, Alexander H, Administrative Law and Policy of the European Union (OUP 2011). Tversky, Amos and Kahnemann, Daniel, ‘The Framing of Decisions and the Psychology of Choice’ (1981) 30 Science 453. Učakar, Tjaša, ‘The Rhetoric of European Migration Policy and Its Role in Criminalization of Migration’ in Neža Kogovšek Šalamon (ed), Causes and Consequences of Migrant Criminalization (Springer 2020) 91–108. Urth, Helene, ‘Building a Momentum for the Integration of Third-Country Nationals in the European Union’ (2005) 7 EJML 163. Vammen, Ida Marie and Brønden, Birgitte Mossin, ‘Donor‐Country Responses to the Migration– Development Buzz’ (2012) 50 Intl Migration 26. van Berckel Smit, Jasper, ‘Taking Onboard the Issue of Disembarkation’ (2020) 22 EJML 492. Van Bueren, Geraldine, The International Law on the Rights of the Child (Martinus Nijhoff 1998). Van Criekinge, Tine, ‘Integration of Migration Issues in EPAs’ in Gerrit Faber and Jan Orbie (eds), Beyond Market Access for Economic Development (Routledge 2009) 173–95. van de Peer, Patricia, ‘Negotiating the Second Generation of the Common European Asylum System Instruments’ in Vincent Chetail, Philippe De Bruycker, and Francesco Maiani (eds), Reforming the Common European Asylum System (Brill 2016) 55–73.

Bibliography 631 van den Brink, Martijn, ‘Justice, Legitimacy and the Authority of Legislation within the European Union’ (2019) 82 MLR 293. ——‘The Relationship between National and EU Citizenship’ in Dora Kostakopoulou and Daniel Thym (eds), Research Handbook on European Union Citizenship Law and Policy (Edward Elgar Publishing 2022) 100–14. van der Woude, Maartje and van der Leun, Joanne, ‘Crimmigration Checks in the Internal Border Areas of the EU’ (2017) 14 EJC 27. Van Elsuwege, Peter, ‘The EU and its Neighbours’ in Ramses A Wessel and Joris Larik (eds), EU External Relations Law (Hart Publishing 2020) 437–60. ——and Vankova, Zvezda, ‘Migration and Mobility in the EU’s Eastern Neighbourhood’ in Ippolito Francesca, Gianluca Borzoni, and Federico Casolari (eds), Bilateral Relations in the Mediterranean (Edward Elgar Publishing 2020) 160–83. van Gestel, Rob and Micklitz, Hans-Wolfgang, ‘Why Methods Matter in European Legal Scholarship’ (2014) 20 ELJ 292. van Houte, Marieke, ‘The Evolution of Governance and Financing of Migration and Development Policy and Politics in Europe’ in Agnieszka Weinar, Saskia Bonjour, and Lyubov Zhyznomirska (eds), The Routledge Handbook of the Politics of Migration in Europe (Routledge 2019) 420–32. van Meeteren, Masja, Engbersen Weitaus, Godfried, and van San, Marion, ‘Striving for a Better Position: Aspirations and the Role of Cultural, Economic, and Social Capital for Irregular Migrants in Belgium’ (2009) 43 IM Rev 881. van Middelaar, Luuk, Alarums and Excursions: Improvising Politics on the European Stage (Agenda 2019). Van Selm, Joanne, ‘Temporarily Protecting Displaced Persons or Offering the Possibility to Start a New Life in the European Union’ (2001) 3 EJML 23. Vandenhole, Wouter, ‘Migration and Discrimination’ in Vincent Chetail and Céline Bauloz (eds), Research Handbook on International Law and Migration (Edward Elgar Publishing 2014) 216–38. Vankova, Zvezda, Circular Migration and the Rights of Migrant Workers in Central and Eastern Europe (Springer 2020). Vauchez, Antoine, Brokering Europe: Euro- Lawyers and the Making of a Transnational Polity (CUP 2015). Vavoula, Niovi, Immigration and Privacy in the Law of the European Union (Brill 2022). ——‘Consultation of EU Immigration Databases for Law Enforcement Purposes’ (2020) 22 EJML 139. ——‘Artificial Intelligence (AI) at Schengen Borders’ (2021) 23 EJML 457. ——‘Information Sharing in the Dublin System’ (2021) 22 German LJ 391. Vedsted-Hansen, Jens, ‘The Legal Condition of Refugees in Denmark’ (1994) 7 J Refugee Stud 249. ——‘Non-Admission Policies and the Right to Protection’ in Frances Nicholson and Patrick Twomey (eds), Refugee Rights and Realities (CUP 1999) 269–88. ——‘Common EU Standards on Asylum’ (2005) 7 EJML 369. ——‘Current Protection Dilemmas in the European Union’ in Carolus Grütter, Sandra Mantu, and Paul Minderhoud (eds), Migration on the Move (Brill 2017) 95–117. ——‘Asylum Procedures Directive 2013/32/EU’ in Daniel Thym and Kay Hailbronner (eds), EU Immigration and Asylum Law: Article-by-Article Commentary (3rd edn, CH Beck/Hart Publishing/ Nomos 2022) 1429–539. ——‘Border Procedure on Asylum and Return’ in Daniel Thym and Odysseus Academic Network (eds), Reforming the Common European Asylum System (Nomos 2022) 99–112. Velluti, Samantha, ‘What European Union Strategy for Integrating Migrants?’ (2007) 9 EJML 53. Vermeulen, Gert, ‘Mutual Instrumentalization of Criminal and Migration Law from an EU Perspective’ (2007) 9 EJML 347–61. Verschueren, Herwig, ‘Employment and Social Security Rights of Third-Country Labour Migrants under EU Law’ (2016) 18 EJML 373. ——‘The Labyrinth of Employment and Social Rights in the EU Intra-Corporate Transfer Directive’ (2021) 12 Eur Labour L J 280. Vliegenthart, Rens, ‘Framing Citizens in the Field of Immigration Politics’ in Marco Giugni and Maria Grasso (eds), Handbook of Citizenship and Migration (Edward Elgar Publishing 2021) 269–83.

632 Bibliography Vogenauer, Stefan, ‘The Prohibition of Abuse of Law’ in Rita de la Feria and Stefan Vogenauer (eds), Prohibition of Abuse of Law (Hart Publishing 2011) 521–72. Voglrieder, Sabine, ‘Die Sanktionsrichtlinie’ [2009] Zeitschrift für Ausländerrecht 168. Vollmer, Bastian A, ‘Policy Discourses on Irregular Migration in the EU’ (2011) 13 EJML 317. von Bogdandy, Armin, ‘The European Union as a Human Rights Organisation?’ (2000) 37 CML Rev 1307. ——‘Founding Principles of EU Law: A Theoretical and Doctrinal Sketch’ (2010) 16 ELJ 95. ——‘National Legal Scholarship in the European Legal Area—A Manifesto’ (2012) 10 ICON 614. Vonk, Gijsbert and Van Walsum, Sarah, ‘Access Denied’ in Gijsbert Vonk (ed), Cross-Border Welfare State (Intersentia 2012) 3–62. Wahl, Rainer, ‘Die Aufgabenabhängigkeit von Verwaltung und Verwaltungsrecht’ in Eberhard Schmidt-Aßmann, Wolfgang Hoffmann-Riem, and Gunnar Folke Schuppert (eds), Reform des Allgemeinen Verwaltungsrechts (Nomos 1993) 177–218. Waldron, Jeremy, Law and Disagreement (Clarendon 1999). Walker, Neil, ‘In Search of the Area of Freedom, Security and Justice: A Constitutional Odyssey’ in Neil Walker (ed), Europe’s Area of Freedom, Security and Justice (Hart Publishing 2004) 3–37. ——‘Legal Theory and the European Union: A 25th Anniversary Essay’ (2005) 25 OJLS 581. ——‘Denizenship and Deterritorialisation in the European Union’ in Hans Lindahl (ed), A Right to Inclusion and Exclusion? (Hart Publishing 2009) 261–74. Walters, William, ‘Mapping Schengenland’ (2002) 20 Environment and Planning D: Society and Space 561. Walzer, Michael, Spheres of Justice (Basic Books 1983). Ward, Natasha, ‘Facilitating the Temporary Movement of Natural Persons’ in Rahel Kunz, Sandra Lavenex, and Marion Panizzon (eds), Multilayered Migration Governance (Routledge 2011) 143–82. Weatherill, Stephen, ‘Why there is No “Principle of Mutual Recognition” in EU Law’ in Kai Purnhagen and Peter Rott (eds), Varieties of European Economic Law and Regulation (Springer 2014) 401–18. Weber, Ferdinand, Staatsangehörigkeit und Status (Mohr Siebeck 2018). Weil, Patrick, How to be French: Nationality in the Making since 1789 (Duke UP 2008). Weiler, Joseph HH, ‘The Transformation of Europe’ (1990/91) 100 Yale LJ 2403. ——‘European Citizenship and Human Rights’ in Jan Winter and others (eds), Reforming the Treaty on European Union (TMC Asser 1996) 57–86. ——‘To Be a European Citizen: Eros and Civilization’ (1997) 4 J Eur Pub Pol 495. ——‘Epilogue’ in Maurice Adams and others (eds), Judging Europe’s Judges (Hart Publishing 2013) 235–54. Welfens, Nathalie, ‘Selection Categories in European Resettlement and Humanitarian Admission Programmes’ in Moritz Jesse (ed), European Societies, Migration, and the Law (CUP 2020) 81–103. Wendel, Mattias, ‘The Refugee Crisis and the Executive’ (2016) 17 German LJ 1005. Wessels, Wolfgang, ‘The European Council’ in Takis Tridimas and Robert Schütze (eds), The Oxford Principles of European Union Law: Vol I (OUP 2018) 490–516. Wibault, Tristan, ‘Making the Case X & X for the Humanitarian Visa’ in Marie-Claire Foblets and Luc Leboeuf (eds), Humanitarian Admission to Europe (Nomos/Hart Publishing 2020) 271–82. Wiener, Antje, Building Institutions (Westview 1998). ——‘Forging Flexibility: The British “No” to Schengen’ (1999) 1 EJML 441. Wiener, Antje, and Diez, Thomas (eds), European Integration Theory (3rd edn, OUP 2019). Wiesbrock, Anja, ‘Granting Citizenship-Related Rights to Third-Country Nationals’ (2012) 14 EJML 63. —— Legal Migration to the European Union (Martinus Nijhoff 2010). ——‘Social Security Rights of Third Country Nationals under the Euro-Mediterranean Association Agreements’ in Elspeth Guild, Sergio Carrera, and Katharina Eisele (eds), Social Benefits and Migration (CEPS 2013) 45–61. Wilderspin, Michael, ‘Article 77 TFEU’ in Manuel Kellerbauer, Marcus Klamert, and Jonathan Tomkin (eds), The EU Treaties and the Charter of Fundamental Rights: A Commentary (OUP 2019). Wilsher, Daniel, Immigration Detention: Past, Present and Future (CUP 2012).

Bibliography 633 ——‘Economic Migration into the European Union’ (2002) 21 Yearb Eur L 163. Wimmer, Andreas and Glick Schiller, Nina, ‘Methodological Nationalism, the Social Sciences, and the Study of Migration’ (2003) 37 IM Rev 576. —— Ethnic Boundary Making (OUP 2014). Wolter, Miriam, Auf dem Weg zu einem gemeinschaftlichen Asylrecht in der Europäischen Union (Nomos 1998). Worster, William Thomas, ‘The Contemporary International Law Status of the Right to Receive Asylum’ (2014) 26 IJRL 477. Wray, Helena, Agoston, Agnes, and Hutton, Jocelyn, ‘A Family Resemblance? The Regulation of Marriage Migration in Europe’ (2014) 16 EJML 209. Wright, Chris F, ‘How Do States Implement Liberal Immigration Policies?’ (2014) 27 Governance 397. Xiang, Biao and Lindquist, Johan, ‘Migration Infrastructure’ (2014) 48 IM Rev S122. Yavuz, Çiğdem Akın, ‘Analysis of the EU-Turkey Readmission Agreement’ (2019) 21 EJML 486. Zaiotti, Ruben, Culture of Border Controls: Schengen and the Evolution of European Frontiers (University of Chicago Press 2011). ——‘The Italo-French Row over Schengen, Critical Junctures, and the Future of Europe’s Border Regime’ (2013) 28 J Borderlands Stud 337. ——‘Border Management’ in Ariadna Ripoll Servent and Florian Trauner (eds), The Routledge Handbook of Justice and Home Affairs Research (Routledge 2018) 99–109. Zaun, Natascha, EU Asylum Policies: The Power of Strong Regulating States (Palgrave 2017). ——‘Why EU Asylum Standards Exceed the Lowest Common Denominator’ (2016) 23 J Eur Pub Pol 136. Zetter, Roger, ‘Labelling Refugees’ (1991) 4 J Refugee Stud 39. Zglinski, Jan, ‘The Rise of Deference’ (2018) 55 CML Rev 1341. Ziebritzki, Catharina, ‘The Integrated EU Hotspot Administration and the Question of the EU’s Liability’ in Markus Kotzur and others (eds), The External Dimension of EU Migration and Asylum Policies (Nomos 2020) 253–94. ——‘The Objective of Resettlement in an EU Constitutional Perspective’ in Marie-Claire Foblets and Luc Leboeuf (eds), Humanitarian Admission to Europe (Nomos/Hart Publishing 2020) 285–340. Zielonka, Jan, Europe as Empire (OUP 2006). Zimmermann, Andreas, and Mahler, Claudia, ‘Article 1A, para 2’ in Andreas Zimmermann (ed), The 1951 Convention Relating to the Status of Refugees and its 1967 Protocol: A Commentary (OUP 2011). ——and Elberling, Björn, ‘Ausweisungsschutz’ in Oliver Dörr, Rainer Grote, and Thilo Marauhn (eds), Konkordanzkommentar EMRK/GG (2nd edn, Mohr Siebeck 2013) ch 27. Zolberg, Aristide R, ‘The Great Wall against China’ in Jan Lucassen and Leo Lucassen (eds), Migration, Migration History, and History (Peter Lang 1997) 291–315. —— A Nation by Design: Immigration Policy in the Fashioning of America (Harvard UP 2006). Zwaan, Karin, ‘The Aznar Protocol’ in Paul Minderhoud, Sandra Mantu, and Karin Zwaan (eds), Caught in Between Borders (Wolf 2019) 303–14.

Index For the benefit of digital users, indexed terms that span two pages (e.g., 52–53) may, on occasion, appear on only one of those pages Tables and figures are indicated by t and f following the page number absconding, risk of 541–42 abusive practices ‘abuse’, language of 112–13 EU legislation 275–76 golden visas 191 academia 115–17, 118, 127 accountability access to documents 94 alternative mechanisms 90–94 Asylum Agency 214–21 CJEU 90–94 financial 92–93 Frontex 214–21 managerial 92–93 ombudspersons 91–92 parliaments 93 for wrongdoing 214–21 beyond courts 90–91 activism, judicial 74–75 actors multiplicity of 107–8, 175–76 of protection 394–96 home state and international organisations 394–95 internal protection alternative 395–96 administrative law 172–96 administrative centralisation 198–99 administrative mindset 76–77 agencies 200, 203–5 inter-state cooperation 177–79 mass administration 176–77 proceduralisation 314, 323 see also accountability; composite system; good administration; inter-state cooperation; money; national procedural autonomy Aegean Sea 207, 311–12, 343 Afghanistan 101, 102–3, 211–12, 280, 285, 297–98, 351–52, 387, 390–91, 392–93, 536, 584, 590–91, 592 Africa 15, 405, 593–94 eastern 101–2 EU Emergency Trust Fund (EUTF) 93, 193, 591 Horn of 592 Northern 463, 465–66, 590 Western 98–99, 100, 207, 295, 574, 586, 590 agencies 197–222 accountability for wrongdoing 214–21

administrative centralisation 198–99 administrative decision-making 203–5 composite procedures, legal remedies for 218–21 constitutional foundations 203–6 exponential growth in practice 202–3 institutional governance 217–18 mandates, formal expansion of 201–2 Meroni doctrine 205–6 overview 221–22 policy design 198–203 territorial scope 206 theory 198–203 see also Asylum Agency; Frontex agenda-setting 28, 29–30, 49, 573–74 agriculture 173, 192, 436, 464, 467–68, 508–9 farm subsidies 176 air travel 102, 210–11, 535–36 airport transit visas 285 Albania 36, 209–10, 550, 574, 582–83 Algeria 20–21, 589 algorithm-based processing 223, 235, 236–38, 247, 249 see also Artificial Intelligence aliens law conventions of the Council of Europe 124–25 historic minimum standards 123–24 as protection qua nationality 123–27 statelessness, reduction of 126–27 amicus curiae submissions 74 amnesty see regularisation Amnesty International 296n.107 Amsterdam Treaty 27–28 Andorra 558–59, 568–69 Anglo-Saxon countries 116, 474–75 case citations 170 casuistic law-making 153–54 common law tradition 4, 115–16, 151 Ankara Agreement (1963) 554, 561–63 annulment actions 54, 141, 438 apartheid ‘European’ 21 ‘passport’ 266 apprenticeships 450 arbitrariness 270, 511, 512 area of freedom, security and justice 13–41 conceptual vagueness 173–74 international to EU migration law 14–24

636 Index area of freedom, security and justice (cont.) overview 40–41 see also colonial migration; emigration; free movement; labour migration; refugees; single market; Schengen area; state sovereignty; third country nationals; tourism; Union citizenship Arendt, Hanna 18, 104, 109, 127, 128 Aristotle 264 Armenia 567, 593–94 arrest 189–90, 306, 355 Artificial Intelligence (AI) 236–42 automated processing 237–39 implementation deficits 239–41 quality concerns 239–41 see also algorithm-based processing; databases; digitalisation artists 506 Asia migrants from East Asia 15 southeast 405 assisted voluntary return (AVR) 530–31 association agreements 550–69 constitutional foundations 551–54 interpretative parallelism 553–54 overview 568–69 status in EU legal order 552–53 supranational competences 551–52 see also European Economic Area (EEA); neighbourhood; Switzerland; Turkey; United Kingdom; Western Balkans asylum applicants see asylum applicants jurisdiction see asylum jurisdiction right to 309–10, 353–55 Article 18 CFR 353–54 Protocol on Union citizens 354–55 supranational competences 349–51 see also non-arrival; reception conditions; refoulement, prohibition of; Refugee Convention; refugee status; resettlement; subsidiary protection; temporary protection asylum applicants application, making 331, 373–74, 424–25, 444 applications, registration 212–13, 331, 367, 373, 374, 398–99, 414, 424–25, 443, 517–18, 540 detention, criteria for 542–43 ‘fiction of non-entry’ 380–81, 414–15, 425–26, 539–40 human rights compliance 412–14 living conditions 411–12 Member States, differences between 410–11 procedures see asylum procedures restrictions 412–14 right to remain 414–15 ‘uniform status’, option of 417–18 see also reception conditions; relocation; vulnerability asylum jurisdiction 357–72 admissibility 376–77

first entry rule, origin of 358–60 jurisdiction, transfer of 370–72 mutual trust, limits of 368–70 procedure 347, 360, 364–65, 376–77 secondary movements 370–72 solidarity 360–62 substantive criteria, hierarchy of 362–64 take charge/back requests 365–68 transfer 75, 133, 187, 357–58, 367, 368–69, 424, 522, 544–45 see also Dublin III Regulation; Dublin system Asylum, Migration, and Integration Fund (AMIF) 192, 194 Asylum and Migration Management Regulation (AMMR) 345, 357–58, 361–62 Asylum Agency 6–7, 194, 202, 206t, 211–13, 214, 215, 217–18, 219–20, 221–22 accountability 214–21 CJEU architecture 72–73 information gathering 211–12 influence 211–14 international cooperation 214 operational support 212–13 policy design 200–1 practical guidance 211–12 supervision 214 theory 200–1 see also agencies; European Asylum Support Office (EASO) asylum procedures 372–86 access to 373–74 Asylum Information Database (AIDA) 212 border procedures 379–81 external processing 385–86 legal remedies 384–85 operational character 176–77 personal interview 374–76 regular and accelerated examination 376–78 subsequent applications and withdrawal 378–79 see also Common European Asylum System (CEAS); ‘hotspot’ approach; safe countries; transit zones audit/auditors 90, 92 see also European Court of Auditors (ECA) Australia 15, 385, 420, 465, 474–75 Austria 38, 54, 89, 274, 317, 320–21, 339–40, 366, 371–72, 387, 433, 444, 449, 488, 536 authoritarian regimes 102–3, 298, 346, 566, 577, 579 autonomy see national procedural autonomy Azerbaijan 124–25, 593–94 Balkan countries 18, 382 see also Western Balkans Baltic states 60 Bangladesh 584–85, 594–95 bans see entry bans Barcelona Declaration (1995) 567–68 ‘bare life’ 110 Belarus 31, 32, 208, 274–75, 311–12, 331, 346–47, 350–51, 381, 406, 586, 593–94

Index 637 Belgium 18, 21, 24, 83, 101–2, 156, 296, 305, 340, 416, 428, 535 beneficiaries of international protection equal treatment, limits of 415–16 free movement and ‘uniform status’, option of 417–18 refugees without protection status 418–19 rights of 415–19 specific guarantees 416–17 see also refugees; subsidiary protection Benelux countries 20, 24, 25, 54, 312–13, 339–40, 358–59, 457 Berne Initiative 572n.8 bilateral treaties 18–19, 122, 124, 555–56, 565 binational couples 442 biometrics 230, 324, 334–35 dactylographic data 232 biopolitics 110, 224 ‘black’ lists 283–85, 287, 300 see also visas Blue Card Directive 450–53 procedure and rights 452–53 scope, personal and thematic 450–52 see also labour migration Bolivia 284, 584 bona fide travellers 15–16, 244–45, 246, 280, 281, 299–300, 584–85 border controls, external 301–36 asylum law, interaction with 309–10, 330–31, 379–81 checks on persons 325–27 collective expulsion, prohibition of 310–11 constitutional foundations 305–11 ‘control gap’ 105–6 databases 229t document security 334–35 entry conditions 324–25 crossing points 324–28 externalisation 3–4, 294–95, 386, 581–88 extraterritoriality and human rights 306–8 incomplete territorialisation of the EU 303–4 legal remedies 327–28 local border traffic 328 operational character of 176–77 overview 335–36 policy design 302–4 refoulement, prohibition of 309–10 refusal of entry 327–28 rejection at the border 309–11 search and rescue at sea 308–9 state practices 65–66, 386, 567–68, 574, 581, 591, 596 supranational competences 305 surveillance of external borders 328–34 symbolism of borders 302–3 territorial scope 311 theory 302–4 see also e-gates; Eurosur; Frontex; pre-arrival measures; ‘pushbacks’; Schengen area; Schengen governance; sea borders; transit areas

border controls, internal 319–22 police checks in border areas 322–23 temporary reintroduction, requirements for 319–20 see also Schengen area Bosnia-Hercegovina 409 Brazil 15, 317 breaches of EU/international law 81, 82–83 of immigration law 513–14 Brexit 38, 84, 206t, 229t, 283t, 288, 312t, 316, 357, 357t, 440t, 479t, 515t, 552 CJEU case law 146–47 Denmark and 64–65 differentiated integration 63–67 effect of 67, 125, 561, 564–65, 569 Franco-British cooperation 299 free movement 577 Gibraltar 558–59 Ireland and 65–67 nationality laws 500 opt-outs 28, 63–67, 68, 282–83 referendum 48, 559 third country nationals 325–26 transition period 495–96 unskilled workers, effect on 560 visa-free travel 288 Withdrawal Agreement 9, 67, 495–96, 552–53, 554, 559–60, 569 budgets accountability 92 European Union 191–95, 493, 503 national 192, 194 buffer zones 297, 582 Bulgaria 84, 96–97, 159, 229t, 287, 312t, 312–13, 317–18, 325–26, 330, 335–36, 355, 416, 484, 536 burden-sharing see solidarity bureaucracy 50, 236 street-level bureaucrats 5–6, 108, 120, 509 burka 471 Canada 15, 287, 420, 423, 453, 465, 474–75 Canary Islands 207, 281, 574, 586–87 Cape Verde 593–94 capitalism 45n.13, 484 care-workers 433–34 Caribbean states 461 carrier sanctions 295–97 see also visas Case law autonomous concepts, limits of 162–63 coherence, promise and limits of 152–55 constitutional essentials in migration law 155–58 doctrinal foundations of 150–71 drafting history, revival of 159–60 effet utile 163 horizontal analysis 150, 253, 278 interpretation, pitfalls of 158–63 interpretative approximation 154–55

638 Index Case law (cont.) ‘legal order’ as doctrinal self-sufficiency 150–52 legislative fragmentation 152–53 methodology 115–19 multilingualism 158–59 overview 170–71 practical tips for dealing with 169–70 teleological interpretation, indeterminacy of 160–62 see also individual rights of migrants; legislation Catalan independence movement 355 centralisation 197 administrative 57, 73, 198–99, 201, 208, 221 cessation change of circumstance 400 of refugee status 351–52, 392–93, 395–96, 400 exclusion and 398–400, 425 Charter of Fundamental Rights of the EU 134–40 added value 134–38 EU not a ‘human rights organisation’ 136–37 individual rights of migrants 166–68 court practice, ambiguity of 166–68 limited scope of application 137–38 rights of the Child as an example 135–36 silence and vagueness 166–68 checks on persons 325–27 children abduction of 439 custody of 443 legal migration and 438–39 rights of the Child 135–36 see also family reunification; unaccompanied minors; vulnerability China Chinese industrialists 191, 502 emigrants 449 Exclusion Act 15 Chirac, Jacques 174n.5 Christian Democrats (EPP) 55 circular migration 100, 101–2, 452, 463–64, 465, 572, 589, 593, 594 citizenship see Union citizenship civic citizenship 473–74, 500 civic integration 469, 471, 476, 488, 489, 490 civil disobedience 321 civil law tradition 4, 6, 115, 117, 120–21, 151, 171 civil servants expertise and politics of 49–50 working parties of 52–53 climate change 8, 142–43, 391, 392, 393–94, 404, 425, 591 co-decision powers 55, 68, 93 coastguard 207, 211, 281, 306, 308–9, 583 Code Napoléon (French) 16–17, 156 codification 59–60, 116, 153–54, 156, 213 coercion 224, 241, 298 coercive law enforcement 107–8, 114 coherence promise and limits of 152–55 collective expulsion prohibition of 310–11

Colombia 283–84 colonial migration 15, 22, 39, 101–2, 106, 428, 429– 30, 445, 467, 470 colonial domination 21, 40, 577 colonies, former 14, 15, 21–22, 40, 101, 130–31, 428–29, 500–1, 537–38, 551 free movement and 20–22 see also imperialism comitology 58, 173, 175 Commission 49–51 civil servants, expertise and politics of 49–50 monopoly of initiative 28n.90, 49 motor of integration 49, 50–51, 74–75 Staff Working Documents (SWDs) 61–62 stakeholder consultations 61–62 technocratic posture on migration law 50–51 see also civil servants; comitology; Directorates General (DG) Common Agendas on Migration and Mobility (CAMM) 594–95 Common European Asylum System (CEAS) 337–426 asylum, right to 353–55 Article 18 CFR 353–54 Protocol on Union citizens 354–55 constitutional foundations 349–57 geopolitics 345–47 harmonisation 341–42 history of 338–39 intergovernmental cooperation 339–41 legislation 357t overview 424–26 policy crisis (2015/16) 342–44 policy design 338–49 recognition quotas, disparate 387–88 reform efforts 344–45 refoulement, prohibition of 351–53 refugee status 386–400 solidarity (Article 80 TFEU) 355–56 structural deficits 347–49 supranational competences 349–51 territorial scope 357 theory 338–49 see also area of freedom, security and justice; asylum applicants; asylum jurisdiction; asylum procedures; beneficiaries of international protection; persecution; refugee status; resettlement; safe countries Common Foreign and Security Policy (CFSP) 579, 587–88 Border Assistance Missions (BAM) 587–88 CSDP missions 587–88 EUBAM Libya 587–88 EUNAVFOR MED 588 Irini mission 343, 583, 588 common market 21, 23, 40, 174n.4 common values fortress Europe and 31–33 Commonwealth 21, 146–47, 500–1 communism 338–39, 420

Index 639 communitarianism 104 compensation 35n.141, 123, 124, 296, 523, 534–35, 583 competition law 59 complaints mechanisms 91 complementary protection 356–404 Europeanisation 400–1 human rights, divergence from 403 indiscriminate violence, protection against 401–3 national laws 403–4 composite system 173–79 actors, multiplicity of 175–76 areas 173–74 asylum and border controls 176–77 systems 173–74 inter-state cooperation 177–79 legal remedies 218–21 compromise-building 53–54, 55, 57, 67, 68 computer-based processing 234–35, 237–38 see also Artificial Intelligence conditionality 111, 120, 286–87, 474, 476, 581, 583–86 confidentiality 94, 185, 188–89, 244 conflict zones 235 conscientious objection 397–98 constitutionalism 106, 130–31, 511, 539 constructivism 110 black letter 116–18 doctrinal 116–18, 120–21 Foucault and 110 consular instructions 288–89 consular staff 282, 284, 289–91, 300, 325, 543 consultancy groups 68 consumer protection 72, 173, 296 control administrative 91–92 of borders see border controls, external; border controls, internal ‘control gap’ 105–6 imperatives 75, 148–49, 474 migration 566–68 state measures, significance of 103–10 state practices 65–66, 386, 567–68, 574, 581, 591, 596 see also externalisation Convention for the Prevention of Torture (CPT) 139–40 see also refoulement, prohibition of Convention on the Rights of the Child (CRC) 135–36 see also children; vulnerability cooperation, international Asylum Agency 214 horizontal see inter-state cooperation informalisation 60–61, 579–81 inter-state see inter-state cooperation intergovernmental 26–27, 339–41 migration-related projects 192–94, 576–77, 586, 591 pragmatic 566–67 see also third states

COREPER (Committee of Permanent Representatives) 53 corridors embedded 97 humanitarian 423 corruption 92, 109–10, 502 cosmopolitanism 50, 281 Cotonou Agreement 537–38, 551–52, 593–94 Council of the EU 52–54 consensus culture 53–54 majority voting 28, 46, 53–54, 361 veto power 28, 53–54, 319–20, 360, 499 working parties of 52–53 Council of Europe conventions, limited impact of 124–25 counterfeiting 292, 334 coup, military 100, 284–85, 346 Court of Justice of the EU (CJEU) access to and output in migration 79–89 privileged access by EU institutions 80 achievements 69–95 annulment, actions of 79–80 administrative mindset 76–77 appeals 169–70 constitutional authority of 74–78 constitutional imagination, reputation of 75 court architecture: Asylum Agency 72–73 different formations 71f Frontex 72–73 influence of 69–74 non-governmental organisations (NGOs) 73–74 specialised tribunals, disadvantages of 72–73 third party intervention, limits of 73–74 direct actions by individuals 79–80 infringement proceedings 80–86 changing dynamics 81–83 discrepancies between Member States 83–85 incentives for loyal cooperation 85 pre-litigation stage, importance of 81 judicial activism 74–75 judicial passivism 77 languages of proceedings 70, 153, 173n.2, 234–35, 293, 493 lawsuits 188–89, 296 lawyer-linguists 70 legislature, interaction with the 77–78 limitations 69–95 number of judgments (on EU migration law) per year 76f overconstitutionalisation 37 overview 94–95 preliminary references 83–86 ‘demand’ side 87–89 EU migration law per country (2006–22) 84f micromanagement 86, 268, 388 vagueness 86 strategic litigation 89 subject areas of judgments (until 2022) 88t validity disputes 79–80

640 Index Covid-19 pandemic asylum jurisdiction, effect on 367 external travel ban 302–3, 324–25 health screenings 374 illegal periods of stay, effect on 404 internal border controls 335–36 international traffic, decrease in 233 lockdowns v permits, issuance of 427 remittances, effect on 589 resettlement, effect on 420–21 Schengen area travel restrictions 315, 320–21 soft law 59, 60 statistics, effect on 100, 281 crimes against humanity 397 criminal sanctions limits for 518–20 criminalisation of humanitarian assistance 308–9, 343, 516–17 of illegal entry or stay 9, 88, 114, 301, 505, 514–24, 528 of illegal presence 518–19 of smuggling 515 see also securitisation criminality custodial sentences 9, 514–15, 518–19, 526, 548 see also public policy exception; security ‘crimmigration’ 114 crisis asylum policy (2015/16) 342–44 Euro (2010s) 46–47 language of 112–13 oil crisis (1973) 428, 561–62 Schengen area 313–15 Croatia 38, 240, 287, 312–13, 317–18, 329, 330 crossing points border 324–28 see also border controls, external cultural diversity 106–7, 449, 472, 503 cultural integration 472, 477, 488–89, 499, 503 customary international law 124, 537, 579–80 customers posted workers 459 of smugglers/traffickers 520 customs checks 25 control 313–14 union 21n.46, 29–30, 577 Cyprus 54, 191, 229t, 287, 312–13, 317–18, 331–32, 502–3 Czech Republic 18–19, 54, 287 data protection 243–48 data minimisation 244–46 human rights assessment 244–46 personal data, correction of 240–41 preventive supervision 247–48 principles of 243–44 privacy, right to 389–90 reactive supervision 247–48

storage periods 226, 228, 232–33, 249 see also databases databases 223–49 alphanumeric information 225, 226, 232, 240–41, 248–49 constitutional foundations 227–30 decentralised network structure 228 Entry/Exit System (EES) 236 Eurodac 232–33 European Travel Information and Authorisation System (ETIAS) 234–35 ‘function creep’ 225–27 infrastructure 224–25 overview 230–36, 248–49 policy design 224–27 Schengen Information System (SIS) 230–31 supranational competences 227–28 territorial scope 228–30 theory 224–27 Visa Information System (VIS) 233–34 visas and 188–89 see also algorithm-based processing; Artificial Intelligence (AI); biometrics; data protection; digitalisation; EU-LISA; information technology (IT); interoperability deaths of migrants at sea 207, 343, 557–58, 581–82, 583 declarations of intent 142, 198, 508, 509, 567–68, 571, 575 deconstruction theory 506 deference, judicial 5, 37, 148–49, 243–44, 268, 446, 500–1 degrading treatment see refoulement, prohibition of delegated acts 58–59 democracy deliberative 62–63 public opinion 106–7 see also politicisation demographic change 423, 427, 431, 432, 434, 448, 467, 471 denizenship 470–71, 475, 488–89, 490, 494, 499–500 Denmark 26–27, 28, 42, 63–65, 68, 84, 146–47, 203, 206, 206t, 228–30, 229t, 282–83, 283t, 296, 297, 305, 311, 316, 339–41, 349, 354–55, 357, 357t, 385, 439–40, 440t, 476, 478–79, 488, 509–10, 514, 515t, 552, 573, 582 Brexit and 64–65 deportation 110, 126n.20, 352–53, 403–4, 477–78, 509, 511, 527, 536 derogation 165, 167, 273, 274, 275, 279, 319, 328, 481, 503 optional 8, 138, 479–80, 481, 486 detention 539–46 absconding, risk of 541–42 alternatives and length 544–45 asylum seekers, criteria for 542–43 conditions of 545–46 EU legislation, delimitation of 539–40 grounds for 540–43

Index 641 human rights 511–12, 545 length 544–45 pre-removal 268–69, 510 returnees, grounds for 543 see also liberty deterrence 332 dictatorship 96–97, 527, 583 differentiated integration see opt-outs digitalisation 159–60, 225, 236, 240, 241, 298–99 see also algorithm-based processing; Artificial Intelligence; biometrics; databases dignity, human 167–68, 352, 413–14 see also refoulement, prohibition of diplomacy 141, 584 diplomas 436–37, 462, 482, 503–4 direct effect 155–56 directives see legislation Directorates General (DGs) 34, 49–50, 53 disabilities 398–99 discrimination 106–7, 111, 246, 263, 477, 493–94 ethnic 140, 266–67 guarantees against 263 indirect 239 migration status, on grounds of 265–66 nationality, on grounds of 482–83, 556 racial 266–67, 278–79 scrutiny levels in judicial practice C10S11 see also equal treatment; equality provisions discursive framing 32, 114–15, 520 disembarkation 213, 232–33, 305, 308, 332, 333, 335, 385, 577 see also search and rescue displacement 243–44, 398 forced 17–18, 400–1 mass 17, 338, 343 see also refugees Djibouti 591–92 DNA profiles 231 doctoral programmes 454–55 doctrinal foundations see Case law documents access to, as a source of public debates 94 authenticity of 254 security 334–35 veracity of travel documents 289, 324, 326, 460 see also data protection domino effect 38, 430 drug trafficking 260, 528 Dublin II Regulation 53–54, 164, 341–42, 360, 366, 368 Dublin III Regulation 7, 47–48, 65, 77, 125, 138, 152–53, 156, 157, 164, 168, 169, 171, 183–84, 189–91, 214, 244–45, 276, 325, 342, 356, 357–72, 357t, 373–75, 376–77, 409, 424, 447, 522, 533, 534–35, 536, 540, 561 see also asylum jurisdiction Dublin IV Regulation 157–58, 361–62, 370–71 Dublin Convention 13, 27, 232, 341, 357–58, 359, 368, 582 Dublin Regulation 253

Dublin system 50–51, 64–65, 67, 87–88, 90, 136, 148–49, 161–62, 165, 166, 168, 169, 177, 184, 221–22, 232, 316, 318–19, 343–44, 345, 347, 348–50, 352, 354, 357–72, 382, 386–87, 405, 409, 416, 417–18, 424, 550, 551, 557–58, 568–69 dynamism, judicial 130–32, 140–41 e-gates 224, 225, 236, 281, 325–26 contactless travel 281 self-service terminals 325–26 smart borders 225–26, 242 Eastern partnership 566–67 economic activities see labour migration; temporary economic activities Ecuador 284, 584 effet utile 75, 151, 158, 162–63, 167, 171, 235, 256, 272, 321, 365, 379, 403–4, 407, 446, 491, 519–20, 534–35, 539–40 usefulness of 163 effective remedy see remedies effectiveness, principle as a corrective vehicle 180–82 Court practice on migration law 181 time limits as a test case 182 see also national procedural autonomy Egypt 169, 465–66, 587–88, 589, 594–95 elections European Parliament 46, 55, 56, 345 municipal 559 TFEU 57, 436–37 Elgafaji formula 86, 351–52, 401–3 emigration country of 18 demographic stability and 470 mass 15, 21, 40, 100, 449 history of 14–16 rates of 101, 590 employer sanctions 514–15, 522–24 employment, illegal 506, 522–23 enhanced cooperation see opt-outs enlargement of the EU 19, 125, 200–1, 508–9 pre-accession treaties 564 entrepreneurship 453, 589, 593 entry bans inter-state consultation, failure of 189–91 irregular presence and return 524–26 refusal for different reasons 524–25 Return Directive 525–26 entry conditions 324–25 see also border controls, external Entry/Exit System (EES) 7, 178, 224, 225, 236, 249, 281, 324, 325–27, 336, 558 environmental protection 87n.86 equal treatment as a constitutional guarantee 262–67 ethnic discrimination 266–67 human rights 262–63 limits of 415–16 racial discrimination 266–67, 278–79

642 Index equal treatment (cont.) scrutiny levels in judicial practice 264–65 Union citizenship 262–63 see also discrimination, equality provisions equality provisions 479–87 disparate effects of 484–86 complexities 479–80 doctrinal characteristics 481–82 disparate effects 484–86 refugees v subsidiary protection 486–87 social security benefits 482–84 social assistance 482–84 human rights compliance 484 see also discrimination, equal treatment Erasmus+ programme 465, 561 Erdoğan, President 32, 346, 577, 580, 592 Eritrea 96–97, 103 Estonia 177, 228, 330, 441 estoppel 275–76 Ethiopia 591–92, 594–95 ethnic discrimination 140, 266–67 ethno-culturalism 16n.15, 18, 472, 473, 489, 501–2, 503 ethnographic research 109, 120 ETIAS see European Travel Information and Authorisation System (ETIAS) EU-LISA 227, 228, 238, 240–41, 247 EUR-Lex portal 159–60, 169–70, 233, 276–77 Eurodac Regulation (2013) 65, 229t, 232–33, 357, 370 Eurojust 204, 226 Euromed Agreements 567–68 European Arrest Warrant (EAW) 189–90, 355 European Asylum Support Office (EASO) 200, 201, 202, 203–4, 212–13, 217–18 Case Law Database 212 European Border and Coast Guard (EBCG) see Frontex European Border Surveillance System (EUROSUR) 210–11, 225–26, 230, 244, 586 European Coal and Steel Community (ECSC) 20, 42, 173 European Convention on Human Rights (ECHR) function 130–34 general principles 132–33 Grand Chamber 132–33 interaction with EU law 133–34 judicial dynamism 130–32 see also Charter of Fundamental Rights of the EU; European Court of Human Rights (ECtHR) European Council policy output 51–52 political programming 29–31 European Council on Refugees and Exiles (ECRE) 62 European Court of Auditors (ECA) 92–93 European Court of Human Rights (ECtHR) 2, 22, 69–70, 105, 126, 146, 148–49, 151–52, 183–84, 243, 263, 282, 306, 340, 429, 474–75, 511, 582–83

European Court of Justice see Court of Justice of the European Union (CJEU) European Data Protection Supervisor (EDPS) 231, 248, 249 European Database of Asylum Law (EDAL) 212 European Development Fund (EDF) 193n.113 European Economic Area (EEA) 173–74, 551, 555–59 free movement of persons 555–56 membership in Schengen and Dublin 557–58 micro states 558–59 European Economic Community (EEC) 20, 26–27, 358, 561 European Free Trade Association (EFTA) 555, 556 European Fund for Sustainable Development Plus (EFSD+) 591 European Migration Network (EMN) 174n.5, 178 European Neighbourhood Policy (ENP) 566, 592 European Parliament 54–56 co-decision powers 55, 68, 93 committees 55 European Parliamentary Research Service (EPRS) 344n.37 rapporteurs 55, 141, 149 shifting positions on migration law 55–56 European Travel Information and Authorisation System (ETIAS) 7, 58–59, 178, 223, 224–26, 229t, 230, 234–38, 240, 241–42, 243–44, 245–46, 247, 249, 280, 560 European Union Asylum Agency (EUAA) see Asylum Agency European Union Emergency Trust Fund (EUTF) 193, 591 Europeanisation ‘added value’ of 400–1 constraining dissensus 46–47 driving forces behind 42–49 institutional blockage 47–49 limits of 433–34 permissive consensus 46–47 venue shopping 25, 44–46, 63, 67, 108, 582 Europol 198, 200, 204, 213, 226, 233, 516 Euroscepticism 23–24, 65 Eurostat (Statistical Office of the European Union) 360n.145, 427n.1, 495–82nn.179–80, 498n.204 EUROSUR see European Border Surveillance System Eurotunnel 299 ex nunc assessment 180, 183–84, 188, 384, 398–99 exclusion from refugee status 398–400 Palestinian refugees 398–99 serious crime 399–400 terrorism 399–400 Executive Committee (ExCOM) 143–44, 145, 283–84, 288–89, 318 exiles 17, 62 exploitation 21, 109–10, 297, 435, 463, 464, 467–68, 577 expropriation 35n.141, 124 expulsion collective, prohibition of 310–11 decisions, mutual recognition of 534–35

Index 643 externalisation 3–4, 294–95, 386, 581–88 shiprider model 586–87 see also border controls, external; cooperation, international extradition 117, 353, 355, 539–40, 556 extraterritoriality 306–8 Facebook 248 fairness 30–31, 35, 244, 353 family bonds, as a critical gateway 429–30 life, right to 438–39 members, persecution of 392 nuclear 365, 392, 438, 441, 443–44, 447, 478 siblings 258, 363, 371, 443–44, 447–48 see also children; family reunification; marriage; spouses family reunification 440–48 beneficiaries of international protection 447–48 conditions under the Directive 445 EU citizens and their family members 440–43 procedure and rights of family members 445–46 scope of Family Reunification Directive 443–44 nuclear family, limitation to 443–44 optional clauses 444 feasibility studies 200–1, 215, 385, 594–95 federalisation 197, 198, 221, 293 federalism 83 executive 191–92, 195, 198, 221 fees application 290 diplomats 584–85 residence permits 256–57 tuition 555 feminism 112, 119 see also gender; women fiction of non-entry 380–81, 414–15, 425–26, 539–40 financial accountability 92–93 financial remittances 18–19, 98, 108, 571–72, 576, 589 financial support 590–91 fingerprinting 224, 225, 226, 230, 231, 232, 233, 234, 236, 237–38, 240, 242, 244–45, 248–49, 290, 292, 293, 324, 334 Finland 38, 60, 177, 190, 313, 484, 557 Finnish-Russian land border 237 firewalls 506–7, 514 flanking measures see Schengen area force majeure 344–45, 366 forced migration 102–3 see also refugees forgery 292 anti-forgery 334 formalism legal 117 Nordic 151 see also positivism, legal fortress Europe 31–33, 41, 44, 109, 207, 241, 301, 420 common values and 31–33 Foucault, Michel 110, 241

France 18–19, 21, 24, 29, 43, 83, 101–2, 151, 296, 299, 312–14, 318, 322–23, 335–36, 339–40, 342, 358–59, 363–64, 368, 428, 441, 470–71, 488, 495–96, 534, 535, 537, 556, 558–59, 561, 567–68, 574, 589, 593–94 fraud 188, 195, 334, 446, 504 acquisition 498 anti-fraud bodies 90, 92, 217–18 identity 246 free movement 20n.37, 127–28, 414, 482–83, 554 colonial migration and 20–22 EEA 555–56 single market 20 third country nationals 456–58 see also Union citizenship friendship treaties 19, 123, 124, 583 Frontex accountability 214–21 CJEU architecture 72–73 control practices, externalisation of 587–88 equipment, procurement of 210–11 Eurosur and 210–11 institutional formation 207–11 Irini mission 343, 583, 588 operational powers 207–8 policy design 200–1 Sophia mission 343, 588 supervision of the Member States 208–9 theory 200–1 third states, cooperation with 209–10 Triton mission 207, 588 see also agencies; border controls, external; ‘pushbacks’ functionalism neo-functionalists 44 fundamental rights see human rights Gambia 584–85, 589 game theory 39n.162 two-level games 44–45, 347 gang violence 394 GATS see General Agreement on Trade in Services (GATS) GDP (Gross Domestic Product) 359–60, 576, 589–90 GDPR (General Data Protection Regulation) 243– 44, 247–48 gender equality 490 identity 396 specific considerations 147, 342, 390, 425 see also feminism, women General Agreement on Tariffs and Trade (GATT) 143–44 General Agreement on Trade in Services (GATS) temporary economic activities 460–61 general principles abusive practices 275–76 equal treatment 262–67 proportionality 267–70 see also Case law

644 Index Geneva Convention see Refugee Convention gentlemen’s agreements 580 geopolitics 311–12, 338, 345–47, 568 Georgia (Eur.) 38, 285–86, 566, 567, 576–77, 582–83, 593–94 Germany 18–19, 20–21, 24, 43, 54, 65, 83, 87–88, 89, 100, 117, 151, 188–89, 191–92, 200–1, 274, 296, 297, 304, 312–14, 322–23, 339–41, 342, 358–59, 366, 367–68, 387, 388, 428, 431–32, 433, 450–51, 456, 464, 465–66, 470–71, 488, 495–96, 517, 524, 535, 546, 556, 557–58, 564–65, 573, 574, 582, 594 Großraum (greater area) 173–74, 303–4 Nazi 17 Ghaddafi, Muammar 574, 583 Gibraltar 64–65, 303, 500–1, 555, 568–69 micro-states and 558–59 Global Commission on International Migration (GCIM) 571–72 Global Compact for Migration 142–43 Global Compact on Refugees 142–43 globalisation 30, 44, 46–47, 100, 104, 108, 130–31, 197, 502 labour migration and 431–32 post-cold war 460 golden passport schemes 191, 285, 498–99, 502–3, 504 golden visas 187, 195–96, 276, 497, 502–3 abusive practice 191 good administration 185–87 counselling services 276, 530–31 giving reasons 185 legal assistance 184–85 maladministration 91 procedural deficits, consequences of 186–87 right to be heard 185 time limits 157, 182 see also national procedural autonomy Google 237–38, 248 governance institutional 217–18 labels as a means of 111–12 law as instruments of 110–15 managerial skills 202–3, 239–40 money as a means of 191–95 multiple actors 107–8 Greece 36, 53–54, 85, 92–93, 95, 100–1, 103, 169, 193, 213, 217, 281, 313, 320–21, 330, 331–32, 360–61, 363, 365, 368, 405–6, 428, 561, 564–65, 574–75, 589–90 guest workers see temporary economic activities habeas corpus 511 Habsburg Empire 14–15, 16–17 hardship clauses 8–9, 39, 165, 430, 445, 467, 487–88, 490, 492 harmonisation procedural 179–80, see also legislation headscarves 471

health 428, 451–52, 458, 483, 492, 567–68, 581–82, 590–91 care 264–65, 384, 408, 412, 413–14, 425–26, 465–66, 493, 513 Covid see Covid-19 pandemic deterioration in 15, 352–53 hospitals 434, 514 hygiene, personal 352, 413 insurance 560–61 medical treatment 281 medicines 215–16 mental 411–12 public 235, 319, 324–25 refoulement, prohibition of 352–53, 392–93, 403 screening 374 services 506–7 systems 392–93, 472, 503 workplace 485 heard, right to be 185, 529–30 hegemony 139, 159 hermeneutics see positivism, legal holidays 281, 429–30 Holocaust 506 homeland security 224 homosexuality see sexual orientation ‘hotspot’ approach 92–93, 203, 214, 221–22 housing basic need, as a 532 equal treatment 479–80, 483–84 EU programmes 428 facilities 410 reception conditions 411 social 264–65, 416 structural integration 472, 476 human rights data protection 244–46 detention, in conformity with 511–12 divergence from 403 effective remedy 182–85 equal treatment and 262–63 expulsion, protection against 477–78 extraterritorial reach of 306–8 family life 438–39 individual guarantees for ‘everyone’ 127–28 individual rights of migrants compared 164–65 interpretation 166 international treaty bodies 139–40 irregular presence 513–14 legal migration and 438–39 legislation 35–37 persecution and 390–91 qua personhood 127–30 religion, freedom of 391 Refugee Convention, and 147 ‘right to have rights’ 128 social security, compliance with 484 special rapporteurs 141 state sovereignty and 122–49 violation of basic 390–91

Index 645 see also Charter of Fundamental Rights of the EU; European Convention on Human Rights (ECHR); individual rights of migrants humanitarian admission 294–95, 420, 421, 423 see also humanitarian visas; sponsorship, community schemes humanitarian assistance 112, 193, 308–9, 343, 405, 514–17, 548, 579–80, 590–91, 593 humanitarian grounds 284, 294, 318, 325, 349–50, 404, 526 humanitarian law 144n.119, 402 humanitarianism 294 humanitarian visas 21, 74, 77, 131, 280, 282, 288, 294–95, 296, 300, 306, 307–8, 350, 420 see also humanitarian admission; resettlement; sponsorship, community schemes humanity, crimes against 397 Hungary 32, 50, 54, 71–72, 80, 82–83, 84, 188–89, 274, 310, 315, 329, 330, 361, 375 Iceland 9, 74, 282–83, 311, 357, 514, 550–51, 555, 556–57, 568–69 ID cards 65, 230, 256, 334 idealism 33, 44 illness see health Covid see Covid-19 pandemic ill-treatment see refoulement, prohibition of illegal entry see irregular entry illegal stay see irregular migration immigrants, immigration non-immigrants v 474–75 residents and 492 students as ‘ideal’ 453–54 terminology, choice of 3–4 immigration code 153–54 immunity 219, 293, 355 impartiality 375 imperialism 14–16 see also colonial migration implementing acts 58–59 imprisonment 367, 391, 518–19, 562 indeterminacy, of legislation 160–62, 392 India 441–42, 461, 594–95 skilled workers 15, 96–97, 449, 457 subcontinent 338 individual rights of migrants 163–69 court practice, ambiguity of 166–68 human rights compared 164–65 interpretation 166 legislation on 164 rights-based approach 114–15 see also human rights individualisation 402 Indo-Chinese boat people 420 industrialised countries 452–53 informalisation 538, 570, 580–81 side-effects of 47–48 information technology (IT) see also Artificial Intelligence; databases

infringement proceedings see Court of Justice of the EU (CJEU) inhuman treatment see refoulement, prohibition of inspections 464, 517–18, 548 institutions, EU 25–26, 42–68, 576 compromise-building 53–54, 55, 57, 67, 68 negotiations, interinstitutional 54–55, 56, 162, 232, 346, 372–73 overview 67–68 policy output, influence on 49–56 see also Commission; Council of the EU; European Council; European Parliament; Europeanisation; law-making; interinstitutional practices; trilogues Instrument for Border Management and Visa (BMVI) 194, 228 Instrument for Pre-Accession Assistance (IPA III) 193–94 instrumentalisation of migration 208, 279, 331, 345, 346, 350–51, 374, 381, 413–14, 521, 575 Integrated Border Management Fund (IBMF) 194 integration 469–504 citizenship 470–71 constitutional foundations 476–79 disputes 473–74 domestic laws, significance of 488–89 equality and see equality provisions human rights protection: against expulsion 477–78 integrationist approach 569, 578–79 language requirements/tests 487–92 legislation, secondary 479–87 mainstreaming 192, 493 mandatory requirements 487–92 measures 492–94 mixed output in migration law 475–76 models of migrant integration 472–73 overview 503–4 policy design 470–76 status change 474–75 supranational competences, scope of 477 supranational legal framework 489–90 territorial scope 478–79 theory 470–76 see also Long-Term Residents Directive; nationality; social security intelligence services 224, 237–38 interdisciplinary perspectives 96–121 academic discourse 115–16 agency, migrant 109–10 ‘control gap’, as a liberal paradox 105–6 ‘crisis’, language of 112–13 critical approaches 118–19 de-migranticisation of research 493 doctrine and critique 115–19 forced v voluntary migration 102–3 infrastructures 101–2 intergroup threat theory 106–7 labels as a means of governance 111–12

646 Index interdisciplinary perspectives (cont.) language use 112–13 law as instruments of governance 110–15 migration as a process 99–101 migratory movements, drivers of 96–103 multiple actors of migration governance 107–8 networks 101–2 ‘normality’, language of 112–13 normative background 104–5 open borders debate 104–5 overview 120–21 public opinion 106–7 ‘push’ and ‘pull’ factors 97–99 rights-based approach 114–15 securitisation 114–15 sociological institutionalism 53–54 state measures, significance of 103–10 see also constructivism; functionalism; realism intergovernmentalism 13, 28 intergroup threat theory 107, 120 internal border controls see border controls, internal interinstitutional agreement 202 International Civil Aviation Organization (ICAO) 225, 281 International Court of Justice (ICJ) 70, 140, 266 International Covenant on Civil and Political Rights (ICCPR) 139–40 International Covenant on Economic, Social and Cultural Rights (ICESCR) 307n.44 International Labour Organization (ILO) 17–18, 19, 40, 102, 482–83, 522 International Maritime Organisation (IMO) 144n.121, 308 International Organization for Migration (IOM) 102, 142 international protection see refugee status interoperability 236–42 automated processing 237–39 implementation deficits 239–41 quality concerns 239–41 spectre of 241–42 see also databases interpretative approximation 154–55 interviews 374–76 Inter-Corporate Transfers Directive 461–63 interinstitutional practices 56–63 delegated acts 58–59 implementing acts 58–59 international cooperation, informalisation of 60–61 private actors 61–63 soft law 59–60 treaty change, relevance of 56–57 see also trilogues inter-state cooperation 177–556, 557–58, 565 databases 188–89 entry bans 189–91 expulsion decisions, mutual recognition of 534–35 golden visas 191 inter-state consultation, failure of 189–91 intra-EU mobility 191

political cross-fertilisation 7, 152, 170–71, 253, 258 return, on 535–36 secondary legislation 188–89 transnational effects of 187–91 visa procedure 292–93 intra-corporate transfers (ICT) 8, 427, 432, 440t, 457–58, 462, 479–80, 489, 560, 565, 569 investment protection treaties 124 state 434 in technology 192 Iran 102–3, 297–98, 390–91, 392 Iraq 209, 298, 351–52, 395–96, 397, 420–21, 584–85 Ireland 28, 42, 63–64, 65–67, 68, 84, 167–68, 203, 206, 206t, 228–30, 229t, 255, 282–83, 283t, 305, 311, 312t, 312–13, 316, 317–18, 325–26, 335–36, 349, 354–55, 357, 357t, 376, 377, 439–40, 440t, 453, 476, 478–79, 479t, 509– 10, 514, 515t, 534–35, 552, 561, 564–65 Brexit and 65–67 Common Travel Area 65 Irini mission 343, 583, 588 Iron Curtain 44–45, 312–13, 345–46, 359, 459, 470, 582–83 irregular entry 144, 514–24 criminal sanctions, limits for 518–20 employer sanctions 522–24 policy 505–9 search and rescue, intimidation of 517–18 smuggling, prosecution of 515–17 trafficking, victims of 520–22 see also irregular migration; regularisation irregular migration 144, 505–49 binary conceptions of (il)legality 506–7 constitutional foundations 509–14 effectiveness 507–8 human rights and: detention 511–12 those irregularly present 513–14 overview 547–49 policy 505–9 dynamics 508–9 regularisation 546–47 supranational competences 510–11 territorial scope 514 theory 505–9 see also detention; entry bans; illegal entry; interstate cooperation; non-removable migrants; regularisation; Return Directive Islam 48 kafala system 443 Islamic State 234, 326 Israel 211, 398 Istanbul Convention 579 Italy 18–19, 25, 54, 83, 92–93, 101–2, 117, 127, 140, 151, 156, 200–1, 213, 217, 256, 296, 304, 306, 313– 14, 318, 331–32, 335–36, 339–40, 342–43, 358, 360–61, 362, 363–64, 366, 367–69, 387, 404, 423, 428, 435, 481, 517, 534–35, 537, 556, 558–59, 567–68, 574–75, 582–83, 589–90, 596

Index 647 ius sanguinis 16n.15, 501–2 ius soli 16n.15, 498n.204, 501–2 Jordan 398, 591–92, 593–94 EU-Jordan Compact 593 journalism 217–18 Juncker, Jean-Claude 32n.118, 50n.46 jurisdiction asylum see asylum jurisdiction Kenya 420 Kohl, Helmut 25 Kosovo 285–86, 297–98, 409, 565–66 Kurdi, Alan 113, 342–43 Kurdish region (Northern Iraq) 395 labour markets 432, 449, 456–57, 461 labour migration Blue Card Directive 450–53 constitutional foundations 435–40 demand-driven systems 448–49, 451, 453, 467 driving forces 448–50 entrepreneurship 453, 589, 593 Europeanisation, limits of 433–34 globalisation and 431–32 mobility, intra-European 456–58 national prerogatives for 437–38 points-based admission systems 111, 433–34 prehistory of EU rules on 18–19 regulation 448–50 research organisations 454–55 Single Permit Directive 455–56 skilled 448–58 Students and Researchers Directive 453–55 supranational competences 436–37 territorial scope 439–40 welfare state, effects on 434–35 see also diplomas; legal migration; temporary economic activities; qualifications, professional; start-up companies laissez passer 537 Lampedusa (Italian island) 96–97, 103, 314 land borders 329–30 see also border controls, external language ‘abuse’ 112–13 ‘crisis’ 112–13 lawyer-linguists 70 linguistics 159, 510, 527 multilingualism of EU migration law 158–59, 510 tests/requirements prior to departure 39, 86, 107–8, 165, 178, 267, 268, 269, 430, 445, 452–53, 475–76, 487–92, 503 transcription 240–41, 334 translation issues 70, 146–47, 153, 158–59, 292–93 language requirements/tests 39, 86, 107–8, 165, 178, 267, 268, 269, 430, 445, 452–53, 475–76, 487–92 Latin America 14–15, 124, 405

Latvia 38, 177, 218–19, 315, 441 Lausanne Treaty (1923) 18n.24 law-making 27, 30, 40, 45, 56–57, 61–62, 68, 75, 93, 126, 136–37, 153–54, 157, 160, 198, 349, 466–67, 519 see also institutions, EU; legislation; trilogues lawyers 69, 89, 116, 150, 232, 365–66, 375, 563–64 League of Nations 17, 40 Lebanon 294, 398, 591–92 legal assistance 184–85 counselling services 276, 530–31 legalisation see regularisation legal migration 427–68 child, rights of the 438–39 CJEU 87f constitutional foundations 435–40 Europeanisation, limits of 433–34 family bonds as a critical gateway 429–30 fitness check 433 globalisation and 431–32 human rights 438–39 legislation 440t national prerogatives 437–38 policy design 428–35 sectoral approach 432–33 simplification 153–54, 170, 366 overview 466–68 theory 428–35 see also family reunification; labour migration; temporary economic activities legal orders 58, 85, 155, 186–87, 552 as doctrinal self-sufficiency 150–52 legal remedies see remedies leges speciales 180–81, 244n.119, 247n.141, 263, 274, 362–63, 408n.514, 412, 416, 458n.203, 483–84, 495–96 Leggeri, Fabrice 217–18 legislation 253–79 agency involvement 215–17 casuistic law-making 153–54 complexities of 479–80 delegated acts 58–59 direct effect 155–56 directives 156–57 exceptional non-compliance (Art 72 TFEU) 273–75 EUR-Lex portal 159–60, 169–70, 233, 276–77 fitness check 433 fragmentation 152–53 equality provisions 479–87 informal trilogues, prevalence of 57–58 general features 253–79 governance, as instruments of 110–15 horizontal analysis 150, 253, 278 human rights and 35–37 implementing acts 58–59 indeterminacy 160–62, 392 individual rights of migrants 164 interpretation, pitfalls of 158–63 immigration code, vision of 153–54

648 Index legislation (cont.) more favourable domestic rules 270–72 overview 278–79 practical tips for dealing with 276–77 preparatory acts 219 preparatory bodies 49, 52, 53, 67–68 primacy 155–56 public policy exception 258–62 regulations 156–57 residence permits 254–57 subsidiarity as political preference 157–58 sufficient resources requirement 257–58 see also codification; law-making; recitals; soft law lex specialis 408, 437, 480n.68, 482, 528 liability, non-contractual 220–21, 222 liaison officers 209, 295, 300, 561, 575–76, 587 abroad 298–99 liberalism 104, 431 liberty 539 deprivation of 511, 512, 545 personal 530–31, 549 restrictions on 415, 507, 539–40 right to 511, 541 statue of 581–82 see also detention Libya 93, 96, 100–1, 193, 309–10, 331–32, 386, 421, 566, 568, 574, 576, 583, 587–88, 589–90 coastguard 211, 306, 308–9, 583, 586–87 Liechtenstein 282–83, 311, 357, 514, 550–51, 555, 556, 557, 558, 568–69 limbo see non-removable migrants; regularisation Lisbon Treaty 29, 40, 66, 270–71 autonomy of migration law 29 literacy 488, 492 Lithuania 32, 207, 214, 274–75, 298, 330, 331, 346, 354, 374, 586 lobbying 62 local border traffic 328 Long-Term Residents Directive 494–98 independence from original purpose 494–96 procedure and loss 497–98 scope and conditions 496–97 longue durée 13, 105 Lukashenko, Alexander 298 Maastricht Treaty 22–23, 24, 34–35, 38, 40, 44, 46–47, 64, 91, 273, 282, 283–84, 283t, 292, 340, 428–29, 430, 494–95, 499, 507, 573 intergovernmental cooperation 26–27 Macedonia 209–10 Maghreb countries 295, 567–68 Magna Carta 143, 149, 338 maladministration 91 Mali 193 Malta 54, 191, 331–32, 502–3, 574–75 EUREMA project 360–61 managerial accountability 90, 92–93 Mare Nostrum (Italian operation) 207, 343 Maritime Safety Committee 308 marketing 98

marriage 109 child 429–30, 444 of convenience 430, 446 dissolution of 441 forced 160–61, 429–30, 444, 489, 490 polygamous 443 see also family; spouses maternity 482 matrimony see marriage Mauritius 286, 288 media 13, 74–75, 90, 113, 208, 312–13, 508, 567–68 see also social media medical cases 352–53, 392–93, 403 medical treatment see health Mediterranean, Union for the 567–68 memoranda of understanding (MOU) 213, 385 memory 14–15, 187, 302, 405, 428 meritocracy 489, 547 Merkel, Angela 360, 582 Meroni doctrine 203, 205–6 methodology see interdisciplinary perspectives micro states 558–59 migrants, migration forced v voluntary migration 102–3 terminology, choice of 3–4 migration status contextual complexity of 589–90 development and migration 463–64, 591, 595, 596 distinctions based on 265–66 status change 474–75 military coup 100, 284–85, 346 missions 397, 400, 587–88 operations 425, 583 service 86, 397–98 takeover 392 minimum standards as inter-state obligations 123–24 minority treaties 18, 128 minors see unaccompanied minors mismanagement 92–93, 195, 203, 217 Mitterrand, François 25 mobile phones 102 mobility intra-EU 191 mobility partnerships 593–95 Moldova 96–97, 285–86, 293, 328, 434, 567, 587–88, 593–94 Monaco 553, 555, 558–59, 568–69 monetary union 29–30, 37, 40, 48n.33, 63, 115, 174n.4, 315 money laundering 502–3 as a means of governance 191–95 monitoring cross-border 200–1 efficient 210 mechanisms 92, 139 systems 191, 464, 531

Index 649 more favourable domestic rules 270–72 mandatory provisions, identification of 270–71 legislation, patterns in 271–72 Morocco 100–1, 193, 209, 297–98, 309–10, 330, 346, 436, 538, 550, 568, 586, 589–90, 591, 593–95 Multiannual Financial Framework (MFF) 191–92, 585 multilateralism 338 multilingualism 158–59, 510 multinational companies 62, 467–68 multiple speeds see opt-outs Muslims 107, 471 mutatis mutandis 322 mutilation see genital mutilation mutual dependency 154–55, 200–1 mutual interest 43, 51–52, 466, 567, 576–77, 583–84, 588 mutual learning 178, 201, 296 mutual recognition 174, 187, 418, 458, 507, 533, 535, 548, 555–56 of expulsion decisions 534–35 mutual trust 133, 177, 187, 191, 313–14, 337 limits of 368–70 national procedural autonomy limits of 179–87 harmonisation 179–80 see also effectiveness; good administration; remedies nationalism methodological 99, 100, 109, 110 nationality acquisition of 498–503 conceptual lacuna of EU migration law 499–500 Golden passport schemes 502–3 state prerogatives, limits of 500–2 Aliens Law 123–27 natural disasters 142–43, 393–94, 404 naturalisation 126, 442, 469, 470–71, 474–75, 478, 488, 498–99, 500–1, 502, 504, 562 negotiations see cooperation, international; diplomacy; institutions, EU neighbourhood Eastern partnership 566–67 Union for the Mediterranean 567–68 see also association agreements Neighbourhood, Development and International Cooperation Instrument (NDICI) 193–94, 591 neo-functionalism 44, 45 neo-liberalism 106, 111, 120, 431, 489 Netherlands 13, 21, 24, 29, 54, 83, 89, 96–97, 101–2, 117, 169, 189, 293, 296, 313, 322–23, 339–41, 342, 367–68, 378, 433, 470–71, 488, 490, 552, 567–68, 573, 589 New Pact on Migration and Asylum (2020) 30, 344–45, 361–62, 409, 572, 575 New Zealand 465 NGOs see non-governmental organisations (NGOs) nicknames 277

Nigeria 211–12, 438, 466n.258, 538, 589, 594–95 nationals 127, 190–91, 295, 296, 317–18, 387, 449, 466 non-arrival (non-entrée) 338, 339, 340, 348, 538 policies 337 non-compliance 7, 144–45, 255, 257, 284, 291, 350– 51, 360, 361, 522, 541 Article 72 TFEU 273–75 exceptional 273–75, 279 refusal of entry 324, 526 sanctions for 143, 322 non-contractual liability 220–21, 222 non-derogability 391 mandatory provisions, identification of 270–71 non-discrimination see discrimination non-entry, fiction of 380–81, 414–15, 425–26, 539–40 non-governmental organisations (NGOs) 45, 363, 375, 507–8, 583 fund-raising 516 third party intervention 73–74 non-refoulement obligations see refoulement, prohibition of non-removable migrants 450–51, 496, 506–7, 513, 526–27, 532 see also irregular migration; regularisation non-state actors 63, 147, 342, 351–52, 368, 390, 394–96, 425, 516–17 Nordic states 54 Norway 9, 74, 282–83, 311, 328, 357, 514, 550–51, 555, 556–58, 568–69 notification 288, 321–22, 366, 458 nuclear family see family nuclear option 60–61, 209, 221 nuisance 188, 234–35, 290 OAU Convention Governing the Specific Aspects of Refugee Problems in Africa 400–1 Odysseus Network v, 116 offences administrative 501 criminal 226, 230, 235, 477–78, 519 terrorist 226 oil crisis (1973) 428, 561–62 oil industry 96–97, 589–90 ombudspersons 5, 90 own-initiative inquiries 91–92, 217 as vehicles of administrative control 91–92 open borders debate 104–5 open method of cooperation (OMC) 62 oppression, political 99, 423 opt-outs, country-specific 28, 63–67, 228–30, 282– 83, 311, 436–37 Brexit 67 Denmark 64–65 Ireland 65–67 Orbán, Victor 32 ordre public see public policy exception Organisation for Economic Co-operation and Development (OECD) 19, 454, 465 ‘otherness’ of migrants 470–71

650 Index oversight see judicial oversight Oxfam International 591n.122 Pacific islands 285, 393–94 Pact on Immigration and Asylum (2008) 30 Pakistan 391, 594–95 Palermo Protocol 332, 333, 507–8, 514–15, 516, 548 Palestine ipso facto recognition 399 Palestinian refugees 398–99 UNRWA 157, 338, 398 pandemic see Covid-19 pandemic panopticon 110, 241 parallelism 133–34, 181, 256, 259, 322, 395–96, 400, 450, 525–26 emigration 14–16 imperialism 14–16 interpretative 553–54 tourism 14–16 parental responsibility 439 parliaments Parliament of the EU see European Parliament political oversight by 93 Passenger Name Record (PNR) 230 passports airport clearance 325–26 ‘apartheid’ 266 application procedures 290 asylum seekers and 232 biometrics and 230, 406 citizenship and 471 commodification of 502 Common Travel Area 65 data storage 236 derived rights and 37 expiry dates 235 golden schemes 191, 285, 498–99, 502–3, 504 ‘invention’ of 16–17, 304 Member State citizens 437, 442 Nansen passport 417 Nordic Passport Union 287, 358–59, 557 passport union, calls for a 23, 25, 302 readmission agreements 537 regulations 66, 312t, 334–35 scanned copies of 248–49 stamping of 236, 326–27, 336 statelessness and 126–27 visa policy 280 peace-building 567–68 penalties 82–83, 296, 300, 515, 518, 519, 523 pensions 486, 560–61, 562–63 permits residence see residence permits persecution 390–94 climate change and 393–94 family members 392 military service, refusal to perform 397–98 poverty and 392–93 punishment, disproportionate or discriminatory 391, 397–98

reasons for 396–98 refugees sur place 392 social group membership 396–97 violation of basic human rights 390–91 personhood, legal 112–13, 523 philosophy 112 political 103–4 social 509 photographs 283t, 290, 292 Plaumann formula 79–80 pluralism, legal 140 points-based admission systems 111, 433–34 Poland 18–19, 32, 38, 54, 84, 127, 155–56, 208, 274–75, 285–86, 298, 315, 328, 329, 330, 331, 346, 350–51, 374, 451, 464, 566–67, 586 police border checks 322–23, 371–72 political salience 47, 86, 115–16, 142 politicisation 5–6, 43, 46–47, 48–49, 50–51, 56, 63, 67, 68, 69–70, 78, 89, 107, 120, 128, 130–31, 132, 424, 428–29 depoliticisation 128 political salience 47, 86, 115–16, 142 populism 56, 67 Portugal 98, 317, 556, 564–65 Poseidon operation 207 positivism, legal 116 continental tradition 117 hermeneutics, legal 4, 5–6, 74–75, 77–78, 96, 115–17, 118, 119, 150, 155, 158, 171 methodology 115–19 see also hermeneutics posted workers 427, 434–35, 458, 460–63, 496, 560, 564, 573, 582 single market 459–60 post-colonial theory 119 post-structural theory 96, 110, 112, 120 pourvoi (appeals) 169–70 poverty 102–3, 109–10, 391 climate change and 391, 392–93 extreme material 352, 413 pragmatism, technocratic 5, 50–51, 67–68, 431–32 pre-arrival measures 26, 280, 295–99, 304 effectiveness of 297 pre-departure integration requirements 299 language tests 39, 86, 107–8, 165, 178, 267, 268, 269, 430, 445, 452–53, 475–76, 487–88, 489–90, 491, 503 preliminary reference see Court of Justice of the European Union preparatory acts 219 primus inter pares 51–52, 139 prisoners’ dilemma 43 see also game theory prisons 546 privacy see data protection procedural deficits consequences of 186–87 see also national procedural autonomy

Index 651 proportionality 267–70 abstract v individual assessment 269–70 context-specific outcomes 267–69 prosecution 507, 514–15, 521, 523–24 of smuggling 515–17 prostitution 520 protected entry procedure (PEP) see humanitarian admission; humanitarian visas; resettlement; sponsorship, community schemes protection see refugee status actors of protection 394–95 see also complementary protection; subsidiary protection; temporary protection Prüm Convention (2005) 535 psychiatric facilities 367 psychology 107 public opinion 106–7 public policy exception 258–62 criminal convictions 262, 267, 528 ordre public 258n.37, 273–74 sector-specific outcomes 260–62 Union citizenship, overlap with 259–60 see also security punishment see persecution ‘pushbacks’ 82–83, 88, 93, 203, 214, 217, 313–14, 330–31, 342–43, 346–47, 348–49, 374, 574, 580, 583, 586–87 qualifications, professional 8, 417, 436–37, 450, 451, 485, 503–4, 555–56, 566 qualified majority voting (QMV) 28, 46, 53–54, 361 qualitative methods 89, 97, 402–3, 478, 496, 531 quantitative methods 47, 78, 97, 230–31, 402–3, 478, 523 quotas/quota systems 350, 359–60 country-specific 15, 18, 436 financial frameworks 585 labour migration 437–38, 451–52, 463–64 micro states 558 recognition quotas 347, 381, 387–88, 425 relocation, of asylum seekers 82, 361 replacements for 361–62 seasonal and posted workers 564, 582 racial discrimination 266–67, 278–79 racial profiling 323, 327–28 racism 111, 120, 471, 591 railways 15–16, 102, 299, 323 rapporteurs 55, 141, 149 ‘shadow’ 58n.87 ratification procedures 20, 141 ratione loci 327, 461–62 ratione materiae 92, 138, 168, 332, 333–34, 357–58, 364, 380–81, 461–62, 526–27, 528, 540 ratione personae 125, 127–28, 246n.136, 407–8, 418, 440, 455–56, 461–62, 494, 496, 531 ratione temporis 545n.298 Rawls, John 104 readmission agreements between Member States 536–37

with third states 537–39 realism 33, 129–30, 345–46 realpolitik 50–51, 586 reasonable doubt 289 reasons, giving 185 reception conditions, asylum 410–15 cooking facilities 412 see also equality provisions; health; housing reciprocity 125, 156, 265–66, 283, 558 third states, relations with 287–88, 577 recitals 63–64, 162, 166, 410, 418, 489 referenda 29, 64–65 see also Brexit référendaires 116–17 refoulement, prohibition of 34, 122, 131, 134–35, 146, 209–10, 275, 305, 307–8, 309–10, 329, 333–34, 349, 351–54, 355, 368–69, 384, 399, 403, 418–19, 529, 532–33, 536 degrading treatment 131, 134–35, 166, 306, 403, 417 ill-treatment 220, 222, 294, 306, 351–52, 368–69, 389, 390, 396, 407–8, 521 inhuman treatment 36, 294, 351 medical cases 352–53, 392–93, 403 right to asylum and 309–10, 353–54 socio-economic deprivation 392, 393–94 torture 131, 134–35, 139–40, 141, 586–87 see also dignity, human; poverty Refugee Convention 143–47, 383, 384, 386–87, 447 human rights, practical priority of 147 interaction with Union law 143–45 supporting role in judicial practice 145–47 refugees history 17–18 Palestinian 398–99 protection in the Global South 591–93 sur place 392 without protection status 418–19 see also asylum applicant; Common European Asylum System (CEAS); displacement; reception conditions; refoulement, prohibition of; solidarity; Refugee Convention; UNHCR refugee status 386–400 Asylum Information Database (AIDA) 212 burden of proof 388–90 military service, refusal to perform 397–98 recognition quotas, disparate 387–88 status of 386–400 subsidiary protection v 486–87 see also asylum procedures; cessation; exclusion; refoulement, prohibition of; Refugee Convention; persecution; protection, actors of; safe countries; subsidiary protection; temporary protection refusal of entry 183, 184, 220, 249, 325, 327–28 see also entry bans Regional Development and Protection Programmes (RDPPs) 592 regularisation 4, 157, 313–14, 318, 349–50, 436, 505, 546–47

652 Index regulations see legislation religion 239, 390, 391, 396–97, 493–94 relocation of asylum seekers 48, 54, 80, 82, 155–56, 157–58, 261, 274, 356 remedies effective remedy: fundamental right to 182–85 human rights 183 judicial oversight, scope of 183–84 scope of the right to 168–69 suspensive effect 184–85 legal remedies: asylum procedures 384–85 border controls 327–28 composite procedures 218–21 refusal of entry 327–28 visas 291–92 removals 533, 536n.218 see also Return Directive remuneration 459–60, 462–63, 506–7, 513, 523, 565 research organisations delegation of responsibility to 454–55 researchers see Students and Researchers Directive resettlement 420–23 asylum policy, soft side of 420–21 EU framework 421–23 see also humanitarian admission; humanitarian visas; sponsorship, community schemes residence permits acquisition of 254–57 constitutive state authorisation 255–56 fee levels 256–57 permits 254 revocation 188, 284, 379, 399–400 short stay 282, 474 visas 254 residency requirements 269 Return Directive entry bans 525–26 exceptions from the scope 528 force, reasonable 530 illegal stay, reasons for 527 inter-state cooperation 533–39 joint return operations (JRO) 91n.101 non-removable migrants 532–33 prospect of removal 166, 532, 544, 546–47, 549 return decision 528–30 return to Member States 536–37 structure of 526–33 voluntary departure and removal 530–31 see also detention; irregular migration; regularisation; readmission agreements returnees 9, 163, 365, 505, 509, 522–23, 531, 535, 538, 542–43, 546, 580–81 additional grounds for 543 non-removable 532–33 reunification see family reunification revocation see residence permit right to be heard 185

right to have rights 128 rightlessness 128, 414–15 right-wing parties 48 rights-based approaches 110, 114–15 securitisation v 114–15 rights of migrants see individual rights of migrants Roma community 266, 287 Romania 38, 83, 84, 100, 229t, 240, 285–86, 287, 312t, 312–13, 315, 317–18, 328, 335–36, 355 Rome Statute of the International Criminal Court 397 Rome Treaty (1957) 19, 40 Russia 311–12, 328, 566 exiles 17 Finnish-Russian land border 237 nationals 423, 441, 502 oligarchs 502 Tsarist 16–17 war of aggression against Ukraine 60, 132, 289, 346–47, 397, 407, 464, 502, 566–67 Rwanda 385–86, 561 safe countries 381–84 common lists 381–82 controversies 383–84 of origin 381–82 rebuttal of the presumption of safety 382 Sahara 101–2, 587, 590 sub-Saharan states 586, 589–90, 591 Western Sahara 346 Sahel region 209, 393, 587–88 salary levels 451–52 salience, political 47, 86, 115–16, 142 same-sex relationships 443 San Marino 558–59 sanctions carrier 295–97 criminal 518–20 employer 514–15, 522–24 sanctuary practices 509 Scandinavia 14–15, 91, 420, 484–85 Schäuble, Wolfgang 582 Schengen area association 557–58 compensatory logic 24–26 flanking measures 4–5, 24, 26, 29, 34–35, 40, 224, 298, 304, 305, 322, 335 free travel for short stays 317–18 illegal state practice 320–22 internal border controls 319–22 internal movements 316–23 police checks in border areas 322–23 secondary movements 318–19 territorialisation of the EU 303–4 third country nationals, circulation of 317–19 see also border controls, external; border controls, internal Schengen governance 311–16 compensatory logic 24–26

Index 653 crises 313–15 Europeanisation 43–44 membership 312–13 phased implementation 312–13 reform efforts 313–15 Schengen acquis, scope of 315–16 territorial scope 312t Schengen Information System (SIS) 230–31 Schinas, Margaritis 33 Schuman Declaration 23 sea borders 331–34, 342–43, 586–87 basic procedural safeguards 333–34 coastal waters 305 coastguard 207, 211, 281, 306, 308–9, 583 Sea Borders Regulation 332–33 shiprider model 586–87 see also border controls, external; search and rescue Seahorse Network 586–87 search and rescue 308–9 deaths of migrants 207, 343, 557–58, 581–82, 583 intimidation of 517–18 shipwrecks 113, 343, 360–61, 574–75 see also disembarkation; sea borders seasonal workers 8, 100, 431, 443, 450–51, 479–80, 573 Seasonal Workers Directive 7, 433, 440t, 458, 463–64, 467–68, 486, 525–26 Seasonal Workers Directive (SWD) 7, 433, 440t, 458, 463–64, 467–68, 486, 525–26 secondary legislation see legislation secondary movements 318–19, 370–72 migratory movements, drivers of 96–103 securitisation 33, 34, 44–46, 67, 110, 114–15, 120, 224, 245, 507–8, 517, 572, 581, 582 rights-based approach v 114–15 see also control; criminalisation security document 334–35 safeguarding of internal 273–74 serious crime 399–400 threats 32–33, 114, 161–63, 188–89, 237, 249, 261, 275, 279, 321, 374, 378, 399–400, 506, 542–43 see also non-compliance, exceptional; public policy exception; terrorism sedentary bias 97, 99, 100, 110 self-employment 419 self-service terminals 325–26 Senegal 586–87, 589–90, 593–95 sensu stricto 114, 134, 267, 308–9, 547 Serbia 285–86, 298, 310, 330 settlement see integration sex 235, 239, 263, 265 same-sex couples 443 sexual abuse 260 sexual orientation 147, 391, 396, 425 Seychelles 286 Sharpston, Eleanor 70, 159 siblings 258, 363, 371, 443–44, 447–48 see also family

Sicily 311–12, 534–35 simplification of EU migration law 153–54, 170, 366 single market free movement 20 posted workers 459–60 Single Permit Directive 455–56 skilled labour see labour migration slavery 520 ‘sliding scale’ approach see Elgafaji formula slogans 31–33, 302, 459, 506, 592 Slovakia 36, 54, 80, 328, 361 Slovenia 220–21, 233, 317, 320–21, 330, 371–72 small and medium-sized companies (SMEs) 462 smart borders 225–26, 242 smartphones 225 smuggling of migrants 47, 109–10, 332, 333, 505, 507–8, 514–15, 520 prosecution of 515–17 snowball effect 38, 88 social benefits see equality provisions; social secondary Social Democrats (S&D) 55 social dumping 459 social media 98 see also Facebook; Twitter social security 9, 18–19, 127, 191, 264, 416, 434, 436–37, 459–60, 462–63, 478–79, 479t, 482–86, 503–4, 513, 523, 552–53, 554, 555–56, 559, 560–61, 562–63, 564, 565, 566, 567–68, 569 exclusion from social assistance 482–84 human rights compliance 484 means-testing 480, 503–4 pensions 486, 560–61, 562–63 public transport 482–83 sickness insurance 447, 451–52, 454, 497 see also equality provisions; welfare state socioeconomic rights 309, 384, 405 sociolegal research 5–6, 108, 509, 547–48 soft law 5, 56, 59, 68, 90, 149, 175–76, 324–25, 348– 49, 421–22, 426, 436, 469, 492, 580, 594 ancillary role of 59–60 solidarity measures 64–65, 356, 534–35 Article 80 TFEU 355–56 burden-sharing 359–60, 409, 420 see also mutual trust Somalia 146, 351–52, 395–96 Sophia mission 343, 588 South Sudan 593 sovereignty European 32, 44, 114 normative counterweight to 129–30, 148 state see state sovereignty Soviet Union 126, 488, 513–14 Spain 13, 54, 84, 208, 274–75, 304, 318, 329, 330, 331–32, 339–40, 342–43, 346, 363–64, 365– 66, 428, 495–96, 537, 547, 558–59, 560–61, 564–65, 574, 582–83, 586–87, 593–94, 596 special rapporteurs 141

654 Index specialised tribunals disadvantages of 72–73 specialists 98, 431, 451, 458, 462, 467–68, 576 spillover effects 13–14, 44, 432 sponsorships 361–62, 420, 447–48 community schemes 423, 547 see also humanitarian admission; humanitarian visas return 534–35 spouses 22, 48, 160–61, 363, 429–30, 438, 440–41, 442, 443, 444, 447, 452–53, 467, 477–78, 562, 563–64 Sri Lanka 189, 293 Stabilisation and Association Agreements (SAAs) 564, 565 Staff Working Documents (SWDs) 61–62 stakeholder consultations 61–62 stamps digital registration v 236 Entry/Exit System 281, 558 passports and travel documents 236, 326–27, 336 start-up companies 433–34, 449, 452 stasis 78, 99, 109 state sovereignty 18, 35–36, 42–43, 104–5, 109, 122–49, 291, 309, 415–16 diplomatic initiatives 141 dynamism, judicial 130–32, 140–41 Global Compact for Migration 142–43 human rights and 6, 122–49 treaty bodies 139–40 international bodies and political fora, role of 139–43 migration law 16–17 overview 148–49 see also aliens law; Charter of Fundamental Rights of the EU; Council of Europe; human rights; Refugee Convention; sovereignty; statelessness statelessness 123, 501 reduction of 126–27, 148 Statewatch 94 statistics 71f, 79, 87, 245–46, 281, 288–89, 339–40, 347, 362–63, 370, 384, 458, 495–96, 500 steamships 102 stereotypes 389–90 stigma 48–49, 410, 425–26 Stockholm Programme 30–31, 41, 51–52, 355, 493, 571 Strategic Committee for Immigration, Frontiers and Asylum (SCIFA) 52 strategic litigation 89 stratification 7–8, 225, 281, 290, 299–300, 325–26 Students and Researchers Directive 453–55 delegation of responsibility to research organisations 454–55 students as ‘ideal immigrants’ 453–54 subcontractors 239–40, 380, 411–12, 460, 523 subsidiarity 132–33, 150, 179–80, 227–28, 436, 457, 464–65, 477, 510, 523 as political preference 157–58

subsidiary protection 356–404 Europeanisation 400–1 human rights, divergence from 403 indiscriminate violence, protection against 401–3 national laws 403–4 Sudan see South Sudan sufficient resources requirement 257–58 supervision preventive 247–48 reactive 247–48 supranationalisation 13–14, 27–28, 43, 44, 45, 64, 303, 314–15, 360, 488 supranationalism 46, 155–56 Surinam 589 surveillance border 8, 58, 62–63, 138, 193, 207, 211, 221, 244, 301, 322, 325, 327–32, 336, 373, 525 of entries, exits, and migration statuses 236–37, 249 integrated systems 586 mass 223, 241 see also European Border Surveillance System (EUROSUR) suspensive effect 184–85 sustainable development 589, 591 Sweden 297, 339–40, 365–66, 420, 459–60, 467–68, 557, 582 Swiss Peoples’ Party 131–32 Switzerland 74, 189, 282–83, 293, 311, 339–40, 357, 375, 428, 514, 550–51, 568–69 association agreements 555–59 courts 189, 293, 555–56, 563–64 EEA and 555–58 symbolism 65 Syria 146, 211–12, 351–52, 398 civil war 393, 566, 579–80 EU Regional Trust Fund in Response to the Syrian Crisis 590–91 military service 86 nationals seeking protection 84–85, 100–1, 102– 3, 285, 288, 294–95, 297–98, 318, 360, 365, 380, 382, 397–98, 405–6, 421, 591–92 systems conceptual vagueness of 173–74 talent partnerships 593–95 Taliban 584, 590–91, 592 Tampere Conclusions (1999) 22, 30, 31, 35, 39, 400– 1, 431, 490, 494–95, 499, 507, 570, 571 taxation, double 125 technocracy 50–51 overreach 361 pragmatism 5, 50–51, 67–68, 431–32 technology border surveillance 221 ‘function creep’ and 225 information see information technology (IT) modern 110, 210–11, 223, 224, 236–37 police checks 322 security 193

Index 655 teenagers 447–48 telecommunication 173 telegraph services 15–16 teleological interpretation 6, 23–24, 76–77, 119, 158, 171 indeterminacy of 160–62 television broadcasts 187 telos 31, 78, 94–95, 118, 145–46, 160–61 temporary economic activities 458–66 GATS and trade agreements 460–61 guest workers (outside EU) 428–29 Inter-Corporate Transfers Directive 461–63 legal pathways for economic purposes 464–66 low-income earners 463 posted workers in the single market 459–60 Seasonal Workers Directive 463–64 see also exploitation; labour migration; subcontracting temporary protection 405–9 blanket recognition, rationale for 405–6 conditions and contents of 406–8 EU migration law, interaction with 408–9 temporary reintroduction internal border controls 319–20 terminals 281, 325–26 terminology 3–4 territorialisation of the EU 303–4 terrorism 114, 260, 320, 399–400 9/11 attacks 224 Basque terrorist suspects 354–55 bombings across Europe 225 cross-border cooperation 535n.216 fight against 30, 41, 230, 248–49, 507–8, 579 IRA 350n.73 student visas 259 Themis mission 207, 588 third country nationals circulation of 317–19 distinct policies for 24–33 ‘fortress Europe’ 31–33 free travel for short stays 317–18 political programming 29–31 Schengen and compensatory logic 24–26 secondary movements 318–19, 370–72 Treaty of Amsterdam 27–28 Treaty of Lisbon 29 Treaty of Maastricht 26–27 Union citizenship distinguished 33–39 legislative process 35–37, 39 overlap 37–38 primary law 34–35 third states capacity building 586–87 conditionality 583–86 constitutional foundations 577–81 CSDP missions 587–88 development cooperation 588–95 externalisation of control practices 581–88 financial support 590–91

foreign policy 575–76 Frontex, cooperation with 209–10, 587–88 informal cooperation frameworks, spread of 579–81 international cooperation with 570–96 legal pathways 588–95 migration-development nexus 589–90 migration management 571–72 ‘mobility’ and ‘talent’ partnerships 593–95 mutual interest 576–77 operational cooperation 586–87 overview 595–96 policy design 571–77 policy discourse 581–83 readmission agreements with 537–39 reciprocity in relations with 287–88 refugee protection in the Global South 591–93 supranational competences 578–79 theory 571–77 threat perceptions 106–7 see also intergroup threat theory time limits 157, 182 tort law 220 torture see refoulement, prohibition of totalitarianism 223, 303–4 tourism 14–16, 281, 285, 464, 467–68 parallelism of 14–16 trade agreements temporary economic activities 460–61 see also General Agreement on Trade in Services (GATS) Trade and Cooperation Agreement (TCA) 67, 288, 461, 551, 560, 569 trade unions 62, 68, 106–7, 459, 523 traffic see border controls, external trafficking child 234 drugs 260, 528 human 109–10, 367, 505, 510–11, 581–82, 588 victims of 514–15, 515t, 520–22, 523–24, 548 transit areas 327, 335, 415, 511, 512 transit countries 100–1, 120, 209, 339, 535 ‘safe’ 376–77 visa policy, export of 297–98 transit zones 82–83, 296, 307, 322, 363–64, 379–80, 511 transit visas, airport 285 transit zones 82–83, 296, 307, 322, 363–64, 379–80, 511 transitional periods 459, 564–65 transnationalism 99, 100 transparency 58, 59, 94, 244 transportation 15–16, 102, 296, 299, 305, 530–31, 537 aircrafts 210–11, 535–36 cargo ships 582–83 cruise ships 331–32 lorries 516, 563 public 482–83 railways 15–16, 102, 299, 323

656 Index transposition 71–72, 76, 81–82, 156, 157, 272, 433, 544–45 travaux préparatoires 160 travel free travel for short stays 317–18 see also tourism Travel authorisation (ETIAS) see European Travel Information and Authorisation System (ETIAS) treaty amendment/change 24–33 limited relevance of 56–57 Treaty of Amsterdam see Amsterdam Treaty Treaty on the Functioning of the European Union (TFEU) 7, 20–21, 72, 102, 124, 155, 173, 227, 253, 282, 303, 337, 435, 474, 505, 551 Treaty of Lisbon see Lisbon Treaty Treaty of Maastricht see Maastricht Treaty tribunals, specialised 72–73 trilogues format 68, 202, 232, 411 informal 5, 55, 56, 57–58 ‘triple-win’ concept 572, 589 Triton mission 207, 588 Trump, Donald 302 tuition fees 555 Tunisia 465–66, 568, 574, 593–95 migrants 96–97, 103, 313–14, 318, 405–6 nationals 441 posted workers 459 Turkey 32, 38, 84–85, 100–1, 102–3, 124–25, 193, 207, 274–75, 284–86, 295, 297–98, 331, 380, 382, 383, 538, 550–51, 552, 570, 580, 589–92 association agreements 561–64 EU-Turkey Statement 77, 79–80, 213, 255, 343, 346, 382, 386, 405–6, 421, 574–75, 577, 580, 582, 583, 592 standstill as dynamism for first admission 563–64 Turkish nationals residing in Member States 561–63 Twitter 360 Uganda 593 Ukraine association agreement 552 border controls 311–12, 328 European Neighbourhood Policy (ENP) 566–67, 582–83 Russian war of aggression (2022) 60, 132, 193, 321 African students, effect on 378–79 golden passport schemes 502 internal displacement of Ukrainians 346–47 military service 397 subsidiary protection 407 temporary protection 405, 406, 409 travel restrictions for Russians 289 Ukraine Solidarity Trust Fund 193 visa requirements 285–86, 297–98 Schengen zone, access to 576–77 unaccompanied minors 27, 136, 162, 166, 169, 363, 365–66, 375–76, 416, 440, 447–48, 512, 532–33 unemployment 431–32, 452, 456, 458, 482, 495, 562

unilateral declarations 143–44 Union citizenship equal treatment 262–63 public policy exception 259–60 theoretical ambiguity of 22–24 third country nationals distinguished 33–39 legislative process 35–37, 39 overlap 37–38 primary law 34–35 Union for the Mediterranean 567–68 United Kingdom (UK) 9, 21, 26–27, 28, 63–64, 66, 67, 68, 84, 89, 109, 206t, 229t, 282–83, 283t, 288, 296, 312t, 316, 339–40, 357, 357t, 363–64, 385–86, 420, 431, 440t, 449, 461, 470–71, 479t, 488, 500–1, 515t, 550–51, 552–53, 554, 559–61, 564–65, 569, 577 Common Travel Area 65 EU withdrawal see Brexit (former) Union citizens, acquired rights of 559–60 sovereign control over future movements 560–61 see also Brexit; Trade and Cooperation Agreement (TCA); Withdrawal Agreement (WA) United Nations Agenda for Sustainable Development (2030) 589 United Nations Convention on the Law of the Sea (UNCLOS) 308–9, 517–18 United Nations Convention against Torture (CAT) 139–40 see also refoulement, prohibition of United Nations General Assembly (UNGA) 142 United Nations High Commissioner for Refugees (UNHCR) 2, 74, 102, 103, 108, 141, 142, 143–44, 145–46, 147, 148, 217–18, 338–39, 349, 353, 375, 378, 386, 395–98, 399–400, 405, 415, 420, 422, 423, 511, 583, 590–92 United Nations International Children’s Emergency Fund (UNICEF) 590–91 United Nations Relief and Works Agency for Palestine Refugees in the Near East (UNRWA) 338, 398–99 United States (US) 9/11 terrorist attacks 224 academia 453 amicus curiae submissions 74 application fees 290 blacklisting 287 carrier sanctions 296 civil rights movement 472 compensation in case of expropriation 124 Congress 15, 54–55 ‘crimmigration’ 114 externalisation 581–82 extraterritorial controls 299 federal law 198 green cards 450 history of mass migration 14–15, 18, 100 homeland security 224 immigrants and non-immigrants, distinctions between 474–75 interdisciplinary approaches 118–19

Index 657 intra-European mobility compared 457 Iraq, military presence in 395, 397 literacy tests 488 outsourcing asylum applications 385 presidency 302 resettlement 420 strategic litigation 89 Supreme Court 69–70, 264, 307 transnational role of courts 146–47 travel authorisation, applications for 234–35 UNHCR budget 591–92 visa requirements 287 visa waiver practices 58–59, 287 Universal Declaration of Human Rights (UDHR) 128, 353 Universal Postal Union (UPU) 15–16n.10 universalism 104 unskilled workers 464, 522, 560 vagabonds 281 vagueness 86, 136 asylum, right to 353 Charter of EU rights 166–68 conceptual 173–74, 276 micromanagement and 268, 388 validity disputes 69, 79–80 Vanuatu, Republic of 285 Vatican City 553, 558–59 veil of ignorance 104 venue shopping 25, 44–46, 63, 67, 108, 582 verification procedures fingerprints 233 identity, as grounds for detention 542 Schengen evaluation procedures 313 visas 288 veto power 28, 53–54, 319–20, 360, 499 victimhood of children 520 persecution, fear of 521 residence permits 521 of women 520 victims of trafficking 520–22 Vienna Convention on the Law of Treaties (VCLT) 130n.38, 159–60 Vietnamese boatpeople 338–39 violence, indiscriminate armed conflict 147, 405, 409 Elgafaji formula 86, 425 protection against 401–3 safety and 351–52 Visa Information System (VIS) 233–34 visas admission criteria 288–90 airport transit 285 application procedure 290–91 ‘black’ and ‘white’ lists 283–85 carrier sanctions 295–97 constitutional foundations 282–83 databases and 188–89 decision 291–92

electronic applications 290 format 292 golden see golden visas humanitarian 35, 74, 77, 131, 280, 282, 288, 294–95, 296, 300, 306, 307–8, 350, 420 inter-state cooperation 292–93 issuance of 235n.66, 286n.36 legal remedies 291–92 liaison officers abroad 298–99 liberalisation 283, 284, 285–87, 290, 297–98, 300, 538, 569, 577, 583–84 multi-entry/multiple-entry 281, 286, 290, 584–85 overview 299–300 policy design 280–300 export to transit countries 297–98 pre-arrival measures 295–99 procedure 288–93 reciprocity and: third states, relations with 287–88 requirements 283–88 residence permits 254 short-term 88n.88 supranational competences 282 territorial scope 282–83 theory 281 visa-exempt travellers 235, 236, 249, 324 visa-facilitation agreements 235 visa-free travel 58–59, 235, 285–87, 297–98, 317, 502, 584–85, 596 see also European Travel Information and Authorisation System (ETIAS); Visa Information System (VIS) Visegrád countries 47–48, 54, 361, 472 vocational education/training 441, 450, 503–4 voluntary migration 17–18, 102–3, 111–12 voluntary return see assisted voluntary return (AVR) von der Leyen, Ursula 32, 113, 572 voting municipal elections 500, 559 qualified majority 28, 46, 53–54, 361 rights 557 rules 217–18 treaty law 436–37 veto power 28, 53–54, 319–20, 360, 499 vouchers 411–12 vulnerability assessments 209, 213, 315, 375–76, 412, 507–8 groups 45–46, 352–53, 375–76, 380, 411, 412, 420, 422, 512 of migrants 520, 545 minority groups 420–21 people/persons 342, 365, 378, 412, 416, 521, 539, 545 refugees 385 wage levels 20, 111, 120, 434–35, 448–49, 455, 459, 556 minimum wage 257 waivers fees for diplomats 584–85 visas 285, 286, 287

658 Index warlords 583 water as a basic need 413–14 territorial 309 usage 393 wave-through approach 343 weapons 216 welfare magnet hypothesis 98–99, 120 welfare state 8–9, 104–5, 106–7, 123–24, 416, 419, 425–26, 428, 464, 469, 472, 479, 483, 484, 486 legal migration, effect of 434–35 see also equality provisions; social security Western Balkans contemporary agreements 565–66 former agreements 564–65 pre-accession guarantees 564–66 transitional periods 564–65 Westphalian state system 18, 303–4 whitelisting 283–85, 287 see also visas Wiesel, Elie 506 women equal treatment of 266 genital mutilation 394 Istanbul Convention on violence against 579

violence against (Istanbul Convention) 579 see also feminism; gender; spouses World Tourism Organisation (UNWTO) 281 World Trade Organization (WTO) 460–62 see also General Agreement on Trade in Services (GATS) World War I 17, 280, 428 World War II 14, 17–18, 19, 21, 40, 105, 124–25, 147, 148, 346–47, 428, 470, 483–84, 499–500, 503 wrongdoing accountability for 72, 214–21 beyond courts 90–91 legal responsibility for 228 prevention of 217 serious 386–87 xenophobia 111, 488–89 Yazidis 420–21 Yemen 96–97 Yugoslavia, former dissolution of 126, 513–14 wars 359–60, 400–1, 405, 406, 530–31, 573 zero immigration 32, 431, 467, 469, 470–71