Deserving Citizenship : Citizenship Tests in Germany, the Netherlands and the United Kingdom [1 ed.] 9789004251076, 9789004251069

Citation preview

Deserving Citizenship

Immigration and Asylum Law and Policy in Europe Edited by

Jan Niessen Elspeth Guild

VOLUME 31

The titles published in this series are listed at brill.com/ialp

Deserving Citizenship Citizenship Tests in Germany, the Netherlands and the United Kingdom

By

Ricky van Oers

LEIDEN • BOSTON 2014

Library of Congress Cataloging-in-Publication Data Oers, Ricky van. Deserving citizenship : citizenship tests in Germany, the Netherlands and the United Kingdom / by Ricky van Oers. pages cm. -- (Immigration and asylum law and policy in Europe ; v. 31) Based on author’s thesis (doctoral)--Radboud University Nijmegen, 2013. Includes bibliographical references and index. ISBN 978-90-04-25106-9 (hardback : alk. paper) -- ISBN 978-90-04-25107-6 (e-book) 1. Naturalization--Germany. 2. Naturalization--Great Britain. 3. Naturalization--Netherlands. 4. Citizenship--Germany--Examinations, questions, etc. 5. Citizenship--Great Britain-Examinations, questions, etc. 6. Citizenship--Netherlands--Examinations, questions, etc. I. Title. KJC6044.O337 2013 323.6’23--dc23 2013029116

This publication has been typeset in the multilingual “Brill” typeface. With over 5,100 characters covering Latin, IPA, Greek, and Cyrillic, this typeface is especially suitable for use in the humanities. For more information, please see www.brill.com/brill-typeface. ISSN 1568-2749 ISBN 978-90-04-25106-9 (hardback) ISBN 978-90-04-25107-6 (e-book) Copyright 2014 by Koninklijke Brill NV, Leiden, The Netherlands. Koninklijke Brill NV incorporates the imprints Brill, Global Oriental, Hotei Publishing, IDC Publishers and Martinus Nijhoff Publishers. All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission from the publisher. Authorization to photocopy items for internal or personal use is granted by Koninklijke Brill NV provided that the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive, Suite 910, Danvers, MA 01923, USA. Fees are subject to change. This book is printed on acid-free paper.

Contents Acknowledgements�����������������������������������������������������������������������������������������������������������������xi 1. Introduction�������������������������������������������������������������������������������������������������������������������������1   1.1. Structure, Research Questions and Methodology��������������������������������������������3   1.2. Delineation of the Research�����������������������������������������������������������������������������������6   1.3. Existing Studies������������������������������������������������������������������������������������������������������� 10   1.4. Structure of the Book�������������������������������������������������������������������������������������������� 13 2. Justifying Citizenship Tests������������������������������������������������������������������������������������������� 15   2.1. Dimensions of Citizenship���������������������������������������������������������������������������������� 17    2.1.1. Citizenship as Legal Status (‘What You Get’)������������������������������������� 17    2.1.2. Citizenship as Activity (‘What You Owe’)�������������������������������������������� 18    2.1.3. Citizenship as Identity (‘What You Feel’)�������������������������������������������� 19   2.2. Theoretical Foundations for Conditions for Naturalisation����������������������� 19    2.2.1. Liberal Citizenship������������������������������������������������������������������������������������ 20    2.2.2. Republican and Communitarian Critiques of Liberalism�������������� 32       2.2.2.1. The Republican Account of Citizenship������������������������������ 33       2.2.2.2. Communitarian Citizenship��������������������������������������������������� 37   2.3. Conclusions�������������������������������������������������������������������������������������������������������������� 40 3. Repairing the Failures of the Past: The Introduction of a Citizenship Test in the Netherlands�������������������������������������������������������������������������������������������������� 41   3.1. From Informal Interview to Formalised Test: Debates on Dutch Nationality Law from 1985 to 2006������������������������������������������������������������������ 42    3.1.1. The 1985 Dutch Nationality Act and the Minorities’ Policy������������������������������������������������������������������������������������������������������������� 42    3.1.2. Revision of the Dutch Nationality Act������������������������������������������������� 45    3.1.3. From Naturalisation Test to Integration Examination�������������������� 51   3.2. Testing Integration as a Condition for Naturalisation���������������������������������� 60   3.3. Conclusions�������������������������������������������������������������������������������������������������������������� 64 4. ‘Deutschland sucht den Superstaatsbürger’: Integration Requirements in German Nationality Law������������������������������������������������������������������������������������������� 67   4.1. Testing Integration as a Condition for Naturalisation Prior to the 2000 Reform������������������������������������������������������������������������������������������������������������ 68   4.2. The Reform of 2000: Further Liberalisation?������������������������������������������������� 71

vi  Contents

  4.3. Amendments following the Zuwanderungsgesetz���������������������������������������� 79   4.4. Proposals to Change the Staatsangehörigkeitsgesetz Put Forward at the IMK of May 2006���������������������������������������������������������������������������������������� 81   4.5. The Government’s Proposals in the Richtlinienumsetzungsgesetz������������ 83   4.6. Debates about the Naturalisation Test������������������������������������������������������������� 87   4.7. Conclusions�������������������������������������������������������������������������������������������������������������� 88 5. On a Journey to Citizenship: Earning the Right to Full Membership in the UK���������������������������������������������������������������������������������������������������������������������������� 91   5.1. Integration Requirements in British Nationality Law���������������������������������� 92    5.1.1. Language Testing Prior to the Adoption of the Nationality, Immigration and Asylum Act����������������������������������������������������������������� 92    5.1.2. Towards More Demanding Ways of Testing Integration����������������� 93   5.2. The New Way of Integration Testing���������������������������������������������������������������100   5.3. Further Proposals to Restrict Access to British Citizenship���������������������103    5.3.1. The Path to Citizenship��������������������������������������������������������������������������104    5.3.2. The Borders, Citizenship and Immigration Bill�������������������������������106   5.4. Conclusions������������������������������������������������������������������������������������������������������������110 6. The Citizenship Tests Compared�������������������������������������������������������������������������������113   6.1. Qualifying the Tests����������������������������������������������������������������������������������������������115   6.2. Qualifying the Dutch Integration Examination�������������������������������������������119    6.2.1. Is the Dutch Citizenship Test Liberal?�����������������������������������������������122    6.2.2. Republican Elements in the Dutch Citizenship Test����������������������127    6.2.3. Communitarian Elements in the Dutch Citizenship Test������������128   6.3. Qualifying the German Einbürgerungstest����������������������������������������������������129    6.3.1. Can the Einbürgerungstest Be Qualified as Liberal?�����������������������130    6.3.2. Can the Einbürgerungstest Be Qualified as Republican?��������������135    6.3.3. Can the Einbürgerungstest Be Qualified as Communitarian?�����135   6.4. Qualifying the British ‘Life in the UK’ Test�����������������������������������������������������136    6.4.1. A Liberal Test?�������������������������������������������������������������������������������������������139    6.4.2. Republican Elements������������������������������������������������������������������������������140    6.4.3. Communitarian Elements���������������������������������������������������������������������140   6.5. Comparing the Tests��������������������������������������������������������������������������������������������142    6.5.1. What Is Tested?�����������������������������������������������������������������������������������������142    6.5.2. How Can the Tests Be Qualified?���������������������������������������������������������144 7. The Quantitative Impact of the Citizenship Tests in the Netherlands, Germany and the UK����������������������������������������������������������������������������������������������������147   7.1. Lower Number of Naturalisations?�����������������������������������������������������������������148    7.1.1. The Absolute Number of Naturalisations�����������������������������������������148

Contents  vii

   7.1.2. The Relative Number of Naturalisations�������������������������������������������154    7.1.3. The Relative Influence of the Tests on the Number of Naturalisations������������������������������������������������������������������������������������������157       7.1.3.1. The Netherlands�����������������������������������������������������������������������157       7.1.3.2. Germany�������������������������������������������������������������������������������������161       7.1.3.3. The UK����������������������������������������������������������������������������������������163    7.1.4. Other Ways of Acquiring Citizenship�������������������������������������������������165    7.1.5. Conclusions�����������������������������������������������������������������������������������������������168   7.2. Gender, Age, Nationality, Level of Education and Handicaps: Differences in Effects?�����������������������������������������������������������������������������������������168    7.2.1. Gender��������������������������������������������������������������������������������������������������������168    7.2.2. Age���������������������������������������������������������������������������������������������������������������171    7.2.3. Nationality�������������������������������������������������������������������������������������������������174    7.2.4. Level of Education�����������������������������������������������������������������������������������184    7.2.5. Medical Impediments and Illiteracy��������������������������������������������������185   7.3. Conclusions������������������������������������������������������������������������������������������������������������186 8. The Effects of Citizenship Testing: The Case of the Netherlands���������������������191   8.1. Increased Uniformity and Less Inequality?��������������������������������������������������192    8.1.1. New Problems�������������������������������������������������������������������������������������������193    8.1.2. New Inequalities��������������������������������������������������������������������������������������196    8.1.3. Conclusions�����������������������������������������������������������������������������������������������199   8.2. Improved Integration?����������������������������������������������������������������������������������������199    8.2.1. Immigrants for Whom the Test Constitutes a Redundant Requirement����������������������������������������������������������������������������������������������200    8.2.2. Persons for Whom the Test Constitutes a Considerable Barrier���������������������������������������������������������������������������������������������������������204    8.2.3. Test Content and Integration����������������������������������������������������������������211    8.2.4. Conclusions�����������������������������������������������������������������������������������������������215   8.3. Conclusions: Intended and Unintended Effects of the Formalised Language and Integration Requirement��������������������������������������������������������216    9. Language and Integration Testing in Practice: The Case of Germany����������219    9.1. A Uniform Application of the Naturalisation Requirements?��������������220     9.1.1. A Uniform Language Test?������������������������������������������������������������������220     9.1.2. A Uniform Knowledge of Society Test?�������������������������������������������224     9.1.3. Conclusions���������������������������������������������������������������������������������������������227    9.2. Improved Immigrant Integration?����������������������������������������������������������������228     9.2.1. Integrated, but no Proof����������������������������������������������������������������������229     9.2.2. Immigrants for Whom the Language and Integration Requirements Constitute Barriers����������������������������������������������������230

viii  Contents

    9.2.3. Test Content and Integration�������������������������������������������������������������236     9.2.4. Conclusions���������������������������������������������������������������������������������������������237    9.3. Other Possible Barriers to Citizenship���������������������������������������������������������237    9.4. Conclusions: Intended and Unintended Effects of the Language and Knowledge of Society Tests��������������������������������������������������238 10. The ‘Knowledge Requirements’ in Practice: The UK�����������������������������������������241    10.1. A More Consistent Application of the Language Requirement?��������242          10.1.1. Hassle with ESOL������������������������������������������������������������������������������243          10.1.2. Differences Between the Two Routes: Costs and Time���������������������������������������������������������������������������������������������246          10.1.3. Conclusions����������������������������������������������������������������������������������������247    10.2. An Improved Immigrant Integration and Increased Community Cohesion?����������������������������������������������������������������������������������247          10.2.1. The Test Route versus the Course Route������������������������������������248          10.2.2. The Value of What is Tested�����������������������������������������������������������249          10.2.3. The New Requirements as Barriers���������������������������������������������250          10.2.4. Conclusions����������������������������������������������������������������������������������������252    10.3. Other Barriers to Citizenship�����������������������������������������������������������������������253    10.4. Conclusions: Intended and Unintended Effects of the Knowledge Requirements�����������������������������������������������������������������255 11. Conclusions�������������������������������������������������������������������������������������������������������������������257    11.1. Why Have Tests been Introduced?��������������������������������������������������������������257          11.1.1. Changing Ideas on Immigrant Integration��������������������������������257          11.1.2. The Process of Path Dependency�������������������������������������������������258          11.1.3. Equal Treatment as a Justification for Raising the Requirements������������������������������������������������������������������������������������260          11.1.4. The Influence of EU Law and of Policies Applied in Other EU Member States����������������������������������������������������������������260          11.1.5. Reducing the Number of Naturalisations����������������������������������262          11.1.6. The Political System�������������������������������������������������������������������������263          11.1.7. Conclusions on Convergence��������������������������������������������������������264    11.2. The Effects of Citizenship Testing: Have the Goals Been Reached?������������������������������������������������������������������������������������������������������������265          11.2.1. Do the Tests Assess What They Are Meant to Assess?�����������266          11.2.2. An Improved Integration?��������������������������������������������������������������269    11.3. Can the Practice of Citizenship Testing be Justified from the Liberal, Republican or Communitarian Perspectives?��������������������������274

Contents  ix

Bibliography��������������������������������������������������������������������������������������������������������������������������279 Annex I    Methodological Justification of Qualitative Research and Overview of Interviewed Immigrants and Stakeholders�����������������������287 Annex II Questionnaire for the Research into the Effects of the Requirement ‘to have Sufficient Knowledge of Language and Life in the UK’ as a Condition for British Citizenship��������������������295 Index����������������������������������������������������������������������������������������������������������������������������������������301

Acknowledgements This book is the outcome of a PhD research which I publicly defended at the Radboud University Nijmegen on 1 March 2013. The text of the book has been updated since. Final cut off date is 1 June 2013. I would not have been able to complete the research without the help of the following people, to whom I owe immense gratitude: Promotor Kees Groenendijk and co-promotor Betty de Hart encouraged me to convert ‘De Naturalisatietoets Geslaagd?’ (Van Oers 2006) into a dissertation. Their continuous support, interest and enthusiasm provided an endless source of inspiration. Without them, this book would not have been written. I am furthermore thankful to all respondents who were willing to share their experiences. To conduct the research in Germany and the UK, I spent some time at the Institut für Migrationsforschung und Interkul­turelle Studien in Osnabrück and the Kent Law School in Canterbury. I  thank my German and British colleagues for their warm welcome and the academic and other support they provided. In this regard, I am especially grateful to Bernard Ryan, Sophie Vigneron, Anne Walter and Jutta Tiemeyer. I am furthermore indebted to Klaus Barwig of the Akademie der Diözese Rottenburg-Stuttgart, who intro­duced me to several key respondents and who provided a roof at the comfortable Tagungszentrum in Hohenheim. During the course of the research, several people provided helpful comments and suggestions, among whom my colleagues and former colleagues at the Centre for Migration Law, the Institute for Sociology of Law, and the Centre for the Administration of Justice of the Radboud University Nijmegen. Jürgen Bast and Bernard Ryan provided the necessary help on the German and British chapters. I thank Ines Michalowski for commenting on the chapter containing the content analysis of the citizenship tests. Berry Tholen provided helpful comments on draft versions of the chapter containing the theoretical framework. Jeremy Bierbach and Andrej Zaslove are probably unaware of the fact that their comments during a Wine & Discussion meeting in the fall of 2010 provided the missing links I did not even know I was looking for myself. ‘Non-academic’ support was provided during coffee breaks with Karina, Sophie and Anoeshka, as well as the members of the ‘read and dine’ club consisting of my cool colleagues from the constitutional and administrative law departments and the philosophy of law section: Janine van Dinther, Lonneke Peperkamp, Saskia Hillegers and Naima Qoubbane. I am also grateful for the pleasant conversations with, and support received from, fellow PhD researchers Sandra and Reyer.

xii  Acknowledgements

The computer skills of Carolus Grütters proved indispensable for the composition of the index and some of the tables in the book. I thank Hannie van de Put for providing all kinds of support in the years she was working as the secretary for the CMR and beyond. Outside the University, I would like to thank friends and family for respecting my policy of not discussing the content of my research and thereby providing the required moments of relaxation. On Mondays, I have always been able to count on the radiant company of my model-friends, who never tire of my tv addiction, despite, or maybe thanks to, its focus on channels 4 and up. On Fridays, Juni en Geer faithfully joined me in café Maxim. Being my ‘paranimfen’, I have also been able to lean on them during the public defence of this research. I thank Jackie for being born, thereby putting things into perspective and providing the necessary distraction. I thank my parents and parents-in-law who have been willing to watch Jackie in cases where too much distraction loomed. In this regard, a special word of thanks is devoted to Jeanne, who has in the past two years been willing to largely surpass the day a week she was supposed to babysit her granddaughter. I am grateful to my parents for allowing me to make my own choices, good and bad, and for providing the required emotional and financial support. I thank my sister Elske for mocking me at the right moments, and for the hilarious childhood (and years beyond) enabled by her presence. I now have my own ‘nuclear family’. Willem and Jackie, now this project is finished, I hope to be able to join you on your trips to the swimming pool and the zoo, in the future together with a new member.

Chapter One Introduction In the past decade, a number of Western European countries have formalised or introduced language and knowledge of society tests for immigrants apply­ing for citizenship, permanent residence or admittance (Van Oers, Ersbøll & Kostakopoulou 2010). In most cases, the language and knowledge of society requirements were initially only required for naturalisation, after which they made their way into other domains.1 In the area of naturalisation, the language and knowledge of society requirements have been formalised: whereas the integration of the applicants used to be tested in an informal interview by a local government official, their integration is now tested in formal examinations. This formalisation coincided with a considerable increase in the knowledge required. The formalisation and reinforcement of the integration tests at naturalisation and the introduction of such tests in the areas of permanent residence and admittance signals that policy makers and politicians have established a direct link between the acquisition of knowledge on the one hand and integration on the other. The idea that knowledge of language and society is a reflection of the degree to which an immigrant has integrated, has become commonly accepted (Bjornson 2009: 10). Despite their recent surge, knowledge tests are not a new phenomenon. In the area of naturalisation, informal integration tests have applied in European Member States for more than half a century (Guild, Groenendijk & Carrera 2009: 8). Digging deeper, one will find that language tests were applied in ‘classical’ immigration countries such as the USA, Canada, and Australia and the British Empire in the 19th century. These tests were often discriminatory in nature, seeking to keep unwanted immigrants from entering the country or to disenfranchise certain parts of the population. In the USA, for instance, between 1890 and 1910, literacy and comprehension tests were used to keep blacks and poor whites from voting (Steiner 2009: 117). Furthermore, the 1906 statute providing for English language skills as a requirement for naturalisation, must be seen in the context of the immigration restrictionists’ continuing efforts to use a literacy require­ment to decrease immigration from Eastern and Southern Europe (Neuman 1994: 263). 1 Before 2000, the only EU Member State which applied a language requirement for permanent residence was Germany (Guild, Groenendijk & Carrera 2009: 8).

2  Chapter One

Canada followed a similar pattern by enacting literacy requirements for naturalisation, designed specifically to exclude Chinese immigrants (Etzioni 2007: 354). In the British Empire, in the South African colony Natal, a language test was introduced in 1897 to exclude immigrants from the Indian subcontinent, who had profited from their imperial free movement rights. In a slightly modified form, the ‘Natal Dictation Test’ subsequently entered Australia’s 1901 Immigration Restriction Act, to stay there until 1957 as an element of the ‘White Australia’ policy (Bast 2011: 99).2 In Europe, the first use of language tests as an instrument for migration control occurred in Germany, which introduced a language test in 1997 to restrict the immigration of ethnic Germans from Eastern Europe, and to gradually terminate the privileged admission policy for Jewish immigrants from Russia (Bast 2011: 96, Groenendijk 2011a: 4).3 The above shows that two distinct functions are attributed to language and integration tests. On the one hand, proponents of such tests argue that the tests contribute to the integration of immigrants whose aim it is to move to the country or to settle there as permanent residents or fully-fledged members. The underlying assumption is that knowledge of the language and the society will enable these immigrants to function independently within the society, which will result in easier acceptance by the host society. Opponents of the tests, however, contend that the tests function as instruments for the selection of immigrants, pointing to the exclusionary effects of formalised tests on lower educated immigrants. While officially aiming at the improvement of the integration of the immigrants, opponents of the tests argue that the tests actually have the effect of excluding immigrants with lower skills, who are generally considered unwanted, as well as hampering the integration of the sponsors of those immigrants who intend to move to the country for the purpose of family reunification. The aim of this book is to assess the explicit and hidden goals the formalised language and knowledge of society tests are meant to achieve, as well as to analyse their intended and unintended effects, thereby focusing on citizenship tests, i.e. formalised language and knowledge of society tests as a requirement for naturalisation. To achieve this aim, the political debates preceding the introduction of the tests, as well as the effects produced by the tests will be analysed. Three countries will be addressed: Germany, the Netherlands, and the United Kingdom.

2 What was required was a 50-word dictation in a ‘European’ language, to be chosen by the immigration official. Together with the absence of control by a neutral authority, this measure was a way of pursuing a racist policy of exclusion (Bast 2011: 99). 3 Germany had previously introduced ‘basic’ and subsequently ‘sufficient’ language skills as a requirement for permanent residence in the 1978 residence regulation (Verfestigungsregelung; Michalowski 2010b: 188).

Introduction  3

1.1. Structure, Research Questions and Methodology The research addresses the issue of citizenship tests from the point of view of liberal, republican and communitarian theory. These theories offer different justifications for the exclusion of legal residents from membership. On the basis of these theories, different justifications for the application of requirements for naturalisation in general, and citizenship tests in particular, can be formulated.4 In Chapter 2, models for naturalisation based on the liberal, republican and communitarian theories will be presented, as well as an analysis of the question of whether a citizenship test can be justified in these models, and, if so, under what conditions. The main question this research aims to answer is 1) Can the citizenship tests in the countries under consideration be justified from the liberal, republican or communitarian perspectives? To answer this question, I will employ a strategy of simultaneously ‘zooming in’ and ‘zooming out’ on the citizenship tests.5 By ‘zooming out’ on citizenship tests, I intend to paint a panoramic picture, focusing on the context, timing, and the official discourses leading to the introduction of the tests on the one hand and the effects the tests have produced in the three countries under consideration on the other. ‘Zooming in’ on citizenship tests implies an analysis of the content of the tests which are presented to those applying for naturalisation. ‘Zooming out’ on citizenship tests leads to two distinct research questions: 2) Why have Germany, the UK and the Netherlands introduced formalised citizenship tests into their respective legislations? 3) What have been the effects produced by these citizenship tests? The first question will be considered on the basis of an analysis of the political debates leading up to the introduction of the citizenship tests and the general political and social context and timing in which these debates took place. Chapter 3 analyses the debates leading up to the introduction of the citizenship test in the Netherlands, Chapter 4 focuses on Germany, and Chapter 5 covers the debate in the United Kingdom. The questions used to examine the reasons why the countries under consideration introduced citizenship tests, employed to 4 It is possible to identify additional normative models. The liberal, republican and communitarian models will however suffice to reach the goal for which I intend to use them. 5 This indication of my way of approaching the issue has been derived from Kostakopoulou’s contribution to the discussion ‘How liberal are citizenship tests?’ on the EUDO-website, downloadable from http://eudo-citizenship.eu/docs/RSCAS_2010_41.pdf (Kostakopoulou 2010a: 15).

4  Chapter One

analyse the political debates, focus on the arguments which were put forward by the proponents as a justification for the introduction of the tests, and the arguments which were put forward against the introduction of such tests. Who favoured the introduction of the tests and who opposed them and why? How does support or opposition fit with other social and political agendas relating to immigration and integration? What do the advocates of such tests say that they are trying to accomplish? Is there evidence to indicate whether the advocates of the tests also intended to achieve other, hidden goals?6 The models for naturalisation based on the liberal, republican, and communitarian theory, as presented in Chapter 2, will be used as an instrument of analysis. In practice, just like the conceptions do not exist in a clearly distinguishable way in the naturalisation laws of individual countries, there will not be one indistinguishable conception that will be used for the introduction of a citizenship test (Kostakopoulou 2003: 94). Using the conceptions of citizenship as an instrument to analyse the debates, as represented in the official parliamentary records, governmental policy documents, literature and newspaper articles, will nevertheless allow me to qualify the arguments in order to identify where the centre of gravity lies in the argumentation of those who support the introduction of citizenship tests and those who oppose them. Arguments which cannot be categorised as belonging to a theoretical citizenship model defined in Chapter 2 will also be reproduced.7 As regards the time frame covered, the period over which the debates have been analysed differs per country researched. In all three countries, the researched period covers the year in which discussions regarding an amendment of the language and integration requirement started, until the actual introduction of the formalised citizenship test. This means that in the case of the Netherlands, the analysis starts in 1993, when a proposal to amend the 1985 Dutch Nationality Act was introduced, until 2007, in which year the Integration Act took effect. In Germany, the researched period runs from the year 1999, when a proposal to reform the nationality law was introduced, until 2007, when the Act for the implementation of eleven Directives on asylum and immigration was adopted. As regards the UK, the period described starts in 2001, when several official reports, urging for amendments to be made in the field of naturalisation policy, were published following the occurrence of riots in Northern England. It ends in 2009, in which year the Borders, Citizenship and Immigration Act was adopted. For all three countries the situation of language and integration testing prior to

6 These questions have been derived from Joseph Carens’ contribution to the discussion on the question ‘How liberal are citizenship tests’ on the EUDO citizenship forum (Carens 2010: 20). 7 This for instance applies to the argument that language and integration tests need to be introduced to keep up with developments in European law and other EU countries.

Introduction  5

the formalisation will be also described, and so will the period after the introduction of the formalised tests, in case of relevant new developments or debates. The second ‘zooming out’ question relates to the effects produced by the tests. This question will be considered in Chapter 7, where statistical information regarding the number of naturalisations in the countries under consideration and data regarding the tests themselves are analysed. Where possible, information has been collected regarding the gender, nationality, level of education and age of the naturalisation applicants and test candidates, to evaluate whether the tests have differently affected different groups of immigrants. The information analysed has been derived from available official statistical information, found on websites and in reports from the countries’ official statistical bureaus and com­ petent governmental agencies and ministries. These are the Central Bureau of Statistics and the Immigration and Naturalisation Service’s Centre for Information and Analysis (INDIAC) in the Netherlands, the Federal Office for Statistics (statistisches Bundesamt) and the Federal Office for Migration and Refugees in Germany (Bundesamt für Migration und Flüchtlinge), and the Home Office statistics and UK Border Agency in the UK. Where necessary, the information gathered was completed with data from EUROSTAT, and information published on the EUDOcitizenship website.8 Furthermore, statistical information found in newspaper articles, parliamentary documents and official evaluations has been used. The question regarding the effects of the tests will furthermore be answered using qualitative empirical research. In all three countries under consideration, semi-structured interviews were held with immigrants, as well as stakeholders: language teachers, staff members of immigrant organisations and municipal officials. Among the immigrants, three categories were interviewed: test candidates, naturalisation applicants, and immigrants who had refrained from applying for naturalisation. In Germany and the UK, interviews were also conducted with policy makers.9 In total, 213 interviews were conducted between January 2006 and April 2009.10 The majority of the interviews, 163 in total, were conducted with immigrants.11 The analysis of the empirical data gathered in the Netherlands will 8 http://epp.eurostat.ec.europa.eu/portal/page/portal/eurostat/home/, http://eudo-citizenship.eu/. 9 Policy makers have not been interviewed in the Netherlands. The reason is that the interviews held in the Netherlands formed part of an earlier research, conducted in 2006, into the reasons for introduction and effects of the naturalisation test (Van Oers 2006). In the light of the limited time which was available for this research, it was decided that policy makers would not be interviewed. Policy makers have however been interviewed within the framework of other research, the results of which will be involved in the analysis (see introduction to Chapter 8). 10 In the Netherlands, a total of 76 interviews were conducted, compared to 82 in Germany, and 55 in the UK. 11 In the Netherlands, a total of 57 immigrants were interviewed, compared to 69 in Germany, and 37 in the UK.

6  Chapter One

be presented in Chapter 8, Chapter 9 will cover the data found in Germany, and Chapter 10 will focus on the data retrieved in the UK.12 One of the aims of analysing the effects produced by the tests is to determine to what extent the goals the tests are supposed to serve, as expressed in the political debates, have actually been achieved. The analysis will furthermore allow us to examine whether other, unexpected, unintended, or latent, goals have been achieved. In the concluding chapter of this book, the question will be answered whether, and to what extent, the produced effects can be justified in the different citizenship models. Whether the effects produced are positive or negative will depend on the conceptions of citizenship that lay behind their introduction. ‘Zooming in’ on the citizenship tests leads to a fourth research question: 4) How can the content of the citizenship tests be qualified? As an instrument of analysis of the tests, the three citizenship models will be applied. This way, the analysis of the content of the tests contributes to answering the principal question of this research, and thereby the question regarding the liberal character of citizenship testing, which has been going on in the literature (Bauböck & Joppke 2010, Joppke 2010b, Michalowski 2011, Orgad 2010a). Further­ more, since implicit in all the citizenship tests are distinct conceptions of what makes a good citizen, analysing the content of the citizenship tests is relevant because it will allow us to gain insight into the image of the elites or those in charge of drafting the tests about the society and the immigrant (Etzioni 2007: 358). The content of the tests thus sends out a message of who belongs, and who does not form part of the community of members. Lastly, a content analysis makes it possible to answer the question of whether the ideas concerning the image of the ‘good citizen’ and the integration of future citizens expressed in the political debates actually recur in the tests. If the content of the actual citizenship test significantly differs from the ideas on what its content should be, as put forward in the democratic debates, the democratic acceptability of the test may be questioned. The content analysis of the Dutch ‘integration examination’, the German Einbürgerungstest, and the British ‘Life in the UK’ test can be found in Chapter 6. 1.2. Delineation of the Research As stated above, as a topic for this dissertation, I have chosen language and integration tests as a condition for naturalisation in Germany, the United Kingdom 12 Detailed information regarding the persons interviewed for the purpose of the research can be found in annex I.

Introduction  7

and the Netherlands. How can the choice for this topic be explained? Why did I not take language and integration tests for newcomers as a starting point, or the controversial pre-entry tests for family reunification, which have also been introduced in the countries under consideration? And why Germany, the United Kingdom and the Netherlands? Let me first explain the interest I have taken in language and integration tests as a condition for naturalisation. The reason why citizenship tests provide an interesting case for research is that the issue is subject to an obvious inherent tension. On the one hand, testing for naturalisation is experienced as self-evident: what is at stake is fully-fledged membership of the host community. By acquiring membership, the individual will also gain rights and duties he or she did not have as a non-citizen, of which the right to remain in the country and to vote are probably the most important examples. Seen in this light, it appears to be only natural that something may be asked in return. On the other hand, however, those applying for citizenship will generally have been living in the society of their country of residence for a considerable period of time, often for so long that it is no longer reasonable to refer to these societies as ‘host’ societies. By living, working, and raising their children, i.e. participating in the society of which they hope to become full members, the potential citizenship applicants have become part of the society in which they have been residing on a more or less permanent basis. Seen in this light, the justifications offered for subjecting these actual members of society to a membership test may not be self-evident.13 This point of view makes the question regarding the justification of citizenship tests more pressing. Apart from the inherent tension in the issue of citizenship tests, the growing concern for naturalisation in recent years, both in politics and in literature, makes the case of language and integration tests as a condition for naturalisation an interesting topic for research. After years of remaining on the political sidelines, nationality law in general and naturalisation in particular have become increasingly politicised and subject to electoral competition. As a consequence, the issue of naturalisation and citizenship has experienced increased attention in the academic literature.14 The rise in the interest in naturalisation can be seen in the light of the transfer of competence in the field of immigration to the European

13 Authors such as Carens (Carens 2010) and Rubio-Marín (2000) also point to the problems excluding (permanent) members of society from voting rights will create from a democratic point of view. 14 See, for instance, the NATAC research on citizenship laws in the EU 15 (Bauböck, Ersbøll, Groenendijk & Waldrauch 2006) and its successor, the EUDO-project (http://eudo-citizenship .eu/), Bosniak 2006, Hansen & Weil 2001, Howard 2009, Janoski 2010, Joppke 2008, Joppke 2010b, Koopmans, Statham, Guigni & Passy 2005, OECD 2011).

8  Chapter One

Union, which implies an ever limited competence of EU Member States to regulate access of EU and third- country nationals to their territories. In order to be perceived as being in charge of regulating immigration, these Member States resort to nationality law, which is still an area of almost exclusive competence for these states. The renewed interest in naturalisation is furthermore explained by the supposed threat to the state’s national identity and internal cohesion, experienced due to the presence of a considerable immigrant population (Cesarani & Fulbrook 1996: 214, Jacobson 1996, Schinkel 2007: 82). In addition, the feeling that past integration policies have not led to the desired result has emerged (De Hart & Van Oers 2006: 318). These developments have led to a re-valuation of citizenship and naturalisation: in order to protect the basic elements that allow for the existence of a liberal democracy, it is felt that control needs to be exercised over who is given political voice and who is allowed to remain unconditionally, in other words, over who is allowed to become a citizen (De Hart & Van Oers 2006: 318, Joppke & Morawska 2003: 1, Rubio-Marín 2000: 60–63). Citizenship tests are meant to tie citizenship more firmly to a shared national identity as an antidote to the presence of the diverse and allegedly disintegrating immigrant population (Joppke 2007: 4). Citizenship and naturalisation are hence perceived as medicines for curing the ‘harms’ created by immigration. This revaluation of the concept of citizenship contradicts post-national theories of citizenship in decline. From the upgrading of the alien status by decoupling rights from citizenship, some scholars have concluded that national citizenship has become less relevant for the integration of immigrants, and that the whole idea of ‘national citizenship’ is increasingly obsolete (Soysal 1994). However, as we will see in this book, recent years have witnessed ‘robust public debates’ about what it means to become a national (Hansen 2003: 88).15 Furthermore, despite the existence of the permanent resident status which entails social, civil, economic and even political rights comparable to citizenship, naturalisation still appears to have an added value. An obvious benefit attached to ownership of the nationality of (most) EU Member States is the possibility to travel to other countries without having to apply for a visa. Moreover, acquiring the nationality of an EU Member State also entails the acquisition of European citizenship, which opens up the possibility to live and work in all 27 EU countries. Immigrants who acquire European citizenship are hence not ‘stuck’ to one Member State, and practice has shown that they actively make use of their free movement rights.16 15 Hansen mentions the examples of the United States, France, Germany and the UK (Hansen 2003: 88). 16 Consider, for instance, the 10,000 Somali refugees, who, after having acquired Dutch nationality, migrated to the UK in the years 2001–2005 (Van Den Reek & Hussein 2003).

Introduction  9

Those not making use of their acquired free movement rights however also benefit from naturalisation. Research has shown that immigrants with citizenship status for instance often have better labour market outcomes than their peers who did not opt for naturalisation (Bevelander & Pendakur 2012: 220, OECD 2011: 3). Naturalisation appears to open up access to employment, especially to the public sector and higher skilled and better paid occupations, and have a positive impact on wages and occupational mobility. Furthermore, naturalisation has potentially important implications for the immigrant’s social integration, and is linked with other domains of immigrant integration such as language, housing, and the like (OECD 2011: 3–4). These findings make one wonder why states, instead of opening up access to citizenship, introduce new hurdles, such as citizenship tests. By analysing the reasons for introduction of citizenship tests, this study intends to contribute to providing an answer to this question. As for the choice of countries, besides Germany, the UK and the Netherlands, other European countries have introduced formalised language and knowledge of society tests. Currently, also Austria, Denmark, Estonia, France, Hungary, Latvia, Lithuania, Moldova and Romania present formalised language and knowledge of society tests to those applying for citizenship (Van Oers et al. 2010, Wallace Goodman 2010, Bertossi & Hajjat 2012).17 In France, the formalisation of the language and knowledge of society requirement has occurred relatively recently: a test on the history and culture of France was introduced by the law of 2011, and since 1 January 2012, applicants for naturalisation are required to present a language diploma which proves that they master French language at level B1 (Bertossi & Hajjat 2012: 15). If France would have been included in this study, an analysis of the political debates and the content of the test could have been presented, but it would have been impossible to say anything about the effects created by the recent formalisations. This however does not apply to the other European countries that use formalised language and knowledge of society tests, as these countries have either introduced these tests around the same time as the countries that have been selected, or well before. Austria introduced formalised language and knowledge of society tests in 2006, whereas Denmark formalised the language requirement for naturalisation in 2002, and introduced a test of Danish culture, history and society in 2007. The introduction of formalised tests in the Eastern European and Baltic states even dates back to before 2000.

17 Hungary does not apply a separate language requirement, but passing the examination of constitutional issues requires a strong command of the Hungarian language, and, as the naturalisation proceedings are conducted in Hungarian, so does the entire naturalisation procedure (Tóth 2010: 217).

10  Chapter One

Latvia and Hungary for instance have been testing language and knowledge of society since 1993 and 1994 respectively (Van Oers et al. 2010: 319–320).18 Why have I not included these states in the comparison? There are several reasons why comparing exactly Germany, the United Kingdom and the Netherlands is interesting from an academic perspective. In the literature, the countries are depicted as different, but they show some striking similarities as well. As far as the similarities go, the three countries under consideration for instance have significant non-national populations as a percentage of the total population.19 As regards the differences, the Netherlands and the UK are found to have responded differently to the issue of immigrant incorporation and integration than Germany. The former two countries have in the past applied explicit multicultural policies. Contrary to the Netherlands, British multiculturalism has however generally been typified as a laissez-faire type of policy, rather than a policy characterised by active attempts to allow ethnic minority groups to experience their cultures (Joppke 2004: 249).20 Germany, on the other hand, until the late 1990s, classified itself as not being a country of immigration. Consequently, its integration policy has been qualified as ‘segregationist’ and its model of immigrant incorporation via citizenship attribution as “ethnic”, “exclusive” or “rather closed” (Howard 2009, Koopmans, Statham, Guigni & Passy 2005: 7, MPG 2006). The citizenship regimes in the UK and the Netherlands have however been qualified as “open” or “traditionally liberal” (UK) or “rather open” (the Netherlands) (Howard 2009, Koopmans et al. 2005: 7, MPG 2006). All three countries have in recent years nevertheless introduced citizenship tests as a condition for naturalisation. How can this development be explained? The answer to this question will be formulated in the answer to the question scrutinising the reasons for introduction of the tests in the countries under consideration. 1.3. Existing Studies As stated above, citizenship legislation in general, and citizenship testing in particular, has counted on a growing interest in academic literature.21 What does the 18 In Hungary, the knowledge of society requirement consists of an examination of constitutional issues, whereas the Latvian test is on basic provisions of the Constitution, the text of the national anthem and the Latvian history (Kruma 2010: 252–253, Van Oers et al. 2010: 319–320). 19 In 2007, 4.2% of the population in the Netherlands had a foreign nationality, compared to 8.2% in Germany, and 6.5% in the UK, www.oecd.org. 20 As we will see in Chapter 3, the qualification of the Dutch integration policy as enabling immigrants to preserve their own identity has generally been exaggerated in the literature. 21 As regards general literature on citizenship legislation, the following studies can be mentioned as examples: Bauböck, Ersbøll, Groenendijk & Waldrauch 2006, Bosniak 2006, Brubaker 1992,

Introduction  11

present research add to what has already been said about citizenship tests? There are two new issues that the present study aims to address. First, the research aims to add to the studies, considered in Chapter 2, which deal with the question of what requirements for naturalisation can be justified. This literature sometimes, but not always, also focuses on the justification of the requirement of passing citizenship tests as a condition for naturalisation, and, if considered justified, which effects such tests are allowed to produce. This literature however does not look at the practice of citizenship testing. In other words, the question of whether the effects which the actual citizenship tests have produced in practice can be justified in the light of the model chosen is not answered. By examining the effects of the tests through an analysis of quantitative (statistics) and qualitative (interviews) empirical research, this study treats the question of the justification, under the citizenship models developed in Chapter 2, of the effects produced, thereby connecting theory with practice. Second, where other studies have focused on one or the other, the present research both zooms in and out on citizenship tests. Previous research either ‘zoomed out’ on citizenship tests by exploring the reasons why citizenship tests have been introduced (see, e.g., Joppke 2007, Van Oers, De Hart & Groenendijk 2006, Spijkerboer 2007), or ‘zoomed in’ on citizenship tests by exploring either their actual content (Michalowski 2011, Perchinig 2010, Prins 2010) or the question of what the content of citizenship tests should be (Bauböck & Joppke 2010, Hampshire 2010, Orgad 2010a). In analysing the question to what extent a citizenship test can be perceived as ‘liberal’, Joppke contends that a citizenship test may ask about factual knowledge, which can be learnt and mechanically reproduced, and knowledge about the applicable legal norms, but he draws a line in cases where the test questions aim to scrutinise the test candidates’ inner motivations (Joppke 2010b: 141). To this distinction between ‘what is right’, which can be tested, and ‘what is good’, from the testing of which a liberal state should abstain, Orgad adds his test of ‘national constitutionalism’, to guarantee that only the issues which are relevant enough for citizenship are asked about (Orgad 2010a). However, whereas both authors have investigated the question of what the content of a citizenship test ought to be, they have refrained from applying their findings to the citizenship tests which have been introduced in practice, in the form of a systematic content analysis.

Calder, Cole & Seglow 2010, Howard 2009, Janoski 2010, Joppke 2008, Koopmans et al. 2005, OECD 2011, Soysal 1994. Examples of literature which focuses entirely or inter alia on integration tests as a requirement for naturalisation are: Bauböck & Joppke 2010, De Hart & Van Oers 2006, Joppke 2007, Michalowski 2011, Orgad 2010a, Peucker 2008, Van Oers 2006, Van Oers, Ersbøll & Kostakopoulou 2010.

12  Chapter One

Michalowski and Perchinig, on the other hand, have conducted such an analysis.22 By classifying the Austrian citizenship tests, applied at federal and provincial levels, into nine thematic categories, Perchinig assessed whether they reflect a liberal, republican or communitarian understanding of citizenship (Perchinig 2010). Michalowski’s study is comparative in nature, since she analysed the content of the citizenship tests applied in Austria, Germany, the Netherlands, the United Kingdom and the USA. Relying on the distinction, also made by Joppke and Orgad, between ‘what is right’ and ‘what is good’, she comes to the conclusion that the content of citizenship tests in restrictive citizenship policy regimes, such as those of Germany and Austria, can be as liberal as the content of citizenship tests in open citizenship policy regimes, such as the regime in the USA (Michalowski 2011: 765). At the same time, Michalowski’s analysis showed that the content of the Dutch citizenship test partly consists of questions relating to ‘what is good’, thereby trespassing the boundaries of liberalism.23 Michalowski therefore concludes that the surrounding citizenship regimes and procedural details of the testing procedure are of no help in predicting the content of citizenship tests, and that equating restrictive citizenship policy regimes with cultural assimilation, as well as open citizenship regimes with political liberalism, is “outright misleading” (Michalowski 2011: 764–765). By conducting a systematic content analysis of the tests, the present research adds to the very scarce literature on the empirical study of citizenship testing. It thereby hopes to contribute to the academic discussion regarding the liberality of citizenship tests. Furthermore, by simultaneously ‘zooming in’ and ‘zooming out’ on citizenship tests, the present study adds to what has already been written about the purpose and character of the tests. Michalowski’s study has shown that a proper evaluation of the character of a citizenship test is impossible on the basis of merely an analysis of the content (Michalowski 2011: 764). And it is true that her analysis, by including the surrounding citizenship regimes and procedural details of the testing procedure, has taken into account more than merely the content of the citizenship tests to interpret their meaning. In this sense, her study is more than a ‘zooming in’ exercise. My ‘zooming out’ exercise however goes further than the exercise conducted by Michalowski. As stated above, I will also

22 Prins has commented on the content of the Dutch citizenship test, but instead of a systematic content analysis, her comments focus on some of the questions in the Dutch examination (Prins 2010). 23 For a test to qualify as liberal in Michalowski’s analysis, they have to fulfil a second criterion, being that questions about ‘the political system, democracy and rights’ should be the biggest thematic category within each test. Michalowski concludes that, of the tests considered in her analysis, only the Dutch test does not fulfill this second criterion (Michalowski 2011: 752, 762).

Introduction  13

take account of the context, timing, and the official discourses leading to the introduction of the tests, as well as their effects. 1.4. Structure of the Book As stated above, Chapter 2 of this book contains the theoretical framework by explaining the liberal, republican and communitarian citizenship theories, and the different models for naturalisation based on these theories. Chapters 3, 4 and 5 analyse the political debates leading up to the introduction of the citizenship tests in the countries under consideration, using the models for naturalisation as an instrument of analysis. The order of the chapters is chronological, starting with the country which first presented a formalised language and knowledge of society test to its naturalisation applicants (the Netherlands). Chapter 6 subsequently contains the content analysis of the citizenship tests. This chapter will answer the question to what extent the content of the tests can be justified under the liberal model, or whether the tests better match the republican or communitarian citizenship ideals. The question of whether the tests actually match the ideas on what their content should be, put forward in the political debates preceding their introduction, will be left for the concluding chapter of this book. The results of the quantitative and qualitative empirical research conducted within the framework of this study will be represented in Chapters 7, 8, 9 and 10. Chapter 7 contains an analysis of statistical data, Chapters 8, 9 and 10 analyses of the interviews. These chapters aim to provide an answer to the question of what effects the tests have produced, and whether the tests have affected some groups more negatively than others. As stated above, the question whether, and to what extent, the produced effects can be justified in the different citizenship models, will be dealt with in the concluding chapter of this book.

Chapter Two Justifying Citizenship Tests What conditions may a state require immigrants to fulfil in order to become fully-fledged citizens?1 This paragraph will examine the literature on citizenship theory, in order to answer this question. States will generally have an interest in their citizens being law-abiding, making a productive contribution to the state’s (economic) well-being and adopting the civic virtues that contribute to its flourishing (Calder & Seglow 2010: 161). It has furthermore been argued that states’ nationality laws “in all cases […] express […] and consecrate […] the conception of the nation and reinforce […] the homogeneity of national populations” (Schnapper 1998: 107). Do these presumptions imply that states are free to throw up whatever hurdles they deem fit in order for their long-term residents to acquire citizenship? Below we will see that the answer to this question depends on which concept of citizenship is applied. For now it will suffice to note that indefinitely excluding all long-term residents from taking part in the process of making laws which also concern them constitutes a problem of democratic legitimacy. Preventing these long-term residents from becoming fully-fledged members is hence hard to justify from a democratic point of view.2 However, at the same time, it is not very realistic to ask states to, for democracy’s sake, simply allow all their long-term residents unconditional access to the status of citizenship. Most, if not all, states at least ask their permanent residents to fulfil residence and public order requirements when they apply for nationality. Assuming that it is justifiable for states to require their aspiring citizens to fulfil certain conditions, the question arises of which conditions states may reasonably ask their permanent residents to fulfil. More specifically, for the purpose of this dissertation, the question is whether states may require their aspiring citizens to

1 For the purpose of this chapter, I assume that it is justifiable to require immigrants to fulfil certain requirements in order to obtain the status of citizenship. I will hence not deal with the question of whether it is justified to throw up a second barrier after the first barrier, i.e. obtaining access to a state’s territory, has been taken. Examples of studies which have treated the question of the justification for the erection of barriers to admission and/or full membership are: Carens 1987, Walzer 1983, Tholen 1997. 2 See for a constructive development of this argument Rubio-Marín, who in her book defends an automatic conferral of nationality after a certain period of residence (Rubio-Marín 2000).

16  Chapter Two

have a certain level of language skills and knowledge of society, to be proven by passing a test. As it is a widely shared assumption in political science that the legal conditions and terms of inclusion and exclusion of immigrants in a state are strongly related to the self-image of the political community, the answer to this question largely depends on how the community perceives itself (Gerdes, Faist and Rieple 2007: 47). And since all three have different views on the justification of exclusion of foreigners from the community, the theoretical perspectives on the community of liberals, republicans and communitarians are a useful means to answer the question on the justification of naturalisation requirements.3 Liberal theory considers all residents of a certain territory, whether native born or foreigner, equal. In this position, there is consequently little room for conditions for becoming a full member of society: since they are considered as barriers to equality, conditions for naturalisation are hard to justify. Once admitted, foreigners should be considered full members as soon as possible. Liberal scholars therefore consider the erection of a second barrier in the form of naturalisation conditions as inappropriate once the first barrier, being admitted to the country, has been taken. On the other side of the spectrum are authors defending a republican or communitarian view of citizenship. These authors reproach the liberals for not taking sufficient account of the particularities of the community into which a foreigner is admitted. By simply allowing everyone full citizenship rights, the survival of the community is threatened. Not the equality of all those present on the territory is put first. Rather, someone is considered to be a citizen if he or she actually forms part of the liberal collective. In order for the community to survive, newcomers will therefore have to fulfil certain requirements before becoming members: they need to have certain capacities and a certain loyalty (Tholen 1997: 197). Normative models based on liberal, republican and communitarian conceptions of citizenship will provide different answers to the question of whether or not the imposition of language and knowledge of society requirements is justified. The answer to this question will depend on the answer to the underlying question of whether, and to what extent, the models allow for the justification of exclusion of outsiders from the community. With which arguments can a state according to the models justify the exclusion of immigrants from the benefits of citizenship? The purpose of this chapter is not to provide a comprehensive answer to the normative question of which conditions for naturalisation are justified. Rather, the way in which the liberal, republican and communitarian models have answered the question regarding the justification of exclusion of outsiders, and 3 In the literature, these models are indeed often used for this purpose. See, for instance, Kostakopoulou 2003, Neuman 1994, Tholen 1999.

Justifying Citizenship Tests  17

the conditions for naturalisation which follow there from, will be described, and in the subsequent chapters used as an instrument of analysis of the political debates (Chapters 3, 4 and 5). For the arguments expressed in the political debates preceding the introduction of citizenship tests by both proponents and opponents to the tests, corresponding considerations will hence be sought in the literature. Apart from using the models as an instrument of analysis of the political debates, the citizenship models will also be used as an instrument to analyse the content of the citizenship tests that have been introduced in the countries under consideration, and to interpret the effects produced by these tests (Chapter 6). I will first elaborate on the different dimensions of citizenship below (2.1). I will subsequently explain the citizenship conceptions, which emphasise one of these dimensions (2.2). 2.1. Dimensions of Citizenship Since citizenship can refer to virtually every aspect of the relationship among citizens or between the citizen and the state, it is useful to distinguish between the different dimensions of citizenship (Fermin 2001: 67). In the literature, three different meanings of the concept of citizenship have been identified: citizenship as legal status, citizenship as (political) activity and citizenship as identity (Bauböck 2001, Bellamy 2004, Carens 2000, Cohen 1999, Entzinger 2008, Fermin 2001, Honohan 2010b, Kostakopoulou 2003, Kymlicka & Norman 2000). 2.1.1. Citizenship as Legal Status (‘What You Get’)4 Citizenship as legal status refers to ‘nationality’, which needs to be understood as a specific link between a person and a state (Vermeulen 2007: 53). Of the three dimensions discerned, it is the most minimal understanding of citizenship, since it merely refers to a legal position (Pluymen 2008: 17). A citizen in the legal sense profits from a privileged relationship with the state, where she enjoys a number of (civil, political and social) rights, but where she may also have (a relatively small number of) duties to that state. The presence of a large community of immigrants and the creation of discrepancies in the legal status of citizens and non-citizens has called for renewed attention for this dimension of citizenship (Fermin 2001: 72).5 4 Kostakopoulou 2003: 86. 5 Citizenship as a bundle of legal rights can, however, also reach beyond the framework of ‘nationality’, since rights have expanded beyond the formal status of citizenship. Tomas Hammar has

18  Chapter Two

2.1.2. Citizenship as Activity (‘What You Owe’)6 This aspect of citizenship refers to citizenship as an attitude (Pluymen 2008: 17). It concerns participation in the political, social and economic life of the community. Communities which emphasise the activity dimension of citizenship will demand more of citizens than communities which stress the legal dimension of citizenship. Next to having a privileged relationship with the state in the form of having rights, a citizen will hence also be required to behave like a ‘good citizen’, i.e. to actively contribute to the flourishing of society. To do this will require citizens to display ‘civic virtues’, which Bauböck has defined as “the disposition of citizens to regard the common good of the polity as an important part of their own interests” (Bauböck 2006: 31). Traditionally, citizenship as activity has put political participation at its centre: a good citizen will take part in elections and engage in political deliberations with fellow citizens. As we will see below, this is the kind of behaviour to which most value is attributed in the traditional republican conception of citizenship. Civic virtues however do not only focus on narrowly defined political behaviour. Recent governmental discourses tend to emphasise participation in political life as much as participation in the economic or social life of the community (for example by doing voluntary work or by sustaining social contacts).7 Some authors claim that several trends have led to a renewed interest in the activity dimension of citizenship. They mention increased voter apathy, the erosion of the welfare state, the failure of environmental policies that rely on voluntary citizen cooperation, as well as ‘moral decay’ (Fermin 2001: 69, Kymlicka & Norman 2000: 5). The rising presence of permanently residing immigrants creating a new form of diversity has also contributed to a renewed interest in the conception of citizenship as activity. The question of whether good Muslims can be good democratic citizens has, for example, been raised several times in the Dutch political debate (Fermin 2001: 69).8

introduced the term ‘denizens’ for foreign immigrants enjoying rights originally tied to citizenship (Bauböck 2001: 42). New Zealand has, for instance, given denizens the right to vote after one year’s residence (Calder & Seglow 2010: 162). 6 Kostakopoulou 2003: 86. 7 It goes without saying that the republican model departs from an ideal image of the citizen. In reality, far from all those who are entitled to cast their vote or participate in the political debate actually do so. Furthermore, only a minority is member of a political party, or engages in voluntary work. 8 As we will see, this question also plays a role in the British political debate on naturalisation requirements (paragraph 5.3).

Justifying Citizenship Tests  19

2.1.3. Citizenship as Identity (‘What You Feel’)9 Citizenship as identity refers to an exclusive social status of membership of a polity and the identification with it. Carens has called this the ‘psychological’ dimension of citizenship (Carens 2000). It focuses on issues regarding the identification with and loyalty and commitment to the culture and tradition of the society of which one forms part. Loyalty to the state is not so much concerned as loyalty to the nation: a group of people of common descent, who share a collective history and cultural characteristics, who derive their identity here from and who show a certain degree of mutual solidarity (Entzinger 2008: 181). The substantial, socio-psychological dimension of citizenship as identity has been attracting renewed attention due to the immigration to Western democracies of immigrants having different ethnic backgrounds and cultures (Fermin 2001: 68, Schinkel 2007: 72). It has made the question of who does and who does not belong to the nation more acute (Entzinger 2008: 181). 2.2. Theoretical Foundations for Conditions for Naturalisation The distinction made between the various dimensions of citizenship can serve as a useful starting point for the description of the different approaches to the concept of citizenship in political-philosophical theory. In liberal, republican and communitarian concepts of citizenship, the way in which the dimensions are emphasised and interpreted differs, since they have a different perspective on the meaning of citizenship for society and for the individual (Bauböck 2001: 40, Bellamy 2004, Fermin 2001: 70). The theories thus represent different views on the ‘all-embracingness’ of citizenship, varying from thin conceptions of citizenship, which regard citizenship as a strictly legal relationship, to thick ones, which emphasise the aspects of identity and participation (Bauböck 2001). Broadly speaking and at the risk of over-simplification, the liberal conception of citizenship focuses primarily on citizenship as a legal status, whereas the dimension of citizenship as activity is prioritised in the republican conception of citizenship (Honohan 2010a: 5). Finally, the communitarian conception of citizenship emphasises the identity dimension. The different models of citizenship thus provide different answers to the question regarding the justifiability of requirements for naturalisation. Even though the liberal, republican and communitarian conceptions of citizenship are based on the liberal, republican and communitarian theories, these 9 Kostakopoulou 2003: 86.

20  Chapter Two

theories are not theories of citizenship. They have been developed to offer insight into the meaning of justice and equality in a community and address the question to what extent the ideal of equality can be realised in the community (Pluymen 2008: 16). The liberal, republican and communitarian theories hence do not provide a clear-cut answer to the question of who forms part of the community of citizens and on the basis of which criteria the community can be joined, if it is possible to join at all. The theories rather consider membership as a given. The ideas put forward in the political-philosophical theories have however been extrapolated to the subject matter of migration and citizenship. Scholars have used the theories to answer the question of whether the formulation of requirements for access to citizenship, such as integration requirements, can be justified, and, if so, which conditions immigrants can be required to fulfil in order to become citizens (Entzinger 2008, Etzioni 2007, Kostakopoulou 2003, Neuman 1994). The different conceptions of citizenship will be described below. I will examine whether, and, if so, how, the various conceptions justify the imposition of integration conditions for naturalisation. If the answer to this question is positive, the second question of whether it is justifiable to require applicants for naturalisation to pass a uniform citizenship test will be answered. 2.2.1. Liberal Citizenship Two characteristics of liberal citizenship have already been mentioned above, the first being that in the liberal conception, citizenship boils down to a legal status that links individuals to states. The liberal model hence uses a ‘thin’ conception of citizenship rather than emphasising the aspects of identity and participation. Secondly, in the liberal model, all residents in a certain state are considered equal, which implies that exclusion is hard to justify. In ‘A Theory of Justice’, John Rawls, an important representative of liberal theory, asks the question whether and under which circumstances exclusion of outsiders from a community can be justified (Pluymen 2008: 20, Rawls 1971). In order to answer this question, he puts himself in the position of a group of people covered by a ‘veil of ignorance’, meaning that they know nothing about their own personal situations (class, race, gender, natural talents, beliefs, etc.). Rawls is convinced that people in this ‘original position’ would follow their own concerns, and choose equality as the default position. From this idea of equality as the highest good, it can be derived that exclusion of persons from rights that, in principle, apply to everyone, requires justification (Dummett 2001: 24–25, Pluymen 2008: 20). Rawls did not answer the question of what obligations a state has towards those who are not its citizens, or under what circumstances these foreigners should be allowed to become citizens. Excluding long-term resident immigrants from full citizenship rights can, however, not be easily reconciled with the liberal core value of equality

Justifying Citizenship Tests  21

of individuals and therefore requires profound justification.10 In the liberal model, naturalisation is accordingly perceived as a right for long-term residents, which should be open to immigrants who fulfil relatively easy naturalisation conditions. The Liberal Model for Naturalisation In the liberal model, what conditions can states ask their permanent residents to fulfil in order to become members? Whereas they agree that naturalisation is a right that should be made available to all residents, liberal scholars propose different modes of acquiring the nationality of the country of residence. ‘Liberal minimalists’ support a very minimum of naturalisation requirements, but there are also authors who defend the imposition of more demanding naturalisation requirements, including citizenship tests, in the liberal model. The first author that can be cited as a proponent of minimal naturalisation requirements is Michael Walzer.11 According to Walzer, requiring immigrants to pass a second barrier in the form of fulfilling requirements to access the status of citizen, after they have overcome the first barrier of gaining access to a state’s territory (admission), amounts to tyranny: the imposition of requirements to naturalisation will risk the creation of a population of permanent aliens, whose condition Walzer compares to that of slaves or in-living servants (Walzer 1983). Walzer hence claims that there is no justification for the exclusion from full membership of aliens who have been admitted to stay on the state’s territory.12 Another important representative of ‘liberal minimalism’ is Joseph Carens. Like Walzer, Carens regards naturalisation as a right that should be available to all those present in a state’s territory. According to Carens, migrants have a moral right to citizenship, since they become ‘de facto’ members of the society after a certain period of time (independent of whether or not they contribute to it) (Carens 1989: 3). Applying this ‘membership argument’, all that may be required to obtain the status of citizen in Carens’ view is for the immigrant to have resided on the state’s territory for a certain period of time. Carens opposes the intro­ duction of language and integration tests as a requirement for naturalisation. 10 As we have seen, citizenship rights can also reach beyond the framework of nationality. Denizens can enjoy rights which used to be attached to citizenship. The status of nationality will however entail more rights and securities that any denizenship status. 11 Even though Walzer is often ranged among authors who have been labelled ‘communitarians’ for his ideas on the value of a community for an individual’s well-being (see below), he clearly objects to a selection procedure for naturalisation. 12 According to Walzer, naturalisation can only be subject ‘to certain constraints of time and qualification, never to the ultimate constraint of closure’ (Walzer 1983: 59). Walzer does not explain the exact period of residence that can be required as a condition for naturalisation, and what he means by ‘constraints of qualification’.

22  Chapter Two

In his opinion, requiring integration beyond residence is to violate the principle of toleration and respect for diversity to which all liberal democratic states are committed and to call into question the equal status of current citizens who differ from the majority (Carens 1989: 40). Language, as well as other tests such as sufficient income or absence of a criminal record, furthermore undermine the inclusivity of the political community and give state officials the power to make arbitrary decisions (Carens 2005: 39, Seglow 2009: 796). According to Carens, someone who has functioned in a society for a considerable period of time without having learned the majority language should be deemed capable of participating in the political process. He thereby parries the assertion that a certain level of language skills needs to be required in order for citizens to exercise their newly acquired political rights (Carens 2002: 111). Though he admits that normative values such as loyalty, patriotism and identity should be encouraged, Carens thinks those values should be thought of as norms or aspirations rather than requirements for citizenship. The argumentation of the liberal minimalists is clear: requirements for naturalisation will lead to the exclusion of those who do not meet them, which cannot be defended with liberal arguments. Naturalisation requirements should hence be as minimal as possible.13 Applying the liberal minimalist theory to the naturalisation policies of Western liberal democratic states shows us that most, if not all, of their naturalisation policies should be disqualified as being morally unjustifiable. None of the EU Member States have institutionalised a procedure of unconditional naturalisation after a certain length of residence, and neither does this apply to ‘classical’ immigration countries such as the USA and Canada.14 Certain liberal theorists do not opt for the minimalist solution either and attempt to justify arguments for exclusive membership in the liberal model. 13 Other authors who can be ranged among the liberal minimalists are Bauböck, Rubio-Marín and Kostakopoulou. According to Bauböck, ‘the capacity to be a citizen should be ascribed in principle to any person who has not given strong evidence of the contrary in speech or deeds’ (Bauböck 1994: 92). Kostakopoulou appears to agree with Bauböck by claiming that no extra naturalisation conditions apart from domicile and the absence of serious criminal convictions can be required (Kostakopoulou 2006). 14 Looking at other forms of nationality acquisition by immigrants than naturalisation however shows us that several states have institutionalised ways for certain categories of immigrants to acquire nationality in a more or less automatic way. All of the three countries under consideration for instance provide for automatic acquisition by children of immigrants, born on their soil. The UK and Germany apply this form of ius soli acquisition to the children of first-generation immigrants, whereas in the Netherlands, the children of the third generation automatically acquire Dutch citizenship at birth. For a clear and thorough comparison of the nationality laws of the EU 27 see http://eudo-citizenship.eu/.

Justifying Citizenship Tests  23

Hampshire, for instance, argues that it is unreasonable to disqualify all current naturalisation policies in Western states for being illiberal (Hampshire 2010). Hampshire argues that, when applying for a status, we generally think it reasonable that we are required to demonstrate some understanding of what that status entails, and some commitment to its constitutive values. Since naturalisation is the process by which the status of citizenship is obtained, this is no different. It is, according to Hampshire, consequently quite legitimate for liberal states to institutionalise the formal rights and responsibilities of citizenship and the ideals of liberal citizenship in their naturalisation policies (Hampshire 2010: 87). The imposition of language and knowledge of society requirements can in Hampshire’s opinion be justified on the basis that we should accept that the stability of liberal institutions and the pursuit of social justice also depends on the dispositions and capacities of citizens (Hampshire 2010: 90). Hampshire shows that liberal theorists, partially forced to do so following civic republican and communitarian critiques of liberalism (see below), have already made this step. There is growing recognition of the fact that the success of liberal government depends as much upon the dispositions of citizens as on free institutions (Patten 2000: 203, quoted in Hampshire 2010: 85). This line of thinking appears to offer room for the imposition of extra naturalisation conditions besides residence in order to safeguard the flourishing of liberal democracies. At the same time, the requirements for naturalisation may not be so high that the naturalisation policy will no longer enable permanent residents to naturalise. Liberal states will hence have to steer a course between these two conflicting imperatives (Hampshire 2010: 87). A Liberal Citizenship Test? What does this imply for the introduction of citizenship tests scrutinising the applicant’s knowledge of the language and society? Let us first examine what this account of liberal citizenship implies for the institution of a language requirement. Even though Carens has dismissed the imposition of a language requirement for the reasons explained above, Hampshire argues that it is difficult to see how a person can participate in politics without a facility in the language of political debate. He claims that the imposition of a language requirement can be justified by reference to the demands of citizenship: the ability to develop cooperative political virtues presupposes an ability to speak and understand the predominant language of fellow citizens (Hampshire 2010: 88). According to Hampshire, a language requirement that is already so common to liberal states’ naturalisation policies is hence not only legitimate, but also desirable (ibid.).15 15 At the time of writing, of the EU 27 only Belgium, Cyprus, Italy, Ireland, Poland and Sweden lacked a language requirement in their naturalisation policies (Wallace Goodman 2010: 14).

24  Chapter Two

Taking into account the second imperative which prescribes that naturalisation policy should remain open and enabling, we can assume that the language requirement may not be so demanding that it will prevent aspiring citizens from fulfilling it. Formalised tests requiring a certain level of language skills are automatically biased towards less educated immigrants. Requiring a fixed level that is to be attained by all applicants, regardless of their individual backgrounds and capacities, hence is a bad idea if the exclusion of a certain proportion of the aspiring citizens is to be prevented. If a uniform level is demanded, then at least sufficient possibilities for preparation should be provided for, as well as possibilities for exemption of those for whom obtaining the required level is simply unfeasible. It goes without saying that these preparation possibilities as well as the test itself must be reasonably priced, or, more ideally, be available at no cost at all, since too high a price attached to passing the tests may, just like a too high level, be experienced as a barrier. Hence, in order to qualify as liberal, a language test should not lead to exclusion on the basis of either intellectual capacity, learning skills, or costs. The justification for a requirement of having knowledge of the society appears to be a more vexed issue: whereas the importance of command of the language for exercising one’s citizenship rights is relatively obvious, having knowledge of the society appears to be of a less immediate concern in this respect. Furthermore, even if a case can be made for testing this kind of knowledge, the question of what knowledge should be required for naturalisation is very difficult to answer. However, whereas liberal minimalists oppose knowledge of society tests, several authors have attempted to provide a justification for these tests in the liberal model. Joppke for instance has contended that it is “incontrovertibly legitimate” for states to ask for “knowledge of the principles and procedures of liberal democracies” and that it is “not as such an unreasonable expectation” to “ask for knowledge of, or even experience with” a country’s “historical specificities” (Joppke 2010a: 1). Even though it is highly questionable what a state in reality can accomplish by testing factual knowledge about these issues, according to Joppke, knowledge of society tests in such cases are an instance of political liberalism, since they pursue the goals of liberalism and, like other obligatory measures, are capacity enhancing (Joppke 2010b: 1, Michalowski 2011: 750). In similar vein, Hampshire has contended that tests of liberal values and civic skills can be consistent with and supportive of liberal justice and democracy (Hampshire 2010: 88). Further­more, if administered properly, citizenship tests can incentivise residents to obtain the appropriate dispositions, and assess whether they have acquired them (Hampshire 2010). Hence, rather than “a fast-handed debunking of the entire genre”, Joppke therefore argues that knowledge of society tests can be liberal, provided their implementation and content meet certain demands (Joppke 2010a: 1).

Justifying Citizenship Tests  25

Regarding its content, several scholars have formulated the requirements which a knowledge of society test needs to meet to qualify as liberal. These requirements are described below. Apart from certain requirements regarding the content, and taking into account the second imperative which prescribes that access to nationality in the liberal model should be guaranteed, a knowledge of society requirement may furthermore not be so demanding that it will prevent aspiring citizens from fulfilling it. In order not to lead to exclusion, a knowledge of society test will hence need to fulfil similar conditions as formulated above in respect of the justification for the language requirement.16 ‘What Is Right’ versus ‘What Is Good’ What kind of questions may a citizenship test contain in order to be justifiable when applying liberal standards? This question has been answered by Joppke, who, relying on Kant, asserts that it is “a key precept of liberalism that law and public policy can only regulate the external behaviour of people, not their inner motivations” (Joppke 2010a: 140). A liberal state hence is liberal in so far that it refrains from prescribing substantive conceptions about ‘what is good’ (Vermeulen 2007).17 This means that liberal states can require their citizens to conform to liberal principles in practice and accept policies that are pursued in the name of such principles as legitimate, but that they should abstain from requiring their citizens to believe liberal principles (Miller 2004: 14).18 In other words: liberalism also contains the freedom to choose not to be liberal, as long as one’s way of life is democratic and legal (Orgad 2010a: 94). Applying this theory to the field of integration requirements for citizenship, this means that questions scrutinising the candidates’ ‘inner dispositions’ should be disqualified for being illiberal (Joppke 2010a, Joppke 2010b): immigrants seeking naturalisation may be asked to accept and respect the liberal principles as the law of the land, but the state cannot require them to morally agree with them (Orgad 2010a: 101). Questions in citizenship tests relating to liberal principles should hence focus on the external respect for these principles, i.e. the knowledge and acceptance of the legal norms flowing from these principles, and ignore internal perceptions (Joppke 2010a: 2, Joppke 2010b: 140, Orgad 2010a: 101). This leads us to the following conclusion: while questions relating to legal commitments are 16 This means that the level of the test may not be so high that it would automatically exclude a certain part of the immigrant population. Furthermore, its price may not be too high, and sufficient possibilities for preparation and exemption need to be provided for. 17 In similar vein, Orgad and Joppke rely on the Kantian distinction between ‘legality’ and ‘morality’ to define the line between liberal and illiberal test contents (Joppke 2010: 2, Orgad 2010a: 101). 18 Miller in the same paper claims that citizens are thus free to advocate alternative arrangements. The liberal state should tolerate those who oppose liberalism, such as fascists, anarchists and communists (Miller 2004: 14).

26  Chapter Two

permissible, those relating to moral obligations and judgments are not. In other words, questions about ‘what is good’ should be disqualified, whereas questions which ask about ‘what is right’ can qualify as liberal (Michalowski 2011). Concretely, this means for instance that a question asking whether men and women are equal under the legislation of the country of which citizenship is sought (legality, ‘what is right’) is allowed, whereas a question asking about the applicant’s opinion of gender equality (morality, ‘what is good’) is not. A knowledge of society test that aims to identify the candidate’s true values or beliefs should therefore be regarded as “pernicious from a liberal point of view”, even if the values and beliefs that are questioned pertain to the rules of liberal democracy (Joppke 2010a: 2). Asking someone to be liberal in order to qualify as a citizen in liberal democracy will furthermore create what has been referred to as a ‘paradox of liberal toleration’: excluding someone based on ideological ideas is illiberal in view of the liberal principles of freedom and tolerance (Entzinger 2003: 80, Neuman 1994: 261).19 By prescribing a shared way of life, Joppke has referred to this form of liberalism as “identity liberalism”, a substantive, less procedural form of liberalism which intends to separate liberal from illiberal people (Joppke 2010b: 139). By leading to “liberal exclusion”, identity liberalism can amount to “repressive liberalism”, a liberalism where the liberal state is considered for liberal people only, and which is brought forward with an unabashedly exclusionary intention against liberalism’s “presumed Other”: Islam and Muslims (Joppke 2010b: 140– 142). Joppke concludes that states may not require a certain identity on the part of the citizenship applicant, even if ‘liberalism’ is the content of that identity (Joppke 2010b: 123). The desire to make sure that future citizens morally agree with the liberal values furthermore creates the problem of assessing whether this is the case: it is highly questionable whether this moral agreement can be tested, let alone that it can be tested in a multiple-choice citizenship test, which only measures whether the test candidate is able to produce the desirable answers. A test will thereby measure intelligence and studying skills rather than convictions. Reasoning ‘a contrario’, Joppke qualifies questions that ask for factual knowledge about a country’s history, culture and institutions as unproblematic from a liberal point of view, as “such matter is merely cognitive, it can be learned and mechanically reproduced” (Joppke 2010a: 141). In her analysis of citizenship tests, Michalowski however shows that in certain cases, questions testing merely 19 Requiring immigrants to adhere to liberal values will create what Orgad has referred to as a Paradox of Liberalism: liberal states, in order to preserve what they perceive as a liberal regime, are resorting to illiberal means to guarantee liberal values. The paradox lies therein that the liberal state must either tolerate illiberal practices, or turn to illiberal means to ‘liberate’ the illiberal. Either choice, as Orgad argues, undermines liberalism (Orgad 2010a: 92).

Justifying Citizenship Tests  27

cognitive knowledge should also be disqualified. She comes to this conclusion using Rawls’ definition of political liberalism (Michalowski 2011, Rawls 1993). As Michalowski explains, Rawls has asserted that political liberalism consists of finding an overlapping consensus among comprehensive doctrines on a political conception of justice, which “should be, as far as possible, presented as independent of comprehensive religious, philosophical, and moral doctrines” (Michalowski 2011: 752, Rawls 1993: 144). Consequently, Michalowski disqualifies questions as illiberal not only if they scrutinise a candidate’s inner dispositions, but also if they ask the candidate to know what is considered ‘good’ by the receiving society (Michalowski 2011: 752). According to Michalowski, such questions are illiberal as they “suggest a homogeneity of values, opinions, or moral attitudes that does not exist in western societies, and that- according to a strict definition of political liberalism – is not even necessary for a liberal democracy to function” (Michalowski 2011: 752). Michalowski in this regard refers to questions relating to social norms which are believed to be generally accepted in the host society, even though they are not laid down in law (Michalowski 2011: 753). As examples of questions falling in this category, she refers to questions in the Dutch test which ask the test candidates to know that the relationship between men and women is equal (in the home and in public) and that they should work out regularly and eat healthily in order to stay healthy (Michalowski 2011: 759, 762). Even though they do not scrutinise the candidates ‘inner dispositions’, Michalowski has disqualified this type of questions for not relating to a legal norm, and for pushing for the acceptance of social norms. They should therefore also be ranged in the category of illiberal questions testing ‘what is good’, and for that reason not form part of a liberal citizenship test (Michalowski 2011: 762). Even if mere cognitive knowledge is tested. What Is Relevant? According to the above, questions testing ‘factual knowledge’ or ‘what is right’ may be allowed in a liberal citizenship test. There is however another condition for citizenship test questions that has been formulated in the literature. Certain scholars have claimed that in order for a question to deserve a place in a citizenship test, more than merely refraining from transgressing the thin line between the regulation of behaviour and the control of beliefs is required. These scholars contend that, apart from ‘merely’ refraining from asking about ‘what is good’ or by testing factual knowledge, the test questions also need to deserve a place in a citizenship test for being sufficiently relevant for citizenship. They have accordingly formulated, what I like to call, ‘relevancy tests’, which aim to prevent the citizenship tests from having an arbitrary content, asking about whatever topics those who happen to design its content deem fit. The relevancy tests developed by Hampshire and Orgad will be explained below.

28  Chapter Two

• Hampshire’s Relevancy Test As we have seen above, Hampshire has defended the introduction of citizenship tests in liberal states. In Hampshire’s opinion, naturalisation tests can constitute means to “assess an applicant’s understanding of citizenship rights and duties, and more controversially they can symbolise the ideals of citizenship in a liberal state” (Hampshire 2010: 89). At the same time, Hampshire however sets limits to the content of a citizenship test by arguing that only topics that are relevant for effective citizenship deserve to be tested (Hampshire 2010). Applying this relevancy criterion, Hampshire argues that questions relating to liberal values and civic skills may be asked. They are essential to effective citizenship as they are consistent with, and even supportive of, liberal justice and democracy (Hampshire 2010: 88–89). But what are these ‘liberal values’ and ‘civic skills’ that may be tested? Hamp­ shire appears to take a narrow view, limiting the knowledge which can be tested to knowledge required to be able to engage with other citizens in political dialogue (Hampshire 2010: 89). In a later writing, Hampshire makes a case for the testing of knowledge which is “essential to participation as a citizen” (Hampshire 2011: 967). Hampshire admits that knowledge of certain historical events or cultural practices might be essential to effective citizenship, and that for that reason questions about such events or practices could form legitimate items of citizenship tests (Hampshire 2011: 967). He however warns that such questions may easily become “a spurious cover for social prejudice”, which makes them suspicious from a liberal point of view. As we have already seen, liberal states may not impose a particular conception of the good, so any requirement that the applicant must assimilate into the national culture veers towards a violation of this liberal core principle (Hampshire 2010: 83, 89). Consequently, “the general presumption in tests should be towards political […] knowledge” (Hampshire 2011: 967). Taking account of Hampshire’s relevancy criterion will probably leave states little room for emphasising their specific identities, as the applicable ‘liberal values’ and ‘civic skills’ will not differ much between liberal democratic states. In this respect, Joppke has referred to the ‘paradox of universalism’: while aiming to make sure that ethnic minorities become a part of this and not any political community, states fail to succeed in naming any particulars that distinguish the ‘here’ from the ‘there’ (Joppke 2010b: 130). • Orgad’s ‘National Constitutionalism’ Does this mean that states have no possibilities of showing their particularities in a citizenship test? Not yet. Orgad has made a case for questions asking about historical facts which reflect the country’s road to becoming a liberal democracy (Orgad 2010a: 104). Asking the citizenship applicant to have knowledge about its

Justifying Citizenship Tests  29

particular road to democracy realises the fact that she seeks entry not in any but in this political community (Joppke 2010a: 3), thereby avoiding the trap of the ‘paradox of universalism’. Such facts form part of what Orgad has labelled a country’s ‘national constitutionalism’, by which principle the definition of the content of citizenship tests in his opinion should be guided (Orgad 2010a). Only information relating to the state’s specific constitution is essential enough to merit testing; all other questions should be dismissed as trivia.20 But when does an item form part of a country’s national constitutionalism? Orgad admits that this is difficult to define. In any case, questions about “values, principles, figures or events that have a tenuous factual connection to the Con­ stitution” must be excluded (Orgad 2010a: 103). Only information having normative significance essential to citizenship in a specific state merits testing. In Orgad’s opinion, the sine qua non principles and values required for obtaining citizenship in a specific liberal democracy need to be identified as principles that all citizens are to be expected to be familiar with and to endorse (Orgad 2010a: 98, 99, 104). But what are these sine qua non principles and values? The question, according to Orgad, and this is where his relevancy test differs from the test applied by Hampshire, should not be whether a certain principle is a liberal core value, but whether this value is an essential constitutional principle in the respective state (Orgad 2010a: 104). As an example, Orgad refers to the principle to protect and improve the environment, which, because it has its basis in the Dutch Constitution, is an essential constitutional principle in the Netherlands, whilst it may not be in other liberal democracies. In developing his test of national constitutionalism, Orgad appears to have departed from the presumption that there is agreement among citizens about the principles they are familiar with and endorse. This presumption however lacks an empirical base. As already indicated above, Orgad’s relevancy test leaves room for questions about history. A condition is that the knowledge asked about reflects a country’s constitutional history, relating to a country’s story of becoming a liberal democracy. Hence, a country’s history in general should not be asked about, but the history of a country’s Constitution (Orgad 2010a: 104).21 Taking Germany as an 20 Orgad has noted that it is also legitimate to ask would-be citizens to recognise and accept essential principles of European constitutionalism, to wit, EU constitutional principles and institutions (Orgad 2010a: 104). 21 Michalowski has criticised Orgad’s proposed separation of items of national history that deserve testing and items that do not. By pointing to the difficulty of separating relevant from less relevant items of history and asking where historiography about a country’s road to liberal democracy starts, she proposes to either consider all history questions as liberal or to only consider those questions liberal that correspond to the history and civic education knowledge that is required in regular school curricula (Michalowski 2010a: 6).

30  Chapter Two

example, questions relating to Kant and Karl Benz are according to Orgad less legitimate parts of history to ask about, whereas Bismarck’s 1871 Constitution, the Weimar Republic and World War II in his opinion are items of constitutional history which a liberal democratic state may legitimately ask citizenship applicants to be familiar with (Orgad 2010a: 104). Orgad admits that separating the less legitimate test items from the more legitimate items can be a challenge. There is a grey area of acceptable questions. Rather than disqualifying the test questions that do not pass the test of national constitutionalism for being illiberal, Orgad proposes to supply such knowledge to newcomers as orientation material, rather than make it mandatory in a citizenship test (Orgad 2010b: 23). The above shows that, in terms of content and effects produced, there are limits to the types of questions that a language and knowledge of society test may consist of in order to be justifiable from a liberal perspective.22 Questions that are objectionable from a liberal point of view may however play a role in tests which are defendable in other citizenship models, discussed below (paragraph 2.2.2). A Rejoinder by Liberal Minimalists As we have seen above, several authors have made an attempt to provide justification for citizenship tests under liberal theory. These authors argue that citizenship tests can be defended in a liberal citizenship regime, as long as their content meets certain demands.23 It has furthermore been argued that the introduction of citizenship tests, rather than being an exclusive phenomenon, are a “net win in liberality”, since they are “far more inclusive than alternative procedures involving subjective criteria” (Klekowski von Koppenfels 2010: 11). In this respect, Joppke has stated that citizenship tests actually increase “the naturalisation procedure’s calculability on the part of citizenship applicants, who are no longer subject to open-ended, individual interview procedure and the state-agent’s unfathomable discretion” (Joppke 2010b: 4). Klekowski von Koppenfels takes this argument one step further by stating that “a liberal citizenship test can in fact even institutionalise naturalisation, making it a normal and expected procedure for immigrants, rather than an exceptional step” (Klekowski von Koppenfels 2010: 12, italics in original text). This line of reasoning appears to depart from the presumption 22 Looking at all the conditions that need to be fulfilled for a test to be acceptable under the liberal model, and taking account of the fact that it is questionable whether tests are a viable means of assessing knowledge and measuring integration, one might ask the question whether it is not better to abstain from the introduction of a test altogether. Different means to improve the language and knowledge of society skills of aspiring citizens can be thought of. 23 As we have seen, to this condition must be added a second condition focusing on the implementation of the tests: their level may not be too high, they must be reasonably priced, and sufficient possibilities for exemption and preparation must be provided for.

Justifying Citizenship Tests  31

that discretion leads to arbitrariness, and that arbitrariness will lead to a stricter policy, and a higher number of refusals.24 Worried by the “narrowing of liberalism’s remit” (Kostakopoulou 2010a: 15), liberal minimalists have opposed this attempt to ‘squeeze in’ citizenship tests in liberal theory. Kostakopoulou for instance rejects the qualification of citizenship testing as liberal, merely by looking at the content of the tests: when determining whether citizenship tests are liberal, it is, in her opinion, equally required to examine their socio-political context, timing, discourse and effects (Kostakopoulou 2010a: 17).25 Taking account of these variables, Kostakopoulou concludes that citizenship tests cannot be reconciled with liberalism. Rather than with the ideal of liberal autonomy, Kostakopoulou associates the introduction of citizenship tests with concerns to reduce (unwanted) migration and number of naturalisations, and with exclusionary practices and restrictive migration control (Kostakopoulou 2010a: 17). As far as the effects of the tests are concerned, according to Kostako­ poulou, it does not matter whether the hurdle created is too high or a bit lower: “the crux of the point is that naturalisation is not liberalised; it is made more restrictive” (Kostakopoulou 2010a: 16). In similar vein, Carrera and Guild also reject the introduction of citizenship tests, whose nature and effects in their opinion are generally illiberal (Carrera & Guild 2010: 29). They observe a paradox in the testing of liberal democratic principles and values, which lies therein that the principles and values “are operationalised in a testing fashion and therefore become exceptions or derogative clauses for foreigners to have access to the very same liberal democratic entitlements and fundamental rights that citizenship tests are said to uphold” (Carrera & Guild 2010: 29). By consequence, they observe the ‘paradox of liberal toleration’ mentioned above even if the testing of the liberal democratic principles and fundamental rights concerns objective knowledge or a ‘what is right’ rather than the candidate’s inner values. In reply to those contending that the formalisation and standardisation are a net win in liberality, Carrera and Guild state that, as the tests are based on the “indeterminate legal concept of integration”, their introduction does not add to legal certainty and objectivity for the individuals subject to them (Carrera & Guild 2010: 32). In their opinion, wide discretionary powers still exist, but they have been passed on from the local official to the private actors and

24 Whether this assumption is correct will be examined in Chapter 7. 25 Kostakopoulou furthermore points out the unfairness of treating individuals as equals and fully integrated into society, when it comes to the extraction of their resources, and unequals, who must fulfil integration requirements, when it comes to the enjoyment of benefits (Kostakopoulou 2010a: 15).

32  Chapter Two

computers who determine “whether or not the applicant meets the ‘national proto­type’’’ (Ibid.). Carrera and Guild furthermore contend that citizenship tests conflict with respect for diversity and the principle of non-discrimination, both core liberal values, by sending the message that foreigners do not share the liberal democratic values which are tested (Carrera & Guild 2010: 31). Together with a reduction in the number of naturalisations, this, according to Carrera and Guild, is one of the core functions of citizenship tests (Carrera & Guild 2010: 33). In sum, Carrera and Guild contend that citizenship tests provide a case where “universalistic liberal democratic principles” function as “illiberal exceptionalism” (Carrera & Guild 2010: 33). Together with the other authors who make a stand against the liberal justification of citizenship tests, they remain faithful to the liberal minimalist stance that “the most liberal citizenship test is none at all” (Carens 2010: 19–20).26 2.2.2. Republican and Communitarian Critiques of Liberalism As stated above, republican and communitarian authors have criticised liberalism for its failure to acknowledge the meaning a community might actually have for its members. They argue that the liberal approach wants to give each resident, stranger or not, an equal position, but does not have a sufficient answer to the question into what immigrants have been admitted. They thereby reproach liberal authors for being blind to the ‘beyond-individual’ aspects of the political community (Tholen 1999: 43). The importance of these aspects has been translated into arguments for the imposition of naturalisation requirements. Rather than emphasising the equality of all residents of a certain territory, as liberal theory does, republicans and communitarians will only accept a resident as a citizen if he or she will also effectively form part of the collective of the community. In order to become a member, a resident will hence have to meet certain requirements; citizens are required to have certain capacities and to display a certain loyalty (Tholen 1997: 197). Requiring from prospective citizens that they have merely resided on the state’s territory for a certain period of time and that they have a clean criminal record will hence in the opinion of most republican and communitarian thinkers generally not suffice. Since republicans and communitarians are mainly concerned with the wellbeing of the community and its survival, the fate of those who do not form part of the community is of a lesser concern. Exclusion of those non-members who do 26 Note that, even though other accounts of liberalism, such as liberal-egalitarianism and libertarianism, have criticised liberalism, in both accounts, a citizenship test remains hard to justify. See Bauböck 1994, Fermin 1999, Kostakopoulou 2003, Kymlicka & Norman 1994, Nozick 1974, Tholen 1997 & De Vries 2008 for explanations of liberal-egalitarianism and libertarianism.

Justifying Citizenship Tests  33

not have the qualities on which the survival of the community depends therefore is more easily justifiable under the republican and communitarian models than it is under the liberal model. It will be discussed below how republicans and communitarians see the community and which requirements can accordingly be formulated for aspiring citizens. 2.2.2.1. The Republican Account of Citizenship As we have already seen, the republican conception of citizenship prioritises the dimension of citizenship as activity. Emphasis is put on political participation. Republicanism sees the community as a society of mutually vulnerable citizens who have voluntarily subjected themselves to a collective coercive authority. Republican citizenship is understood as membership in this self-governing community (Honohan 2010b: 91). Citizens have a joint responsibility for the common good, from which it follows that they need ‘public spirit’ or ‘civic virtue’. These virtues are a willingness to acknowledge and assume responsibilities entailed by interdependence, self-restraint in pursuing individual or sectional interests rather than the common good, and the inclination to engage open-mindedly with the viewpoints of others. These political virtues do not prescribe a comprehensive vision of the good life (Honohan 2010b: 94). Rather, republican citizens, through deliberation and debate, aim to strive for the promotion of the common good in the morally and culturally diverse contemporary society. Republicans criticise the passive conception of citizenship in the liberal model. Rather than thinking of citizenship as a right, republicans primarily see it as a role or an office (Fermin 1999: 108). According to Van Gunsteren, the ‘duty’ of a citizen is to “transform a community of fate into a republic that can be willed by all who are involved as citizens” (Van Gunsteren 1998: 26). Citizenship hence entails powers and responsibilities that cannot be defined entirely in terms of legal or binding requirements flowing from the legal bond between a citizen and a state, but that depend on the broader attitudes and inclinations of citizens as well. Compared to its liberal counterpart, republican citizenship consequently involves more public orientation, more active engagement, and a more significant degree of solidarity. Republicans furthermore support a more militant form of tolerance than liberals, since in their opinion, diversity and difference of opinion are conditions for public deliberation and political activity. Citizenship consequently concerns acceptance of and competence in ways of dealing with cultural and social differences, not acceptance of one style of life that happens to be dominant (Van Gunsteren 1998: 93–96). This also means that citizens do not necessarily share a deeper or overarching (cultural) identity with all other citizens, which is characteristic in citizenship in the communitarian sense (see below). Or, in Mason’s words, rather than sharing a sense of “belonging

34  Chapter Two

together”, citizens should share a sense of “belonging to the polity” (Mason 1999).27 According to Mason, ‘belonging to the polity’ is different from a ‘sense of belonging together’, that is, the citizens’ belief that there must be some special reason why they should associate together, which might be provided by the belief that they share a history, religion, ethnicity, mother tongue, culture or conception of the good (Mason 1999: 272). As a consequence, future citizens need not abandon their particular ethnic identification and cultural commitments; all that is required is that they embrace the civic republican ideal of the common good. Consequently, members of the republican community are ‘patriots’ rather than ‘ethnics’ (Kostakopoulou 2003: 92). Whereas the bond between citizens in the communitarian model can be thought of as the bond between family members, the bond between citizens in the republican model can be compared to the bond between colleagues: it may exist between those who are relatively diverse, separate and distant, but interdependent, semi-voluntarily, in a way that gives them common concerns (Honohan 2010a: 7). Accordingly, citizenship is strong rather than thick: it is a bond strong enough to unite the members of a polity who are otherwise thoroughly divided by their private interests (Bauböck 2001: 43). It is not thick in the way that citizenship in the communitarian sense is, since the emphasis is on political socialisation; the polity does not reject the participation of individuals who have not adopted a fuller set of dominant cultural values (Neuman 1994: 241). ‘Classical’ republicanism has been criticised for its idolisation of dedication to the public cause, which sharply contrasts with the contemporary practice in which people predominantly look for happiness in their private lives (Fermin 2001: 72). Contemporary accounts of republicanism acknowledge that public participation has instrumental as well as intrinsic value, rather than seeing it as the ultimate value in human life. From this it follows that there is no requirement for any particular level of constant participation for citizens (Honohan 2010b: 93). The Republican Model for Naturalisation What does the above imply for conditions for naturalisation? One could say that a regime founded on republican principles would appear to have the greatest stake in encouraging aliens to seek naturalisation, since higher rates of naturalisation mean increasing the number of people able to fully participate and deliberate in

27 According to Mason, a person has a sense of ‘belonging to the polity’ “if and only if she identifies with most of its major institutions and some of its central practices and feels at home in them.” ‘Identification with most of its major institutions’ entails the citizens’ perception of them as valuable, conducive to their flourishing and reflective of their concerns (Mason 1999: 272). ‘Feeling at home in them’ refers to the ability of the citizens to find their way around institutions and to experience participation in them as natural.

Justifying Citizenship Tests  35

the political life of the country (Schuck 1994: 330). This however does not mean that citizenship is granted to anyone who happens to be present on the state’s territory, since, as we have seen, republican citizenship presupposes a certain level of commitment. Since a higher degree of commitment can be expected from republican citizens than from liberal citizens, the republican framework offers more room for the imposition of extra naturalisation requirements than those justifiable under the liberal model (Neuman 1994: 260). Because the purpose of naturalisation is to admit the applicant to the political project of republican self-government, the question of whether the naturalisation requirements are justifiable under the republican model depends on whether the criteria focus on the essential features of the project of self-government (Neuman 1994: 261). Van Gunsteren has formulated three basic requirements: a) general citizen competence defined as capability of dialogic performance; b) competence to act in this particular community, operationalised as knowledge of the language and respect for the laws; c) access to an ‘oikos’, i.e. a reasonably secure access to means of existence in order not to be forced to sell oneself or one’s judgment (Van Gunsteren 1988: 736).28 That there is no agreement among republican authors regarding the imposition of naturalisation requirements becomes clear from Honohan’s disqualification of stringent conditions for naturalisation in the name of republicanism. In her account of republican requirements for naturalisation, residence can be taken as the shorthand for interdependence and the sharing of a common future, by virtue of living, working, paying taxes and sending children to school. Since the primary basis of citizenship is subjection to a common authority, those who are long-term residents are already, in most cases, in the same predicament in this respect as citizens (Honohan 2010b: 99–100). Naturalisation should hence take place almost automatically after residence. Honohan on the other hand also argues that “it will be desirable” that applicants for citizenship are provided with the knowledge of structures of society and economy in their new country, and that the state may require participation in language classes and in certain practical political exercises (Honohan 2010b: 91). But applying a republican standard, Honohan disqualifies requiring applicants to achieve a certain fixed standard in tests of knowledge, skills or values: “few conditions not required of native born citizens should be required of those naturalising, and these should be more a matter of participation than of skills or identity” (Honohan 2010b: 91). This implies that, apart from maybe the length of the period of residence required, in her view there is no significant difference from the criteria for naturalisation justifiable under the liberal model. 28 ‘Oikos’ refers to the basic unit of society in ancient Greece.

36  Chapter Two

A Republican Citizenship Test? Based on those republican authors who have argued for the justifiability of criteria for naturalisation, besides residence, does the republican model allow for the justification of a citizenship test as a condition for naturalisation? Let us first see whether a language requirement is defendable with republican arguments. In republican theory, active participation in public life is the most important virtue (Orgad 2010a: 87). Citizens should be capable of communicating and engaging with others in political deliberations. Without a doubt, absence of language skills will hamper participation in the political decision-making process and the integration of immigrants into common public institutions (Kostakopoulou 2003: 99). Consequently, from the republican perspective, prospective citizens’ investment in acquiring knowledge that will improve their political participation represents the exercise of civic virtues (Neuman 1994: 266). Given these considerations, the republican model appears to justify the imposition of a language requirement for naturalisation. Does it follow from this that testing language skills at a certain level in a formalised citizenship test is also justified? Not yet. The level of language skills required should be determined by the needs of the republican project (Neuman 1994: 265). And since contemporary accounts of republicanism do not require all citizens to be involved in public and political debate to the same degree, requiring a uniform level of language skills appears to be hard to defend. For those who are ‘merely’ interested in casting their votes in local and national elections, basic oral language skills which will suffice to understand the national television and radio will probably do.29 Asking all applicants to pass a uniform test in which written language skills are also required can hence hardly be justified. Consequently, the republican model appears to either only allow for a relative language test, developed for the needs of each individual citizenship applicant, or a test at a basic level high enough to make sensible use of one’s voting rights. What about testing knowledge of society? If such a test focuses on the essential features of the republican project, a knowledge of society requirement could be defended. This means that questions could for instance focus on knowledge necessary to make a sound political judgment, or to understand the country’s 29 Kostakopoulou however signals that it has never been demonstrated that lack of linguistic competence undermines participation, or that lack of linguistic competence makes participation impossible. A historical look at migration and settlement according to Kostakopoulou shows that immigrants, having no or only a basic knowledge of the host language, have contributed effectively in the workplace and society, and by speaking and writing in their home language can be both active and concerned members of society (Kostakopoulou 2006: 102–103). According to Kostakopoulou, there hence is no justification for the imposition of a language test in the republican model, not even if the level required is very basic.

Justifying Citizenship Tests  37

political structures and laws. Furthermore, since citizenship in the republican model is an attitude, questions figuring in a knowledge of society test may aim at more than testing mere objective, cognitive knowledge. Contrary to a liberal citizenship test, a republican test may hence aim at realising a certain desirable behaviour in citizens, which goes further than merely requiring them not to transgress the applicable laws. Since citizens are jointly responsible for the common good and share a future, questions aimed at encouraging future citizens to start caring about the preservation of the environment, such as taking time to recycle waste, or to engage in voluntary work or activities of care, can, for instance, be justified. Since future citizens cannot be required to abandon their particular ethnic and cultural identification, questions relating to desirable behaviour which aim at wiping out or ignoring differences originating from this identification, on the other hand, do not merit justification. Whereas a citizenship test can, under certain circumstances, be justified in the republican model, certain authors have pointed to the limited value of citizenship tests as the appropriate means to make sure that immigrants have the attitudes required for republican citizenship. Honohan has claimed that it seems more effective to promote the attitudes and political values required for republican citizenship in another way, for instance by making it a participatory exercise, comparable to a citizen’s jury (Honohan 2010b: 102). Orgad has mentioned the Danish Declaration of Active Participation or the British idea of active citizenship as “more reasonable” qualifications for citizenship in the republican model (Orgad 2010a: 88). 2.2.2.2. Communitarian Citizenship As stated above, communitarians, like republicans, reproach liberals for not paying sufficient attention to the value and meaning actual communities have for individuals (Tholen 1997).30 Communitarian theory therefore offers an alternative to an understanding of community which only has an instrumental meaning for individuals (Tholen 1997: 155). Unlike liberals, communitarians contend that meaningful life is only possible inside the community: only in the community can individuals reach full development. In order to preserve their own development, individuals hence need to safeguard the community’s existence (Pluymen 2008: 316). And this requires actual devotion and commitment to the 30 Tholen (1997) indicates that it is not entirely clear which authors belong to the canon of communitarian authors. Most surveys in any case mention Michael Sandel, Alasdair MacIntyre, Charles Taylor and Michael Walzer. Whereas Walzer is often rated among communitarian authors, as we have seen, his ideas on naturalisation of those who have been admitted to the territory of the community are qualified as ‘liberal’. His ideas on the community itself and its importance for individuals can however be qualified as communitarian.

38  Chapter Two

community (MacIntyre 1984, Sandel 1984, Taylor 1989, Tholen 1999: 40). Conse­ quently, in the communitarian perspective, naturalisation policy should be dictated by the ideals of the community: only if the newcomers can identify with these ideals will they be admitted as members. What these ideals are depends on how the community perceives itself. But generally, communities in the communitarian sense are ‘communities of fate’ with a common past and an ideal for the future. Communitarians argue that citizenship is rooted in a culturally defined community, which forms the foundation of the political community (Delanty 2002: 159–160). Members of the community hence share a culture, in which language, religion and norms and values can play an important role, and a history.31 A state can therefore appeal to ‘the national culture’ (Tholen 1997: 173). What does the above mean for the formulation of naturalisation requirements and the justification of a citizenship test? The Communitarian Model for Naturalisation Communitarian citizenship emphasises the dimension of identity. Since the survival of the community depends on the preservation of the community’s identity, only those who share the identity the state intends to materialise may naturalise. (Potential) members of the community are therefore expected to adhere to a set of shared values, serving as a framework or ‘glue’ (Etzioni 2004: 171). Consequently, rather than offering all permanent residents a right to naturalisation, applicants for citizenship will have to meet rather strict conditions and even be asked to ‘assimilate’ into the majority culture (Kostakopoulou 2006: 93). The communitarian model hence considers naturalisation to be the end point of a completed integration process, rather than a step in this process (Entzinger 2008). From the communitarian perspective, the government’s stance on naturalisation is not one of encouraging aliens to become citizens, as it is in the liberal and republican models. Rather than seeing the alien’s wish for naturalisation as sufficient evidence that she possesses that identity, it is at most regarded as an indication of a degree of solidarity felt with the communal identity. At the same time, immigrants’ failure to seek naturalisation may be perceived as strong evidence that they do not share that identity (Schuck 1994: 330–331).32 A Communitarian Citizenship Test? Does the communitarian model offer room for a citizenship test in which knowledge of the language and of society are tested? Since in the communitarian model 31 The communitarian authors see the modern state as such a community, or in any case present such a community as an ideal to the modern state (Tholen 1997). 32 According to Schuck, this appears to be the view taken by many Germans of the unwillingness of many long-term resident Turks and other ethnic minorities to naturalise despite efforts by Germany to ease the preconditions for their doing so (Schuck 1994: 330–331).

Justifying Citizenship Tests  39

only completely integrated or even assimilated immigrants will be able to naturalise, a case could be made for the application of demanding integration conditions. However, the communitarian standard also sets limits: applying the communitarian test, requirements for naturalisation have to be required for participation in the identity of the community. As the communitarian model sees language as a prominent defining factor in national identity, and mastering the language as a sign of allegiance to the nation’s identity, a language requirement can be justified in the communitarian model. Because of the importance of language for the national identity, the requirement may be more strictly applied in the communitarian model than in its republican counterpart, where language skills are merely functional. A uniform level of minimum language skills can hence be justified, and so can probably also the requirement to have reading and writing skills. What about testing knowledge of society? If a knowledge of society examination is seen as a means of testing whether the future citizen adheres to the state’s identity, the requirement to have sufficient knowledge of society also fits well in the communitarian model. There are however limits to the content of the knowledge required: only questions relating to the state’s identity are justified in a communitarian citizenship test. This leaves the state a broad range of subjects from which to choose. Since immigrants are required to have socially and culturally adapted to the community and identify with it before they can be admitted as citizens, the knowledge tested may range from the country’s history, geography, or customs, but may also focus on prevailing norms and values.33 Furthermore, in a communitarian citizenship test, questions may aim at scrutinising a candidate’s inner dispositions, i.e. ask ‘what is good’, to see whether these match the identity of the state, as perceived by the state. Faced with the presence of a supposedly ‘illiberal’ immigrant community, which has displayed illiberal behaviour in the form of honour killings, female genital mutilation and forced marriages, and following the terrorist attacks on American and European soil, Western states tend to perceive this identity as ‘liberal’. To separate the liberal from the illiberal, future citizens are supposed to identify with or adhere to liberal norms and values. Liberalism has thereby transmuted into an identity, an ethical way of life to which everyone is expected to conform (Joppke 2010b: 138). 33 Testing whether future citizens adhere to or are at least aware of the prevailing norms and values however presupposes an unambiguous code of conduct or set of norms or rules adhered to by all citizens (Prins 2010: 3). This presupposition however will often be false, since for a community as large and diverse as modern states, such a code of conduct will be lacking. This means that there will no obvious right or wrong answers to questions about the code of conduct. The problem with citizenship testing in the communitarian model will hence be that the diversity and heterogeneity in society are underestimated.

40  Chapter Two

This leaves room to exclude those immigrants whose supposedly illiberal ideas are at odds with the state’s liberal values. The above shows that, in the communitarian model, states may use citizenship tests as a way to emphasise or express their identities. Identity can however also be expressed by not allowing someone full access to the political community. Hence, by showing a high degree of national self-determination in terms of citizenship admission, states can express a communitarian understanding of citizenship even without resorting to formalised citizenship tests. This way, states that use informal rather than formal tests of aspiring citizens’ language skills, knowledge of society and values can have low naturalisation rates. 2.3. Conclusions The question of which theoretical models for citizenship allow for the introduction of citizenship testing, and under what circumstances, has been answered above. In the overview, the liberal model for naturalisation has received more attention than the republican and communitarian models. The reason for this discrepancy in the amount of attention is obvious: whereas the introduction of a citizenship test is more or less self-evident in the republican and communitarian models, provided that the tests meet certain conditions, the imposition of barriers to naturalisation is much harder to justify in the liberal model. As we have seen above, liberal authors have sought to provide the arguments required to fit the issue of citizenship testing in the liberal model, inter alia by formulating standards to which a liberal citizenship test must comply. The large part of the academic debate on citizenship tests has focused on the formulation of these standards. The attempt by several scholars to try to ‘squeeze in’ the requirement of passing a formalised citizenship test as a condition for naturalisation in the liberal model has resulted in criticism by liberal minimalists. The point of view that I hold in this study is that a citizenship test, depending on its content and the reasons why it was introduced, may be defendable in the liberal model, provided the test does not lead to exclusion. For this reason, and as explained in the introduction, I also examine the actual effects produced by the citizenship tests in the countries under consideration, an issue which so far has been largely ignored in the research on citizenship testing.

Chapter Three Repairing the Failures of the Past: The Introduction of a Citizenship Test in the Netherlands A formalised language and knowledge of society test was introduced into Dutch nationality law on 1 April 2003. This ‘naturalisation test’ replaced the integration interview that used to be conducted by local officials. Four years after its introduction, on 1 April 2007, the naturalisation test was replaced by the ‘integration examination’, the passing of which is also a condition for a permanent residence permit. The introduction of the naturalisation test into Dutch nationality law can be seen as the marker of a restrictive turn in Dutch citizenship policy, which until then had figured among the more open in Europe. The relatively easy naturalisation requirements together with the attribution to ethnic minorities of the right to maintain their cultural identities led Koopmans et al. to refer to the Netherlands as an “almost ideal typical example of a multicultural citizenship regime” (Koopmans et al. 2005: 8).1 The introduction of formalised tests, focusing on Dutch language skills and knowledge of Dutch norms and values however appears to indicate a rejection of multiculturalism. How can this turn in Dutch citizenship and integration policy be explained? And is it as radical as sometimes presumed, or has the introduction of a language and integration test only led to a marginal restriction in an otherwise still very open citizenship policy? This appears to be the conclusion of the Migrant Integration Policy Index (MIPEX), in which, of the EU-15, the Netherlands occupies fourth place in a ranking of policies regarding access to citizenship.2 From an analysis of Dutch citizenship policy from the 1980s until 2008, Howard even concluded that the Netherlands is a country that can be categorised among those 1 We will however see that Dutch integration policies focused far less on preservation of ethnic minorities’ identities than often claimed. Whether these policies are as ‘multicultural’ as generally assumed is hence highly questionable. 2 With a score of 66, the Netherlands is less open than Belgium, Portugal and Sweden, but more open than the other states (except for Luxembourg, which also scores 66 points; MIPEX results of 2010, available at http://www.mipex.eu/play/bar.php?chart_type=bar&countries=9,10,15 ,17,18,19,20,22,23,27,29,32,36,37,39&objects=180&periods=2010&group_by=country, site accessed on 8 April 2013).

42  Chapter Three

countries that liberalised their citizenship policies in the 1990s and the early 2000s (Howard 2009: 73). Apparently, Howard does not see the naturalisation test and its successor as interfering with the image of the Netherlands as having an open citizenship policy.3 This chapter will assess whether and to what extent these qualifications of Dutch citizenship policy require adjustment. To answer the above questions, the political debates preceding the changes in the Dutch Nationality Act will be analysed, starting with an explanation of the 1985 Act, which opened up access to Dutch nationality. 3.1. From Informal Interview to Formalised Test: Debates on Dutch

Nationality Law from 1985 to 2006

3.1.1. The 1985 Dutch Nationality Act and the Minorities’ Policy On 1 January 1985 a new Dutch Nationality Act (DNA) came into force, replacing the former 1892 Act.4 The text of the 1985 Act was heavily influenced by the integration policy which had been adopted shortly before its entry into force. This policy had been implemented after the publishing of the Scientific Council for Governmental Policy’s (WRR) report ‘Ethnic Minorities’ in 1979, in which the Council acknowledged that most ‘guest workers’ would settle in the Netherlands, instead of returning to their home countries.5 Until then, the presence of minorities in the Netherlands had been considered to be temporary, and the government had repeatedly firmly declared that the Netherlands was not an immigration country (De Heer 2004: 178, Lechner 2008: 152). This attitude constituted the basic ideology of a ‘thin’ minorities’ policy, which aimed at preserving the immigrant’s identities (De Heer 2004: 178). Hence, the strategy of ‘integration with preservation of the own identity’ was a pragmatic one, aimed at preparing guest workers for their return (Vink 2007: 345). However, now it had become clear that most immigrants would not return to their countries of origin, the WRR argued that the disadvantages these permanent members of society were suffering were no longer acceptable (Heijs 1995: 179).

3 Even though Howard notes that the Netherlands “have been moving in a somewhat more restrictive direction in the 2000s”, he concludes that, observing the time horizon spanning the 1980s to the end of 2008, “the Dutch story is still one of liberalising change” (Howard 2009: 83). 4 Act of 19 December 1984, Staatsblad 628. 5 Guest workers are labour migrants who, between approximately 1960 until 1973, the year of the oil crisis, had been recruited from Mediterranean countries such as Spain, Portugal, Yugoslavia, Greece, Turkey, and Morocco. In 1973, 129,500 labour migrants from the countries where recruitment had taken place lived in the Netherlands (Penninx, Martens & Smeets 1998: 1050–15).

Repairing the Failures of the Past  43

In order to promote the equal participation and emancipation of ethnic minorities, the Council advised the government to adopt a policy aimed at improving their legal position. The notion of ‘preservation of own identity’ according to the Council needed to be replaced by a more active encouragement of immigrants to participate in society (WRR 1979: xxi, Vink 2007: 344). In reaction to the WRR report, in 1983 the government published the ‘memorandum on minorities’ (Minderhedennota), which announced the introduction of an actual integration policy, the so-called ‘minorities’ policy’ (Minderhedenbeleid), the overall objective of which was achieving a society in which all members of minority groups in the Netherlands, individually and also as groups, are in a situation of equality and have full opportunities for their development.6

A central element of the new policy was strengthening the legal position of nonDutch minorities, and thereby their integration. The Dutch Nationality Act was mentioned as being of special importance for the policy (Heijs 1995: 179). The minorities’ policy hence heavily influenced the final text of the 1985 Dutch Nationality Act. By stipulating the conditions for naturalisation, it became clear that, rather than a favour, naturalisation was seen as a right. The new Act mentioned the following requirements: • being at least 18 years of age, • having been granted a residence permit for a purpose not limited in time, • having been residing in the Netherlands for at least five consecutive years prior to the application, • having made an effort to renounce one’s foreign nationality, unless renunciation could not reasonably be demanded, • not constituting a danger to public order, • having reasonable knowledge of the Dutch language and being accepted into Dutch society. Fulfilment of the last condition, referred to as the language and integration requirement, was tested by a local official in a short interview. Since naturalisation was seen as a means for integration, the official manual with instructions for municipal officials handling naturalisation applications provided that “the concept of integration serves to indicate a process of development” and that “at the moment of naturalisation, it is not required that the future citizen has completed the process leading to a feeling of solidarity with Dutch society”.7 According to 6 TK 1982–1983, 16 102, no. 21, p. 10, translation provided in Entzinger 2003: 63. 7 Manual for the Application of the Dutch Nationality Act 1985, June 2001, A1 - Article 8, p. 19.

44  Chapter Three

the manual, applicants who spoke Dutch adequately and who had social contacts with Dutch nationals would fulfil the language and integration requirement. In practice, only the language test was used. Written language skills were explicitly not demanded.8 The new naturalisation policy appeared to be successful, as numbers for naturalisation rose considerably after the adoption of the 1985 Act. As far as the new minorities’ policy itself was concerned, the government in 1987, i.e. four years after its adoption, concluded that it had been insufficiently successful, thereby again turning to the WRR for advice (Heijs 1995: 198). In its 1989 report Allochtonenbeleid (Immigrants Policy), the WRR stated that the minorities’ policy pursued so far had failed to provide immigrants the possibilities to become selfsupportive by treating them too much as ‘care categories’. It advised intensified policies in the fields of employment and (adult) education. For the first time, the need for compulsory language and education courses for immigrants was mentioned (Klaver & Odé 2009: 44). The improvement of the immigrants’ legal position in the 1989 report occupied a far less central position (Heijs 1995: 199). The WRR nevertheless made some important recommendations to ameliorate the immigrants’ legal position, one of these being not to make naturalisation more difficult than strictly necessary. Consequently, the government, consisting of the Christian democratic CDA and the social democratic PvdA, in the memorandum ‘multiple nationality/voting rights for aliens’ of May 1991, decided to abolish the renunciation requirement.9 The memorandum represented a compromise between the government parties: where the Christian Democrats gave up their objections to dual nationality, the Social Democrats gave up their wish to extend voting rights for non-national residents to the national level (Van Oers, De Hart & Groenendijk 2010: 15). In November 1991, the Minister of Justice, with the Second Chamber’s consent, used his statutory discretion to abolish the renunciation requirement.10 The measure, which was meant to increase the number of naturalisations, led to a considerable rise, in particular among Turkish and Moroccan nationals and refugees (Van Oers, De Hart & Groenendijk 2010: 15).11 Its goal had hence been achieved. A peak was reached in 1996, when almost 54,000 immigrants acquired Dutch nationality via naturalisation. 8 Manual for the Application of the Dutch Nationality Act 1985, June 2001, A1- Article 8, p. 20. 9 TK 1990–1991, 21 971, no. 14. 10 Circular of Secretary of State of Justice of 20 December 1991, Migrantenrecht 1992, p. 62. 11 After the abolition of the renunciation requirement, the naturalisation rate of Turkish nationals quickly increased, reaching a peak in 1996 with 20% (Böcker, Groenendijk & De Hart 2005: 163). After the reintroduction of the renunciation requirement, the Turkish naturalisation rate dropped again, to 5% in 1999–2001. See paragraph 7.2.3 for more detailed information on the development of the Turkish naturalisation rate.

Repairing the Failures of the Past  45

3.1.2. Revision of the Dutch Nationality Act The 1993 Proposal of Law: Naturalisation as a means for Integration A proposal to formally delete the renunciation requirement from the Dutch Nationality Act was put forward in 1993. Now the renunciation requirement as an instrument for testing integration had disappeared, the government intended to see whether “special measures needed to be taken to give the research into the knowledge of Dutch language something to go on”.12 In the government’s view, the renunciation requirement and the language and integration requirement apparently functioned as communicating vessels. Another reason for amending the language and integration requirement was that the requirement had been criticised for its vagueness (Jessurun D’Oliveira 2005: 39).13 The manual for the application of the naturalisation requirements for instance provided for a more lenient treatment of several categories of immigrants. To prevent disproportionate harshness, poor knowledge of the Dutch language should not stand in the way of naturalisation applications from elderly people, those with limited or no education, illiterates, and married women whose integration “had fallen behind compared to the integration of their husbands”.14 But what exactly was understood by ‘a more lenient treatment’ or ‘limited education’? The vagueness of the requirement led to a divergent application of the requirement in practice. Research conducted in 1988 showed that in more than 10% of cases, reading and writing Dutch were tested in addition to speaking and understanding (Heijs 1988). A more detailed formulation of the requirement would, according to the government, lead to a more uniform application, which would in turn enhance equality and fairness among the applicants. The bill thus provided for a reformulation of the language and integration requirement by asking for “a reasonable knowledge of Dutch language in the light of general social intercourse”.15 During the debates on the proposed amendments, the Christian Democrats in parliament appeared to have developed a more restrictive view on naturalisation than the government, of which the Christian democratic party also formed part.

12 Memorandum multiple nationality/voting rights, TK 1990–1991, 21 971, no. 14, p. 19. 13 This led Jessurun D’Oliveira in 1983 to conclude that “a barrier which was introduced as an ideological condition, based on social-political considerations can grow out to become an impossible obstacle, comparable to voting rights for blacks” (1983: 1302). As we will see, Jessurun D’Oliveira’s statement was rather prophetic. 14 Manual for the application of the Dutch Nationality Act 1985, June 2001, A1, p. 21. According to the Manual, “the beliefs of certain minority groups regarding the position of the woman” should not stand in the way of naturalisation. 15 Article 8 Paragraph 1 sub d of the proposed Act.

46  Chapter Three

Rather than seeing naturalisation as a means of integration, they saw the acquisition of Dutch nationality as the “legal and emotional completion of the integration”.16 Since acquiring Dutch nationality entailed obtaining certain rights and duties, applicants could, in their opinion, be expected to be committed to Dutch society.17 According to Christian Democratic parliamentarian Verhagen, Dutch nationality should therefore not be handed out “too easily”: If Dutch citizenship is desired for the rights and duties that come along with it, the person who becomes a Dutch citizen should be able to fully make use of these. What use are voting rights for him, in case he is unable to exercise them? What use is the right to be appointed in certain functions, in case he can’t execute this right due to an insufficient knowledge of language or society?18

Consequently, in 1994 Christian Democratic parliamentarians proposed requiring written language skills, as well as knowledge of the Dutch society and state, to be assessed in a formalised test. The Christian Democrats were eventually supported by the Conservative Liberals (VVD), who also felt that Dutch citizenship was “not something that can be given away just like that” and that something might be asked in return.19 Parliamentarians on the left of the political spectrum on the other hand opposed a strengthening of the language and integration requirement. Dittrich, member of the progressive liberal party D66, for instance argued that acquiring reading and writing skills would pose problems for illiterates, persons coming from countries using another alphabet and the handicapped.20 Dittrich furthermore pointed out the problems that women, who under the current practice would profit from a more lenient application of the language and integration requirement, would face when this ‘compensation criterion’ no longer applied.21 Together with Rabbae (GreenLeft) he fiercely opposed testing knowledge of the Dutch state and society. Dittrich referred to such a requirement as “undesirable”, while Rabbae reproached the Christian Democrats for only wanting to allow “top citizens” to live in the Netherlands.22 Despite initial opposition from the left, the Second Chamber eventually accepted that a formalised language and knowledge of society test would be introduced, in which a professional language teacher would assess an applicant’s

16 TK 1993–1994, 23029 (R1461), no. 9. 17 TK 1993–1994, 23029 (R1461), no. 22. 18 MP Verhagen, TK 1995–1996, 23 594 (R 1496), no. 23, p. 4. 19 TK 1995–1996. 23 594 (R 1496), no. 23, p. 3. 20 Hand. TK, 22 February 1995, 51–3272, TK 1995–1996, 23 594 (R 1496), no. 23, p. 3. 21 TK 1995–1996, 23594 (R 1496), no. 23, p. 3. 22 TK 22 February 1995, 51–3278, TK 1995–1996, 26 September 1995, p. 4–230.

Repairing the Failures of the Past  47

knowledge of Dutch society in a conversation.23 This test, which would not test written language skills, would, however, never be put into practice, and neither would the proposed Act, as the First Chamber could not agree on the abolition of the renunciation requirement. The CDA, which after the 1994 elections no longer formed part of the government, no longer felt obliged to keep to the compromise on dual nationality reached with its former coalition partner PvdA. Together with the Conservative Liberals, the Christian Democrats did not interpret the rise in the number of naturalisations as an indication that the goal of the policy of not applying the renunciation requirement had been reached, but as a proof that acquiring Dutch nationality had become too easy.24 Without ever having been officially abolished, the renunciation requirement was re-applied in 1997. In February 1998, the government introduced a new bill for the amendment of the Dutch Nationality Act.25 The 1998 Proposal of Law: Naturalisation as Reward In the new bill, both the renunciation requirement and the reformulation of the language and integration requirement were upheld. The requirements thereby apparently no longer functioned as communicating vessels. In the debates, which had become more emotional and ‘nationalist’ (Entzinger 2003: 75), the language and integration requirement called the tune. Several explanations can be given for the change in the tone in the debates, which had already been set by the Christian Democrats in parliament during the discussions on the previous bill. One of the reasons is the outcome of the 1994 elections, which were won by the VVD. The Conservative Liberal victory is often ascribed to a speech which prominent VVD party member Frits Bolkestein held at the Liberal International in Zürich in 1991. In this speech, Bolkestein referred to the supposed incompatibility of Islam and Western values. Bolkestein opposed the ‘culture-relativism’ which had characterised the supposedly noncommittal integration policy of the 1980s, and stated that immigrant integration needed to be addressed ‘with guts’. It is often said that a ‘silent majority’ among voters who agreed with Bolkestein revealed itself at the 1994 elections (Entzinger 2003: 71). Whereas it is unclear to what extent the VVD’s electoral gain in 1994 directly resulted from Bolkestein’s speech in 1991, it did provoke a debate which put the integration issue high on the political agenda (Bonjour 2009: 193). Shortly before the 1994 elections, and perhaps in the hope of preventing voters from switching to the VVD, the PvdA/CDA Government published an agenda-setting memorandum entitled ‘Ethnic Minorities Integration Policy’ (Contourennota Integratiebeleid 23 Hand. TK, 21 February 1995, 50–3200. 24 TK 1995–1996, 23 January 1996, 45–3420. 25 TK 1997–1998, 25 891 (R 1609), no. 1–2.

48  Chapter Three

Etnische Minderheden). By stressing the immigrants’ own responsibility for their integration, the ability to manage independently, and ‘active citizenship’, the memorandum marked a change, and introduced the so-called ‘Integration Policy’. From now on, immigrants would be required to participate in the labour market and educational system, to learn the Dutch language and to become acquainted with Dutch society (Klaver & Odé 2009: 44). Six years after Bolkestein’s speech, at the beginning of 2000, publicist and prominent PvdA member Paul Scheffer published an influential article entitled ‘The Multicultural Drama’ (Het Multiculturele Drama) in a renowned newspaper.26 In this article, Scheffer signalled the development of an ‘ethnic subclass of immigrants’, notably Muslims, who did not feel a bond with Dutch culture and society and who could not and did not want to integrate. The existence of this ‘subclass’ would, according to Scheffer, eventually lead to the undermining of social cohesion and liberal democracy, especially due to the illiberal ideas Muslims were said to have. Immigrants should hence be required to adapt to the principles of liberal democracy, and to have a better knowledge of Dutch culture and history (Entzinger 2003: 78–79). The publication of ‘the multicultural drama’ triggered debates on the role of culture in integration (Klaver & Odé 2009: 48). During the so-called ‘Scheffer debate’ in the Second Chamber, the CDA was particularly critical vis-à-vis the integration policy, which, despite the more recent focus on duties against rights, according to this party had long been too permissive. They stressed the poor cultural immigrant integration, including the shortcoming in language skills.27 It was in this atmosphere that the debates on the 1998 proposal of law to amend the DNA took place. During the debates, Verhagen continued to strive for the introduction of a stricter language and integration requirement. To reinforce his arguments, he referred to a 1996 report of the Social Cultural Planning Bureau, from which he drew the conclusion that half of all naturalised citizens did not feel Dutch, which he thought was undesirable. Instead of being a consumer article, Verhagen thought Dutch nationality should be something to be proud of, and that the acquisition of Dutch nationality should be the crowning of the integration.28 Verhagen furthermore used the introduction of a formalised test within the framework of the Newcomers Integration Act (NIA), introduced in 1998, as an argument for the introduction of a formalised language and integration test for naturalisation. The NIA obliged immigrants from outside the EU who had settled in the Netherlands for a non-temporary purpose to follow an integration course,

26 NRC Handelsblad, 29 January 2000. 27 Report ‘Building Bridges’, Blok Commission, TK 2003–2004, 28 689, no. 9, p. 53. 28 TK 1999–2000, 16 February 2000, p. 50–3635, TK 1999–2000, 17 February 2000, p. 51–3685.

Repairing the Failures of the Past  49

provided by the government at no cost, which consisted of 500 hours of language education and a 100-hour orientation course regarding the employment market and society.29 The level of language skills strived for during the courses was level 3, which was more or less equal to level A2 of the Common European Framework (CEF), developed by the Council of Europe, which in the future would be frequently used to indicate the language skills required, and to make comparisons with language and integration policies for immigrants in other European countries.30 The language skills acquired would be measured in a ‘profile test’, in which merely 40% of the test candidates managed to attain the level strived for.31 Even though the test was only meant to measure language skills acquired, Verhagen claimed that the fact that more was required from newcomers than from future citizens was “the world turned upside down”.32 Together with Niederer (VVD), Verhagen put forward an amendment requiring applicants for Dutch citizenship to have oral and written knowledge of the Dutch language, as well as knowledge of the Dutch state and society. GreenLeft and PvdA again opposed the requirement of writing skills, but the Progressive Liberal D66 made a U-turn by stating that it was willing to accept the amendment, provided that ‘written knowledge’ only implied reading skills and that exceptions would be provided, especially for those applicants aged 65 and over.33 With the U-turn by D66, a majority in Parliament favoured a toughening of the language and integration requirement (Groenendijk & Barzilay 2001: 58). The bill was subsequently amended to provide that the requirement of knowledge of the Dutch language and of the Dutch state and society would be regulated in a decree. After the adoption of the bill in the Second Chamber, it was sent to the First Chamber where it was adopted on 19 December 2000.34 The lobbying by the three coalition parties that would a few years later form the second coalition Government of Christian Democratic PM Balkenende (CDA, VVD, D66) eventually resulted in the introduction of the requirement of written language skills for naturalisation. The Royal Decree concerning the language and integration requirement, the so-called Naturalisation Test Decree, was however 29 In practice, this meant that the Act mainly targeted immigrants who had come to the Netherlands for family reunification or –formation from the traditional immigration countries (Odé & De Vries 2010: 18). 30 For a description of the different language levels of the CEF, consult http://www.coe.int/t/dg4/ education/elp/elp-reg/Source/Key_reference/Overview_CEFRscales_EN.pdf, p. 5, site visited on 13 March 2013. 31  Report of the Temporary Advice Commission Standardisation Integration Requirements’, TK 2004–2005, 29 543, no. 2, p. 31. 32 TK 1999–2000, 16 February 2000, p. 50–3635. 33 TK 1999–2000, 16 February 2000, p. 50–3637. 34 In the Second Chamber, all parties, except for the Christian Democrats, voted in favour of the Act.

50  Chapter Three

still drafted under the responsibility of Social Democratic PM Kok’s Purple cabinet, of which, besides VVD and D66, the PvdA formed part. This Decree, which, like the revised Dutch Nationality Act, entered into force on 1 April 2003, provided that naturalisation applicants had to pass a ‘naturalisation test’, measuring whether applicants had sufficient knowledge of Dutch society and were able to speak, understand, read ànd write Dutch at level A2 of the CEF. The decree hence went further than what parliament had agreed to, namely that only reading skills would be required, and not the ability to write. Furthermore, contrary to D66’s demand that immigrants aged 65 and over would be exempt from passing the test, an exemption regulation for elderly immigrants was not provided. The decree would, however, enter into force unaltered, as parliament did not make use of the opportunity to comment on the draft decree before its introduction. Slightly less than a year before the entry into force of the revised Dutch Nationality Act and the Naturalisation Test Decree, the Purple Government was replaced by a centre right Government, led by PM Balkenende (CDA). This Government consisted of the CDA, VVD, and newcomer LPF (List Pim Fortuyn), a right-wing party with an explicit anti-immigration programme. The latter party also delivered the Minister for Aliens’ Affairs and Integration, a new post, who chaired a department which was accommodated in the Ministry of Justice. The participation of a populist, anti-immigrant party in the government, which also delivered the minister with the competence to implement the naturalisation test, probably explains the high price of the test (€ 260), and the absence of possibilities for preparation. This strict implementation of the naturalisation test can furthermore be explained by the fact that the responsibility for integration policy was moved from the Ministry of the Interior, which had been dealing with the issue for two decades, to the Ministry of Justice, which traditionally had been in charge of border control, residence permits and expulsion.35 Integration policy was thereby put in the service of immigration and foreigners’ policy, meaning that integration and tests would be used by the government as instruments of restriction of immigration and to prevent immigrants from obtaining a secure status (Groenendijk 2004: 113, 117).36 35 On 14 October 2010, the overall responsibility for integration was however returned to the Ministry of the Interior. Previously, under the centre-left cabinet Balkenende IV (CDA-PvdAChristian Union), which took office on 22 February 2007, integration issues were dealt with by the Ministry of Housing, Spatial Planning and the Environment, within which the Minister for Housing, Neigbourhoods and Integration was accommodated. Hence only in the first three Balkenende cabinets (22 July 2002–22 February 2007) was integration policy the responsibility of the Ministry of Justice. 36 A similar move has taken place in Germany and at the European level, where the Ministers of Social Affairs were replaced by Justice and Home Affairs Ministers to negotiate the EU immigration and integration directives (Michalowski 2010b: 189–190).

Repairing the Failures of the Past  51

It is clear that the centre right Balkenende Governments, unlike the government which had introduced the 1993 bill which aimed to facilitate access to Dutch nationality, and which also contained the CDA, viewed naturalisation as the end point of the integration process. The idea of naturalisation being an instrument for integration had been thrown overboard, without research having been conducted into the question of whether or to what extent the former assumption of naturalisation benefiting integration had been correct. Verdonk (VVD), Minister for Aliens Affairs and Integration in the second and third Balkenende administrations, repeatedly referred to Dutch nationality as “the first prize”.37 Groenendijk consequently concluded that naturalisation had become a trophy which was reserved for only a few immigrants, the winning of which was unsure (Groenendijk 2005: 10). Statistics indeed showed that a quarter of all candidates failed the naturalisation test.38 After the introduction of the formalised test, the number of naturalisations dropped by almost 50%.39 Members of parliament, both belonging to coalition parties and the opposition, however hardly paid attention to the discouraging effect the naturalisation test apparently had on the potential applicants, or to the question of whether or to what extent this effect was desirable.40 3.1.3. From Naturalisation Test to Integration Examination A Second Paradigm Change Leading to a New Integration Policy After the first paradigm change in immigrant integration policy, which took place in the 1980s with the realisation that the guest workers had settled in the Netherlands, a second paradigm change took place after the turn of the 21st century, implying the introduction of a more obliging, result oriented integration policy. This paradigm change can be connected to the reasons mentioned before, but among the reasons also was growing popular and scientific concern about the representation of immigrants in ‘wrong’ statistics on unemployment, crime, and school drop-out rates (Schinkel & Van Houdt 2010: 703). Furthermore, a major influence on the changes in the political perspectives on immigration and integration was constituted by the 9/11 terrorist attacks (Bonjour 2009: 237–238, 242, 37 TK 29 200 VI, no. 7, p. 3, Hand. TK 10 December 2003, p. 2486, TK 27 083, no. 63, p. 15 and Hand. TK 2 September 2004, p. 6075 and 6096. 38 See paragraph 7.1.3.1. 39 In 2004, the first full year the test was in force, the number of naturalisations was almost half as low as the number of naturalisations in 2002, the year prior to the introduction of the test. See paragraph 7.1.1 for more information on the number of naturalisations in the Netherlands. 40 Only MP Dijsselbloem (PvdA), a year after the introduction of the test, asked Verdonk about the possible relation between the decline in applications for naturalisation and the level of the test. Verdonk ducked this question by stating that it was too soon to draw any conclusions about the drop in the number of applications. TK 2003–2004, Appendix Proceedings no. 1282.

52  Chapter Three

Groenendijk 2008: 116). Various opinion polls conducted at the end of September 2001 showed that most Muslims living in the Netherlands could understand the attacks, which both the press and government representatives such as Prime Minister Kok and Minister for Big Cities Policy Van Boxtel interpreted as proof that their integration in the Netherlands had failed dramatically (Bonjour 2009: 242). The post-9/11 atmosphere proved to be fertile soil for the LPF, which won the 2002 parliamentary elections, guaranteeing it government participation.41 Internal conflicts between LPF Ministers however caused the only government in which the LPF ever took part to resign after less than 90 days.42 The party has nevertheless profoundly influenced the immigration and integration discourse in the Netherlands. Ever since its rise, the larger political parties sought to regain the voter’s trust: from now on, everything needed to be different (Bonjour 2009: 244). The pursuance of a different path in integration policy was explicitly stated in the coalition agreement of Balkenende’s second cabinet (CDA, VVD and D66), which read that whoever wants to permanently settle in our country will have to actively participate in society and master Dutch language, be aware of Dutch values, and observe the norms.43

By announcing a new line on integration policy, the government judged the policy which had been conducted so far as having been insufficiently successful. Parliament had apparently passed a similar judgment, since, at the initiative of the Socialist Party, it installed a parliamentary commission whose task it was to investigate why, rather than whether, the integration policies pursued so far had not been successful. This commission, led by Blok (Conservative Liberals), presented its final report early in 2004, which revealed the surprising conclusion that the integration of a large part of the immigrant population had actually been successful.44 The commission could however not establish causal links between immigrant integration and pursued integration policies, apart from the improvement in the immigrants’ legal position.45 Political parties for this reason refused to take the report into account and questioned its conclusions, while demanding a new policy be put in place (Groenendijk 2008: 117). The fact that the need was 41 Participating for the first time, the LPF acquired 26 out of 150 seats in Parliament. The victory is often partly ascribed to party leader Pim Fortuyn’s murder a few days prior to the elections. 42 In the following parliamentary elections, the LPF obtained disappointing results, to completely disappear from the Second Chamber after the 2006 elections. 43 ‘Meedoen, Meer Werk, Minder Regels’, Coalition Agreement for the CDA, VVD and D66 cabinet, 16 May 2003, p. 11. 44 Report Bruggen Bouwen (Building Bridges), TK 28689, no. 8–9, 19 January 2004. 45 TK 2003–2004, 28 689, no. 9, p. 522.

Repairing the Failures of the Past  53

felt for the introduction of a new integration policy, roughly six years after the introduction of the NIA, is an expression of ‘democratic impatience’ (Vermeulen & Penninx 2000: 229): the ‘old’ policy had not produced tangible results, and for that reason had been declared a failure. For similar reasons, the minorities’ policy, four years after its introduction, had been declared insufficiently successful (see above paragraph 3.1.1). This line of reasoning however fails to take into account that integration is a long-term process, which takes at least several generations (Lucassen & Penninx 1994: 196). In the fall of 2003, Minister Verdonk presented the ‘New Style Integration Policy’ (Integratiebeleid Nieuwe Stijl). In a letter to parliament, the Minister stated that in future, rather than cultivating cultural identities, emphasis should be put on ‘shared citizenship’: speaking Dutch, keeping to basic Dutch norms and being prepared to actively contribute to society.46 A year later, on 2 November 2004, controversial filmmaker Theo van Gogh was murdered by ‘home-grown’ Moroccan-Dutch Muslim extremist Mohammed B. Following the murder, two parliamentarians who had criticised Muslims and Islam, Hirsi Ali (VVD) and Wilders (former VVD member and founder of the rightwing Freedom Party (PVV)), received special protection. A year before, criminal proceedings had been started against a small group of Islamist extremists called the Hofstad group, which inter alia had led to three fundamentalist imams being expelled (Groenendijk 2008: 127). These events only reinforced the feeling that immigrant integration had failed, and that a new, firm, integration policy was required. They offer an explanation for the relative ease with which Minister Verdonk managed to get her proposal for a new Integration Act, which would replace the 1998 NIA and which was introduced in parliament at the end of 2005, adopted.47 In the explanatory memorandum to the proposed Act, Verdonk stated that the “continuity of the society would be threatened” if “large parts of the population would not participate actively in society and the economy”.48 Therefore, an integration obligation would be introduced, which would “repair the flaws of the old policy”.49 The government thereby appeared to have forgotten that the legislation in place at the time already contained such an obligation: those falling under the scope of the NIA were after all obliged to follow an integration course, and a 46 TK 2003–2004, 29 203, no. 1. 47 TK 2005–2006, 30308, no. 2. The events described also led to the rise and the electoral gains of Geert Wilders’ Freedom Party in the 2006 and 2010 parliamentary elections. The Freedom Party as such did not form part of the parliament in the period the Integration Act was discussed. As the only member of the ‘Group Wilders’, Geert Wilders himself however did take part in the debates. 48 TK 2005–2006, 30308, no. 3, p. 25, 40. 49 TK 2005–2006, 30308, no. 3, p. 25, 40.

54  Chapter Three

system of fines had been put in place to punish those who intended to back out of this obligation.50 Sanctions were however hardly applied in practice, partly because most newcomers stopped following an integration course because they had found a job, and partly because imposing a fine was time-consuming, would require a lot of administration, and hardly led to the desired effect (Odé & Brink 2002: 156). The government, and later on also parliament, apparently was not willing to learn from these lessons as it insisted on the introduction of a far-reaching and complex system of administrative fines. The proposed Act would furthermore introduce a new language and integration examination. Failing this examination would result in the denial of a permanent residence permit, a naturalisation application, or in administrative fines, something the government presented as incentives for the immigrant to take his integration seriously and “a guarantee for Dutch society that he actually will”.51 That a test would also represented an instrument with which to select and exclude immigrants was not put forward as a goal of the new policy. Another presumed incentive embedded in the proposed legislation was the fact that immigrants in future would be required to pay for the courses and tests themselves. The average costs of an integration course were estimated to be around € 4,300.52 The government would offer only limited financial support as the past ‘service-oriented’ approach had supposedly “diminished the initiative of those concerned”.53 It would furthermore no longer carry responsibility for the provision of courses, as this responsibility would be transferred to the market, i.e. certified language schools and education institutions. Besides the reparation of the supposed failures of the past, the government referred to the European framework on immigration and integration as a justification for the new policy.54 The bill’s explanatory memorandum for instance referred to the European Council which had repeatedly stressed the need for a 50 Immigrants, who failed to report to the local authorities for an ‘integration interview’, on the basis of which it would be decided whether or not they would be required to integrate, would be fined a maximum of € 2,269 (Entzinger 2003: 77). Immigrants who stopped following a course without having lawful reasons could also be fined, or suffer a cut in their social security remittances (Odé & Brink 2002: 156). 51 TK 2005–2006, 30 308, no. 7, p. 117. The new examination would hence replace the existing naturalisation test. Initially, passing the examination would only be required for those with an integration obligation under the IA and for those filing for permanent residence. It was decided in May 2006 that the new examination would replace the naturalisation test. TK 2005–2006, 30 308, no. 16, p. 50. 52 The Minister could not indicate what amount an individual person with a duty to integrate would minimally or maximally have to pay for his or her integration. TK 2005–2006, 30 308, no. 16, p. 10. 53 TK 2003–2004, 29 543, No. 2. 54 TK 2005–2006, 30308, no. 3, p. 6–7.

Repairing the Failures of the Past  55

more powerful immigrant integration policy, as well as to the possibilities to introduce integration conditions offered by Directive 2003/109/EC. It also mentioned one of the Common Basic Principles (CBPs) on integration, drafted and adopted by the European Council under the Dutch Presidency, which stated that a basic knowledge of the language, the history and the institutions of the host society is essential for integration.55 That the CBPs also emphasise that integration is a two-way process was not mentioned. Neither did the government explain to what extent the proposed policy, which would make it harder for immigrants to attain a secure legal position, would interfere with the conclusions of the 1999 European Council of Tampere, containing a commitment to fairer treatment of third-country nationals, including the granting of rights and obligations comparable to those of Union citizens (Kostakopoulou 2010b: 12).56 Discussions in the Second Chamber Discussions on the new Integration Act initially mainly focused on the integration obligation for naturalised Dutch citizens the Act in the first instance also contained.57 After the Council of State had judged this obligation to be racially discriminatory and contradictory to the principle of equality, it was deleted from the bill.58 Afterwards, the discussions mainly centred around the way in which the proposed provisions would be implemented, and the level of language skills that would be required. Other issues, such as the content of the new examination, its form and the exemption regulations, or the financial implications the new integration obligation would entail for immigrants, could not count on receiving much attention.59 After the adoption of the Act, these barely debated issues 55 CBP number 4, which states that “basic knowledge of the language, the history and the institutions of the host society is indispensable to integration; enabling immigrants to acquire this basic knowledge is essential to successful integration”, European Council, 2618th meeting of the Justice and Home Affairs Council, ‘Common Basic Principles for Immigrant Integration Policy in the European Union’, 14615/04, Brussels, 19 November 2004. 56 Council of the European Union (1999), Presidency Conclusions, Tampere, 15–16 October 1999. 57 The initial proposal of law contained an integration obligation for certain categories of Dutch nationals of immigrant descent (i.e. those on welfare, religious ministers and immigrants taking care of children). This idea met with fierce opposition in society and parliament. Minister Verdonk asked parliamentarians for “political courage” to adopt the integration obligation for Dutch citizens, but this appeal was unsuccessful. 58 TK 2005–2006, 30 308, no. 108. Consequently, the Act’s original target group was halved (reduced to about 260,000 residents) (Extra & Spotti 2009: 72). 59 The Socialist Party (SP) did criticise the Act for the high level of bureaucracy its implementation would require. The regulations contained in the Act would limit the possibilities for officials at the local level to implement the rules according to their own discretion.TK 2005–2006, 30 308, no. 6, p. 3, p. 41. The criticism of the SP members was however not taken up by other parliamentarians.

56  Chapter Three

would however give rise to many problems in practice, which in turn would lead to at least twelve amendments to the law and secondary legislation (Significant 2010: 81). Changes were for instance required because of the disappointing results the new Act produced: in the first year after its introduction, and due to the fact that immigrants were required to bear the costs for the integration courses themselves, the number of immigrants starting an integration course fell to onethird of the numbers in previous years when the NIA had been in force.60 This in turn led to numerous dismissals at local educational centres (Groenendijk 2011b: 337).61 The discussions on the content and level of the new examination will be examined below, as the examination replaced the naturalisation test as a condition for naturalisation. • Content of the Knowledge Tested The new examination, in the words of Minister Verdonk, would ensure that immigrants would have enough knowledge of the Dutch language to communicate in daily life and have knowledge of prevailing norms and values.62 The Minister defined these basic Dutch norms as “those norms which make it possible to in freedom and according to one’s own perception build up an autonomous life and take part in society”. Norms such as doing your best to be able to support yourself, keeping to the applicable laws and rules and taking care of one’s own surroundings were mentioned as examples, as well as norms flowing from liberal core principles, such as respecting the physical integrity of others (also within marriage), accepting everyone’s right to express their opinion, accepting other people’s sexual preferences and the equality of men and women. Minister Verdonk presented these norms as being typically Dutch. Verdonk thereby fell into the trap of what Joppke has referred to as ‘the paradox of universalism’: while aiming to make sure that ethnic minorities become a part of this and not any political community, Verdonk failed to succeed in naming any particulars that distinguish the ‘here’ (the Netherlands) from the ‘there’ (other liberal democratic societies) (Joppke 2010b: 130). Furthermore, by incorporating them into the knowledge that

60 In 2007, around 10,000 persons started an integration course, whereas in the years before, three times as many persons had started a language trajectory; TK 2007–2008, 31 143, nos. 14 and 38 (Odé & De Vries 2010: 25). The number of immigrants who started a course was considerably lower than the number of immigrants that were expected to start a course. 61 For the dismissals at language schools see ‘Scholen moeten stoppen met inburgeringscursus’ [schools need to stop organising integration courses], Brenda Peeters, Elsevier, 30 June 2008. Several language schools sued municipalities for making false promises about high numbers of students. 62 TK 2005–2006, 30308, no. 3, p. 25, 40.

Repairing the Failures of the Past  57

would be tested, Verdonk presumed that the norms referred to were not sufficiently adhered to by the immigrant population. Several parliamentarians asked the Minister to inform parliament of the exact content of the test that was being developed.63 The Minister eventually sent a draft curriculum of (a part of) the new examination to parliament on 2 June 2006.64 The curriculum would, however, never be discussed in parliament.65 Visser (VVD) deemed parliament unfit to be “a kind of expert in the field of examination”, thereby giving a reason for the omission to address the examination’s content.66 The curriculum of the examination as it was sent to parliament in June 2006 can now be found in exactly the same form as an appendix to the Ministerial Regulation on Integration.67 Language Skills The level of language skills required to pass the examination would be A2. The choice for this level was inspired by a disqualification of the lower level A1 for its inability to guarantee participation, and the higher level B1 for the costs that would be involved in making sure that lower educated immigrants would attain this level.68 Level A2 on the other hand was deemed reachable within the given time frame.69 The choice for level A2 thereby appeared to be the result of a choice between the desired level and a level which would be realistically attainable at reasonable costs. The level of the proposed examination equalled the level of the naturalisation test and the level strived for under the NIA. As we have seen, the majority of 63 Nawijn (LPF), Huizinga (Christian Union) (TK 2005–2006, 30 308, no. 12, p. 13). 64 TK 2005–2006, 30 308, no. 25. 65 Only two parliamentarians indicated they were concerned about the content of the new examination. Vergeer (SP) and Varela (LPF) worried about the lack of clarity regarding the examination’s curriculum (TK 2005–2006, 30 308, no. 63, p. 14.). In reply, Verdonk referred to the curriculum (final achievement levels) for the examination, which had been sent to Parliament on 2 June 2006. 66 TK 2005–2006, 30 308, no. 63, p. 15. 67 The curriculum has been published as an appendix to article 2.5 of the Ministerial Regulation on Integration (Regeling van de Minister voor Vreemdelingenzaken en Integratie van 6 december 2006, nr. 5456790/06, tot uitvoering van de Wet inburgering, het Besluit inburgering en tot wijziging van de Regeling verstrekkingen asielzoekers en andere categorieën vreemdelingen 2005). A small change was made in July 2011, when knowledge of the Dutch anthem, the Wilhelmus, was added to the curriculum, upon a request by Christian Democratic and SGP parliamentarians. Stscrt. 2011, no. 7426, 29 April 2011. 68 TK 30308, no. 6. 69 TK 30308, no. 3, p. 19. Those with an integration obligation under the Integration Act would need to pass the examination within a period of five years. This period was shorter (three and a half years) for those who had already passed the integration abroad examination. Since 1 January 2013, all immigrants with an integration obligation have three years to pass the examination.

58  Chapter Three

the newcomers with an integration obligation under the NIA however did not manage to achieve this level after following an integration course, whilst 25% of candidates failed the naturalisation test. By choosing level A2 as the level for the new examination, the government therefore knowingly accepted the possibility that a large part of the immigrant population would either be condemned to paying substantial fines, or be excluded from naturalisation or acquiring a permanent residence permit. This makes one wonder whether reducing the number of permanent residents and naturalisations was a latent aim of the new examination. The level of the language skills required received a fair amount of attention in the parliamentary discussions. It was criticised for being too low as well as for being too demanding. Furthermore, the fact that a uniform level of language skills were required was criticised. This led Azough (GreenLeft) to conclude that the proposed Act punished illiteracy as well as ambition: whereas some immigrants would be scared off by the high level of the examination, highly educated or more ambitious immigrants would not be allowed to start learning Dutch at a higher level which would benefit them in the employment market.70 Vergeer (SP) consequently suggested abolishing the examination, putting forward a motion for the introduction of an ‘obligation to make an effort’ instead of an obligation of result.71 Vergeer thereby in fact pleaded for a continuation of the policy of the NIA. The motion was however withdrawn, and never voted upon.72 Christian Democratic parliamentarians condemned the level of language skills required as being too low in general. They asked the Minister to set level B1 as a minimum level, and to formulate exceptions for those who could not obtain this level due to their learning capacities.73 Triggered by the discussions on the level of the integration examination, the Christian Democrats, together with the small Christian parties (Christian Union and SGP), asked the Minister to raise the level of language skills required for naturalisation.74 The Minister was however not willing to do this, as the test for naturalisation should not “unintentionally” function as a selection instrument.75 The Minister thereby ignored the fact that the test in fact already functioned as a selection mechanism. From the decrease in

70 TK 2005–2006, 30 308, no. 101, p. 21. 71 TK 2005–2006, 30 308, no. 63, p. 6. 72 TK 2005–2006, 30 308, no. 52. Vergeer withdrew the motion, because she stated that the proposal to replace the obligation of result with an obligation to make an effort was referred to the Council of State for advice (Handelingen TK 98, 7 July 2006, p. 5). When the proposal of law, in an adapted form, was presented to the First Chamber, the obligation of result was however still upheld. 73 TK 2005–2006, 30 308, no. 12, p. 12. 74 TK 2004–2005, 29 800 VI, no. 101, p. 9–10, 25, p. 26. 75 TK 2004–2005, 29543, no. 7, p. 8.

Repairing the Failures of the Past  59

the number of applications for naturalisation after the introduction of the naturalisation test, Verdonk instead however drew the conclusion that immigrants had integrated poorly.76 She thereby ascribed the decrease to a lack of effort on the part of the immigrants, instead of to the introduction of a formalised naturalisation test, which was significantly more demanding than the old integration interview.77 The Minister was not willing to adapt the level of the proposed examination either. She dismissed the complaints regarding the lack of differentiation in the courses by stating that trajectories tailored to the needs of the immigrants could be drawn up.78 Those for whom the examination would be too demanding could rely on the exemption possibilities.79 In reply to a query by Social Democratic parliamentarians, Verdonk refused to open up possibilities for municipalities to use discretion to grant exemptions to immigrants suffering from mental or physical disabilities, or to those who had sufficient language skills, but could not prove this with a diploma.80 Verdonk preferred testing along the same, objective norms, as opposed to a subjective impression of an official. She thereby rejected what was in fact a plea for a (partial) return to the situation prior to the introduction of the naturalisation test, when the judgment of whether someone was sufficiently integrated or incapable of fulfilling the language and integration requirements was left to municipal officials.81 Lastly, Vergeer, fearing that the new examination would be too difficult, asked the Minister to provide an analysis of why so many people failed the naturalisation test. Since this test had the same level as the proposed examination, Vergeer deemed this information essential in order to estimate to what extent the new examination would produce exclusionary effects.82 The Minister never answered Vergeer’s question. Neither Vergeer nor other parliamentarians subsequently reminded the Minister to elaborate on the exclusionary effects the new test could produce. Like the government, parliament hence implicitly accepted that a number of permanent residents would be excluded indefinitely from full membership. 76 TK 2005–2006, 30308, no. 3, p. 13, 15. 77 See paragraph 3.2 below for a comparison between the integration interview and the naturalisation test. 78 TK 2005–2006, 30 308, no. 7, p. 68. 79 TK 2005–2006, 30 308, no. 7, p. 70. 80 TK 2005–2006, 30 308, no. 16, p. 23, 26–27. 81 TK 2005–2006, 30 308, no. 16, p. 23. As we will see in Chapter 8, the fact that municipal officials are no longer allowed to exempt people from fulfilling the language and integration requirement has led to embarrassing situations and frustration on both the part of the officials and the immigrants. 82 TK 2004–2005, 29 800 VI, no. 101, p. 34.

60  Chapter Three

On 7 July 2006, all members of the Second Chamber, except for Koşer Kaya (D66), voted in favour of the Act.83 The First Chamber approved the Act on 28 November 2006. There had been more opposition to the Act than five months earlier in the Second Chamber, as D66, GreenLeft, SP and one member of the independent OSF voted against it.84 The Integration Act entered into force on 1 January 2007. Three months later, on 1 April 2007, the integration examination replaced the naturalisation test as a condition for naturalisation. This meant that the naturalisation test was abolished only four years after it had been introduced. Below, both tests will be described and compared to each other and to the integration interview. 3.2. Testing Integration as a Condition for Naturalisation In the course of the past ten years, the way in which integration has been tested as a condition for naturalisation has changed significantly. In the table below, the integration interview, which was conducted until 1 April 2003, and its successors, the naturalisation test and the integration examination, are compared. From table 3.1, some striking differences between the integration interview and the formalised language and integration tests that succeeded it become apparent. First of all, with the formalisation of the language and integration test, the costs have risen considerably. Whereas the integration interview was free of charge, € 260 was charged for the naturalisation test, and the costs payable for the Table 3.1. Comparison between integration interview, naturalisation test and integration examination. Integration interview (until 1 April 2003)

Naturalisation test (until 1 April 2007)

Integration examination

An interview was conducted with all applicants for naturalisation.

Applicants could be Applicants can be exempted on the basis exempted on the basis of a diploma or a doctor’s of a diploma or a doc­ tor’s certificate. Special certificate. Special exemption procedure exemption procedure for illiterates. for illiterates.

83 TK 98–6084, 7 July 2006. Koşer Kaya gave no reasons why she voted against the Act. 84 EK 10–407, 28 November 2006. The increased opposition to the bill in the First Chamber could be partially explained by the fact that 25 university professors had sent a letter to the members of the First Chamber, warning them of the flaws in the bill they were about to adopt.

Repairing the Failures of the Past  61

Table 3.1. (Cont.) Integration interview (until 1 April 2003)

Naturalisation test (until 1 April 2007)

Integration examination

More lenient treatment Illiterates who wished to Illiterates who wish to of women, the elderly, qualify for exemption qualify for exemption persons who had needed to undergo a need to undergo a ‘feasibility inves­ followed no or a ‘feasibility investigation’ tigation’ in Amster­ limited education in Amsterdam costing dam costing € 290. and illiterates. € 290. Duration of the Total duration of the test: Total duration: central part ± two hours; prac­ interview: about 15 about 4 hours. tice part, depending minutes. on route chosen, ± one after­noon or several weeks.85 Knowledge of Dutch Candidate passed the Central part of the state and society not Society Orientation part examination contains required. of the test if he or she a ‘knowledge of Dutch answered 70% of the society’ examination, questions correctly. of which 62% needs to be answered correctly. General guidelines Test content was secret. Curriculum for the  for conversation Government offered ‘knowledge of Dutch  contained in the no possibilities for society’ part of the  Manual. Room for preparation. examination has been  interpretation by published; immigrants  municipal official. can follow courses. Interview conducted Test could be taken at Examination can be  in all municipalities.  nine Regional Education  taken at seven  locations.  Centres (RECs). (Continued) 85 Until 1 January 2013, the integration examination consisted of two parts, a central part and a practice part. As of 1 January 2013, following the adoption of a proposal to amend the Integration Act (TK 2011–2013, 33086, adopted by the First Chamber on 11 September 2012), the practice part of the examination, which candidates could either pass by submitting a portfolio of twenty proofs that someone had used the Dutch language in the practice, by taking part in an assessment, or a combination of both, has been abolished. Furthermore, the central part of the examination has changed. Until 1 January 2013, this part of the integration examination consisted of three parts: an electronic practice examination, asking candidates to answer questions about situations which might occur in practice, a test in spoken Dutch, and an examination regarding knowledge of Dutch society (KDS examination). As of 1 January 2013, the integration

62  Chapter Three

Table 3.1. (Cont.) Integration interview (until 1 April 2003)

Naturalisation test (until 1 April 2007)

Integration examination

Costs of interview: € 0.

Costs of naturalisation Costs for central part: test: € 260 (more if € 126, for practice re-examination required). part: varying between € 104 and € 1,200.86 Total costs (following course + taking exam­ ination) estimated to lay around € 5,300.87

integration examination vary between € 230 and € 1,200.88 Secondly, the formalisation of the test has led to a rise in the level of the language skills required, and has put a stop to the possibility to treat certain categories of immigrants more leniently than others. Those for whom acquiring the level required will prove too difficult can be exempted, but, like the language and integration check itself, this procedure has been formalised. Those who require exemption on grounds of mental or physical disabilities need to submit doctors’ certificates, whereas illiterates need to undergo a ‘feasibility investigation’ at the Regional Education Centre of Amsterdam, for which € 290 is charged.89 Thirdly, until 1 April 2003 immigrants could have their language and integration checked at the municipalities in which they lived. However, since the introduction of the formalised tests they have had to travel to one of nine (naturalisation test) or seven (integration examination) examination locations. Travelling to these locations to take a test takes significantly more time and money than going to the municipality for a short interview. examination consists of five parts: the test spoken Dutch, the KDS examination, and exams testing the candidates’ reading, writing and listening skills. 86 Stouten & Brink 2010: 16. 87 Costs estimated by the Dutch Association of Municipalities. “Integratie de nek omgedraaid”, Jonathan Witteman, de Volkskrant, 16 September 2010. 88 With the abolishment per 1 January 2013 of the practice part of the integration examination, the costs have decreased, as the costs of the five parts in the integration examination amount to €250. 89 In the feasibility investigation, it is checked whether someone is able to reach the level required to pass the naturalisation test, or, as of 1 April 2007, the integration examination, within a period of five years. If so, exemption will not be granted. Only those who can demonstrate that they have made an effort in the past to learn Dutch, above what has been legally required, have the possibility to be exempt. Illiterates will in any case have to pass the test in spoken Dutch, one of the three parts of the central part of the integration examination.

Repairing the Failures of the Past  63

It is clear that with its formalisation, fulfilling the language and integration requirement has become much harder, not only because of the rise in the level of skills required. Parliamentarians hardly scrutinised the extra barriers that would be thrown up once the language and integration checks would be formalised. Attention was however paid to groups for whom passing a test would be too difficult, but the exemption regulations have proved stricter than parliament probably had in mind.90 It did however not make use of the possibility to comment on the decree implementing the naturalisation test, nor did it address the naturalisation test or the integration examination after they had been introduced, or the effects they produced. The integration examination however lowered some of the barriers thrown up by the naturalisation test, namely those relating to the undisclosed test content and the absence of possibilities for preparation.91 As was the case with the naturalisation test, the questions in the integration examination have not been published, but the curriculum for the knowledge of Dutch society test in the central part of the examination has. Furthermore, a report containing materials which can be used for self-study has been published.92 And last but not least, those for whom self-study is unsuitable can prepare for the test by following an integration course. To emphasise their own responsibility for the integration, immigrants were initially required to pay for the courses and test themselves, if necessary with money obtained from a loan for this purpose.93 As the costs appeared to form a barrier for immigrants to start following courses and taking examinations, in 2007 the government introduced the ‘Deltaplan Integration’ which included a decision to start paying for the integration courses in all cases.94 The plan was 90 See paragraph 8.2.2 for the application in the practice of the exemption regulation for those for whom the level of the examination is too high. 91 At the time, the fact that the test content was kept secret was defended by the leader of the project to introduce the test in the following way: “There will be no booklet containing sample questions, as is the case for the driver’s exam, since one does not only want to test the proper knowledge of the applicant for naturalisation, but also his proper attitude. And this cannot be learnt by heart” (Groenendijk 2005: 30). 92 ‘Onderzoek Zelfstudiemethoden Inburgering’, CINOP, October 2009, downloadable from www .rijksoverheid.nl. In the report, books with and without CD-roms and e-learning methods are listed which can help immigrants to prepare for the different parts of the integration examination. In the materials listed, all the topics mentioned as ‘final achievement levels’ for the integration examination are considered. The materials will not only help future candidates to obtain the necessary knowledge in terms of content, but also teach them to make assignments on the computer. 93 A loan of maximum € 5,000 could be obtained. If an immigrant fulfilled the integration obligation (i.e. passed the examination) within the given time frame, he or she would get a refund of 70% of the costs to a maximum of € 3,000. 94 Deltaplan Integration: TK 2006–2007, 31143, no. 1. A change in the law, which was introduced on 29 December 2008, but which entered into force retrospectively as of 1 November 2007, allowed

64  Chapter Three

introduced to give integration a new impetus, as the number of immigrants following courses was disappointingly low. The new plan appeared to be successful, as after its introduction, the number of immigrants following courses and taking part in examinations increased (Significant 2010: 21, 23). Nevertheless, the centre right coalition Government, consisting of CDA and VVD, and receiving support from Geert Wilders’ Freedom Party (PVV), in power from 14 October 2010 until 21 April 2012, decided to reduce the integration budget to zero in 2014.95 To this end, it put forward a proposal to amend the IA, which was adopted on 11 September 2012, and which entered into force on 1 January 2013.96 Now immigrants again have to pay for the courses, it is expected that the number of course participants will once more decrease considerably.97 The revised policy will therefore probably hamper rather than foster immigrant integration. 3.3. Conclusions In the past decades, Dutch citizenship policy has taken a restrictive turn. The goals ascribed to the 1985 DNA can be characterised as liberal, as one of the primary aims of the Act was to increase the number of naturalisations, and thereby contribute to the equality of the foreigners living in the Netherlands. The idea was that an equal legal position would enable integration. The discussions on proposals to amend the DNA, starting in 1993, the Act which eventually came into force in 2003, as well as the discussions on the Integration Act a few years later however show that communitarian ideals have succeeded in making their mark on Dutch citizenship and integration policy, at the expense of liberal ideals of equality. Rather than being a means for integration, naturalisation was more and more

municipalities to pay for the integration courses for those with an integration obligation under the IA (Staatsblad 2008, no. 604). Until then, only the courses and examinations for those who integrated on a voluntary basis would be paid for. 95 The coalition agreement of PM Rutte’s (VVD) first cabinet stated that immigrants will in future need to take care of their integration themselves, and that a system of loans will be set up for those lacking the financial means to do so (‘Freedom and Responsibility’, coalition agreement VVDCDA, 30 September 2010, accessible via http://www.rijksoverheid.nl/documenten-en-publicaties/ rapporten/2010/09/30/regeerakkoord-vvd-cda.html, site visited on 11 August 2011). 96 TK 2011–2012, 33086, nos. 1–3. 97 A system of loans has been set up for those who cannot afford to pay for the courses. Family migrants and religious servants can obtain a loan of maximum € 5000, whereas asylum seekers can obtain a loan of maximum €10,000. The coalition agreement of PM Rutte’s second cabinet however states that the government will pay for the integration courses of asylum seekers (‘Bruggen Slaan’, coalition agreement VVD-PvdA, 29 October 2012, p. 30, http://www.rijksoverheid .nl/documenten-en-publicaties/rapporten/2012/10/29/regeerakkoord.html).

Repairing the Failures of the Past  65

seen as the completion of the integration process, only accessible for those immigrants who felt Dutch, and would be proud of their Dutch nationality. After the adoption of the 2003 Act, naturalisation was only accessible to those who had passed a formalised test. In the time which passed during the beginning of the discussions on the alteration of the Dutch Nationality Act in 1993 and the adoption of the new Act in 2000 an inversion of the relationship between integration and naturalisation hence took place. How can this change be explained? A first explanation is the change in the thinking about immigrant integration, which kicked off after Bolkestein’s speech in 1991. The idea emerged that immigrants had been pampered without having to bear clear obligations. The supposedly poor immigrant integration was more and more ascribed to the social and cultural differences between immigrants and the autochthonous population, rather than to the immigrants’ disadvantageous legal position. The integration debate also influenced the naturalisation debate, as both policy fields had been inter-connected since the minorities’ policy. A second explanation is the political process. After the 1994 elections, the Christian Democrats no longer felt committed to the compromise they had reached as part of the government with their former coalition partner regarding the abolition of the renunciation requirement. They started to follow a new course which focused on stricter rather than easier conditions for the acquisition of Dutch citizenship. Furthermore, as the issue of naturalisation, from being primarily an elite-controlled process, received more and more attention in parliament, and consequently also in the media, there was something at stake for the politicians (Böcker, Groenendijk & De Hart 2005: 161, Howard 2009: 92). Striking a harsher tone in the debates would prevent voters from going over to another political party with more outspoken viewpoints on integration. A third explanation is the process of path dependency. After the coming into force in 1998 of the Newcomers Integration Act, which introduced a formalised test for newcomers with an integration obligation, some politicians felt that the requirements for naturalisation needed to be reinforced. After all, it could not be that more would be required from newcomers than from future Dutch citizens. The ‘profile test’ of the NIA hence exerted upward pressure on the language and integration requirement for naturalisation. The question remains whether the turn in Dutch citizenship and integration policy is as radical as sometimes presumed or whether the introduction of a formalised language and integration test has only led to a marginal restriction. Whereas it seems inappropriate to state that the Netherlands belongs to a group of countries that liberalised their citizenship policies in the 1990s and the early 2000s (Howard 2009:73), the changes in citizenship and integration policy are not as drastic as sometimes claimed either. If the Netherlands is seen as having pursued an explicit multicultural integration policy since the 1980s (see, for instance,

66  Chapter Three

Joppke 2007: 5), the changes which have occurred since 2000 are radical indeed. As we have seen however, the creed of ‘integration with preservation of identity’, which is often cited to underline the explicit multicultural intentions of Dutch integration policy, has played a much more marginal role than has been assumed (Vink 2007). As early as the beginning of the 1990s had a more obliging integration policy replaced the Minorities’ Policy, leading to the introduction of the Newcomers Integration Act in 1998. Even though it was only for educational purposes, the Netherlands thereby was the first country in the European Union to introduce a formalised test in the legislation relating to the integration of immigrants (Michalowski 2009: 264–265). Once a formalised integration test had been introduced, the step to introducing a similar test for naturalisation was only small. Seen in this light, the image of the Netherlands as being a country that went from almost naively protecting the identities of immigrants to a country that requires immigrants to adapt by learning the language and the values requires adaptation, in the minds of both politicians and academics.

Chapter Four ‘Deutschland sucht den Superstaatsbürger’:1 Integration Requirements in German Nationality Law It is often said that in the study of citizenship law, no country has received nearly as much attention as Germany (Howard 2009: 119). One of the reasons for the interest in the German case is the purported revolution which has taken place in German citizenship law and policy at the turn of the 20th century. Traditionally, German citizenship law has been classified as “ethno-cultural” or “exclusive” for its almost exclusive reliance on ius sanguinis citizenship attribution and a reluctance to attribute nationality to immigrants (Brubaker 1992, Koopmans et al. 2005: 7). In 2000, a great reform of German citizenship law took place: a form of ius soli citizenship attribution was introduced for qualified immigrants of the second generation, and the residence requirement for naturalisation was reduced from fifteen to eight years. When the new citizenship law was introduced, it was claimed that the changes demonstrated a new public and political culture of openness (Michalowski 2010b: 208). Consequently, the reform is generally seen as a liberalising change (Howard 2009: 141, 147). It has led some scholars to conclude that it is the expression of a more republican understanding of citizenship, in which not descent, but participation is of pivotal importance when it comes to citizenship attribution and belonging (Gerdes, Faist and Rieple 2007). Part of the liberalisations that took place in 2000 has however been accompanied and reversed by changes in the law by which new requirements were added and existing requirements reinforced. Examples of these changes are the introduction and subsequent formalisation of the language requirement in 2000 and 2007 res­ pectively, and the introduction of a knowledge of society test, the so-called Einbürgerungstest, in 2008. The question now is whether and, if so, to what extent ethno-cultural elements have been reintroduced, and whether the statement that an ethnic understanding of nation no longer offers an explanation for changes in

1 Wolfgang Wieland (Bündnis 90/Die Grünen) during discussions in the 174th session of the Bundestag, Plepro 16/174, p. 18613 (B). This statement is based on the name of the German tv show ‘Deutschland sucht den Superstar’, which is the German equivalent of ‘American Idol’, the aim of which is to find the country’s next pop icon.

68  Chapter Four

German citizenship policy still prevails (Gerdes et al. 2007: 48). Are we still all ‘Republican’? This question will be answered via an analysis of the political debates on German nationality law, starting with the debates leading up to the reform of 2000. The chapter will start with a description of integration testing as a condition for naturalisation prior to the reform. 4.1. Testing Integration as a Condition for Naturalisation Prior to the

2000 Reform

In Germany, until the early 1990s, immigrants who wished to acquire German nationality via naturalisation could rely on Article 8 of the Reichs- und Staatsangehörigkeitsgesetz (RuStAG) of 1913, which provided for discretionary  naturalisation (Ermessenseinbürgerung). Naturalisation of immigrants however was the exception rather than the rule. The Naturalisation Guidelines (Einbürgerungsrichtlinien) of 1977 provided that “the Federal Republic of Germany is not a country of immigration”.2 The guidelines consequently stated that the “Federal Republic of Germany […] does not strive to increase the number of German citizens by way of naturalisation” and that “the granting of German citizenship can only be considered if a public interest in the naturalisation exists” (Hailbronner & Renner 1998: 865f, translation provided in Anil 2006: 449).3 The guidelines required the applicant for German nationality to have oral and written command of the German language in a way which could be expected from persons in their social environment (Dornis 2001: 163, Davy 2008). Since the German Länder implement the federal laws in their own right, the language requirement was very diversely applied throughout the Federation.4 Bavaria for instance applied a very extensive language test, requiring applicants for naturalisation to read a German text aloud and to write a few sentences in German (Dornis 2001). Most other Länder however preferred an indirect test of language 2 Einbürgerungsrichtlinien von 15. 12. 1977, GMBl. 1978, S. 16. 3 Ibid. 4 To provide for more uniformity in the application of the provisions of the federal laws, Article 84 of the Basic Law provides for a procedure for the government to issue general administrative guidelines (Allgemeine Verwaltungsvorschriften), with consent of the Bundesrat (Article 84 paragraph 2 of the Basic Law). Guidelines which are realised with the consent of the Bundesrat are binding on the Länder (but they do not bind the courts). Until federal guidelines have been adopted with the consent of the Bundesrat, the Länder may issue their own administrative  guidelines (Hailbronner 2006: 239). In the meantime, the Minister can issue vorläufige Anwendungshinweise (temporary application instructions), which are non-binding on the Länder, to pave the way until the federal guidelines are put in place.

‘Deutschland sucht den Superstaatsbürger’  69

skills on the basis of the oral and written communication which was required for completing the application. Besides having language skills, the 1977 guidelines required the applicants for naturalisation to have “culturally assimilated” (Joppke 1999: 201). Naturalisation was only granted if an applicant had a “voluntary and permanent orientation (Hinwendung) to Germany” which was assessed by his or her “basic attitude to German culture” (Hailbronner & Renner 1991: 626). This assimilation was not derived from objective facts, such as the length of residence or absence of a criminal record, but evaluated on a case by case basis via an “examination of the life course and the personality of the applicant” (Hailbronner & Renner 1991: 628). In practice, naturalisation officers would visit applicants’ homes and question their neighbours (Joppke 1999: 203).5 In addition, applicants for naturalisation were required to have a basic knowledge of the order of the German state, and to acknowledge Germany’s free and democratic basic order (Dornis 2001: 128). It was clear that naturalisation could only occur at the end of a successful integration process and was not considered as a means to achieve a fuller integration (Green 2007: 99). The uninviting citizenship regime with a naturalisation process being largely at the discretion of the authorities resulted in a low number of naturalisations. Until the late 1980s, the naturalisation rate remained under a half per cent, in absolute numbers less than 10,000 naturalisations a year, the lowest in Europe (Joppke 1999: 202). At the same time, immigrants of German descent (Aussiedler) were provided with an almost automatic right to citizenship (Hailbronner 2006: 216).6 The dissonance of de facto foreigners automatically classifying as Germans, a group that witnessed a massive increase after the liberalisation and democratisation of the Eastern bloc (Hailbronner 2006: 216), and of de facto Germans classifying as foreigners, i.e. a third generation of immigrants born to foreign parents who had themselves grown up in Germany, called for a change in German nationality law. This call was underlined by a decision of Germany’s Constitutional Court (Bundesverfassungsgericht) of 31 October 1990, which stated that Länder laws granting local voting rights to foreigners in order to diminish Germany’s democratic deficit were unconstitutional (Hailbronner 2006: 220).7 According to 5 According to Joppke, the Bavarian assimilation test “unsurprisingly went the extra mile” by asking applicants to sing the first verse of the Bavarian national anthem, how to recognise a German Volkszugehöriger, or to name the German territories lost after the second World War (Joppke 1999). 6 Under Article 116 of the German Basic Law (Grundgesetz), ethnic Germans possessing the legal status of a German without German nationality were entitled to German nationality on the basis of their admission to German territory. 7 Bundesverfassungsgericht, judgment of 31.10.1990, volume 83, page 37, 51f.

70  Chapter Four

the Court, the only solution to solving the gap between permanent population  and democratic participation lay in amending the nationality law, for instance via facilitating the acquisition of German nationality (Hailbronner 2006: 220). Laws liberalising the naturalisation policy were passed in the early 1990s. In 1990 a simplified naturalisation procedure (erleichterte Einbürgerung) was introduced in the Ausländergesetz (Foreigners Act) for two groups of immigrants: those with a long period of residence (Article 86) and those aged between 16 and 23 (Article 85). When fulfilling the necessary conditions, these applicants for naturalisation were to be naturalised ‘as a rule’ (Regelanspruch). At the same time, the fees for naturalisation were reduced from 5,000 to 100 DM (Carle 2007: 152). The Regelanspruch was transformed into a definitive right to naturalise (Anspruchseinbürgerung) for both groups in 1993 (Green 2000: 111). As regards the first group (first-generation immigrants with a long period of residence), a right to naturalisation existed if an immigrant had lived in Germany for a period of fifteen years or longer. Further requirements were renunciation of the original nationality, absence of a criminal record and the ability to earn a living without any recourse to social welfare or unemployment benefits (Dornis 2001).8 Young foreigners aged between 16 and 23 had a right to naturalise under similar conditions if they had resided in Germany for a period of eight instead of fifteen years. Furthermore, this category did not have to prove the ability to earn a living. In neither case were language skills required. It was assumed that young foreigners would generally have a good command of the German language since they had attended German schools, and long-term resident foreigners were supposed to have integrated into German society during their time of residence (Joppke 1999: 203). The above shows that the early 1990s witnessed a breach with two core principles of the German naturalisation policy: absolute state discretion and cultural assimilation as a precondition for German citizenship. The liberalisation of nationality law led to a considerable increase in the naturalisation rate, which went from 0.4% in 1990, to 2% in 1999 (Green 2004). The liberalisation proved especially successful for Turkish nationals. Whereas in 1989, 1,713 of the 1,612,600 Turkish nationals living in West Germany naturalised, this number rose to 103,900 naturalisations in 1999, the last year the 1993 law was in effect (Nathans 2004: 248).

8 Article 86 Foreigners Act (applicable until 2000). A hardship clause prevented that those who could not be blamed for the fact that they profited from social security benefits would be unable to benefit from the right to naturalise.

‘Deutschland sucht den Superstaatsbürger’  71

4.2. The Reform of 2000: Further Liberalisation? After the social democratic SPD won the 1998 parliamentary elections, the time seemed right for a further liberalisation of German citizenship law. Together with Alliance 90/the Greens, the SPD formed a coalition, which enjoyed a comfortable majority in both houses of parliament. The newly formed government admitted that in the past, an irreversible immigration process had taken place and that, consequently, “the integration of those immigrants who live here on a permanent basis and who accept our constitutional values” needed to be addressed.9 The centre of the coalition’s integration policy would be formed by the creation of a modern nationality law. The coalition especially envisaged the introduction of two simplifications. The first would be the automatic conferral of German nationality to children of qualified immigrant parents born on German soil.10 The second would be the introduction in the law of a general tolerance of multiple citizenships (Green 2004: 97). At a press conference on 13 January 1999, Interior Minister Otto Schily (SPD) presented a first draft for the reform of the nationality law.11 This controversial draft would however never be discussed. Its chances of passing through both houses of parliament had vanished after the Christian Democratic Union (CDU) won the elections in the state of Hesse, causing the government to lose its majority in the Bundesrat, parliament’s upper house representing the Länder (Hailbronner 2006: 223). Furthermore, in a petition drive against dual nationality, organised jointly with the Christian Democratic Union (CSU), the CDU managed to obtain five million signatures from German citizens. The combination of the success of the petition drive and the loss in the Hesse elections forced the coalition to amend the bill to get support from the liberal Free Democrats (FDP) (Green 2004: 102).12 The Union parties were able to use the mechanism of getting the public involved to force the government to adapt its proposals now the conditions for naturalisation had been specified in the law. Had these conditions, as

  9 Coalition agreement between the SPD and the Alliance 90/Greens, signed in Bonn on 20 October 1998. Quoted from Hansen 2003. 10 During the negotiations between the coalition partners, it was decided that only children of parents who had been born in Germany or who had moved to Germany during childhood, i.e. before the age of fourteen, would acquire German nationality at birth (Green 2004: 97). 11 Faced with the threat of the CDU/CSU petition drive against dual citizenship, Schily introduced some “decidedly conservative markers” in the Bill, which required naturalisation applicants to fulfil a language requirement and to declare loyalty to the Constitution (Green 2004: 99). Furthermore, the draft raised the naturalisation fees from 100 to 500 DM. 12 The success of the CDU in the Hesse elections has actually been ascribed to the success of the petition drive (Green 2004: 101).

72  Chapter Four

had been the case until 1990, remained specified in the administrative guidelines (Verwaltungsvorschriften), involving the public would have been much harder, as changes in the conditions would have been the result of negotiations between federal and state officials.13 A ‘watered down’ bill was sent to the Bundestag in March 1999.14 Rather than abolishing the renunciation requirement, as it had first proposed, the coalition government now envisaged making dual nationality the exception rather than the rule.15 The Bill maintained the proposal to introduce a form of ius soli citizenship attribution, but required that children who had acquired German nationality this way would have to choose between German nationality and the nationality they had acquired via their parents before reaching the age of 24.16 The explanatory memorandum stated that the bill intended to achieve two aims. As explained above, the first aim was to increase the number of naturalisations of the large number of permanent residents in Germany. The second aim of the bill was to improve the integration of immigrants who were permanently living in Germany and of their German born children, by facilitating access to German nationality.17 According to Minister of the Interior Schily, integration could only succeed when citizens of foreign descent had the possibility to participate equally in social life in Germany.18 Hence, as was the case in the Netherlands, which had liberalised the nationality law in 1985 to increase the number of naturalisations, naturalisation was seen as a means for integration, and nationality law was to be used as a means to pursue the aims of integration policy, something which the explanatory memorandum stated explicitly.19 Apart from opening up access to German nationality, the bill also introduced new requirements for naturalisation. A language requirement was introduced, as integration required future German citizens to be able to understand the German media and to communicate with the German people. Without language skills, integration, which included participation in the process of political opinion making (Beteiligung am politischen Willensbildungsprozess), would, according to 13 I thank Professor Jürgen Bast for pointing this out to me. 14 Entwurf eines Gesetzes zur Reform des Staatsangehörigkeitsrechts, Bundestag Drucksache (BT-Drs.) 14/533. 15 The exception clauses to the renunciation requirement were extended (Gerdes, Faist & Rieple 2007: 47). 16 Furthermore, rather than being born out of parents (or: a parent) who had migrated to Germany before the age of fourteen, it would be required that at least one of the parents would have been legally residing in Germany for at least eight years. It would moreover be required that the parent(s) had a permanent residence permit. 17 BT-Drs. 14/533, p. 2. 18 Plepro 14/40, p. 3417. 19 BT-Drs. 14/533, p. 10.

‘Deutschland sucht den Superstaatsbürger’  73

the explanatory memorandum, not be possible. ‘Sufficient knowledge of the German language’ would therefore be required, but in principle not formally tested.20 Eventually, in practice, a language check would however be carried out in some of the Länder, who after all had already built up experience in the testing of language skills within the framework of discretionary naturalisation under Article 8 of the RuStAG. Those applying for citizenship would furthermore be required to have an “inner orientation towards the Federal Republic of Germany” (innere Hinwendung zur Bundesrepublik Deutschland). The bill thus required applicants for naturalisation to “acknowledge the free and democratic basic order of Germany’s Basic Law”.21 By introducing these requirements, the bill provided for a partial return to the 1977 guidelines, which also required naturalisation applicants to have language skills and a ‘Hinwendung’ to Germany. Rather than via visits to applicants’ homes, signing a declaration of loyalty would this time suffice as proof for the Hinwendung.22 Apart from the introduction of a language requirement and a declaration of loyalty, the bill strengthened the naturalisation requirements by raising the naturalisation fee fivefold, from 100 to 500 DM, thereby partially undoing the lowering of the fees which had occurred nine years before in 1990. In the explanatory memorandum, the amount of 500 DM was referred to as being ‘cost-effective’.23 Eventually, rather than liberalising the rules relating to naturalisation, the proposal of law reinforced some conditions. The only liberalisations consisted of the reduction in the residence requirement from fifteen to eight years, and the extension of the exemption clauses to the renunciation requirement. The Union parties nevertheless opposed the bill, which, in their opinion, did not require enough, and submitted an alternative bill.24 In their proposal, the Christian parties attributed another place to citizenship and naturalisation in the integration process than the government: rather than seeing naturalisation as a necessary means to obtaining full equality and a condition for integration, they saw acquiring German citizenship as the expression of a successful integration and socialisation in 20 BT-Drs. 14/533, p. 18. 21 “Ein Ausländer […] ist auf Antrag einzubürgern, wenn er sich zur freiheitlichen demokratischen Grundordnung des Grundgesetzes bekennt.” Article 85(1)(1) of the bill. 22 Applicants would be required to sign a one-page declaration, in which they would agree to particular elements of the Constitutional Law, such as free and democratic elections, the rule of law, the right to parliamentary opposition, the right to change the government, the independence of the courts, the rejection of any form of despotism and the human rights as laid down in the Constitutional Law. Applicants would also declare that they had never and did not currently pursue any activities contrary to the free and democratic basic order, the public security or impeding public administration and authorities (Michalowski 2010b: 195). 23 BT-Drs. 14/533, p. 20. 24 BT-Drs. 14/535.

74  Chapter Four

Germany. Naturalisation should therefore only be available for those immigrants who had obviously adapted to the German way of life and who had permanently chosen Germany as the centre of their lives.25 This could be assumed if the applicant had sufficient knowledge of the German language and of the constitutional order of the Federal Republic of Germany. The Christian Democrats hence had proposed to (re)introduce knowledge other than language as a condition for naturalisation as early as 1999. As we will see, ten years later, they would see their wish granted. Discussing the Reform Bill in the Parliament Parliament devoted most attention to the issues of birthright citizenship and dual nationality, since these were the most controversial between the political parties. Far less attention was paid to the introduction of the language and integration requirements. Two points of view can be discerned. On the one hand, left-ofcentre political parties criticised the new bill for introducing new obstacles in the form of a language requirement and a declaration of loyalty, which would form barriers to equal treatment. Jelpke, member of the socialist PDS, criticised the language requirement for being undefined and for its creation of a very high hurdle for naturalisation, especially for the elderly.26 The PDS introduced proposals for amendment aiming inter alia at the deletion of the declaration of loyalty and the language requirement, and at keeping the fees for naturalisation at 100 DM.27 All proposals were rejected. On the other hand, the bill, unsurprisingly, received criticism from CDU/CSU parliamentarians, who blamed the government for giving away naturalisation for free (Einbürgerung zum Nulltarif), and for thinking that integration would follow automatically after a passport had been issued.28 The Christian Democrats in the Bundestag were supported by FDP member Westerwelle, according to whom naturalisation required a conscious turn towards German society.29 Whoever wished to obtain German nationality via naturalisation would need to “convert” to Germany: the German pass was “not just a paper, which one likes to accept additionally”.30 In the Bundesrat Edmund Stoiber (CSU), minister-president of the Land Bavaria, opposed the bill by emphasising the importance of nationality law as a means to maintain the identity and cohesion of German society, established  by a common history, culture and coherence in nation and homeland (Heimat). He claimed that integration consequently not only required the foreign 25 BT-Drs. 14/535, p. 1. 26 Plepro 14/28, p. 2296 (D). 27 BT-Drs. 14/992, BT-Drs. 14/993, BT DS 14/996. 28 Plepro 14/28, p. 2310, Plepro 14/40, p. 3459 (A). 29 Plepro 14/40, p. 3436 (C). 30 Plepro 14/40, p. 3436 (C).

‘Deutschland sucht den Superstaatsbürger’  75

population to learn the language, but also the decision to “clearly join our nation, being a cultural community, a community of values and solidarity.”31 Despite votes against the Act by certain CDU/CSU and PDS parliamentarians, the bill was adopted in the Bundestag on 7 May 1999. The Bundesrat approved the bill a few weeks later on 21 May 1999. The new Nationality Act (Staatsange­ hörigkeitsgesetz, StAG) entered into force on 1 January 2000. The polarisation in the debates, caused by the clash between the positions taken by the SPD, Greens and PDS on the one hand and that of the CDU/CSU and, to a certain extent, FDP, on the other, has eventually led to the coming about of a law containing many contradictions. A fairly liberal form of ius soli acquisition was introduced, whereas dual nationality was rejected at the same time (Triadafi­ lopoulos & Faist 2006: 9). Moreover, the reduction of the residence requirement from fifteen to eight years was accompanied by the introduction of a language requirement, a declaration of loyalty and a five-fold increase in the naturalisation fees. As the new Act reinforced rather than liberalised the conditions for naturalisation, the number of naturalisations quickly decreased after 2000. Whereas 190,000 immigrants acquired German nationality in 2000, this number decreased to 113,000 in 2007.32 This means that one of the primary goals of the new Act, being to increase the number of naturalisations, was, unsurprisingly, not achieved.33 The Interpretation of the Language Requirement and the Declaration of Loyalty As we have seen above, the new Nationality Act introduced a language requirement for naturalisation and the requirement to acknowledge the free and democratic basic order of Germany’s Basic Law. Below, we will see how these requirements were interpreted and applied in the Länder. • The Language Requirement As we have seen above, the language requirement for discretionary naturalisation under Article 8 of the RuStAG was differently applied in the Länder. Differences of opinion concerning the interpretation of the language requirement were also expressed in the negotiations on the guidelines on the application of the nationality law of 2000 (Allgemeine Verwaltungsvorschrift zum Staatsangehörigkeitsrecht (StAR-VwV)) (Green 2000: 114, Hofhansel 2008: 177).34 In the discussions, it came 31 738th session of the Bundesrat, 21 May 1999, p. 181 ff. 32 See paragraph 7.1.1. The high number of naturalisations in 2000 can partially be explained by the rise in the number of applications prior to the introduction of the new Act, filed by immigrants who wanted to avoid the stricter requirements. 33 As about 40,000 immigrants per year acquire German nationality via ius soli, this provision appears to have nevertheless facilitated access to German nationality. 34  “Deutschland: Verwaltungsrichtlinien für das neue Einbürgerungsrecht”, Migration und Bevölkerung, Ausgabe 9, Dezember 1999. Accessible via http://www.migration-info.de/mub

76  Chapter Four

to a clash between the government and the CSU-led Land of Bavaria, supported by Baden-Württemberg, which was also Union-led (Davy 2008: 33). Bavaria proposed requiring both oral and written language proficiency and basic knowledge of the free and democratic basic order, whereas the federal government, as we have seen above, preferred to assume rather than test language skills. The nonbinding provisional application instructions issued by the federal Ministry of the Interior (vorläufige Anwendungshinweise) eventually stated that sufficient knowledge of the German language existed if an applicant for naturalisation could cope for him/herself in his/her German surroundings in daily life and in contacts with institutions by being able to hold oral conversations on a level which corresponds to his/her age and level of education. This also requires that the applicant is able to read, understand, and reproduce the essential contents of a German text about daily affairs. Impediments which permanently hamper the applicant’s ability to speak or read should be taken into account. Simple oral language proficiency is not sufficient.35

If the applicant could not submit proof of knowledge of German in the form of a diploma, the local naturalisation official would be required to test whether the applicant’s language skills were indeed ‘sufficient’.36 The application instructions provided for a more lenient application of the language requirements for spouses of German citizens, who would be required merely to “be able to express themselves in daily life without experiencing considerable difficulty”.37 The above shows that the Union-led Länder eventually succeeded in leaving a considerable mark on the interpretation of the language requirement. After they had managed to curtail the liberalising changes in German nationality law by the petition drive in 1999, they succeeded in further diminishing the liberalisation of German nationality law by influencing the text of the application instructions. Since the Länder were left a considerable margin of appreciation, the application of the language requirement continued to vary considerably from state to state. Differences existed in the way the Länder took account of the applicant’s age and level of education as required by the application insructions (Renner artikel.php?Id= 990901. Site visited on 1 July 2009. Differences of opinion also existed regarding the question of when an applicant for naturalisation was not responsible for his/her dependence on social assistance and the testing of an applicant’s loyalty to the Constitution. 35 Vorläufige Anwendungshinweise des BMI, 10 December 2004, no. 11.1.1.1. 36 The application instructions stated that sufficient knowledge of German would generally be assumed if an applicant had obtained the ‘Zertifikat Deutsch’, requiring language skills at level B1 of the CEF, or a comparable (language) diploma. 37 Vorläufige Anwendungshinweise des BMI, 10 December 2004, No. 9.1.2.1. and No. 9.1.2.1 Abs. 3 StAR-VwV. A similar language requirement had applied to spouses of German nationals applying for naturalisation under Article 9 of the RuStAG prior to the reform of 2000 (Renner 2002: 344).

‘Deutschland sucht den Superstaatsbürger’  77

2002: 346), and the way in which language was tested: whereas language skills were normally assessed in an interview between the applicant and an official of the municipal foreigner’s office (Ausländerbehörde), some Länder directed naturalisation applicants to an adult education centre (Volkshochschule).38 • Acknowledging Germany’s Free and Democratic Basic Order Not only the language requirement led to divergent practices in the Länder. This also applied to the requirement to acknowledge the free and democratic basic order of the Germany’s Basic Law and the declaration of loyalty. The year 2006 witnessed the introduction in the state of Baden-Württemberg of the so-called ‘interview manual’ (Gesprächsleitfaden), a list of questions which would be used if doubts existed as to whether the applicant had understood the declaration of loyalty, or whether the declaration matched the applicant’s inner convictions (Wolfrum & Röben 2008: 4). Examples of questions were “What is your attitude towards criticising a religion? Do you think that is allowed?”, “What do you think about the statement that a woman should obey her husband and that her husband is allowed to beat her, if she is disobedient?” and “What is your opinion about forced marriages? Do you think such marriages are compatible with human dignity?”. The interview manual, labelled ‘Muslim test’ in the press, was initially to be used as a rule in all cases where an applicant for naturalisation originated from a country being a member of the Islamic Conference.39 After research conducted under the authority of the city of Heidelberg established that this was discriminatory (Wolfrum & Röben 2008: 15), the interview manual was subsequently applied in all cases where doubts existed regarding the applicant’s acceptance of Germany’s order of values (Werteordnung), regardless of the applicant’s belief. The interview manual nevertheless remained subject to controversy. Also in 2006, the Land Hesse, like Baden-Württemberg CDU-led, introduced a proposal for a naturalisation test in the form of a questionnaire of 100 questions.40 Like the interview manual, this test was introduced within the framework of the declaration of loyalty, as “only those persons who know how the state and the 38 Two judgments from the Bundesverwaltungsgericht, the German Federal Administrative Court (FAC), of 20 October 2005 aimed to end the differences in application. The Court ruled that the requirement to have ‘sufficient knowledge of German language’, next to oral language skills, also implied having certain written language skills. Whereas an applicant was not required to be able to write German him/herself, he or she should be able to read and dictate a German text on daily affairs and to check the accuracy of other people’s writings. BVerwG 5 C 8.05 and BVerwG 5 C 17.05, www.bundesverwaltungsgericht.de. 39  “Keine Diskriminierung islamischer Einbürgerungsbewerber”. http://www.innenministerium .baden-wuerttemberg.de/de/Meldungen/111612.html. Site visited 21 April 2009. 40 Leitfaden Wissen & Werte in Deutschland und Europa. Hessisches Ministerium des Innern und für Sport, 14 March 2006.

78  Chapter Four

society operate, can understand, judge and accept its values”.41 The test proposed by Hesse contained questions regarding the applicant’s inner values (Werte) similar to those asked in Baden-Württemberg, such as “What do you say if someone characterises the Holocaust as a myth or a fairytale?” and “A woman should not be allowed to show herself in public or to travel without being accompanied by a male member of the family: what is your opinion?”. The Hesse test however also contained questions relating to objective knowledge (Wissen) by for instance asking candidates to name German philosophers and low mountain ranges, and to state what is depicted on Caspar David Friedrich’s portrait of Rügen. Both the test proposed by Hesse and the interview manual introduced by Baden-Württemberg were heavily criticised by federal and local parliamentarians, as well as academics and societal organisations.42 Whereas the Hesse test would never be put into practice because of the introduction of a federal naturalisation test in 2008, the interview manual in Baden-Württemberg was eventually officially abolished in July 2011. It however appears that it is still applied in the practice.43 The differences in the application of the naturalisation requirements resulted in differences between the naturalisation rates of the Länder. The naturalisation percentage in Schleswig-Holstein, where the guidelines were applied to 41  Thus Hesse’s Interior Minister Volker Bouffier: “Innenminister Bouffier legt Konzept für Einbürgerung vor, ‘Wissens- und Wertetest’ soll für alle Bewerber Pflicht werden”’, Presse­ mitteilung des hessischen Innenministeriums vom 14.03.2006; Einbürgerungstest-Fragenkatalog in Hessen, Innenminister Bouffier, www.migrationsrecht.net, 16 March 2006, accessed on 2 September 2008. 42 For criticism on the interview guidelines of Baden-Württemberg, see for instance: “Deutschland: Streit um Einbürgerungsleitfaden”, in Migration und Bevölkerung, Ausgabe 1, Februar 2006, “Extrafragen an Muslime umstritten; Parteien kritisieren Leitfaden des Innenministeriums für Einbürgerung”, Stuttgarter Nachrichten, 7 January 2006, “Germans to put Muslims through loyalty test”, UK Telegraph, 31 December 2005, Stellungnahme des Bischöflichen Ordinariats Rottenburg-Stuttgart und des Erzbischöflichen Ordinariats Freiburg, and Hanschmann 2008, Joppke 2010a, 2010b, Orgad 2010. For criticism regarding the test proposed by Hesse see: “Organspenden für Europa; Vorsicht bei Frage 78 des hessischen Einbürgerungstests: Es ist entweder ein Fehler – oder eine perfide Fangfrage”, Die Tageszeitung, 5 May 2006, “20 Rüsselsheimer im Einbürgerungstest”, www.taz.de, 16 March 2006 (site accessed on 10 March 2008), “Hessen will bundesweiten Fragenkatalog”, www.spiegel.de/politik/deutschland/0,1518,405970,00.html (site accessed on 5 March 2009), “Merkel kocht Hessen runter”, Stern.de, 20 March 2006 (site visited 7 July 2009). In the Bundestag, the tests of both Baden-Württemberg and Hesse were criticised by FDP, Green and SDP parliamentarians, inter alia during the debates concerning the plans for the federal budget of 2006, 29th session of the Bundestag, 30 March 2006, Plepro 16/29, p. 2435 (A). 43 Information obtained from Professor Kees Groenendijk, who acquired the information from the leader of Heidelberg’s foreigners’ service (Ausländerbehörde) during the Hohenheimer Tage zum Ausländerrecht of January 2012, on file with the author.

‘Deutschland sucht den Superstaatsbürger’  79

the letter, for instance was above 3% in the first seven years after the coming into force of the new law (Renner 2002: 426, Tränhardt 2008: 16–18). The percentages in Bavaria however never rose above 2%, and have been below 1.5% since 2002.44 Because of the differences in the application of the naturalisation requirements, Union-led Länder feared ‘naturalisation tourism’: in order to avoid the strict naturalisation policy in their Land of residence, immigrants were suspected of filing their applications in a Land which applied the naturalisation requirements more leniently.45 They therefore called for uniform naturalisation requirements. 4.3. Amendments following the Zuwanderungsgesetz On 7 February 2003, roughly three years after the coming into force of the reformed Nationality Act, the federal government submitted a proposal for an Act on the ‘Regulation and Limitation of Immigration and of the Regulation of the Residence and Integration of Union Citizens and Immigrants’, the so-called Zuwanderungsgesetz (Immigration Act).46 The Act entered into force on 1 January 2005.47 It amended the Nationality Act by replacing the articles on naturalisa­ tion  from the Foreigners Act, which was repealed, to the Nationality Act (Staatsangehörigkeitsgesetz). Secondly, upon the initiative of the opposition party CDU/CSU, a new article was introduced in the Nationality Act, which required the naturalisation authorities to submit the personal data of any applicant who has reached the age of sixteen to the secret services (Bundesamt für Verfas­ sungsschutz, Hailbronner & Renner 2005: 836, Hailbronner 2006: 227). According to the CDU/CSU fraction, the terrorist attacks of 9/11 and other attacks showed that it was absolutely necessary to fully make use of all possibilities in the naturalisation process to exclude terrorists from acquiring German nationality.48 Lastly, the Nationality Act required adaptation because of the new residence titles provided for in the new Residence Act (Aufenthaltsgesetz): the limited-term residence permit (Aufenthaltserlaubnis) and the unlimited-term settlement permit (Niederlassungserlaubnis).

44 See paragraph 7.1.1 for a further comparison of the naturalisation percentages in the Länder. 45  “Deutschland: Streit um Einbürgerungsleitfaden”, Newsletter Migration und Bevölkerung, Ausgabe 1, February 2006. The phenomenon would also be referred to by federal Interior Minister Schäuble in defence of formalisation of the naturalisation requirements (Plepro 16/29, p. 2433, 30 March 2006). 46 BT-Drs. 15/420. 47 BGBl. 2004 Teil I, Nr. 41. 48 BT-Drs. 15/955, p. 44.

80  Chapter Four

One of the major features of the Immigration Act has been the emphasis on integration requirements (Hailbronner 2006: 225). Consequently, the discussion surrounding the adoption of the Immigration Act took the requirement to have knowledge of the German language out of the context of naturalisation: knowledge of the language became the ‘key to integration’ in general (Davy 2008: 34). The new law introduced integration courses, modelled after the Dutch example, of 600 hours of language tuition and 30 hours of civic education.49 Only those who participated successfully in the integration courses, meaning that they attained level B1, would be able to acquire a permanent residence permit.50 Furthermore, the requirement of having “basic knowledge of the juridical and social order and the living conditions in Germany” was introduced as a condition for obtaining a permanent residence permit.51 Many MPs hoped that knowledge of German language would hinder the arising of ‘parallel societies’ (Parallelgesellschaften), thereby forming an antidote to segregation (Davy 2008: 34). Compared to a few years earlier, a change in thinking becomes apparent: during the discussions on the naturalisation reform, segregation was thought to be best prevented by including immigrants by offering them full membership rights. This is now rather the other way round: in order to prevent the arising of parallel societies, immigrants should first integrate, after which they will be rewarded with a stronger residence right. The Immigration Act which was eventually introduced differs from the intentions of the Commission for Immigration which had been installed in 2000 to make policy recommendations on immigration policy. The recommendations of this so-called Süssmuth Commission had actually kicked off the debates on introducing a German integration programme. According to the Commission, learning German was the key to a successful integration. It proposed that language and societal orientation courses, modelled on the Dutch model, would be offered (Michalowski 2010b: 187). The Commission had however rejected the use of sanctions and did not even mention the possibility to link residence rights to successful participation in the integration programme (Michalowski 2010b: 187).

49  After the adoption of an Act implementing EU Directives on asylum and immigration (Richtlinienumsetzungsgesetz), immigrants who need a more extensive training can, since 1 January 2008, follow 900 hours of language education. The number of hours dedicated to civic education increased from 30 to 45 hours. 50 Verordnung über die Durchführung von Integrationskursen für Ausländer und Spätaussiedler (Integrationskursverordnung), of 13 December 2004, article 3(2). 51 Residence Act (Aufenthaltsgesetz) Article 9 paragraph 2 number 8.

‘Deutschland sucht den Superstaatsbürger’  81

4.4. Proposals to Change the Staatsangehörigkeitsgesetz Put Forward at the

IMK of May 2006

The changes brought about by the Immigration Act did not put a stop to the perceived need for an increased harmonisation in the application of the naturalisation requirements. They would, however, eventually influence the outcomes of new debates regarding the naturalisation requirements, which kicked off after the conference of Ministers and Senators of the Interior (Konferenz der Innenminister und -senatoren, IMK) of 4 and 5 May 2006.52 The need for harmonisation of the naturalisation requirements stood high on the agenda of the IMK. It was furthermore mentioned as a goal in the coalition agreement of the ‘great coalition’ government consisting of the Union parties and the SPD which had been in power since November 2005.53 Nevertheless, at the IMK, the Union-led Länder clashed with those led by SPD-led governments. Whereas the former Länder proposed to introduce civic courses and tests for nationalisation applicants, the latter opposed an extensive knowledge and values test after the examples of Hesse and Baden-Württemberg.54 Eventually, an agreement regarding the standards for naturalisation was however unanimously agreed upon. In future, all applicants for naturalisation would be required to prove mastery of the German language at level B1 both orally and in writing in a language test.55 Furthermore, citizenship applicants would be required to demonstrate basic civic knowledge

52 The IMK consists of the 16 Ministers of the Interior of the Bundesländer, joined by the federal Minister of the Interior. It is a consultative organ where the Länder and the Bund meet on equal footing. In reality, the IMK generally serves the purpose of reaching a compromise between the parties forming the ‘great coalition’: the Social Democrats (SPD, ‘A-Seite’) and the Christian Democrats (CDU/CSU, ‘B-Seite’). Resolutions adopted in the IMK are hard to ignore, and will generally be accepted by the federal government, especially if the government consists of the parties forming the great coalition. 53 ‘Gemeinsam für Deutschland. Mit Mut und Menschlichkeit.’, Koalitionsvertrag von CDU, CSU und SPD, 11 November 2005. Accessible via http://www.bundesregierung.de/Content/DE/__ Anlagen/koalitionsvertrag,property=publicationFile.pdf, site visited 3 July 2009. 54 The minutes of the Conference are undisclosed; only the resolutions are published. In order to describe the process of decision making in the IMK, I therefore relied on the following articles: “Innenminister einigen sich nicht auf Einbürgerungstest”, Hamburger Abendblatt, 2 May 2006, “Noch mehr Hürden; SPD gibt nach: Einbürgerungswillige müssen künftig eine Prüfung zur Staatsbürgerkunde ablegen”, Die Tageszeitung, 6 May 2006 and “Einbürgerung; Testfrage Integration; Nach kurzem Wahlkampf-Streit suchen Union und SPD einen Kompromiss bei der Prüfung von Neubürgern”, FOCUS, 13/2006. 55 Sammlung der zur Veröffentlichung freigegebenen Beschlüsse der 180. Sitzung der Ständigen Konferenz der Innenminister und –senatoren der Länder am 5. Mai 2006 in Garmisch Partenkirchen, p. 13.

82  Chapter Four

and knowledge of the basic principles and values of the German Constitution.56 The aim of the new requirements was thought to lie in promoting the loyalty of future German citizens. As Interior senator of Hamburg, Udo Nagel, declared: “the goal is clear: we do not want to naturalise enemies of our Constitution”.57 After the IMK, chair Beckstein (CSU, Bayern) and the Federal Commissioner for Migration, Maria Böhmer (CDU) explicitly stated that the new requirements  were not meant to create new hurdles.58 Beckstein distanced himself from the naturalisation test developed by Hesse by stating that questions regarding geographical and historical peculiarities would not be asked.59 According to Beckstein, the level of the new requirements would not be too high: “We would not want that only academics will be able to naturalise.”60 By making their statements, Beckstein and Böhmer appeared to take account of the fact that the new language and integration requirements could in reality form barriers to naturalisation. A Bundesrat bill containing amendments to the conditions for as-of-right naturalisation was subsequently put forward.61 The explanatory memorandum of the bill stated that naturalisation could only be the “end point” (Schlußstein) of a successful integration, because only then would a foreigner be able to participate in the social and political life.62 It was hence clear that the Bundesrat no longer saw naturalisation as a means for integration, but that it had adopted the position previously mainly held by the Union parties. As only integrated immigrants who “accept the Constitution and its order of values as obligatory principles for life in Germany” should be able to naturalise, the bill required “elementary” knowledge 56 Ibid. 57  “Innenministerkonferenz – Einbürgerungskurs und Eidesleistung zur Einbürgerung”, www .migrationsrecht.net, 16 May 2006, site accessed 10 March 2008. Apart from agreements on the language and knowledge of society requirements, the Interior Ministers decided on a reinforcement of the public order requirement, and agreed that the applicants for naturalisation should in future be questioned regarding their membership in or support of extremist organisations. (Sammlung der zur Veröffentlichung freigegebenen Beschlüsse der 180. Sitzung der Ständigen Konferenz der Innenminister und -senatoren der Länder am 5. Mai 2006 in Garmisch Partenkirchen, p. 13–14). 58 “Sprachtests und Kurse vor der Einbürgerung; Innenminister einigen sich auf einheitliche Standards/ Ausführung bleibt den Ländern überlassen; Kompromis im monatelangen Streit”, Süddeutsche Zeitung, 6 May 2006; Press release “Durch Einbürgerung und Integration zum Staatsbürger”, 5 May 2006, REGIERUNGonline, site accessed 20 January 2009. 59 “Sprachtests und Kurse vor der Einbürgerung; Innenminister einigen sich auf einheitliche Standards/ Ausführung bleibt den Ländern überlassen; Kompromis im monatelangen Streit”, Süddeutsche Zeitung, 6 May 2006. 60 “Innenminister einigen sich nicht auf Einbürgerungstest”, Hamburger Abendblatt, 2 May 2006. 61 25 April 2007, BT-Drs. 16/5107. 62 BT-Drs. 16/5107, p. 1.

‘Deutschland sucht den Superstaatsbürger’  83

of the Constitution and its principles and values. Furthermore, naturalisation applicants would need to demonstrate oral and written language skills “orientated towards” level B1.63 Requiring level B1 was thought to be too demanding, since experience had taught that only a small number of the participants in the integration courses, introduced within the framework of the 2005 Immigration Act (Zuwanderungsgesetz), actually reached this level. As the pass rate in comparable language tests for naturalisation which were already employed by certain Länder was 70%, the proposed language test was “hence” thought to be “reasonable”.64 The Bundesrat thereby took for granted the fact that the new require­ ment  would exclude about one-third of the immigrant population from full membership. The proposals in the Bundesrat bill sharply contrasted with the recommendations of federal, state and communal foreigners’ and integration commissioners made a year before. In a declaration, these commissioners had stated that efforts needed to be made in order to increase the number of naturalisations, since many of Germany’s permanent residents had not obtained German nationality, and the number of naturalisations had been decreasing since the year 2000. They advised that the language requirement be eased, not strengthened, especially for the first generation of immigrants.65 4.5. The Government’s Proposals in the Richtlinienumsetzungsgesetz On 23 April 2007, the great coalition Government introduced a bill for the implementation of eleven EU Directives on asylum and migration (Entwurf eines Gesetzes zur Umsetzung aufenthalts- und asylrechtlicher Richtlinien der Europäischen Union).66 Despite the bill’s title, its Article 5 provided for important amendments in the Nationality Act, along the lines of the IMK resolutions, but requiring more from naturalisation applicants than the Bundesrat bill.67 As a justification for requiring more, the federal government referred to the requirements for obtaining a permanent residence permit. The Residence Act of 1 January 2005 required applicants for permanent residence to have language skills at level B1, as

63 BT-Drs. 16/5107, p. 1. 64 BT-Drs. 16/5107, p. 10. 65 “Einbürgerung erleichtern und fördern”, declaration of the federal conference of immigration and foreigners commissioners on the subject ‘naturalisation’, Düsseldorf 24 and 25 May 2005. The foreigners’ and integration commissioners furthermore pleaded for a less strict application of the renunciation requirement. 66 BT-Drs. 16/5065. 67 The bill put forward by the Bundesrat was not adopted.

84  Chapter Four

well as a basic knowledge of the juridical and social order and the living conditions in Germany.68 To observe the “gradual rise in integration policy”, the proposed Act required applicants for naturalisation to have “knowledge of the juridical and social order and of the living conditions in Germany” instead of “basic knowledge”.69 For the same reason, it refrained from asking for language skills ‘orientated towards’ level B1, by asking for oral and written language skills at level B1 instead. The requirements for permanent residence hence exerted an upward pressure on the requirements for naturalisation. As in the Netherlands, the requirements for naturalisation would hence partially be the result of a process of ‘path dependency’. According to the government, introducing the requirement for knowledge of Germany’s juridical and social order would improve immigrant integration.70 To fulfil the requirement, applicants for naturalisation would need to pass a test. For the first time, reference was made to a ‘naturalisation test’ (Einbürgerungstest).71 As a means of preparation, naturalisation courses would be offered, participation in which would not be obligatory. In the explanatory memorandum of the bill, the government explicitly distanced itself from the interview manual applied in Baden-Württemberg, by stating that scrutinising the inner dispositions of naturalisation candidates would be problematic from a constitutional law point of view.72 The naturalisation test would consequently not contain any questions resembling those asked to immigrants applying for naturalisation in BadenWürttemberg. About a year later, the government would however justify testing within the framework of the declaration of loyalty by stating that Länder could themselves decide how they wanted to obtain certainty about the applicant’s loyalty to the Constitution.73 Discussion of the Proposed Changes in the Parliament During the parliamentary debates on the Richtlinienumsetzungsgesetz in May and June 2007, very little attention was paid to the proposed strengthening of the conditions for naturalisation.74 This lack of attention can be partially explained by 68 Article 9 paragraph 2 number 7 of the Residence Act (Aufenthaltsgesetz) read in conjunction with Article 3(2) of the Integrationskursverordnung and Article 9 paragraph 2 number 8 Residence Act. 69 BT-Drs. 16/5065, p. 228. 70  “Bundesinnenminister Dr. Wolfgang Schäuble: ‘Reform des Zuwanderungsrecht fördert Integration in unserem Land’”, www.bmi.bund.de, site accessed 29 March 2007. 71 Article 10 paragraph 5. The required knowledge can also be demonstrated in another way, such as via equivalent German schooling; BT-DS 16/5065, p. 229. 72 BT-Drs. 16/5107, p. 14. 73 Plepro 16/168, p. 17799 (D), 18 June 2008. 74 The Bundesrat however did criticise the fact that Article 5 of the government proposal did not serve the purpose of implementing a directive and therefore constituted a strange element

‘Deutschland sucht den Superstaatsbürger’  85

the fact that members of parliament were preoccupied with the rather far-reaching alterations that had been proposed in the field of immigration law which ‘over-shadowed’ the proposed changes to nationality law.75 The government’s decision to propose far-reaching amendments in nationality law in a bill containing controversial changes to immigration law was criticised by academics, who accused the government of deliberately choosing this tactic in order to more easily get the changes to nationality law accepted (Berlit 2007: 457, Hofmann & Hofmann 2008: 2045).76 It could however also be that parliamentarians chose to direct their energy and criticism towards those changes regarding which no agreement had been reached at the IMK between the parties forming the great coalition, as the chances of reversing these changes would be rather limited.77 During the second and third readings of the bill in the Bundestag in June 2007, left-of-centre political parties criticised the alterations proposed by Article 5 of the bill.78 In a proposal for decision (Entschliessungsantrag), Die Linke proposed a withdrawal of the bill, amongst other reasons because of the “numerous restrictions” in the nationality law, which were not required by the EU Directives.79 In similar vein, the FDP introduced a proposal for amendment of the bill, as they feared that the restrictions would lead to a further decrease in the number of naturalisations, which would be “counterproductive from an integration perspective”.80 Their proposal aimed at deleting the requirement of the Einbürgerungstest, which they feared would especially put off elderly immigrants from making a (Fremdkörper) in the bill, which had not played a role during the negotiations which took place before the start of the legislative procedure. The Bundesrat therefore proposed to replace Article 5 of the government bill by the Bundesrat bill. BT-Drs. 16/5527, p. 12 and BT-Drs. 224/07. 75 The bill for instance proposed the introduction of a pre-entry language requirement for family reunification as well as a reinforcement of the available sanctions (in the form of fines or denials of applications for permanent residence permits) for those who did not fulfil their integration obligations. 76 Geyer referred to this decision as being a “tactic […] to take the sting out of the expected differences of opinion regarding a planned law” (Hofmann & Hofmann 2008: 2045). Berlit has criticised the Richtlinienumsetzungsgesetz for being merely a mantle veiling heterogeneous changes to nationality law: under the lee of the regulations which require implementation and numerous controversial themes, changes in nationality law have been accepted which would have less chance of being adopted when discussed separately (Berlit 2007: 457). 77 Similar considerations seemed to have played a role for immigrant organisations, which in an open letter criticised the language requirements for family reunification and the sanctions for not complying with integration obligations, but did not focus on the increased language requirement and the introduction of a naturalisation test. “Zuwanderungsrecht: Offener Brief an Merkel”, www.migrationsrecht.de, 28 March 2007, site accessed 16 February 2009. 78 Plepro 16/103. 79 BT-Drs. 16/5634, p. 3–4. 80 BT-Drs. 16(4)228.

86  Chapter Four

naturalisation application, and the reference to level B1. Instead, they proposed to offer free naturalisation courses, and a continuation of the more lenient application of the language requirement for spouses which the proposed Act aimed to terminate. As a majority in parliament supported the bill, it was adopted. The Act implementing EU Directives on Asylum and Migration (hereinafter: the 2007 Act) came into force on 19 August 2007, with the paragraph introducing the naturalisation test as a condition for naturalisation entering into force on 1 September 2008.81 The wish for more uniformity in the application of the naturalisation requirements eventually resulted in a strengthening of these requirements, obliging the Länder that applied a lenient interpretation of the naturalisation requirements to adapt to the Länder which applied a restrictive interpretation of the requirements. This makes one wonder whether the non-uniform application was used as an excuse to further restrict access to German nationality. The government has however always denied that the new requirements were meant as hurdles to naturalisation: “The wish that no one should fail the test is intentional.”82 The 2007 Act not only made naturalisation harder by reinforcing the language requirement and by introducing a knowledge of society requirement, it also reinforced the income and public order requirements. With regard to the income requirement, the 2007 Act put a stop to the privileged treatment of naturalisation applicants aged 16 to 23, who until 27 August 2007 had been exempt from the requirement to provide their own means of subsistence. With regard to the public order requirement, the threshold regarding fines and sentences was halved: as of 27 August 2007, fines of up to 90 daily rates (Tagessätze) or three months imprisonment will stand in the way of a successful naturalisation implication.83 The provision that it could be decided in individual cases whether a higher sentence could be disregarded was abolished. Lastly, sentences can be cumulated, something which had not been allowed under the old Act. As a consequence, several individually ‘harmless’ sentences can now lead to a refusal of the naturalisation application (Hailbronner 2011: 92). Following the coming into force of the changes made to the nationality law by the 2007 Act, the number of naturalisations continued to decrease. For the first time since 1998, the numbers were below 100,000 in each of the years 2007 to 2009.84 81 Gesetz zur Umsetzung aufenthalts- und asylrechtlicher Richtlinien der Europäischen Union vom 19. August 2007 (BGBl. 2007, Teil I Nr. 42). 82 BMI spokesperson in “Einbürgerungstest tritt unverändert in Kraft; Wer deutscher werden will, muss mindestens 17 richtige Antworten geben”, Die Welt, 30 August 2008. 83 The threshold used to be 180 daily rates or six months imprisonment. 84 Despite the restrictive changes introduced by the 2007 Act and starting in 2009, the number of naturalisations has however been rising. In 2010, the number of naturalisations again passed

‘Deutschland sucht den Superstaatsbürger’  87

4.6. Debates about the Naturalisation Test Whereas the naturalisation test could not count on much attention during the parliamentary debates on the 2007 Act, the test was criticised by Die Linke, the Greens, and SPD member Edathy, once the Act had been adopted. Die Linke and the Greens feared that the test would lead to a lower number of naturalisations, since it would constitute a hurdle, especially for those who had not been educated in Germany.85 An interpellation by the Greens which lamented the fact that parliament had not been consulted about the content of the test and which recommended that parliament not vote for the Decree for the execution of the naturalisation test without its preceding approval, was rejected.86 Once the questionnaire for the test was published, the Greens criticised its content for reasons such as the high level of the questions, the use of language exceeding level B1, the irrelevance of many questions for the integration of immigrants, and the fact that the questionnaire “transmitted the conservative idea of a Christian and German guiding culture”.87 SPD member Edathy listed 72 mistakes and blamed the test for requiring candidates to learn facts by heart, instead of serving the purpose of asking the candidates about German history, culture and democratic structure.88 In reply to the criticism, the government stated that it would do its best to ensure that naturalisation would not be hindered.89 It referred to the publication of the answers to the test questions on the internet, the exemption possibilities, the possibilities for preparation, and the possibility to repeat the test as often as necessary.90 To face the criticism on the content of the test, the government took refuge behind the fact that the questionnaire had been developed by the Institute for Quality Development (IQB), “an internationally renowned institute at the 100,000. See paragraph 7.1.1 for more detailed information on the development in the number of naturalisations in Germany. 85 Plepro 16/168, p. 17796 (C), p. 17797 (B), Plepro 16/168, 16 June 2008. Those with German school diplomas can be exempt from passing the test, provided they obtained a diploma from a school offering ‘general education’ (allgemeinbildend). 86 “Das Parlament bei der Ausgestaltung des Einbürgerungstests beteiligen”, BT-Drs. 16/9602, 18 June 2008 and BT-Drs. 16/9945, 9 July 2008. 87 BT-Drs. 16/10183, p. 2. 88 “72 Fehler. Einbürgerungstest soll schlampig gemacht sein“, Die Welt online, 1 September 2008, http://www.welt.de/politik/article2378588/Einbuergerungstest-soll-schlampig-gemacht-sein .html, site visited 7 July 2009, Plepro 16/174, p. 18621 (D). 89 Plepro 16/168, p. 17796 (C). 90 Plepro 16/168, p. 17797. “Einbürgerungstest tritt unverändert in Kraft; Wer deutscher werden will, muss mindestens 17 richtige Antworten geben”, Die Welt, 30 August 2008. An example of a site which published the answers to the questions is www.deutsch-werden.de, site accessed 7 July 2009.

88  Chapter Four

Humboldt University of Berlin, financed by all 16 German states”, and behind the fact that parliament had itself agreed to the fact that the naturalisation test would be based on the themes of the integration courses within the framework of the Residence Act.91 Despite the criticism, the naturalisation test would eventually enter into force in an unchanged form. 4.7. Conclusions German nationality law has moved from restrictive, to more open, back to restrictive. Prior to the 2000 reform, the introduction of as-of-right naturalisation meant a liberalisation of German nationality law. With the entry into force of the 2000 Act, access to German citizenship was opened up further with the introduction of a form of ius soli citizenship attribution, albeit in a less far-reaching manner than the government had originally proposed. The conditions for naturalisation were however sharpened, and would continue to be reinforced in the years to come. The 2007 Act eventually formalised the language requirement, and introduced the Einbürgerungstest. How can this evolution in German nationality law be explained? And is the statement still correct that not descent, but participation is of pivotal importance when it comes to citizenship attribution and belonging? The analysis of the debates on the 2000 reform law indeed confirms the conclusion that republican arguments of participation were important arguments to support the liberalisation of German nationality law: one of the main reasons why the government required a liberalisation of nationality law was that naturalisation was thought to enable integration, which included participation in the political process. An improved integration and participation would however also require improved language skills, which is why a language requirement was introduced. Republican arguments were hence at once used to open up and restrain access to nationality. Even though arguments based on an ethno-cultural version of nationhood had disappeared from the debates, as “a modern surrogate for a former ethnic nationalism” and in order to promote the introduction of more far-reaching integration requirements, right-of-centre parliamentarians from the Union parties and parliamentarians from the liberal FDP put forward communitarian arguments of belonging and identity (Gerdes et al. 2007: 69). These parliamentarians referred to naturalisation as the end point of the process of integration and socialisation in Germany, which should only be available to those who had obviously adapted 91 Antwort der kleinen Anfrage zum Einbürgerungstest, 17 September 2008, to be accessed via http://www.migrationsrecht.net/component/option,com_docman/Itemid,127/task,cat_view/gid, 70/?phpMyAdmin=6db17010ee4ecd685ef0b09ff5016085, site visited 8 July 2009.

‘Deutschland sucht den Superstaatsbürger’  89

to the German way of life, or had ‘converted’ to Germany. Next to republican arguments, communitarian arguments hence started to play a role in the debates. In the discussions on German nationality law which continued after the entry into force of the 2000 Act, arguments of republican origin were again put forward to justify the introduction of more demanding naturalisation requirements: since future German citizens should be provided with the tools to participate in Germany’s social and political life, a more demanding language requirement and a knowledge of society test were required. Contrary to the debates within the framework of the 2000 reform, republican arguments of participation were this time almost exclusively put forward to restrict access to citizenship. Moreover, probably under influence of 9/11 and subsequent terrorist attacks on European soil, the communitarian dimension in the argumentation received an impetus. During the discussions in the 2006 IMK and in the Bundesrat proposal of law materialising the IMK’s decisions, the opinion was expressed that only those who accepted Germany’s social and juridical order should be able to naturalise, and that naturalisation should be the end point of a completed integration. Similar concerns had previously driven the state of Baden-Württemberg to introduce a communitarian Gesinnungstest. As communitarian arguments played a considerable role, the statement that “we are all ‘Republican’ now” no longer prevailed. At the same time, it goes too far to state that arguments based on an ethnocultural understanding of the nation have returned in the debates on German nationality law. Those applying for naturalisation are not required to have ‘culturally assimilated’ as was the case prior to the liberalisation of Germany’s nationality law in 1990.92 They do nevertheless need to fulfil most of the other requirements that applied when the 1977 naturalisation guidelines were in force. Hence, without resorting to ethno-cultural arguments, Christian Democratic parliamentarians opposing an opening up of German citizenship law managed to curtail the liberalisations which had taken place in the early 1990s. By (re)introducing tests of linguistic abilities and knowledge of Germany’s social and juridical order, immigrants are currently not excluded from full membership based on descent, but on intellectual capacity and financial means. The claim that the changes in German nationality law demonstrated a new public and political culture of openness has hence proven to be premature, at least when access to German citizenship via naturalisation is concerned.

92 It is safe to say, however, that the test applied in Baden-Württemberg within the framework of the declaration of loyalty comes close to the practice of testing whether someone had culturally assimilated, as applied under the 1977 guidelines.

Chapter Five On a Journey to Citizenship: Earning the Right to Full Membership in the UK This chapter focuses on the language and knowledge of society requirements for naturalisation in the UK. In this country, a language requirement has applied since 1915 (Ryan 2010: 4). It was formalised and strengthened in 2004. A year later, a ‘knowledge of life in the UK’ requirement came into force. The strengthening of the language requirement and the introduction of the ‘knowledge of life’ requirement, together referred to as the ‘knowledge requirements’, represent an attempt to restrict the openness of the British citizenship policy, which had traditionally been typified as being “relatively generous”, or “liberal” (Green 2007: 97, Howard 2009: 148, 160). Like the Dutch regime, the British citizenship regime has furthermore been qualified as “multicultural” by providing for both easy access to citizenship for immigrants and recognition of the right of ethnic minority groups to maintain their cultural differences (Koopmans et al. 2005: 7).1 At the same time, one of the principal aims of the 1981 British Nationality Act (BNA) was to curtail immigration from non-white, ‘new’ Commonwealth countries by awarding residency rights only to ‘patrials’, i.e. those with direct family connections to the UK (Dummett 2006: 568). Whilst patrials were awarded British citizenship, nonpatrials were left with lesser types of British nationality in the form of British Dependent Territories Citizenship and British Overseas Citizenship.2 The Act was consequently referred to as racially discriminatory, and labelled “a disguised racial Immigration Act” in the press (Dummett 2006: 568). The question now is whether, and to what extent, the introduction of the knowledge requirements interferes with the image of the British citizenship regime as being ‘open’ and ‘multicultural’, representing a further attempt to curtail immigration via nationality law. To answer this question, the reasons for reinforcing the language requirement and for introducing the ‘knowledge of life’ requirement will be discussed by analysing the political debates which led to

1 Contrary to the Netherlands, British multiculturalism has however generally been typified as a ‘laissez-faire’ type of policy, rather than a policy characterised by active attempts to allow ethnic minority groups to experience their cultures (Joppke 2004: 249). 2 The Act also provided for the smaller categories of ‘British subject’ and ‘British Protected Person’.

92  Chapter Five

these changes, starting with an explanation of the practice of language and integration testing prior to the coming into force of the stricter language requirement in July 2004. 5.1. Integration Requirements in British Nationality Law 5.1.1. Language Testing Prior to the Adoption of the Nationality, Immigration and Asylum Act The 1981 BNA, which laid the basis for the current law (Dummett 2006: 569), repeated earlier naturalisation rules by requiring applicants to have five years of residence, the last year being free of immigration control, good character, and future intention, meaning that the applicant would intend to make his or her home in the UK (Fransman 1998: 317). The Act also contained a language requirement. All persons who wished to naturalise as British citizens were required “to have sufficient knowledge of English, Welsh or Scottish Gaelic language” (Ryan 2009: 279).3 Only oral language skills were required. According to an explanation given by the Home Office Minister in parliament, the language requirements were operated in “a fairly easy-going way”, as it would be sufficient for naturalised citizens “to play a full role in this country as citizens, especially perhaps in voting and in other manifestations of democracy” and “to be accepted as a sufficiently integrated member of our society”.4 The language requirement was highly discretionary in nature. It generally sufficed if someone could complete the application form. Enquiries regarding the language requirement were only made if there was some indication that it might not be met, for instance in cases where applicants did not complete the form themselves, or made their mark instead of signing the application (Fransman 1998: 329). If the choice to make an enquiry was made, this usually took the form of a personal interview or a telephone conversation in which it was tested whether applicants could speak and understand English to the level where they could make simple conversation about themselves and their family and way of life, communicate sufficiently to deal with everyday situations such as travelling, shopping, visiting the doctor’s surgery or a child’s school, and conducting dealings with officials.5 3 Schedule 1 paragraph 1(1)(c) of the British Nationality Act 1981. 4 House of Commons debates, Standing Committee F, 19 March 1981. 5 The British Nationality Act 1981 – caseworking instructions, Chapter 18, Annex E, section 3, 3.1.1. Accessible via http://www.ukba.homeoffice.gov.uk/sitecontent/documents/policyandlaw/

On a Journey to Citizenship  93

The language requirement was not applied to naturalisation applicants who were married to British citizens.6 Furthermore, the requirement could be waived if the applicant was aged 65 or over, or suffered from a mental or physical disability. It is unclear whether or how often the Home Secretary made use of his discretion to waive the language requirement in practice. In any case, the number of applications refused on the basis of the language requirement was low: in 1982, 2.7% of all applications were denied on the basis of the language requirement and this percentage decreased to 0.15% in 1996.7 5.1.2. Towards More Demanding Ways of Testing Integration Debates to reform the rules for acquisition of British citizenship via naturalisation started in 2001, after a series of riots took place in the northern England towns of Bradford, Oldham and Burnley in the summer of that year.8 The Home Office report ‘Community Cohesion’, drafted by a Review Team chaired by Ted Cantle, ascribed these riots to “a lack of community cohesion arising from racial divisions and social exclusion” (Home Office 2001: 10). The Cantle review team concluded that communities in the UK towns and cities “operate on the basis of a series of parallel lives” and that there had been “little attempt to develop clear values which focus on what it means to be a citizen of a multi-racial Britain” (Home Office 2001: 9). The policy proposals in the report included the establishment of “a greater sense of citizenship” informed by “common elements of ‘nation-hood’ [including] the use of the English language” (Home Office 2001: 19, Meer & Modood 2007: 6) with a view to promoting community cohesion (Home Office 2001: 10). In the 2002 White Paper Secure Borders, Safe Haven; integration with diversity in modern Britain, PM Blair’s Labour Government embraced the conclusions reached in the Cantle report by ascribing the riots to a lack of social cohesion (Home Office 2002). Immigration and ethnic diversity were held primarily responsible. According to the government, migration had led to changes in national identity and culture and posed a challenge to “our concepts of national nationalityinstructions/nichapter18/ch18annexe?view=Binary. Site visited on 5 August 2009. The case-working instructions are used as guidance by the UK Border Agency staff deciding applications for citizenship. 6 Schedule 1 paragraph 3 of the BNA 1981 (old). 7 Home Office Minister David Waddington, written answer, House of Commons Debates, 6 May 1983 col 163w; Immigration Minister Mike O’Brien, written answer, House of Commons Debates, 2 June 1997, col 3w, quoted in Ryan 2010. 8 House of Commons Library, The Nationality, Immigration and Asylum Bill: Nationality and Citizenship, research Paper 02/25, p. 17.

94  Chapter Five

identity and citizenship” (Home Office 2002: 9). Even though the UK had in the past “responded successfully to difference”, the riots had proved that fractured and divided communities had arisen (Home Office 2002: 10). As an antidote to the lacking “sense of common values or shared civic identity to unite around”, allegedly arising because of the laissez faire multiculturalism of the past, “a stronger understanding of what citizenship really means” needed to be developed.9 To promote community cohesion and create a “shared civic identity”, strengthening the requirements for becoming a British citizen via naturalisation was chosen as an instrument. Instead of remaining a “bureaucratic exercise”, the procedure of acquiring British nationality should be changed in order to “engage new members of the community with the fundamentals of our democracy and society” (Home Office 2002: 11). Concretely, the government proposed the introduction of a citizenship ceremony, the modernisation of the oath of allegiance and a toughening of the integration requirements for naturalisation: the language requirement would in future be taken “more seriously” and “light touch education for citizenship” would be provided (Home Office 2002: 11, 32, 34). Specific evidence of language skills would be required in all cases (Home Office 2002: 32, Ryan 2008: 304). The language requirement would furthermore be extended to spouses of British citizens, as reliance on a spouse’s knowledge of the language was considered insufficient to ensure that everyone in British society would be able to take a full and active part in society (Home Office 2002: 34). Language teaching would be offered (Home Office 2002: 12). In the government’s opinion, learning the language and citizenship education would strengthen the ability of new citizens to participate in society and to engage actively in our democracy. This will help people understand both their rights and their obligations as citizens of the UK, and strengthen the bonds of mutual understanding between people of diverse cultural backgrounds. It will also help to promote individuals’ economic and social integration.10

As was the case in the Netherlands and Germany, the government stressed that the new requirements were not meant to reduce the number of naturalisation applications: We have no wish to see applicants fail the requirements. We want to see them meet the requirements and become British citizens.11

For this reason, the government proposed two separate trajectories. For lower skilled immigrants, the government suggested embedding the proposal of 9 Secure Borders, Safe Haven, p. 11. 10 Secure Borders, Safe Haven, p. 12. 11 Secure Borders, Safe Haven, p. 12.

On a Journey to Citizenship  95

citizenship education in the existing framework of ESOL (English for Speakers of Other Languages) language courses, for which learning programmes in a specific citizenship context would be developed. For those who already had a good command of English, suitable study materials containing information about British society and the rights and responsibilities of becoming a British citizen would be created. All applicants for citizenship would be required to pass “a simple examination […] similar to that which exists in many other countries” (Home Office 2002: 11). The Nationality, Immigration and Asylum Bill, introduced in the House of Commons on 12 April 2002, materialised the government’s proposals to amend the requirements for naturalisation.12 The bill added “sufficient knowledge of life in the UK” as a condition for naturalisation and proposed to make regulations to determine whether a person would have sufficient knowledge of English, Welsh or Scottish Gaelic.13 With its proposals, the government hoped to prevent riots from re-occurring. However, as the riots were caused by people who had been born and bred in the UK, the proposals were criticised for representing an “unwarranted simplification of the policy of belonging and citizenship” (Kostakopoulou 2003: 108) and for failing to deal with the issues that were actually at stake: social deprivation and the failure to protect Asian communities from racial violence (Kundnani 2002: 67, Sales 2010: 130).14 A few decades earlier, next to strengthened immigration legislation, race riots had triggered the introduction of anti-discrimination legislation in the form of the Race Relations Act and the Conservative government to prescribe more ‘positive action’ as a way out of failing minority integration (Joppke 2004: 250, Layton Henry 1984). As the “multiculturalist settlement”, from the government’s point of view, was declared insufficient to tackle the current problems, similar proposals to combat racism, discrimination and xenophobia were this time lacking (Kundnani 2002: 67). The Nationality, Immigration and Asylum Bill in the Parliament On 24 April 2002, the Nationality, Immigration and Asylum Bill was debated in the House of Commons. Two months later, on 8 July 2002, it was sent to the House of Lords. As the bill’s main purpose was to deal with immigration and asylum issues, the debates rarely touched upon the nationality provisions (Dummett 2006: 574). When they were discussed, the MPs and Lords mainly put forward concerns regarding the content and scope of the integration requirements, not so 12 Bill 119 of 2001/02. 13 Clauses 1(1) and 1(2) of the Nationality, Immigration and Asylum Bill. Clauses 2(1) and 2(2) of the bill extended the knowledge requirements to spouses of British citizens. 14 See also JCWI research paper 02/25, p. 22.

96  Chapter Five

much against their introduction itself, nor about the justification behind their introduction. The question of whether the goal of increased community cohesion could be achieved by the proposed measures, while simultaneously problematising the entire immigrant community was hence not subject to debate. The fact that achieving the goal of an increased community cohesion via strengthening the requirements for naturalisation seemed to be rather ambitious was not subject to parliamentary scrutiny either. The formalisation of the language requirement and the introduction of ‘knowledge of life in the UK’ as a condition for citizenship were supported by the Conservatives. They criticised the “very light touch” with which the language requirement for naturalisation had been administered so far, which had led to very few applications being denied.15 The government, via Under-Secretary Eagle, assured the Conservatives that the new clause would guarantee that the requirement would be applied more consistently and that evidence of a particular standard of achievement would in future be required.16 As regards the ‘knowledge of life’ requirement, the Conservatives thought that the bill did not go far enough. MPs therefore proposed to also test ‘basic’ knowledge of the history and government of the UK, whereas Conservative Lords proposed an extension of the requirement by adding the criteria ‘political, civic, economic and cultural’ before ‘life in the UK’.17 Like the Conservatives, the Liberal Democrats in the House of Lords supported the ‘knowledge of life in the UK’ requirement. According to Lord Avebury, it was a good idea that those applying for citizenship would have “the fullest possible knowledge”: it would be excellent if all those people had knowledge of our electoral system, of the way in which democracy works in this country, of the rights of individuals, of how individuals access those rights and so on as they will need that knowledge if they are to play their full part as citizens.18

At the same time, however, he expressed his doubts about “one size fits all” tests, because of major differences between people’s needs to know about life in the country according to their professions and occupations.19 Liberal Democrats furthermore wondered whether life in the UK could be talked about, taught and

15 As an example, the number of applications denied in 1996, 27 applications of a total of 17,600, was referred to. House of Commons Standing Committee E, 30 April 2002, Column 031. 16 House of Commons Standing Committee E, 30 April 2002, Column 017. 17 MP Malins (Conservatives), House of Commons Standing Committee E, 30 April 2002, Column 015, Viscount Bridgeman (Conservatives), House of Lords, 8 July 2002, Column 442. 18 Lord Avebury (Liberal Democrats), House of Lords, 8 July 2002. 19 Ibid.

On a Journey to Citizenship  97

assessed at all, the UK being “multi-faith, multi-racial and multi-lingual”, and because of the regional differences:20 Life in the UK has some commonality but varies greatly. Life in Northern Ireland is fundamentally different because of the Irish dimension. Life in the islands of Scotland is fundamentally different, too.21

Concerns were also expressed about the possible exclusionary effects the new requirements might produce. In the House of Commons, Hughes (LD) for instance stated that conventional tests would be inappropriate for “some people”.22 He feared that the requirement would operate as a barrier to citizenship and pleaded for a similar process for ensuring that British born adolescents were equipped with the same sufficient knowledge of life in the UK. Gerrard (Labour) put forward an amendment to provide for specified courses and to make payment for travel and childcare costs for those attending the courses. Other MPs and Lords feared a shortage of language classes and stressed the importance of availability of sufficient English language and citizenship courses. The government tried to appease the concerns by stating that fulfilling the ‘knowledge of life’ requirement was not meant to be exclusionary or hugely expensive.23 On behalf of the government, Lord Filkin stated that the goal of the requirement of ‘knowledge of life in the UK’ would be to help applicants to better integrate, not to make sure that British nationality would only be available to intellectuals, academics, or those with a deep understanding of the constitution.24 He did not expect a shortage of classes, as the government thought that Colleges of Further Education (FE colleges) would develop relevant courses to handle the demand. Before, the government, by way of Under-Secretary Eagle, had promised the House of Commons that citizenship courses would be provided at no cost and that the government would be willing to check in more detail how to accommodate people with children who wanted to attend classes. As regards the content of the new requirements, Lord Filkin promised there would be wide consultation on the curriculum before the bill would be implemented, and that the government would take advice from a group of external experts in education. The bill however eventually received royal assent on

20 Baronness Uddin (Labour), House of Lords, 8 July 2002, Column 446. 21 MP Hughes (Liberal Democrats), House of Commons 30 April 2002 Standing Committee E, Column 011. 22 House of Commons Standing Committee E, 30 April 2002. Hughes did not specify what kind of people he had in mind. 23 Whether the requirement in reality leads to exclusion will be examined in Chapters 7 and 10 of this book. 24 House of Lords, 8 July 2002, Column 455.

98  Chapter Five

7 November 2002, before any concrete statements about the content of the new requirement had been made. The New and the Old The group of external experts in education, the ‘Life in the UK Advisory Group’, chaired by Professor Sir Bernard Crick, published its report The New and The Old in 2003 (Home Office 2003). The appointment of Crick, who had acted as the chair of a similar advisory group on the development of citizenship education in schools a few years earlier, suggests that it was expected that education would be central to the arrangements which would emerge (Ryan 2010: 22). The group saw its work “as falling within the Government’s broader policy aims including enhancing the significance of British citizenship, encouraging community cohesion [and] valuing diversity” (Home Office 2003: 3). In the eyes of the advisory group, the knowledge requirements needed to be implemented in such a way that they would encourage immigrants “to live up to the best and generally accepted elements in the political traditions and values of the UK and the democratic practices of free citizenship”, thereby serving the general aims of public policy to increase participative citizenship and community development. New citizens in Britain’s multicultural society should hence be equipped to be active citizens (Home Office 2003: 9). As different categories of people with different attainments and different language skills seek naturalisation, the group recommended a flexible programme of study. Hence, similar to the government’s proposals in ‘Secure Borders, Safe Haven’, the group proposed that the new citizenship requirements could be satisfied in different ways depending on the applicant’s initial level of language proficiency.25 Applicants having good language skills, i.e. those mastering the English language at or above ESOL Entry Level 3, a level which more or less equals level B1 of the CEF, would be required to either complete a short course or self-study to develop a portfolio of evidence of civic learning, volunteering, or civic participation, and take a short written test.26 Applicants with language skills below ESOL Entry 3 would follow an accredited ESOL with citizenship course (Home Office 2003: 22–23). In both cases, an applicant’s progress in developing skills needed to 25 “A common standard could well prove too demanding for some with little or no English but who will be contributing to the needs of the economy through unskilled employment. For others, such a common standard might equally prove an insufficiently demanding incentive for further language learning that could enhance employment skills and active participation” (Home Office 2003: 20). 26 For a description of the different language levels of the CEF, consult http://www.coe.int/t/dg4/ education/elp/elp-reg/Source/Key_reference/Overview_CEFRscales_EN.pdf, p. 5, site visited on 13 March 2013. .

On a Journey to Citizenship  99

be assessed, as this would serve the “educative and integrative” object of the new requirements (Home Office 2003: 20). The advisory group welcomed the Home Secretary’s stress on “practical knowledge”. As a programme of study, it proposed to teach “knowledge relevant to the everyday problems of settling in and practical information about UK society and civic structures” (Home Office 2003: 10). The group listed six topics to be taught and learnt in whatever order and depth would be appropriate: British national institutions in recent historical context, Britain as a multicultural society, knowing the law, employment, sources of help and information, and everyday needs. The group advised publishing a handbook in which these topics would be elaborated, in bilingual versions in “as many languages as practical” and free of charge for all those interested: applicants for naturalisation as well as new arrivals (Home Office 2003: 17). A handbook entitled ‘Life in the United Kingdom; A Journey to Citizenship’, written by a sub-committee of the Advisory Group, was eventually published in 2004 (TSO 2004). According to one of the members of the Advisory Group, the word ‘journey’ in the handbook’s subtitle portrayed citizenship “as a continual process, and the formal acquisition of the legal status of citizenship as only the starting point” (Kiwan 2008: 66). The advisory group hence did not see the acquisition of citizenship as the completion, or ‘crowning’, of the integration process. Like the government, the advisory group did not intend the knowledge requirements as a means to diminish the number of people already settled and employed in the UK: rather than a hurdle, the new requirements needed to seen as an entitlement (Home Office 2003: 8, 20). To make education attractive to individuals, the group stressed the need for free assessments and tuition (Home Office 2003: 23–24).27 The government accepted most of the recommendations of the Crick group.28 It however indicated that it could not fund the advice regarding free assessments and tuition, thereby breaking the promise that courses would be organised at no cost which it had made in the House of Commons prior to the adoption of the Act. The government made no statements on how applicants with good language skills, i.e. exceeding ESOL entry three level, would fulfil the requirements. Eventually, it was decided that a citizenship test would function as both language test and test of ‘Life in the UK’. The government did not accept the advisory group’s recommendation for candidates with good language skills to submit portfolios (Kiwan 2008: 71). Fulfilling the knowledge requirements would

27 The group for instance strongly recommended abolition of the rule, which applied in England, that people would only qualify for free ESOL classes after three years of residence. 28 Lords Hansard text for 2 February 2004, Columns WS13 and WS14.

100  Chapter Five

for these immigrants hence be a rather solitary exercise, not leading to integration through participation, which was an opportunity ascribed to the knowledge requirements by the Crick group. Nor would the progress that candidates with good language skills were supposed to develop be measured. Hence, even before the introduction of the knowledge requirements, it became apparent that, at least as far as immigrants with higher language skills were concerned, the requirements would not play the educative, activating role the Advisory Group had in mind. 5.2. The New Way of Integration Testing The different sections of the Nationality, Immigration and Asylum Act (NIAA) had different commencement dates. As of 28 July 2004, the language requirement was formalised. Until the ‘knowledge of life’ requirement entered into force on 1 November 2005, interim measures regarding the language requirement applied (Ryan 2009: 280).29 As of 1 November 2005, both ‘knowledge requirements’ merged.30 Those with language skills below ESOL entry 3 level can fulfil the knowledge requirements by successfully completing an ESOL with citizenship course: a language course, involving learning materials which incorporate information about citizenship.31 Those with language skills above ESOL entry level 3 can pass the so-called ‘Life in the UK’ test to fulfil the knowledge requirements. Immigrants suffering from mental or physical impediments can be exempted, as well as immigrants aged 60 and over.32

29 Those who applied for citizenship between 28 July 2004 and 31 October 2005 needed to have language skills equivalent to ESOL entry 3 level (B1 of the CEF). They needed to prove this by either showing that they had passed a language course, or else have a ‘designated person’ certify that the applicant had the language skills at this level. From 1 September 2004, a further option was certification by a designated person, on the basis of an interview, that the individual had “knowledge of the English language to the level reasonably to be expected of a person of full age and capacity whose native language is English”. Immigration staff and consular officers were ‘designated persons’ for both purposes, while notaries and solicitors could certify ‘native language’ competence in English alone. The British Nationality Act 1981 – case-working instructions, Chapter 18, Annex E, section 2, paragraphs 2.2, 4.2 and 5.2. 30 The ‘knowledge of Life in the UK’-requirement was introduced in Schedule 1 paragraph 1(1)(ca) of the British Nationality Act 1981. 31 Since there are no courses that focus solely on citizenship, citizenship materials are embedded in the ESOL courses. Starting from ESOL entry level 3, the courses can have a vocational content, such as ESOL with childcare, business or health and beauty. 32 British Nationality Act 1981, Schedule 1, paragraph 2(e) and paragraph 4, UK Border Agency, Nationality Instructions, Chapter 18, Annex E, paragraph 1.4.4.

On a Journey to Citizenship  101

• The Course Route In order to meet the requirements for naturalisation, course participants will need to show that they have progressed from one ESOL level to the next.33 In the certificate which needs to be enclosed with the naturalisation application, mention only needs to be made of progress in speaking and listening skills. But since passing the reading and writing parts of the course often is a precondition for progressing to the next level, reading and writing skills are (implicitly) also required. Prior to their introduction, the government promised that it would check how to accommodate people with children who wanted to attend courses, and that it expected that FE colleges would develop sufficient courses to handle the demand for adequate courses. No special regulations have however been provided for people with children, and increased shortages of ESOL courses and long waiting lists have been reported.34 As the government also did not live up to its promise to fund the courses, all promises it had made to those MPs who feared that the new requirements would function as barriers to naturalisation have eventually been broken. Surprisingly, this has not led to new debates. • The Test Route Those who master English at or above ESOL entry level 3 can fulfil the knowledge requirements by passing the ‘Life in the UK’ test, consisting of 24 questions, 18 of which (75%) need to be answered correctly. The questions are based on the handbook Life in the United Kingdom: A Journey to Citizenship.35 By making it the basis for the test and designating it as appropriate study material, the government assigned a role to the handbook which the Crick advisory group had not intended for it. This is the reason why many questions in the test have become subject to criticism, both in parliament and outside.36 Various test questions were judged too difficult, even for citizenship teachers.37 None of 100 Britons participating in a

33 Those with no workable English need to acquire at least a basic level of competence. The Border and Immigration Agency, http://www.ind.homeoffice.gov.uk/, site visited on 9 November 2007. 34 See ABNI second annual report April 2006 – October 2007, p. 17 and paragraph 10.1.1 of this book. 35 The handbook on which the test questions are based and which can be used to prepare for the test has in the meantime been rewritten twice. Currently, those preparing for the test have to study the latest (third) edition of the handbook, entitled Life in the United Kingdom; a Guide for New Residents (TSO, £12.99). 36 In the House of Lords, Lord Lester of Herne Hill (Liberal Democrats) questioned the relevance in relation to citizenship of the questions whether women have the same pay as men, how many young people there are in the UK, what the dates of the national holidays are and how many people say they have a religion and attend religious services. According to the government, such knowledge would enable people to understand the fundamentals of what it means to be a British citizen. Lords Hansard text for 28 November 2005. 37 “Testing passport to UK citizenship”, Alan Travis, The Guardian, 1 November 2005.

102  Chapter Five

quiz organised in a London pub succeeded in passing the test, which according to the organisation implied that “there is a significant gap between the exam and real life in Britain”.38 In defence of the test content, Tony McNulty (Labour), then Minister of State for Immigration, stated that the ‘Life in the UK’ test is not a test of someone’s ability to be British or a test of their Britishness. It is a test of their preparedness to become citizens.39

From McNulty’s statement another goal the test was supposed to fulfil, which was not explicitly mentioned prior to the adoption of the knowledge requirements, became apparent: reassuring the public, whose opinion of immigration had worsened since the 2001 riots, that British citizenship was not given away for free. The actual content of the test in this discussion is of lesser importance. From November 2005 to December 2009, the average pass rate in the test was around 70%. The fact that the test had the potential to exclude a significant part of the test candidates from citizenship did not give rise to debate in parliament.40 On the contrary: from the fact that 13,000 immigrants a month successfully passed the test, Conservative MP Clappison concluded that the test was too easy, consequently suggesting that the level of the test should be raised.41 Whereas Clappison welcomed “people […] learning about our country”, he had concerns about the volume of migration, and the strains which this was placing on public services.42 By directly linking the control of immigration to the content of the test, the Conservative MP attributed to the test the function of managing immigration.

38 “Q. How many of 100 Britons passed the citizenship exam? A. Not one”, Justin Gest, The Times, 29 September 2007. Critics in another newspaper article stated that the Life in the UK test, by focusing on knowledge of criminal courts, union rights and benefits, portrays a picture of Britain “far removed from the weather-fixated, queue-loving ale drinkers of the national stereotype”. “Citizenship test runs into flak for lack of perspective”, Simon Freeman and agencies, Times Online, 31 October 2005. 39 “Citizenship test runs into flak for a lack of perspective”, Simon Freeman and Agencies, Times Online 31 October 2005. 40 The fact that 30% of test candidates does not pass the test does not necessarily mean that they will be excluded from obtaining citizenship, since these candidates can choose to fulfil the knowledge requirements by following the course route. Interviews with immigrants analysed in Chapter 10 however show that immigrants have good reasons to prefer the test over the course, among other reasons because following the course route can be quite impossible for immigrants in employment or taking care of children. 41 “The test that’s letting in one migrant every three minutes”, Daily Mail, 22 January 2008. To raise the level of the test, Clappison proposed to also test knowledge of British history. 42 Ibid.

On a Journey to Citizenship  103

5.3. Further Proposals to Restrict Access to British Citizenship Shortly before the introduction of the new knowledge requirements, the UK was hit by terrorist attacks. On 7 July 2005, bombs killed 52 people in London’s public transport system. Four British-born Muslim men turned out to be responsible for the attacks. The concerns with social cohesion, which had previously been given a boost with the 2001 riots, were now “immeasurably strengthened” (Sales 2010: 128). Many people interpreted the attacks as evidence that some sections of the British population had loyalties that conflicted with those of British citizens (CRE 2005: 10). Questions were raised as to what it meant to be a British citizen (Sales 2010: 128). On 8 December 2006, Blair gave a speech on multiculturalism and integration, in which the issue of community cohesion played a prominent role. In his speech, Blair addressed the question of how to react to the alienation and separation of “certain groups” from the values that define “what we hold in common”.43 Blair stood up for multicultural Britain by stressing everyone’s right to differ, but against this right he placed the duty “for all groups living in Britain” to integrate, and to express difference in a way which was consistent with the values “that bind us together”. He referred to the cultural practices of “one group” which contradicted the British values of equality of respect and treatment for all citizens, mentioning the practice of forced marriages and women being debarred from entering certain mosques. Hence, while officially wanting to promote community cohesion, the government problematised Muslim citizens, thereby undermining the trust which is required for cohesive communities and shared citizenship (Sales 2010: 129). Whilst Blair nevertheless made clear that integration was a two-way process, and that action needed to be undertaken to tackle disadvantages and promote equality, as the only concrete measure he announced that, as of 2 April 2007, the knowledge requirements for citizenship would also be applied as a condition for settlement (Independent Leave to Remain, ILR). Now the settlement requirements were identical to the citizenship requirements, the government expected that “over time the need for citizenship testing will reduce.”44 The introduction of the knowledge requirements for everyone who wanted to stay in the UK permanently had already been announced in the government’s five-year strategy on asylum and immigration, in which a “managed migration

43 Speech on Multiculturalism and Integration, 8 December 2006. www.number10.gov.uk/Page 10563. Site accessed 6 October 2009. 44 Knowledge of Life in the UK: new requirements for Settlement Applications from 2 April 2007, Home Office Border and Immigration Agency.

104  Chapter Five

policy” was announced to more effectively control immigration (Home Office 2005). Exercising control over who would be allowed to stay in the country was now officially attributed to the knowledge requirements as a policy goal. This goal had however not been officially attributed to these requirements prior to their introduction. 5.3.1. The Path to Citizenship Apart from the extension of the knowledge requirements to the ILR status, Blair’s focus on shared values and obligations did not envisage any changes in Britain’s naturalisation policy. This changed under PM Brown (Labour), his successor, in whose opinion the naturalisation requirements were too meagre for British citizenship. Referring to values such as liberty, tolerance and fairness, being British according to Brown was something Britons could be “proud of”.45 Just like had been the case after the 2001 riots, citizenship was seen as the solution to create more cohesive communities after the suicide bombings. In order for citizenship to become “the common glue”,46 British nationality should in future be earned, instead of “handed to newcomers as they get off the plane or coach”.47 Earning citizenship, it was thought, would enhance a future citizen’s sense of Britishness.48 In the consultation document The Path to Citizenship, changes to the foreigners’ and integration policy and the requirements for naturalisation were announced (Home Office 2008a). The Path to Citizenship referred to integration policies that had been implemented in other European countries, such as Germany, France, Italy, Spain and the Netherlands, as well as in the USA and Australia (Home Office 2008a: 17–18). It was time for the UK to also introduce “arrangements for integrating new citizens with a much greater accent on the need for newcomers to more visibly demonstrate a commitment to their new home” (Home Office 2008a: 17). The UK apparently wanted to copy the practices which already existed in other EU countries, of which it had become well aware during the negotiation process on Directive 2003/109/EC, which allows for the 45 “Remarks at a seminar on Britishness at the Commonwealth Club, London”, 27 February 2007, http://www.hm-treasury.gov.uk/speech_chex_270207.htm, site accessed 19 March 2009. 46 “The future of Britishness”, new year conference, 14 January 2006. http://fabians.og.uk/events/ new-year-conference-06/brown -britishness/speech. 47 Liam Byrne, then Immigration Minister, and then Communities Secretary and Minister for Women Ruth Kelly, in A Common Place, Fabian Society Freethinking Paper, downloadable via http://www.fabians.org.uk/publications/freethinking-papers/. The Fabian Society is a left-of -centre think tank; www.fabians.org.uk. 48 “Remarks at a seminar on Britishness at the Commonwealth Club, London”, 27 February 2007, http://www.hm-treasury.gov.uk/speech_chex_270207.htm, site accessed 19 March 2009.

On a Journey to Citizenship  105

introduction of integration conditions as a requirement for the European permanent residence status. The policy proposals in The Path to Citizenship entailed that, in future, “only those with the skills and the talents, the assets and ideas that we need to remain one of the world’s leading nations” would be offered a route to citizenship (Home Office 2008a: 6). Three routes to naturalisation (work, family or protection) were proposed, as well as three stages: temporary residence, probationary citizenship, and citizenship or permanent residence.49 The temporary residence stage would last five years for ordinary immigrants and two for family members of British immigrants. The current English language and ‘knowledge of life in the UK’ requirements for settlement would be applied to those seeking probationary citizenship. Subsequently, immigrants would need to spend three years in the probationary citizenship stage before they could apply for citizenship, and five years before they would be able to apply for permanent residence. By making the residence period shorter for citizenship than for permanent residence, the government intended to encourage immigrants to opt for citizenship, which in its eyes was “much more desirable”: since citizenship entailed having equal rights and responsibilities, taking up citizenship would enable an immigrant to become fully integrated (Home Office 2008a: 23). The government claimed to take account of those immigrants, who, due to the legislation in their home countries, would feel “unable to apply for British citizenship” by referring them to the route to permanent residence (Home Office 2008a: 23). Nevertheless, by creating a shorter route to citizenship, those immigrants who wanted to file for permanent residence would be punished. The government thereby sent out the signal that these immigrants were less loyal than those opting for citizenship.50 Committing minor crimes would slow down the progress through the system, and more serious crimes would trigger deportation. Immigrants would at the same time be offered the possibility to speed up their journey through the system by “active citizenship”, such as volunteering within a recognised organisation, fund-raising activities for charities and schools or running or helping to run a playgroup (Home Office 2008a: 29–30).51 By introducing the concept of active citizenship, the government wished 49 The status of ILR would disappear. 50 As we have seen in Chapter 2, many Germans appeared to have taken the unwillingness of many long-term resident Turks and other ethnic minorities to naturalise as evidence that they did not share the country’s identity, and hence were less loyal than citizens (Schuck 1994: 330–331). By punishing those immigrants who, for whatever reason, choose the permanent residence status instead of citizenship, the British government gives evidence of a similar way of reasoning. 51 ‘Active citizens’ would be able to apply for naturalisation after six years instead of eight, extending the total residency period for naturalisation in any case by one year compared to the current regulations. As said, the residency period would become longer in case of criminal convictions.

106  Chapter Five to do more to encourage all immigrants who wish to stay in the UK permanently to integrate fully in society, including bringing probationary citizens in greater contact with the wider community, showing current British citizens that those seeking to join them as British citizens are earning citizenship by being active rather than passive participants in UK life (Home Office 2008a: 29)

As was previously the case with the knowledge requirements, the proposed measures aimed at reassuring the public that “the UK is [not] a soft touch”, and that citizenship was not given away for free (Home Office 2008a: 5). By ‘volunteering’, immigrants would be able to show that they were contributing to society, instead of merely being a burden. Respondents in a public consultation, consisting mainly of NGOs supporting immigrant rights, criticised the proposed stage of probationary citizenship for its association with the criminal justice system and the creation of a ‘second class’ citizen status (Home Office 2008b).52 A majority (70%) of consultation respondents disagreed with the proposals on active citizenship, by stating that voluntary activities should not be forced on people, that most Britons did not engage in voluntary activities either, and that immigrants with work and family commitments would have difficulty finding time to undertake voluntary activities (Home Office 2008b: 3, Home Office 2008c: 17). Despite the opposition, the government proposals were enacted in the Borders, Citizenship and Immigration Bill, which was introduced in Parliament in January 2009.53 5.3.2. The Borders, Citizenship and Immigration Bill During the debates in the House of Lords, where the bill began its parliamentary passage, Lord West of Spithead (Labour), on behalf of the government, defended the proposals in the bill as a way to achieve the government’s wish to “strike the correct balance” between the economic benefits that immigration brings and the impact that it has “on those already here”: We want to encourage those with the right values to become citizens. With rights come responsibilities, and those responsibilities must first be demonstrated, ensuring that the benefits of British citizenship are earned.54

Lord West of Spithead thereby stressed that citizenship, in his opinion, was a reward for a completed integration, rather than a ‘journey towards’ inte­ gration,  as it had been for the members of the expert group that advised on the 52 For an overview of the responding organisations, see appendix C of the government response to the consultation (Home Office 2008c). 53 HL Bill 15. 54 Lords Hansard, 11 February 2009, Column 1130.

On a Journey to Citizenship  107

implementation of the knowledge requirements. Lord West of Spithead’s party member Goldsmith (Labour) claimed that more attention should be paid to “citizenship as sharing a sense of belonging and aspirations”.55 He supported the concept of earned citizenship by stating that the new proposals would make […] it clearer that becoming a citizen carries with it responsibilities, not just the ability to stand in a shorter queue at Heathrow.56

The government proposals, in particular the activity condition and the introduction of the status of probationary citizenship, were however also criticised. Parliamentarians disapproved of the activity condition for its obligatory character, which constituted “blackmail”,57 or would lead to people volunteering for the wrong reasons.58 They expressed concerns regarding those immigrants who would experience difficulties in volunteering.59 These immigrants’ journey to citizenship would be prolonged, and this would hamper rather than foster integration, and make the naturalisation process costlier and more confusing.60 Future Immigration Minister Damian Green (Conservatives) referred to the idea of compulsory volunteering as the “ultimate absurdity”.61 In defending the concept of earned citizenship, Minister of State for Borders and Immigration Phil Woolas stated that the proposals should not be seen as punitive on the prospective citizen, but as a route to help those people to integrate into our society. […] This means reassuring our indigenous population that that immigrant positively wants to be a member of our community so that we can have better cohesion and better relations in all our communities. […] The idea of earned citizenship is […] beneficial not just to the immigrant but to the wider community.62

With this statement, it again becomes clear that, apart from the official goals mentioned, the new proposals were also meant to appease the public, by reassuring them that the government was ‘managing migration’, and that citizenship was not given away for free. 55 Lords Hansard, 11 February 2009, Column 1146. 56 Lords Hansard, 11 February 2009, Column 1146. 57 Baroness Hanham, Conservatives, Lords Hansard, 2 March 2009, Column 550. 58 Baroness Butler-Sloss (Crossbench), Lords Hansard, 2 March 2009, Column 568. 59 Single parents, people suffering from an illness (Baroness Quin, Labour), traumatised refugees (Baroness Howe of Idlecote, Crossbench), and people taking care of a family or having a job (Baroness Hanham, Conservatives) were mentioned as examples. Lords Hansard, 11 February 2009, Columns 1165 and 1190. 60  Lord Avebury, Lords Hansard, 2 March 2009, Column 539, Baroness Howe of Idlecote (Crossbench), Lords Hansard, 11 February 2009, Column 1190. 61 Commons Hansard, 2 June 2009, Column 232. 62 Commons Hansard, 2 June 2009, Columns 236 and 238.

108  Chapter Five

Whereas the Conservatives condemned the active citizenship condition, at the same time they criticised the supposedly automatic link between being granted access to the UK and obtaining British citizenship or permanent residence. This alleged link was also debated in a parliamentary group on balanced migration that argued for a reduction in immigration (Symonds 2010). Conservative MPs repeatedly stressed that citizenship ought to be a privilege, rather than a right, and worried about the number of citizenship grants being at an “all time high”, referring to the rise in naturalisations as “significant”.63 Home Secretary Jacqui Smith subsequently claimed that the government proposals on citizenship intended to break the supposedly automatic link between coming to the UK for work or study and settling there permanently.64 Rather than an arbitrary cap on entry, as suggested by the Conservatives, Smith proposed that what we know about the architecture that has been created [should be used] to control the number of those granted citizenship at the next stage.65

Reducing the number of citizenship grants, which was not officially mentioned as an aim of the new policy in The Path to Citizenship, was now hence mentioned as an aim the proposed system was meant to achieve. The bill was given royal assent on 21 July 2009. Minister for Borders and Immigration Woolas announced that the provisions on earned citizenship would be introduced in July 2011. As soon as the Act entered into force, the Labour government put forward a proposal to further elaborate on the ‘earned citizenship’ framework and to make it more restrictive. In the consultation document ‘Earning the Right to Stay: a new Points Test for Citizenship’ (Home Office 2009), a points-based system for citizenship was proposed, with an explicit view to “improve the management of the numbers of people allowed to stay permanently in the UK” (Home Office 2009: 4). According to Immigration Minister Phil Woolas, earned citizenship “would give the Government more control over the number of people permitted to settle in the UK permanently, with the bar raised or lowered according to need”.66 One of the proposed measures included a double test for immigrants: once at

63 See for instance, Conservative MP Grayling, Commons Hansard, 2 June 2009, Column 182, Conservative MP Charles Walker, Commons Hansard, 2 June 2009 Column 222, and Damian Green (Conservatives), Commons Hansard 14 July 2009, Column 223. For the comments on the number of citizenship acquisitions, see statements made by MP Green, Commons Hansard, 11 June 2009, Column 78. 64 Commons Hansard, 2 June 2009, Column 175. 65 Commons Hansard, 2 June 2009, Column 176. 66 “New Migrants to Britain face ‘points test for Citizenship’”, Alan Travis, The Guardian, 4 August 2009, quoted in Kostakopoulou 2010c: 836.

On a Journey to Citizenship  109

probationary citizenship stage, and once at citizenship stage. Whereas the test at probationary citizenship stage would focus on practical issues from the handbook, the test at citizenship stage would consider more challenging topics including history, how the UK is governed, relations with Europe and the rest of the world, voting rights and community engagement (Home Office 2009: 22).67 In May 2010, the Labour government was replaced by a Conservative Government, led by PM Cameron, receiving support from the Liberal Democrats. The new government decided not to implement the provisions on earned citizenship, such as the probationary citizenship status and the activity condition, as they were thought to be “too complicated, bureaucratic, and, in the end, ineffective”.68 Neither did it adopt the plans for a points system for citizenship. In a speech delivered in October 2011, Cameron however announced plans to revise the Life in the UK test in order to “put British history and culture at the heart of it”.69 A new version of the ‘Life in the UK’ test, placing more emphasis on British history and achievements than the ‘old’ version, eventually entered into force on 25 March 2013.70 Furthermore, in April 2013 the government issued a statement of intent saying that all those applying for citizenship will from October 2013 be required to pass an English language course, proving that they speak and understand the English language at level B1, as well as the ‘Life in the UK’ test.71 These changes imply that the ‘two route system’ for fulfilling the knowledge requirements will be abolished. The government justified the changes by referring to similar changes that will be made to the knowledge requirements for ILR.72 67 Persons following the course route would need to show further progress when they would apply for citizenship, beyond language progression demonstrated at probationary citizenship stage, which equals the language skills previously required for ILR or citizenship. 68  Government announcement on settlement reforms, 5 November 2010, http://www.ukba .homeoffice.gov.uk/sitecontent/newsarticles/2010/nov/15-settlement-reforms, site accessed 13 June 2012. 69 Alain Travis, “David Cameron launches immigration crackdown”, The Guardian, 10 October 2011,  http://www.guardian.co.uk/uk/2011/oct/10/david-cameron-immigration-crackdown,  site accessed 22 November 2011. 70 The new test reportedly signals a move away from the ‘old’ one, which, according to a Home Office official, focused on “rights and practical info that has little to do with British culture” (Alain Travis, “UK migrants to face ‘patriotic’ citizenship test”, The Guardian 1 July 2012, http:// www.guardian.co.uk/uk/2012/jul/01/uk-migrants-patriotic-citizenship-test). The new test has been criticised for being comparable to “an entry examination for an elite public school”, Migrants’ Rights Network in “UK Citizenship test to cover Britain’s Greats”, BBC News UK, 28 January 2013, www.bbc.co.uk/news/uk-21221773, site accessed 24 April 2013. 71 h ttps://www.gov.uk/government/uploads/system/uploads/attachment_data/file/182545/ statement-of-intent-koll.pdf, site accessed 21 May 2013. 72 http://www.ukba.homeoffice.gov.uk/sitecontent/documents/news/soi-fam-mig.pdf, p. 28, site accessed 24 April 2013.

110  Chapter Five

According to Home Secretary Theresa May, “it would clearly be wrong for people to become British citizens with a lower level of English than that expected from permanent residents”.73 Just like in the Netherlands and Germany, (proposed) changes to the requirements for naturalisation are hence partially the result of a process of ‘path dependency’. Contrary to what the Labour government had stated upon the introduction of the knowledge requirements as a condition for ILR, the final stage in the introduction of the language and knowledge of life in the UK requirements has therefore probably not yet been reached. 5.4. Conclusions The journey British citizenship law made from being relatively generous to more selective has been described and explained above. Does the introduction of the knowledge requirements represent a further attempt to curtail immigration via nationality law, thereby requiring a revision of the qualification of the British citizenship regime as being ‘open’ and ‘multicultural’? The answer to this question appears to be ‘yes’. Even though the expert group that advised on their implementation intended the knowledge requirements to be an entitlement rather than a hurdle, and the government’s denials that the knowledge requirements were meant to reduce the number of applications for naturalisation, their introduction has made the British citizenship regime less ‘open’. This may not have come as a surprise, given the fact that the government broke most of the promises it had made when reassuring parliament that exclusionary effects would be reduced as much as possible. As a consequence, courses were not funded, facilities were not provided for people taking care of children, and a lack of courses has been reported. Pass rates in the test have shown that a significant part of the immigrant population has been barred from obtaining citizenship. That this exclusionary effect was intended became apparent when, after their introduction, the knowledge requirements were attributed the goal of ‘more effectively controlling immigration’. The public needed to be reassured that citizenship was not given away for free, but that something was being asked in return, which would also be one of the reasons for introducing the proposals on ‘earned citizenship’. Hence, without resorting to concepts like ‘patriality’, the UK, after the adoption of the 1981 BNA, has continued its policy of controlling immigration via naturalisation policy. The conclusion can furthermore be drawn that, at least as far as British citizenship law is concerned, the multicultural ideal has been abandoned. 73 Letter to Keith Vaz, MP, quoted in “British Citizenship Test tightened to include English Test”, http://www.bbc.co.uk/news/uk-politics-22158482, site accessed 24 April 2013.

On a Journey to Citizenship  111

Whereas Blair’s Labour governments repeatedly stood up for multicultural Britain, the subsequent Labour government led by PM Brown framed citizenship in communitarian terms, justifying new proposals with a stronger appeal to British values. Even though Blair had already counter-posed the supposedly British liberal democratic values to those of Muslim communities, he had also, at least in words, valued diversity and Britain’s multicultural society. With the 2005 attacks functioning as a catalyst, his successor Brown however further centralised the British liberal democratic values, sending the message that only those who adhered to these values belonged to the British community. Consequently, the requirements for citizenship were judged too meagre and a ‘more challenging’ test was proposed. This idea has been taken up by PM Cameron who has in the meantime introduced a revised ‘Life in the UK’ test. As changes to the knowledge requirements for both ILR and citizenship, implying the end of the ‘two-route’ system to fulfilling these requirements, are expected to take effect from 1 October 2013, we have probably not yet witnessed the end of the journey of citizenship testing in the UK.

Chapter Six The Citizenship Tests Compared In the previous three chapters, we have seen with which arguments the citizenship tests have been introduced in the countries under consideration, and which goals they were supposed to serve. This chapter will analyse the content of the citizenship tests that are the results of these debates: the ‘Knowledge of Dutch Society’ (KDS) test, forming part of the integration examination in the Netherlands, the Einbürgerungstest in Germany, and the ‘Life in the UK’ test in the United Kingdom.1 As explained in the introduction, systematic analyses of the content of the citizenship tests have so far hardly been conducted.2 Exceptions are the study conducted by Perchinig, which analyses the content of the Austrian citizenship tests, and Michalowski’s analysis of the content of citizenship tests in five countries, including the countries which serve as case studies in this dissertation (Perchinig 2010, Michalowski 2011). In her analysis, Michalowski assesses whether the tests in Austria, Germany, the Netherlands, the UK and the USA are an attempt at cultural assimilation, or whether they respect the pluralist limits of liberal democracies, by using the distinction between ‘what is right’ and ‘what is good’ based on Rawls’ definition of political liberalism. Applying this definition of political liberalism has led Michalowski to formulate a second criterion citizenship tests need to meet, namely that they should focus on subjects that refer to the ‘overlapping consensus’ (that is to say, basic civic rights and freedoms) and to the political system that is supposed to guarantee these rights. Consequently, questions about ‘the political system, democracy and rights’ should be the biggest thematic category within each test (Michalowski 2011: 752). Regarding the first criterion, she concludes that, of all the analysed tests, only the Dutch test contains questions related to ‘what is 1 The analysis of the ‘Life in the UK’ test is based on the version which has been applied until 25 March 2013, when a revised, more ‘patriotic’ version of the test was introduced (see paragraph 5.3.2). 2 Furthermore, as Michalowski has rightly contended, most existing content analyses constitute examples of “anecdotal evidence analyses”, which focus on odd and sometimes even non-existing test questions, thereby yielding biased impressions of the tests they claim to analyse (Michalowski 2011: 751).

114  Chapter Six

good’ by asking about social norms in a general sense and what is deemed to be ‘good’ in the Netherlands (Michalowski 2011:762). Michalowski furthermore concludes that the Dutch test neither meets the second criterion she applies to decide whether a test can be considered liberal, as the biggest thematic sub-category of questions is not politics, rights and democracy.3 Michalowski therefore comes to the conclusion that, among the tests compared, only the Dutch test does not correspond to the definition of political liberalism as applied in her analysis (Michalowski 2011: 763, 765). This chapter builds on Michalowski’s study by asking whether the tests in the countries under consideration can be qualified as liberal. However, unlike in Michalowski’s analysis, for a question to qualify as liberal in the present analysis, more is required than refraining from asking ‘what is good’, either by scrutinising the applicant’s inner convictions or asking about ‘what is considered good by the receiving society’ (being questions not relating to the ‘overlapping consensus’). As explained in Chapter 2 (paragraph 2.2.1), to qualify as ‘liberal’, a question must also meet a second requirement: be relevant enough to deserve a place in a liberal citizenship test. Next to determining to what extent the tests can be defended in the liberal model, this chapter will answer the question to what extent the tests match the republican and communitarian models. As the models are ideal typical representations of citizenship in political-philosophical theory, the models do not exist in a clearly distinguishable way in the naturalisation laws of individual countries. And as we have seen in the previous chapters, there has not been one indistinguishable model that has been used for the introduction of citizenship tests in the countries under consideration either. It can hence not be expected that the tests will completely match either one of the models. The models will therefore mainly be used as normative evaluation criteria. However, in case a test cannot be convincingly qualified under any of the models, the conclusion should be drawn that it is an inappropriate means of testing whether immigrants can be good citizens in either the liberal, republican or communitarian sense. Furthermore, by dividing the test questions into four thematic categories, this chapter will make a comparison of the content of the citizenship tests in terms of where the accent lies in the test questions: is it on the issue of a) rights, freedoms and duties, b) democracy, c) dealing with cultural differences, or d) general knowledge of the country under consideration?4 Answering these questions will allow us to further elaborate on the comparison between the citizenship

3 In the tests in the other countries included in Michalowski’s analysis, ‘politics, rights and democracy’ is the biggest thematic sub-category (Michalowski 2011: 762). 4 The division in categories has been partially derived from Michalowski 2008.

The Citizenship Tests Compared  115

tests. As explained in the introduction of this book, the question of whether the tests actually match the ideas on what their content should be, put forward in the political debates preceding their introduction, will be left for the concluding chapter of this book. 6.1. Qualifying the Tests Liberal, Republican or Communitarian Tests? In paragraphs 2.2.1 and 2.2.2, the standards that will be used to determine whether a question can be qualified as liberal, republican or communitarian have been given. These standards are briefly repeated below. • Liberal Citizenship Tests In order to qualify as liberal, questions in a citizenship test should meet two requirements: 1) not relate to ‘what is good’ and 2) be relevant enough to deserve a place in a liberal citizenship test. If both requirements are met, the questions meet the ‘liberal justification test’, and deserve a place in a liberal citizenship test.5 As regards the first requirement, a question should be disqualified for being illiberal if it either a) scrutinises the candidate’s ‘inner dispositions’, or b) asks about what is considered ‘good’ by the receiving society. The requirement that questions may not aim at testing the candidate’s inner dispositions flows from the idea that liberalism requires that law and public policy can only regulate the external behaviour of people (Joppke 2010a: 140). Whereas liberal states can require their citizens to conform to liberal principles in practice, and accept policies that are pursued in the name of such principles as legitimate, they cannot ask their citizens to believe such principles. As long as their behaviour does not violate the law, all is well from a liberal point of view. Consequently, (questions in) citizenship tests which aim to test whether a future citizen is a ‘true’ liberal should be disqualified for being illiberal, as these questions aim to control beliefs, rather than regulate behaviour (Joppke 2010a: 141). Reasoning a contrario, Joppke reaches  the conclusion that questions asking ‘what is right’, i.e. relate to legal 5 As explained in Chapter 2, liberal minimalists contend that all citizenship tests should be disqualified from a liberal perspective, no matter how ‘liberal’ their content may be: since they function as hurdles to acquiring citizenship, they are barriers to equality. The point of view that I hold in this study is that a citizenship test can qualify as liberal depending on two conditions: 1) its questions must pass the ‘liberal justification test’, and 2) it may not exclude immigrants from full membership. To see whether the tests in the countries under consideration meet the second requirement, I examine the actual effects produced by the citizenship tests in Chapters 7 to 10 of this book.

116  Chapter Six

commitments flowing from liberal values, are permissible in a liberal citizenship test. He furthermore concludes that questions testing factual knowledge about for instance history, culture or institutions are “unproblematic from a liberal point of view”, as “such matter is merely cognitive”, and “can be learned and mechanically reproduced” (Joppke 2010a: 141). Michalowski however opposes to questions testing cognitive knowledge if such knowledge relates to “what is considered ‘good’ by the receiving society” (Michalowski 2011: 752). She comes to this conclusion applying Rawls’ definition of political liberalism. As Michalowski explains in her analysis of citizenship tests in five countries, Rawls has asserted that political liberalism consists of finding an overlapping consensus among comprehensive doctrines on a political conception of justice, which “should be, as far as possible, presented as independent of comprehensive religious, philosophical, and moral doctrines” (Michalowski 2011: 752, Rawls 1993: 144). As questions that relate to social norms believed to be generally accepted in the host society, even though they are not laid down in law, “suggest a homogeneity of values, opinions, or moral attitudes that does not exist in western societies, and that – according to a strict definition of political liberalism – is not even necessary for a liberal democracy to function”, they cannot form part of a liberal citizenship test (Michalowski 2011: 752).6 Not even if merely cognitive knowledge is tested. Michalowski makes a difference between social norms in a general sense, referring to ‘adequate behaviour’ and ‘unwritten rules’, and to ‘what is deemed to be good in the receiving society’. In conclusion, in this analysis, questions relating to legal norms, i.e. ‘what is right’, or to factual knowledge fulfil the first requirement of the liberal justification test. Questions relating to ‘what is good’, scrutinising the candidate’s inner dispositions, or referring to social norms which are not laid down in law do not, even if merely cognitive knowledge is required. To fulfil the second requirement, a question will need to pass the liberal relevancy tests, based on the theories of Hampshire and Orgad. Hampshire qualifies a question as sufficiently relevant if it is “essential to effective citizenship”, that is, “assess[es] an applicant’s understanding of citizenship rights and duties […] symbolis[ing] the ideals of citizenship in a liberal state” (Hampshire 2010: 89). As an example, Hampshire mentions questions relating to liberal values and civic skills. He thereby takes a narrow view, limiting the knowledge that can be tested 6 De Vries has rightly noted that the “impossibility of saying that certain rules are right from a moral point of view obviously applies to legal norms as well as to non-legal conventions” (De Vries 2011: 69). However, even though legal norms are a reflection of ‘what is considered good’ by a majority of the population at a particular time, compared to social conventions, legal norms have a stronger degree of objectivity, consistency and accessibility (De Vries 2011: 69). Testing such norms is therefore unproblematic from a liberal point of view.

The Citizenship Tests Compared  117

to “knowledge required to be able to engage with other citizens in political dialogue” (Hampshire 2010: 86). In a later writing, Hampshire makes a case for the testing of knowledge which is “essential to participation as a citizen”, stating that “the general presumption in tests should be towards political knowledge” (Hampshire 2011: 967). The testing of knowledge of “certain historical events or cultural practices” might be allowed, but Hampshire warns that testing such knowledge may be “a spurious cover for cultural prejudice” (Hampshire 2010: 89, Hampshire 2011: 967). Such questions thereby risk violating the first criterion of the liberal relevancy test via the back door. To define what merits testing for citizenship and what not, Orgad has developed the test of ‘national constitutionalism’, which prescribes that only knowledge relating to a state’s essential constitutional principles should be tested, the controlling factor being the state’s Constitution (Orgad 2010a). These ‘constitutional essentials’ which merit testing concern liberal principles and values enshrined in the country’s Constitution, but may also relate to a country’s road to becoming a liberal democracy, reflecting its “constitutional uniqueness” in matters of history, development, traditions and contextual background (Orgad 2010b: 19). According to Orgad, all issues that are not ‘constitutional essentials’ should be disqualified as testing for “trivia”: information which may be useful, especially for newcomers, but which does not deserve a place in a citizenship test, which should not be “information.com” for life in the state one wants to become a member of (Orgad 2010a: 103).7 • Republican Citizenship Tests For a knowledge of society test to be justified in the republican model, its questions should focus on the essential features of the republican project, i.e. the working of democracy in the political system. Furthermore, since republican citizenship also entails joint responsibility for the common good and the sharing of a common future, questions regarding issues such as taking care of the environment or engaging in voluntary or caring activities can be allowed. Lastly, since citizenship in the republican model is an attitude, questions may aim at realising a certain behaviour in citizens, even in cases where showing this behaviour is not required by law, without necessarily being disqualified, as would be the case in the liberal model. • Communitarian Citizenship Tests Communitarian theory prioritises the identity dimension of citizenship. Consequently, only those immigrants who share the country’s identity are allowed 7 Orgad however notes that it is not per se ‘illiberal’ to ask questions relating to trivia. He nevertheless suggests not to make such knowledge mandatory in a citizenship test, but rather to supply it to newcomers as orientation material (Orgad 2010b: 23).

118  Chapter Six

full membership. In the communitarian model of naturalisation, a knowledge of society test can be used to examine to what extent the candidate adheres to or is aware of the identity of the state of which he or she hopes to become a member. The test thereby is a reflection of the way the state, or at least the elites or those responsible for developing the test, sees its identity. As long as they reflect or relate to the state’s identity, all sorts of questions may be allowed, even those relating to ‘what is good’. Hence, if a country perceives itself as liberal, questions scrutinising whether the test candidate shares its liberal identity, and the social norms emanating from this identity, can be allowed. As almost any topic can, in some way or another, be related to identity, the communitarian category runs the risk of becoming a ‘catch-all’ category, able to allow for all types of questions. I have however chosen to explicitly exclude certain types of questions from relating to a country’s identity. Consequently, in none of the countries under consideration can the applicable citizenship test for the full 100% be justified in the communitarian model. On the contrary: as we will see in table 6.5, only two fifths of all questions in the British test qualify as communitarian, whereas only a small majority of questions in the Dutch test do so. Generally speaking, questions that in this analysis have been qualified as ‘communitarian’ relate to issues of ‘democracy’, ‘rights, freedoms and duties’ and ‘dealing with cultural differences’, three of four thematic categories I discern to analyse on which topic the accent lies in the citizenship tests (see below). Questions relating to ‘general knowledge of the country concerned’ have however only been qualified as communitarian in case they refrain from asking merely about practical knowledge to ‘get by’ in daily life. Examples of such questions are for instance those asking the test candidates whether they understand the main points of an annual electricity bill (Netherlands), whether they know where to go to register a dog (Germany), or what the speed limits are (UK). Questions concerning ‘general knowledge of the country concerned’ that pass either Orgad’s or Hampshire’s liberal relevancy test will on the other hand also be justifiable in the communitarian model. Questions that do not fit into one model may deserve a place in another model. It is however also possible that a question cannot be justified in any of the models, or in more than one model. Questions relating to voting rights are for instance justified in the republican model, but can also be justified in the liberal model, provided that the question is formulated in such a way that it does not refer to ‘what is good’ or ‘what is considered good by the majority society’.8 And questions 8 As knowledge about voting rights can be considered political knowledge required for participation, questions testing such knowledge fulfil Hampshire’s relevancy test. Provided voting rights find their base in the Constitution, questions relating to voting rights will also meet the second requirement in the ‘liberal justification test’ based on Orgad’s relevancy test.

The Citizenship Tests Compared  119

about rights, freedoms and duties can be justified in the liberal model (provided both requirements in the liberal justification test are met), but also in the republican and communitarian models.9 Thematic Categories As already mentioned above, to analyse on which topic the accent lies in the citizenship tests, I have categorised the test questions along four thematic categories, being ‘democracy’, ‘rights, freedoms and duties’, ‘dealing with cultural differences’ and ‘general knowledge of the country concerned’. Questions that have been filed under the latter category relate to practical issues of everyday life and other general knowledge of the country concerned, such as its history, geography, symbols, customs, and traditions. Questions qualified under the topic ‘democracy’ concern issues relating to the working of democracy in the country concerned, such as the voting system, the parliamentary system, and the tasks of the government. Questions filed under ‘rights, freedoms and duties’ concern the liberal rights and freedoms applicable in the countries under consideration, as well the duties related to citizenship. Lastly, questions are categorised under ‘dealing with cultural differences’ if they aim to target lifestyles or behaviour supposedly connected to certain ethnic minorities. They concern topics such as the (in)equality of men and women, child rearing, (forced) marriage, the prohibition of polygamy, and the acceptance of homosexuality. As Michalowski has noted, even though it is not explicitly stated, “it is relatively clear from the current European public debates on Islam that these questions target conservative Muslims” (Michalowski 2011: 759). 6.2. Qualifying the Dutch Integration Examination The curriculum for the Knowledge of Dutch Society (KDS) part of the integration examination was developed by Bureau ICE and Cito, (semi-)private testing agencies that had previously been responsible for developing the naturalisation test. As we have seen, the curriculum was not discussed in parliament. In (almost) unchanged form, it can now be found as appendix to the Integration Regulation.10 9 A question regarding the freedom of opinion is allowed in the republican model for teaching future citizens that they can be politically active by, within the limits of the law, voicing a dissenting opinion without being censored. Such a question can also be justified in the communitarian model as it reflects a country’s liberal identity. 10 The curriculum has been published as an appendix to Article 2.5 of the Ministerial Regulation on Integration (Regeling van de Minister voor Vreemdelingenzaken en Integratie van 6 december 2006, nr. 5456790/06, tot uitvoering van de Wet inburgering, het Besluit inburgering en tot wijziging van de Regeling verstrekkingen asielzoekers en andere categorieën vreemdelingen

120  Chapter Six

The questions in the test concern the following eight topics: work and income; manners, norms and values; housing; health and health care; history and geography; authorities; the state and the rule of law, and, finally, education and upbringing. In establishing the curriculum, the test makers determined so-called ‘crucial practice situations’, ‘crucial acts’ and ‘crucial knowledge’.11 Finally, norms for behaviour, indicating when the ‘crucial act’ can be considered to be successful, have been formulated, the so-called indicators.12 Since the actual test questions are not published, I have chosen to take these indicators, 200 in total, as a starting point for determining the content of the actual KDS examination.13 The reason for this choice is that these indicators, according to the designers of the test, are the ‘guiding concepts’ for the content of the KDS examination.14 The indicators hence provide a good insight into the content of the test. Examples of indicators are: the test candidate “looks for training in case of insufficient application skills” (topic: work and income), “informs him/herself about implicit rules of social intercourse (by observing, asking advice, asking for a confirmation or correction, etc.)” (manners, norms and values), and “concludes the most common insurances for the house” (housing). In the analysis made below, each indicator is treated as a possible question in the KDS examination. In the above overview, six indicators (3% of the total) have been qualified under the heading ‘rights, freedoms and duties’. These indicators all relate to the prohibition of discrimination on the basis of religion, gender, ethnicity and sexual preference, if suffered by the test candidate. The eleven indicators that relate to 2005), and has been presented to the Second Chamber in June 2006 (TK 2005–2006, 30308, no. 25). As of 1 July 2011, knowledge of the Dutch anthem, the Wilhelmus, has been added to the curriculum, upon request by Christian Democratic and SGP parliamentarians. Stscrt. 2011, no. 7426, 29 April 2011. 11 ‘Crucial practice situations’ are situations where the immigrant is required to be able to function ‘adequately’. ‘Crucial acts’ are acts which are essential to function adequately, and for each act, ‘crucial knowledge’ is required. An example of a ‘crucial practice situation’ is “being able to function in the employment market”, to which the ability to “be able to quickly and efficiently look for work” has been formulated as a crucial act. Crucial knowledge in this regard concerns knowing about the services the Centre for Work and Income offers to those looking for work. 12 Examples of indicators belonging to the crucial act of being able to quickly and efficiently look for work, mentioned as an example of a crucial act in the previous footnote, are that the test candidate “will consult own network and sources for vacancies”, and that he or she “directly contacts the Centre for Work and Income to register in case of unemployment”. 13  Whereas the questions are not published, there is a practice examination available at http://www.inburgeren.nl/inburgeraar/examen/oefenen_met_examens/voorbeeldexamen _kennis_nederlandse_samenleving.asp; site accessed 1 March 2012. 14 Preface of the curriculum. In her analysis, Michalowski has chosen to take both the crucial knowledge and the indicators as point of departure. In my opinion, this however means that certain topics are counted more than once, which distorts the analysis.

The Citizenship Tests Compared  121

Table 6.1. Questions in the Knowledge of Dutch Society examination, along four thematic categories. Category

Number of indicators

Percentage of total of 200 indicators

Democracy Rights, freedoms and duties General knowledge of the  Netherlands Dealing with cultural differences

14 6 124

7% 3% 62%

56

28%

the prohibition to discriminate, i.e. prescribe that test candidates should themselves give others an equal treatment, have not been categorised under ‘rights freedoms and duties’, but under ‘dealing with cultural differences’. The reason is that these indicators are formulated in such a way that they specifically target candidates with a certain (read: Muslim) background. Examples are the indicators asking test candidates “not to bother people who openly come out for their homosexuality” and to “acknowledge women’s right to self-determination”. The same applies to the indicators, eight in total, relating to the freedom of religion, the freedom of opinion, the inviolability of the human body, the separation of powers, and the liberal core value of secular laws trumping religious laws. Examples are the indicators that ask the candidates to “respect people with a different opinion than their own” and to “leave others free to celebrate the most important religious and political holidays”. Apart from the above-mentioned indicators on liberal rights, 37 other indicators which depart from the assumption that test candidates, because of their different cultural backgrounds, will be unaware of the situation in the Netherlands, have been classified under the heading ‘dealing with cultural differences’. These include indicators about social norms, which for instance ask the test candidates to “not take immediate offence to directly uttered opinions and direct questions”, to “take a number, take a shopping cart, stand in line, wait for one’s turn”, to “sport, exercise and eat healthily in order to stay healthy” and to “work independently and show initiative”. Furthermore, the indicators asking test candidates to “show understanding for Dutch sensitivity regarding statements of anti-Semitism”, and to “show respect for the celebrations and customs of 4 and 5 May” (commemoration of the dead and liberation day of World War Two) have been classified under ‘dealing with cultural differences’. With 56 indicators qualified under the heading ‘dealing with cultural differences’ this topic makes up almost 30% of the total number of indicators upon which test questions are based.

122  Chapter Six

Fourteen indicators have been categorised under ‘democracy’. These indicators concern voting rights, the Dutch form of government and the separation of powers, and the indicators which aim to convey knowledge on how to call the government to account.15 Lastly, a majority of the indicators (62%) have been categorised under “general knowledge of the Netherlands”. The 124 indicators in this category relate to history (four indicators which for instance ask the test candidates to “explain the most important eras in Dutch history and their characteristics” and to “explain the historical, economical and cultural ties between one’s country of origin and the Netherlands”), geography (four indicators asking the test candidates to, inter alia, “estimate the distances and travel time between places in the Netherlands (for example for work, school, family visits)” and to “explain the relationship between regions and the regions’ economic features”), and practical issues of everyday life, including health care and work and income (totalling 116 indicators).16 The indicators relating to practical issues of everyday life, constituting the largest category of indicators among the indicators qualified under “general knowledge of the Netherlands”, for instance ask the test candidates to “explain why wage taxes are required”, to “bring a referral letter from your G.P. when visiting a medical specialist” and to “register changes in case of birth, marriage, divorce, death, moving, migration”. Based on the above ranking, we will now analyse under which citizenship model the Dutch test can be best qualified. 6.2.1. Is the Dutch Citizenship Test Liberal? Can the Dutch citizenship test be justified from a liberal perspective? Since the test content is based on the indicators, i.e. norms for successful behaviour, which prescribe the test candidate to show a certain type of behaviour, only the questions that concern behaviour which relates to a legal norm, i.e. ‘what is right’, rather than to behaviour which does not lead to a violation of the law, i.e. is norms-based and relates to ‘what is good’, can be qualified as liberal. This would 15 As stated above, indicators relating to the separation of powers have also been qualified under ‘dealing with cultural differences’. This applies to the indicators asking test candidates to “in the first instance apply the laws and rules of the state” and to “accept the judgments of the courts”. Both of these indicators aim to make sure that test candidates know how to deal with the separation between church (religion) and state, and specifically target test candidates with a certain cultural background. The indicator on the separation of powers asking test candidates to “explain the importance of the separation of the three powers in the Netherlands” however aims to convey knowledge on the working of democracy in the Netherlands. For being more neutrally formulated, it has been ranged under ‘democracy’. 16 Indicators relating to customs and traditions are absent in the Dutch test.

The Citizenship Tests Compared  123

mean that a majority of the indicators would need to be disqualified. However, looking past the way the indicators have been formulated, instead of directly disqualifying them, three different types of indicators emerge: indicators aiming to test factual knowledge (129 indicators), indicators relating to social norms (46 indicators), and indicators relating to liberal core values (25 indicators). Based on this categorisation, I will determine whether the indicators can be qualified as liberal. • Indicators about Factual Knowledge In total, 129 indicators relate to objective knowledge. Among these indicators are 116 of 124 indicators relating to ‘general knowledge of the Netherlands’ (examples are the indicators asking test candidates to “follow the rules for applying for housing benefits” and to “directly contact the centre for work and income in case of unemployment”), twelve indicators that have been qualified under the category ‘democracy’ (test candidates should be able to “explain situations in which a vote can be cast” and to “explain the most important Dutch political parties”), and one indicator that has been categorised under ‘dealing with cultural differences’ (test candidates should “dispose of waste in designated places”). Rather than asking the test candidate to show a certain type of behaviour which does not relate to a legal norm, these indicators in fact aim at making sure that the candidate has sufficient factual knowledge of life in the Netherlands. For not relating to ‘what is good’, these questions can be justified in a liberal citizenship test, provided they are relevant enough to merit testing. Applying Hampshire’s relevancy test, twelve of the 129 indicators on factual knowledge can be justified, namely those that relate to voting rights and democracy. These indicators for instance ask the applicant to mention the importance of active and passive voting rights, to explain the most important political parties, and to explain the importance of the separation of the three powers in the Netherlands. In Hampshire’s test, all other indicators, 117 in total, are not essential to citizenship, and should for that reason not form part of a citizenship test. Orgad’s relevancy test allows for a total of eight of the 129 indicators on general knowledge of the Netherlands, namely those which have their basis in the Dutch Constitution. This applies to the indicator relating to the freedom of education, asking test candidates to explain why the Dutch school system is pillarised, and to seven indicators on democracy which also pass Hampshire’s relevancy test.17 17 These indicators ask the test candidate to explain in which situations a vote can be cast, in which elections non-naturalised citizens can participate, to explain the position of the queen, to explain the most important tasks and responsibilities of the central government towards the citizens, to explain the most important tasks and responsibilities of the provincial government towards the citizens, to explain the main points in the parliamentary system (government, first and second

124  Chapter Six

Applying Orgad’s and Hampshire’s relevancy tests means that a majority of either 117 (Hampshire) or 121 (Orgad) of the indicators relating to objective knowledge do not deserve a place in a liberal citizenship test. • Indicators Relating to Social Norms in a General Sense Forty-six indicators in the curriculum for the Dutch citizenship test concern social norms in a general sense, by asking candidates to show ‘adequate behaviour’ or to follow ‘unwritten rules’. Among these indicators are eight indicators that have been categorised under “general knowledge of the Netherlands”, which for instance ask the test candidates to “work independently and to show initiative”, to “periodically visit a dentist for a check up or treatment” and to “deliberate with a child’s teacher about the progress the child has made”. Two of 46 indicators concerning social norms have been categorised under ‘democracy’ and ask the test candidates to “make use of one’s voting rights” and to “take up an individual position on the basis of the main points in the electoral programmes of political parties”. The majority (36) of the indicators relating to social norms have been categorised under “dealing with cultural differences”. Examples are the indicators asking test candidates to “not take directly uttered criticism personally”, and to “indicate to others that they have reached ‘the limit’ (but to do this in a calm way)”. Michalowski has qualified these questions as a “kind of ‘how-to-guide-onetiquette’”, which “push for the acceptance of social norms” (Michalowski 2011: 760). As Michalowski has contended, they need to be disqualified for being illiberal questions, since they suggest a moral consensus on the issues tested, which in reality does not exist, and which is not necessary for a liberal democracy to function (Michalowski 2011: 752, 762). Whereas the majority of the questions relating to social norms need to be disqualified from a liberal perspective for reasons just explained, four indicators can be qualified as liberal for having a connection to the content of Dutch laws, thereby relating to legal acceptance or ‘what is right’ rather than ‘what is good’. These indicators relate to the law regulating the obligation to follow education for persons aged between 4 and 16, and the indicators relating to statements of antiSemitism and showing respect for the customs and festivities of 4 and 5 May.18 Of two indicators it is doubtful whether they can be justified, namely the ones asking the test candidate to “keep one’s garden and surroundings clean” and to “keep to the prevailing rules relating to the environment”.19 If these indicators chambers, ‘Prinsjesdag’, multi-party system), and to explain the importance of the separation of the three powers in the Netherlands. 18 These indicators for instance ask the test candidate to “see to it that his/her children go to school” and to “show respect for the Dutch sensitivities for statements of anti-Semitism and the celebration and the habits on 4 and 5 May (remembrance day and memorial day)”. 19 These indicators can be found in the topic ‘housing’ (wonen).

The Citizenship Tests Compared  125

refer to legal commitments, in this case the prohibition of public littering, they can be justified, but this is unclear from their formulation. For the moment, I will give these indicators the benefit of the doubt. Do the six indicators relating to social norms that do not relate to a ‘what is good’ also fulfil the second requirement in the liberal justification test? Applying Hampshire’s relevancy test leads to the conclusion that the majority does not, as the knowledge tested does not relate to liberal values and cannot be considered ‘political knowledge’ required to be able to engage with other citizens in political dialogue. The only indicator which does pass Hampshire’s relevancy test is the one relating to statements of anti-Semitism, as not taking account of the “Dutch sensitivities regarding outings of anti-Semitism” will hamper one’s possibilities for political dialogue. Since they do not have their basis in the Dutch Constitution, or relate to the road the Netherlands has travelled to becoming a liberal democracy, the six indicators do not pass Orgad’s test of national constitutionalism either.20 Based on the above, the conclusion should be drawn that, apart from the one indicator which passes Hampshire’s relevancy test, the 46 indicators relating to social norms do not deserve a place in a liberal citizenship test. • Indicators about Liberal Core Values The remaining 25 indicators relate to liberal principles. The majority (nineteen) of the indicators have been categorised under ‘dealing with cultural differences’ and for instance ask the test candidates to “not bother people practicing other forms of living together than marriage” and to “not interfere with educational- or carrier paths of children and/or partner”. The remaining six indicators that relate to liberal principles have been categorised under “rights, freedoms and duties” and concern the prohibition to be discriminated against. These indicators for instance ask the test candidates to “report instances of discrimination at the appropriate office”, to “discuss a suspicion of discrimination with one’s employer”, and to “use appropriate strategies to tackle discrimination”. Most of the indicators relating to liberal principles, a total of seventeen, concern the principle of equal treatment, as captured in Article 1 of the Dutch Constitution which prohibits discrimination. The remaining eight indicators relate to the freedom of religion, the freedom of opinion, the inviolability of the human body, the separation of powers, and the liberal core value of secular laws trumping religious laws. 20 This also applies to the indicators relating to the environment, even though protecting and improving the environment is an essential Dutch constitutional value, protected in Article 21 of the Dutch Constitution. But as this article states that it is the authorities’, and not the citizen’s, concern to “keep the country habitable and to protect and improve the environment”, the indicators relating to the environment cannot be defended by reference to the Dutch Constitution.

126  Chapter Six

Regarding the seventeen indicators about the principle of equal treatment, six relate to dealing with discrimination if suffered by the test candidate. The other eleven indicators relate to the prohibition to discriminate, i.e. prescribe the test candidates that they are obliged to give others an equal treatment. Five indicators specifically concern the equal treatment of women. Asking whether the questions on liberal core values relate to a ‘what is right’ or a ‘what is good’ leads us to the conclusion that fifteen of 25 indicators are illegitimate items for aiming to test the latter rather than the former. The reason is that the knowledge about the liberal core values is framed in such a way that there is only a partial link to the legal situation. These indicators, which relate to the provision of equal treatment and the freedom of opinion, for instance ask the candidates not to bother people who openly come out for their homosexuality, or not to bother those who are, in their own opinion, dressed inappropriately or display unknown or undesirable behaviour in the fields of religion, philosophy of life, political preference, race or gender. Even though these indicators target the external respect for liberal core values, and hence refrain from exploring the candidate’s moral perceptions or inner dispositions, the situations described are formulated too broadly to necessarily lead to a violation of Dutch laws: as long as the ‘bothering’ of homosexuals and people who are dressed ‘inappropriately’ remains within the borders of the Dutch laws, there is nothing illegal about it. Instead, these indicators are tied to social norms. They ask the candidate to know ‘what is considered good’ by the receiving society, thereby suggesting the existence of a moral consensus on these issues of which it is doubtful whether it exists, and which is not necessary for a liberal democracy to function (Michalowski 2011: 752, 762).21 They should for this reason be considered illiberal test questions. The indicators are furthermore illiberal in another way: they limit one’s freedom of opinion. The remaining ten indicators on liberal core values can be justified in the liberal model, for they aim to convey factual knowledge about these values, which might be necessary for the candidate to ‘get by’ in the Netherlands, or focus on the legal obligations flowing from the liberal values. Examples are the indicators which ask the candidate to “discuss a suspicion of discrimination with the employer” and to “respect the physical integrity of others”. Are these ten indicators also relevant enough to merit testing? Applying Hampshire’s relevancy test leads to the conclusion that all ten questions are legitimate questions to be asked, since they all relate to liberal values. Furthermore,

21 Orgad rightly states that, as there is no moral consensus on the issues tested, there cannot be a “moral right or wrong answer” to the questions asked either (Orgad 2010b: 21). This is one among five ‘liberal concerns’ Orgad has uttered towards citizenship tests in European Member States.

The Citizenship Tests Compared  127

since they all have a basis in the Dutch Constitution, the questions can also be justified using Orgad’s test of national constitutionalism. Applying the liberal justification test to the 25 indicators relating to liberal core values hence teaches us that a total of ten of these indicators deserve a place in a liberal citizenship test. The above analysis has shown that over a quarter (27%; 55 indicators) of the indicators that are used as guiding concepts for the Dutch knowledge of society test needs to be disqualified for testing ‘what is good’. Part of these indicators aim to test social norms emanating from liberal core values. These indicators are expressions of what has been labelled as ‘the paradox of liberal toleration’: they seek to exclude people because of their supposedly illiberal ideas, thereby breaching the liberal principles of freedom and tolerance (Entzinger 2003: 80, Neuman 1994: 261, Orgad 2010a: 94). In the light of the centrality of the liberal values applicable in the Netherlands, Orgad has labelled the ideological exclusion these questions aim to accomplish, and which is based on the premise that certain behaviour is considered ‘un-Dutch’, as being ‘un-Dutch’ in itself (Orgad 2010b: 22). Of the 145 indicators which fulfil the first requirement of the liberal justification test by relating to factual knowledge or ‘what is right’, either only 23 (Hampshire) or eighteen (Orgad) indicators fulfil the second requirement of the liberal justification test by being sufficiently relevant to merit testing. This means that the large majority of the 200 indicators do not merit testing in the liberal model. 6.2.2. Republican Elements in the Dutch Citizenship Test Asking whether the indicators deserve a place in a republican citizenship test for relating either to the joint responsibility for the common good, the working of democracy or by contributing to test candidates being able to participate politically, teaches us that a total of 19 indicators merit testing in the republican model: the five indicators which ask about the joint responsibility for the common good and the fourteen indicators relating to democracy. The former indicators ask about the collection of waste and the obligation to take care of one’s surroundings. The latter indicators ask test candidates about the working of democracy, thereby providing them the tools to participate in the political process. Whilst 19 indicators can be justified in the republican model, the remaining 181 indicators cannot. This also goes for the three indicators relating to the freedom of opinion, which for instance teach test candidates to “give their opinion, without going overboard (knowing where the boundaries lie)” and to “indicate to others that they are reaching ‘the limit’ (but to do this calmly)”. Instead of teaching test candidates that it is allowed to voice a dissenting opinion without being arrested or censored and thereby (possibly) contributing to their being politically

128  Chapter Six

active, these indicators rather aim at avoiding political discussions between citizens, thus undermining the basis of the republican project. These indicators furthermore appear to be ‘un-republican’ in another way: they aim at ignoring or wiping out cultural differences, whereas, in the republican model, citizenship concerns acceptance of and competence in ways of dealing with cultural and social differences (see paragraph 2.2.2.1).22 6.2.3. Communitarian Elements in the Dutch Citizenship Test As we have seen above, a communitarian citizenship test may contain all sorts of questions, provided they reflect the country’s identity. What conclusions can be drawn for the Dutch test? A first conclusion is that the 56 questions qualified under ‘dealing with cultural differences’ are allowed under the communitarian model: since they aim to teach the candidates how Dutch people generally behave and expect others to behave, or what is considered to be civilised or morally appropriate conduct, they have a link to the Dutch identity.23 Secondly, the fourteen indicators on ‘democracy’ qualify as communitarian for aiming to stress the Dutch democratic identity and the importance which is attached to being politically active. Lastly, the six indicators relating to the prohibition to be discriminated against, qualified under ‘rights, freedoms and duties’, can be justified in the communitarian model for aiming to stress the liberal identity of the Netherlands. Whereas a link can be established between the Dutch identity and the indicators referred to above, establishing such a link sometimes appears to be more difficult for most of the 124 indicators relating to general knowledge of the Netherlands. After all, it is difficult to see what “being able to estimate one’s financial possibilities in case of a mortgage”, knowing that you are required to “make an appointment with your GP in order to get acquainted” or being able to “formulate

22 Two other indicators have been disqualified for being ‘un-republican’ in two ways. This applies to the indicators prescribing the test candidates to “react at least neutrally to persons who behave in an unknown or undesired way as regards abortion, euthanasia, homosexuality and sexuality” and the one that prescribes candidates “not to bother people with different religious beliefs, political preferences, race or gender”. Like the indicators on the freedom of opinion, these indicators aim at avoiding both discussions between citizens and ignoring or wiping out cultural differences. 23 The problem that it is impossible, or very hard to say the least, to identify a Dutch code of conduct however remains. In the case of the Dutch KDS examination, the answer to the question of what is morally justified or considered appropriate conduct has been provided by the test makers (ICE and Cito). Whether their opinion is shared by the rest of the Dutch population is questionable.

The Citizenship Tests Compared  129

a wish relating to housing matching the situation of his family and his income” (sic.) have to do with the Dutch identity. Nevertheless, some of these indicators can be justified under the communitarian model. This applies for instance to the eight indicators on history and geography, and to the indicator relating to the freedom of education, which asks test candidates to explain why the Dutch school system is pillarised. A case could furthermore be made for the 27 indicators relating to general knowledge of the Netherlands in the topic ‘work and income’.24 These indicators for instance ask the test candidates to “discuss one’s wishes regarding education with the Centre for Work and Income” and to “take circumstances which influence one’s chances on the employment market (education, experience, availability, motivation, place of residence, knowledge of the Dutch language) into account [when looking for employment]”. Since these indicators aim to guarantee that test candidates know how to participate in the labour market, they aim to make sure that they can live up to the norm which prescribes that you do your best to be able to support yourself. In the debates preceding the introduction of the test, this norm was identified by Minister Verdonk as forming part of the Dutch range of norms and values, thereby, at least in Verdonk’s opinion, forming part of Dutch identity.25 The conclusion can be drawn that a total of 112 indicators merit a place in a communitarian citizenship test. Seen from a communitarian angle, the remaining 88 indicators however need to be disqualified for having only a tenuous connection to the Dutch identity. 6.3. Qualifying the German Einbürgerungstest The German questionnaire for the Einbürgerungstest, consisting of 300 questions, was developed by the Institute for Quality Development (IQB) of the Humboldt University of Berlin. The content of the questionnaire builds on the themes of the Orientierungskurs of the Residence Act, being ‘living in a democracy’, ‘history and responsibility’, and ‘man and society’.26 The schedule below categorises the 300 questions in the federal questionnaire.27 24 In total, 34 indicators have been formulated for the topic ‘work and income’. 27 indicators relate to factual knowledge. The remaining indicators relate to liberal rights (the candidate “discusses a suspicion of discrimination with the employer”) and social norms (the candidate “enters into contact with colleagues and managers in an appropriate way”), and for that reason have already been qualified as communitarian above. 25 TK 2005–2006, 30308, p. 20. 26 Rahmencurriculum für den Einbürgerungskurs, p. 5. 27 The questions in the questionnaires of the Länder do not form part of this categorisation. These questionnaires each consist of ten identical questions. The questions concern the symbols of the

130  Chapter Six

Table 6.2. Questions in the German Einbürgerungstest along four thematic categories. Category

Number of questions

Percentage of total of 300 questions

Democracy Rights, freedoms and duties General knowledge of Germany Dealing with cultural differences

115 28 145 12

38.33% 9.33% 48.33% 4%

The table above shows that most questions in the German test relate to the theme ‘general knowledge of Germany’, followed by the topics ‘democracy’, ‘rights, freedoms and duties’ and, finally, ‘dealing with cultural differences’. What does the above-mentioned division say about the German naturalisation test? 6.3.1. Can the Einbürgerungstest Be Qualified as Liberal? As we have seen in Chapter 4, the introduction of the Einbürgerungstest has to a certain extent been the result of ‘wild testing’ that took place, or had been proposed, in respectively Baden-Württemberg and Hesse. The government has always claimed that the federal test would not resemble its regional predecessors.28 Apart from two oddly formulated questions, the federal test, unlike the interview manual applied in Baden-Württemberg and part of the ‘Wissen und Wertetest’ proposed by Hesse, indeed refrains from asking about the candidates’ inner dispositions or from asking about social norms, and thereby from asking about ‘what is good’.29 Land (flag and coat of arms), topographical knowledge of the Land (where the Land is situated, what the Land’s capital is, what a Landkreis is), and the Land’s polity (which Minister the Land does not have, what age one needs to have to vote in the communal elections, what the president of the Land’s government is called, which agency one can address in the Land to inform oneself about politics, for how many years the Land’s parliament is elected). 28 See, for instance, Plepro 16/168, p. 17796 (C). 29 These questions are “what should you do if you are badly treated by an official” (topic: general knowledge of Germany) and the question asking what the parents of a 22-year-old women can do if they do not like the boyfriend their daughter is living with (topic: dealing with cultural differences). The correct answers to these questions, respectively being “talk to the manager” and “they should respect the decision of their daughter, who has reached the age of majority”, push for the acceptance of a social norm which is not embedded in the law. In doing so, the questions address a ‘what is good’ and not a ‘what is right’, and should therefore be dismissed from a liberal point of view.

The Citizenship Tests Compared  131

By not relating to a ‘what is good’, all but two questions fulfil the first requirement of the liberal justification test. We now need to determine whether they are also relevant enough to merit testing. This question will be answered per thematic category below. • General Knowledge of Germany 145 questions have been grouped under the denominator of ‘general knowledge of Germany’. Questions in this topic thereby represent almost half of all the questions in the questionnaire. The questions focus on German history (70 questions), practical issues of everyday life (42 questions), geography (nine questions), Landeskunde (nine questions), Germany and the European Union (eight questions), the German culture (six questions), and economy (one question).30 Examples of questions are “which country has a border with Germany?”, “the economical order of Germany is called…”, and “what is the content of the treaties of Rome?”. Questions on culture relate to German customs and traditions, based on Germany’s Christian tradition. These questions for instance ask about what Germans do at Easter, on which occasion they wear colourful costumes and masks, and what the four weeks prior to Christmas are called. The 70 questions on history cover two topics: World War II, national socialism and the post-War era on the one hand, and the division of Germany into a western part and an eastern part and the subsequent unification on the other. Applying Hampshire’s relevancy test to the questions relating to general knowledge of Germany leads to the conclusion that the majority of the questions does not merit testing, since they do not relate to liberal values and civic skills. As knowing the answers to the questions related to general knowledge on Germany is not required for political dialogue or participation they are not relevant for effective citizenship and do not merit being asked as a condition for naturalisation. A case could however be made for the testing of the knowledge of Germany’s history. Hampshire has himself qualified questions related to Nazism or reunification as falling in a ‘grey area’, as such knowledge “could be essential to effective citizenship in contemporary Germany” (Hampshire 2011: 967). As Hampshire has however also claimed that the “general presumption in tests should be towards political rather than cultural knowledge”, I have decided that the questions on Germany’s history also fail Hampshire’s relevancy test (Hampshire 2011: 967).

30 The category ‘democracy’ also contains questions relating to the EU. Unlike the questions on the EU grouped under the topic ‘democracy’, the questions on the EU categorised under ‘general knowledge of Germany’ focus on general knowledge of the EU, rather than its democratic aspects, i.e. the European Parliament and the elections. The topic Landeskunde addresses such subjects as the German flag, the arms of the Federal Republic of Germany and the national anthem.

132  Chapter Six

Does applying Orgad’s relevancy test lead to the same conclusion? This question will first be answered for the 70 questions on history, and subsequently for the remaining 75 questions on general knowledge of Germany. The 70 questions on history pass Orgad’s relevancy test if they concern Germany’s road to becoming a liberal democracy, reflecting its ‘constitutional uniqueness’ (Orgad 2010b: 19). Applying Orgad’s test of national constitutionalism, but without further elaboration, Joppke comes to the conclusion that the questions on history in the German test indeed merit testing (Joppke 2010b: 128).31 By referring to World War Two as a ‘constitutional essential’, but also failing to analyse the actual content of the German test, Orgad appears to agree with Joppke, at least as far as the questions relating to the Second World War are concerned (Orgad 2010a: 104). For equally reflecting Germany’s road to becoming a democracy, the other topic considered in the history section of the test, being Germany’s division into an eastern and a western part and the unification, can probably also be defended for constituting a ‘constitutional essential’. Hence, applying Orgad’s test of national constitutionalism, all 70 questions in the history section of the Einbürgerungstest merit testing.32 An analysis of the remaining 75 questions in the topic ‘general knowledge of Germany’ shows that the majority of the questions asked fail Orgad’s test of national constitutionalism for either having only a tenuous connection to the German Constitution or for lacking such a connection completely. Most questions should, therefore, be disqualified. The only exception seems to be provided by the question asking about the control exercised by the German state on the school system, which actually has its basis in the German Constitution.33 In conclusion, 71 questions on general knowledge of Germany pass Orgad’s relevancy test, whereas none of the questions pass the relevancy test developed by Hampshire. • Questions relating to the Theme ‘Democracy’ A substantial part of the questions (115 in total) in the German questionnaire concern the theme ‘democracy’. Examples of questions are “In Germany, the rule of 31  According to Joppke, the questions in the section ‘history and responsibility’ of the Einbürgerungstest ‘come close’ to Orgad’s national constitutionalism. 32 Whereas we will take this conclusion on the relevance of the questions on history as a starting point for the present analysis, it should be noted that, for some questions, it is doubtful what contribution they will make to improving a candidate’s understanding of Germany’s road to liberal democracy and the content of its Constitution. This for instance applies to questions which concern mere facts, such as “when were the national-socialists with Adolf Hitler in power?” and “when did World War Two end?” 33 Question 68. Article 7(1) of the Basic Law provides that “the entire school system shall be under the supervision of the state”.

The Citizenship Tests Compared  133

law applies. What does this mean?”, “Who chooses the members of the federal parliament?” and “What is not a feature of a democracy?”. Are these questions relevant enough to merit testing? As we have seen above, Hampshire allows for questions related to civic skills and liberal values. Applying Hampshire’s relevancy test, one could argue that all questions in the topic ‘democracy’ deserve a place in the German citizenship test for enabling the candidate to participate politically and to engage in political dialogue. By aiming to teach the test candidate how the German democratic system works and by which liberal democratic principles it is guided the questions about democracy hence constitute legitimate components in a citizenship test. It has to be said though, that the (possible) contribution some questions make to achieve this aim appears to be higher than the (possible) contribution made by other questions.34 Applying Orgad’s test of national constitutionalism leads us to the conclusion that a total of 68 of 115 questions on democracy merit testing on the basis of the topic being enshrined in the Constitution. These questions concern the German polity, voting rights, Germany as a constitutional state, popular sovereignty, the trias politica, and conflict solving in a democratic society.35 Cultural Differences and ‘Rights, Freedoms and Duties’ Lastly, 40 questions (13.33%) in the German questionnaire deal with either the topic ‘dealing with cultural differences’ (12 questions; 4% of the questionnaire) or ‘rights, freedoms and duties’ (28 questions; 9.33%). Can these questions be qualified as relevant enough to deserve a place in a liberal citizenship test? The twelve questions that have been categorised under ‘dealing with cultural differences’ address issues relating to differences between the German and, judging from the content of the questions, (orthodox) Muslim culture. The questions concern topics such as the prohibition of polygamy, the prohibition to beat children, the equality of men and women, the toleration of homosexuality, the possibility to marry someone with another faith, and child rearing. These questions 34 Some questions appear to be more useful for the purpose of helping someone exercise his or her voting rights than other questions. The question “Why is there more than one political party in a democracy?” (question 41) is probably more likely to contribute to someone”s understanding of democracy than the question “where does the German Government work” (question 71) or “what does the abbreviation ‘CDU’ stand for” (question 76), provided of course that citizenship testing attributes to candidates being able to make (better) use of their democratic rights at all. 35 Not all the questions relating to conflict solving in a democratic society and the German polity however have their basis in the Constitution. These questions have, for that reason, been disqualified. This for instance concerns the questions regarding the role of the ‘Schöffe’ (lay jurist; questions 140, 149 and 150) and the question on Germany’s administrative structure (question 69).

134  Chapter Six

aim to confer knowledge of the existence of the liberal principles and freedoms in Germany, and do not refer to the prohibition on the test candidates to infringe these principles and freedoms themselves. Examples of questions are for instance “who in Germany can register a wedding” (equality between men and women) and “who in Germany cannot live together as a couple” (toleration of homosexuality). The questions in the German test hence differ from the questions categorised under ‘dealing with cultural differences’ in the Dutch test, where the questions concern the obligation of the test candidates to treat others (women, homosexuals) equally, or to keep the conflicting ideas they supposedly have on the equality of others to themselves. Applying Hampshire’s relevancy test, the majority of the questions on ‘dealing with cultural differences’ deserve a place in a liberal citizenship test for concerning liberal values.36 The only exception appears to be provided by the question on child rearing, which asks whose responsibility it is to raise children. In Orgad’s test of national constitutionalism, the questions relating to topics that are contained in the Constitution can be regarded as relevant, thereby allowing for the questions on the equal treatment of men and women and homosexuals, the freedom of religion, and the prohibition to beat children.37 The two questions on polygamy and child rearing do not have a basis in the Constitution, and therefore should not be asked in a citizenship test. Twenty-eight questions deal with the topic ‘rights, freedoms and duties’. Twenty-five of these questions relate to rights and freedoms, including the freedom of opinion, the freedom of religion, and the right to equal treatment. Examples are: “In Germany, people can openly say something to the Government, since…”. and “Which right belongs to the basic rights, guaranteed in the German Constitution? The right to…”.38 The three questions on duties concern the duty to perform military service. Since the questions on rights and freedoms concern liberal core values, which are enshrined in the Constitution, these questions pass both Hampshire’s and Orgad’s relevancy tests. The same goes for the three questions on duties: they pass Hampshire’s test because they relate to a duty which is attached to citizenship, and Orgad’s test for having a basis in the German Constitution.

36 It is, however, questionable to what extent the prohibition of polygamy concerns a liberal value, since prohibiting someone from being married to more than one person at the same time could also be judged as interfering with this person’s freedom of religion or opinion. 37 The latter question can be seen as having been derived from article 1 of the German Basic Law stating that “Die Würde des Menschen ist unantastbar.” I thank Ines Michalowski for pointing this out to me. 38 Question 7. A) Freedom of religion, B) Care, C) Employment, D) Housing.

The Citizenship Tests Compared  135

Hence, depending on which test is applied, either 39 (Hampshire) or 38 (Orgad) questions relating to ‘dealing with cultural differences’ and ‘rights, freedoms and duties’ deserve a place in a liberal citizenship test. The relevancy tests developed by Orgad and Hampshire lead to different verdicts about the German Einbürgerungstest. Applying Hampshire’s test leads to the conclusion that more than half (52%) of the 298 questions that fulfil the first requirement of the liberal justification test, i.e. refrain from asking ‘what is good’, are also relevant enough to merit testing.39 Orgad’s test of national constitutionalism leads to a more positive outcome for the German test by allowing almost 60% of all test questions a place in a citizenship test.40 Either way, a substantial proportion of the questions in the German test still need to be disqualified from a liberal perspective. 6.3.2. Can the Einbürgerungstest Be Qualified as Republican? As table 6.2 shows, 38.33% of all the questions in the questionnaire relate to the working of democracy and democratic rights of German citizens. One of the most important aims of the German citizenship test hence appears to be to teach naturalisation candidates how democracy works in the German political system (Michalowski 2011: 758). By focusing on one of the essential features of the republican project, these questions can be defended in the republican model. This also applies to the five questions on the freedom of opinion and the three questions on freedom of the press, since these rights relate to the republican project: they aim to teach the test candidates that they can freely express their (political) opinions without being censored. Hence, in total, 40% of the questions in the German Einbürgerungstest can be qualified as ‘republican’. 6.3.3. Can the Einbürgerungstest Be Qualified as Communitarian? As questions in a communitarian citizenship test need to relate to the identity of the state (community), a case could be made for the questions on cultural differences and those on rights, freedoms and duties, since these questions aim to stress Germany’s liberal identity. These questions constitute 13.33% of the total. A case could furthermore be made for the questions in the category ‘democracy’, 39 These 154 questions are the 115 questions on democracy, the 28 questions on rights, freedoms and duties, and eleven questions on dealing with cultural differences. 40 These 177 questions are the 70 questions dealing with history, one question on the German school system, 68 questions about democracy, 28 questions on rights, freedoms and duties and ten questions on dealing with cultural differences.

136  Chapter Six

which emphasise Germany’s democratic identity, and which make up another 38.33% of the total. As regards the questions focusing on general knowledge of Germany, a difference should be made between the questions regarding history, culture and Landeskunde (88 questions), and those on economy, geography, Germany and the EU and practical issues of everyday life (57 questions). The latter need to be disqualified from a communitarian perspective for lacking a link to the German identity: it is after all hard to establish a relationship between the German identity and knowing in which year the Euro was introduced as means of payment in German bars or what you need to do if you receive a false invoice from the German authorities. Questions on history, culture and Landeskunde on the other hand have a link with German identity: one could argue that, in order to share the identity of a community, a future member should be aware of its past, which allows the questions on history a place in a communitarian citizenship test. The questions on German culture merit testing as, in the communitarian model, it can be expected from those who intend to become members of the community to be aware of its customs and traditions. Lastly, the questions on Landeskunde can be defended in the communitarian model for relating to markers of the German identity. All in all, the majority (81%) of the questions in the Einbürgerungstest merit testing in the communitarian model. The German test can hence be best qualified as ‘communitarian’. 6.4. Qualifying the British ‘Life in the UK’ Test The ‘Life in the UK’ test is based on the text of the handbook ‘Life in the United Kingdom; A Journey to Citizenship’ (TSO 2007). The first version of the book consisted of eight chapters, of which only the chapters on ‘A Changing Society’ (Chapter 2), ‘Britain today: a profile’ (Chapter 3), and ‘How Britain is governed’ (Chapter 4) needed to be studied. Since 1 April 2007, on which date the knowledge requirements also became obligatory for settlement, a second version of the handbook was used as a basis for the questions in the test. Since then, the chapters on ‘everyday needs’ (Chapter 5) and ‘employment’ (Chapter 6) also require studying.41 Studying Chapter 1 on British history is not required because of the fear that this chapter would be “too contentious and too difficult” (Kiwan 2008: 68). 41 As of 25 March 2013, the test questions are based on a third version of the handbook entitled Life in the United Kingdom; a Guide for new Residents. The analysis of the content of the ‘Life in the UK’ test is based on the content of the second version of the handbook.

The Citizenship Tests Compared  137

In the table below, the topics addressed in the ‘Life in the UK’ test have been categorised. Since the test items contained in the test bank are, apart from the examples mentioned on the ‘Life in the UK’ test website and in the online sample test, not published, the number of paragraphs dedicated to a topic in the handbook will be taken as a starting point to determine the content of the ‘Life in the UK’ test. Each of the 318 paragraphs has been categorised under one theme only. Only the paragraphs in the chapters of the guidebook which are tested are included. The assumption underlying the analysis is that the percentage of questions in the test bank concerning a certain topic matches the percentage of paragraphs in the chapters which are tested.42 Table 6.3. Questions in the British ‘Life in the UK’ test along four thematic categories. Category

Number of paragraphs

Percentage of total of 318 paragraphs

Democracy Rights, freedoms and duties General knowledge of the UK Dealing with cultural differences

67 14 217 20

21% 4% 68% 6%

The above table shows that the bulk of all questions in the ‘Life in the UK’ test are likely to focus on topics relating to general knowledge of the UK (217 paragraphs). For the large part (177 paragraphs; 55% of the total number of paragraphs that require studying), questions categorised under ‘general knowledge of the UK’ concern ‘practical issues of everyday life’, including issues such as employment (“interviews for lower paid and local jobs can often be arranged by telephone or in person”), health (“healthcare in the UK is organised under the National Health Service (NHS)”), housing (“two-thirds of people in the UK own their own house”), education (“at age 11 (12 in Scotland) children go to secondary school”), and leisure (“films in the UK have a system to show if they are suitable for children”).43 The remaining 40 paragraphs on general knowledge of the UK concern customs  and traditions (ten paragraphs stating inter alia that “throughout the year there are festivals of art, music and culture, such as the Notting Hill Carnival 42 In her analysis of the British test, Michalowski has made use of a set of 97 official sample questions published by the Home Office (Michalowski 2011: 754). The basis of her analysis thereby hence differs from mine. 43 The majority of the questions on practical issues deal with employment (44 paragraphs), health and health care (32 paragraphs) and education (31 paragraphs).

138  Chapter Six

in west London and the Edinburgh Festival” and that “in Scotland Hogmanay is a bigger holiday for some people than Christmas”), the Church of England and the patron saints (three paragraphs saying for instance that “the church of England is called the Anglican church in other countries and the Episcopal Church in Scotland and in the USA”), children, family and young people (eighteen paragraphs; “in the UK, there are almost 15 million children and young people up to the age of 19”, “many parents worry that their children may misuse drugs and addictive substances”), the population (six paragraphs; “the population has grown by 7.7% since 1971, and growth has been faster in more recent years”), and the UK nations and regions (three paragraphs; “most of the population live in towns and cities”). Covering 21% of the paragraphs that require studying, the second largest topic concerns issues of ‘democracy’. These paragraphs are the paragraphs of Chapter 4 of the handbook entitled How the United Kingdom is governed, and the three paragraphs on young people’s political attitudes in Chapter 2 on ‘a changing society’. They explain the British polity and how democracy works in the UK, for instance by stating that “there must be a general election to elect MPs at least every five years, though they may be held sooner if the Prime Minister so decides” and that “the Whips are a small group of MPs appointed by their party leaders”. Six per cent of the paragraphs focus on the topic of ‘dealing with cultural differences’. These 20 paragraphs are the paragraphs on immigration (eight paragraphs stating for instance that “many people living in Britain today have their origins in other countries” and that “in the 1980s, the largest immigrant groups were from the United States, Australia, South Africa, and New Zealand”), ethnic diversity (five paragraphs; “people of Indian, Pakistani, Chinese, Black Caribbean, Black African, Bangladeshi and mixed ethnic descent make up 8.3% of the UK population”), religion (one paragraph containing a table representing the different religions practised in the UK and the percentages of believers), and equality between men and women (six paragraphs). The latter paragraphs can be found in the section on ‘the changing role of women’ in Chapter 2 (‘A Changing Society’) and inter alia state that “research shows that very few people today believe that women in Britain should stay at home and not go out to work” and that “there is evidence that there is now greater equality in homes and that more men are taking some responsibility for raising the family and doing housework”. Finally, about 4% of the paragraphs have been classified under ‘rights, freedoms and duties’. These fourteen paragraphs concern the prohibition of discrimination (thirteen paragraphs) and the freedom of religion (one paragraph stating that “even though the UK is historically a Christian society, everyone has the legal right to practise the religion of their choice”). The thirteen paragraphs on the prohibition of discrimination only concern the prohibition to be discriminated against, and not that the test candidate him- or herself is obliged to give others an

The Citizenship Tests Compared  139

equal treatment.44 In this sense, the British test, like the German test, hence differs from its Dutch counterpart. What does this categorisation say about the ‘Life in the UK’ test? 6.4.1. A Liberal Test? As becomes apparent from the examples given above, the content of the ‘Life in the UK’ test can be qualified as liberal for not aiming to scrutinise the true convictions of candidates. Questions that relate to what is considered good by the majority society are however included, an example being the paragraphs which state that “very few people believe that women should stay at home” and that “many parents are involved with their child’s school”. However, rather than forcing on acceptance of these social norms, the paragraphs merely aim to make sure that the test candidate knows about their existence. Hence, since the questions on social norms refrain from suggesting a “homogeneity of values, opinions or moral attitudes” (Michalowski 2011: 752) and therefore from pushing for the acceptance of those norms, these questions do not ask about ‘what is good’, and thereby pass the first requirement of the liberal justification test. Do the questions in the ‘Life in the UK’ test also meet the second requirement of the liberal justification test? Applying Hampshire’s relevancy test, the fourteen paragraphs qualified under the thematic category of ‘rights, freedoms and duties’ merit testing. This also applies to the six paragraphs on equality between men and women in the category ‘dealing with cultural differences’. Furthermore, for concerning knowledge which enables the candidate to participate politically and to engage in political dialogue, the 67 paragraphs on ‘democracy’ explaining the British polity and how democracy works in the UK merit testing.45 The bulk of the questions, 231 in total, should however be disqualified in Hampshire’s relevancy test for not testing subject-matter which is essential for effective citizenship. Does applying Orgad’s test lead to a less harsh judgment of the content of the  British test? Asking whether the content of the paragraphs on which the test  questions are based constitute ‘constitutional essentials’ leads us to the conclusion that a case could be made for some of the paragraphs of Chapter 4 on  How the UK is Governed, which has as subtitle “The British Constitution”.46 44 Twelve of the thirteen paragraphs on the prohibition of discrimination concern discrimination in the work place; one paragraph concerns the prohibition of discrimination by landlords. 45 Six paragraphs concern voting and voting rights and two paragraphs concern the right to stand for public office. 46 The UK has no written Constitution. In talking about ‘the British Constitution’, the formal institutions (parliament, the office of the Prime Minister, the judiciary, etc.), laws and conventions are implied.

140  Chapter Six

Ten paragraphs in this chapter explain the constitutional link between the Church of England and the state, provide information on the constitutional monarchy and the role of the Queen, and the British polity. A case could furthermore be made for the fourteen paragraphs qualified under the thematic category of ‘rights, freedoms and duties’ and the six paragraphs on equality between men and women in the category ‘dealing with cultural differences’. Lastly, one of the paragraphs in Chapter 3 (‘UK today: a profile’) equally merits testing for addressing the constitutional link between church and state in England. The bulk of paragraphs that require studying, i.e. the remaining 287 paragraphs, will however need to be disqualified for not relating to British constitutionalism in any way. This judgment seems to be shared by Orgad himself, who has criticised the British test for containing items which are not essential for settling in the UK (Orgad 2010a: 103). Whereas the liberal relevancy tests allow for the testing of some of the paragraphs that require studying, the large majority of the information that is tested on should be dismissed for either not forming part of British national constitutionalism (Orgad) or not being relevant for effective citizenship (Hampshire). Depending on which test is applied, either 90% (Orgad) or 73% (Hampshire) of the questions in the British ‘Life in the UK’ test do not merit testing and therefore do not deserve a place in a liberal citizenship test. 6.4.2. Republican Elements As shown in table 6.3, around 21% of all paragraphs in the handbook that require studying concern the issue of ‘democracy’ for focusing on life in a democracy. Their aim is to inform test candidates of the features of the British Constitution and democratic system. The paragraphs in Chapter 4 of the handbook entitled How the United Kingdom is governed, and the three paragraphs on young people’s political attitudes in Chapter 2 on ‘a changing society’ can therefore be classified under the republican model. Moral statements on the importance of being an active democratic citizen are however absent.47 6.4.3. Communitarian Elements We have already concluded that the ‘Life in the UK’ test refrains from forcing on a certain conception of ‘what is good’ by not scrutinising the adherence of the test candidate to liberal values. The aim of the test is hence not to impose on future 47 Regarding the right to vote, information is for instance given on who is entitled to vote, and how one can register, but statements regarding the importance of voting are not included.

The Citizenship Tests Compared  141

British citizens a liberal identity. Neither does the text of the handbook aim to force on any other aspects of British identity. As stated, the majority of the paragraphs (177 paragraphs, making up 55% of the total number of paragraphs which require studying) are about practical issues of everyday life, in which case a link to British identity is lacking. This also applies to 24 of the remaining 40 paragraphs on general knowledge of the UK: a link between British identity and the eighteen paragraphs on children, family and young people, which for instance state that “children in the UK do not play outside the home as much as they did in the past” and that “young people have different identities, interests and fashions to older people”, is hard to establish. This also applies to the six paragraphs on the population (“a census is a count of the whole population”). Sixteen of the paragraphs on general knowledge of the UK nevertheless appear to be defendable in the communitarian model. The paragraph on the constitutional link between church and state in England for instance relates to the UK’s identity. This equally applies to the other paragraph on the Church of England, as well as the one on the patron Saints, which inter alia states that “England, Scotland, Wales and Northern Ireland each have a national saint called a patron saint”. A case could furthermore be made for the three paragraphs on the UK nations and regions (“there are many variations in culture and language in the different parts of the United Kingdom”, “the English language has many accents and dialects”), and the ten paragraphs on customs and traditions, which, apart from referring to British holidays, also focus on Britain’s multicultural character by mentioning holidays and festivals which are not typically British, such as Eid ulFitr, Diwali and Hanukkah. The latter paragraphs thereby stress Britain’s multicultural identity. Next to the 16 paragraphs on ‘general knowledge of the UK’ referred to above, the 67 paragraphs on democracy can be defended in the communitarian model for focusing on Britain’s democratic character. The fourteen paragraphs on ‘rights, freedoms and duties’ can be justified in the communitarian model for emphasising Britain’s liberal character. What is left are the twenty paragraphs on ‘dealing with cultural differences’. Instead of prescribing immigrants to behave in an appropriate way, like most of the indicators on dealing with cultural differences in the Dutch test, the paragraphs that have been classified in this category in the British test amply refer to Britain’s multicultural society. Even though these paragraphs do not require future citizens to adopt and assimilate to the British culture, they relate to Britain’s identity, which, at least in the eyes of those who drafted the study guide, is multicultural. In total, 117 paragraphs, equalling 37% of the total number of paragraphs that require studying, qualify as communitarian.

142  Chapter Six

6.5. Comparing the Tests 6.5.1. What Is Tested? The table below presents a categorisation of the test questions along the four thematic categories. This categorisation will allow us to answer the question whether or to what extent the tests actually match the ideas, put forward in the political debates preceding their introduction, on what their content should be. As stated before, this question will not be answered here, but in the concluding chapter of this book (paragraph 11.2.1). For now, the categorisation along the thematic categories will provide further insight into the content of the tests compared, and offer help to answer the question of how the differences between the tests can be explained. Table 6.4. Qualification of tests along thematic categories. Category

The Netherlands

Germany

The UK

Democracy Rights, freedoms and duties General knowledge on the respective  country Dealing with cultural differences

7% 3% 62%

38.33% 9.33% 48.33%

21% 4% 68%

28%

4%

6%

The above table shows that, in all three countries under consideration, the tests for the largest part consist of questions concerning general knowledge of the respective country. In the Netherlands and the UK, most questions in this category are dedicated to practical issues of everyday life. In both countries, almost 60% of all test questions concern this topic.48 The German test, on the other hand, pays most attention to history: two thirds of the 105 paragraphs on general knowledge of Germany concern this issue, which amounts to almost a quarter of the total number of questions. The Dutch test pays much less attention to history, and in the British test, the topic of history is not considered at all.49 And whilst the British and German tests pay attention to culture, i.e. customs and traditions, this issue is absent from the Dutch test. 48 55% of the paragraphs (177 paragraphs) on which questions in the Life in the UK test are based concern practical issues of everyday life, compared to 58% of the indicators (116 indicators) in the curriculum for the Dutch test. 49 Four of 200 indicators in the Dutch curriculum concern history. The Life in the UK handbook pays attention to the subject, but it is not tested on.

The Citizenship Tests Compared  143

The second largest topic considered in the British and German tests is ‘democracy’. This topic however comes in third place in the Dutch test, where almost a third of all questions are dedicated to the issue of ‘dealing with cultural differences’. The British and German tests pay much less attention to this topic. Lastly, in the Netherlands and the UK, least attention is paid to the issue of ‘rights, freedoms and duties’, whereas this issue occupies third place in the German test. How can the content of the tests be explained? As for the German questionnaire, part of the explanation lies in the fact that the federal test needed to clearly distinguish itself from its regional predecessors, i.e. the interview guidelines applied in Baden-Württemberg and the Wissen und Werte test proposed by Hesse (Michalowski 2011: 764). The test content needed to be as uncontroversial as possible, and the government promised that it would remain outside the scope of every criticism.50 An explanation for the focus on practicalities in the Dutch and British tests could lie in the fact that these tests are not only used in the process of naturalisation, but also for permanent residence, and, in the Dutch case, for newcomers with an integration obligation under the Integration Act (Michalowski 2011: 763). Furthermore, in the case of the UK, an explanation for the focus on practicalities lies in the fact that the advisory group that drafted the handbook never intended this book to function as the basis or study material for the test, and specifically also thought of newcomers, who will not be studying the handbook to pass the test to be able to apply for ILR or naturalisation, when drafting its content (see paragraph 5.1.2). Lastly, the focus on practicalities might be explained by the wish for the test to not be too contentious or difficult. The reason why more than one fourth of all questions in the Dutch test relate to ‘dealing with cultural differences’ is probably the influence exercised by Minister Verdonk on the eventual test content. Many of these questions concern social norms emanating from liberal values, supposedly not adhered to by the test candidates, and push the test candidates to accept these norms, or at least to not openly come out for the fact that they are not adhered to. As we have seen, by aiming to exclude people on the basis of their ideas, these questions fall into the trap of ‘the paradox of liberal toleration’. Questions like those asked in the Netherlands are lacking in the British and German tests, where all questions on liberal values merely aim to confer factual knowledge of these values, informing the test candidates of their existence. The British and German tests however seem to fall into another trap, namely that of what Joppke has referred to as the ‘paradox of universalism’: while aiming to make sure that ethnic minorities become a part of these, i.e. the British and German communities, and not any political community, the tests, by referring to general liberal values, fail to succeed in naming 50 Plepro 16/168, p. 17796 (C).

144  Chapter Six

any particulars that distinguish the ‘here’ from the ‘there’ (Joppke 2010b: 130). The Dutch test, with its illiberal questions, to a certain extent manages to avoid this trap. 6.5.2. How Can the Tests Be Qualified? The table below presents the content of the tests along the three citizenship models. Of the tests under consideration, the Dutch test contains the smallest amount of liberal questions: slightly under or slightly over 10% of questions fulfil both requirements of the liberal justification test, depending on whether Hampshire’s or Orgad’s relevancy test is applied. Part of the explanation lies in the fact that the Dutch test has the highest share of questions which do not fulfil the first requirement in the liberal justification test for relating to a ‘what is good’. For this reason, a total of 55 indicators, equalling almost 30% of the total, have been disqualified. In comparison, only two of the questions in the German test have been disqualified for relating to a ‘what is good’, compared to none of the paragraphs on which the questions in the ‘Life in the UK’ test are based. Table 6.5. Qualification of tests along citizenship models. Liberal (Orgad) The 9%  Netherlands Germany 60% UK 10%

Liberal (Hampshire)

Republican

Communitarian

11%

9%

56%

52% 27%

40% 21%

81% 37%

The German test by contrast contains the highest share of liberal questions: 60% when Orgad’s test is applied, and 52% in Hampshire’s test. Nevertheless, rather than qualifying the German test as liberal, the qualification ‘communitarian’ suits this test better, since more than 80% of all test questions can be defended under the communitarian model. With 56% of the questions qualified as such, the label ‘communitarian’ also fits the Dutch test best. However, for a test to convincingly qualify as communitarian, more is required than a small majority of its questions being labelled accordingly. This also applies to the British test, which, even more than is the case for the Dutch test, could not be convincingly justified in any of the models either. The above conclusion could imply that the models are not fit to be used as instruments to qualify the content of citizenship tests. It could however also

The Citizenship Tests Compared  145

mean that (parts of) the tests which are actually employed by the Netherlands, the UK and Germany are inappropriate means to test whether immigrants can be good citizens in either the liberal, republican or communitarian sense. Instead, they are testing issues which do not belong in citizenship tests, no matter whether these are liberal, republican or communitarian.51 This conclusion questions the general suitability of the tests as instruments to assess whether or not an outsider can be granted access to the community of citizens.

51 This applies to a lesser extent to the German test, where most paragraphs have been qualified under one of the models, and to a larger extent to the British test, where many paragraphs could not be qualified under any of the models.

Chapter Seven The Quantitative Impact of the Citizenship Tests in the Netherlands, Germany and the UK As we have seen in chapters 3, 4 and 5, in all the countries under consideration, the governments that argued for the introduction of citizenship testing have, at least prior to their introduction, always contended that the tests were not to become hurdles, functioning as a means to lower the number of naturalisations. Furthermore, as we have seen in Chapter 2, it has been argued that, rather than being an exclusive phenomenon, citizenship tests are “far more inclusive than alternative procedures involving subjective criteria”, such as interviews conducted by local officials (Klekowski von Koppenfels 2010: 11). It has moreover been claimed that citizenship tests actually increase “the naturalisation procedure’s calculability on the part of citizenship applicants, who are no longer subject to open-ended, individual interview procedure and the state-agent’s unfathomable discretion” (Joppke 2010b: 4). This line of reasoning appears to be based on the presumption that discretion leads to arbitrariness, and that arbitrariness will lead to a stricter policy, and a higher number of refusals. Following this line of reasoning would suggest that in the countries under consideration, the number of naturalisations were lower prior to the introduction of formalised language and knowledge of society tests, when integration was tested by local officials or legal practitioners in interviews, than afterwards.1 Whether the citizenship tests indeed lead to more openness, or whether their introduction rather restricts access to citizenship, will be examined in this chapter via an analysis of statistics relating to both the number of naturalisations and the tests. As there will be other reasons beside the introduction of new requirements that will affect the numbers of naturalisations, alternative explanations for changes in numbers of naturalisations will also be mentioned. But because the analysis focuses on the effect of the introduction of citizenship tests, I have not conducted a systematic analysis of all other possible (demographic, administrative or political) explanations for changes in the number of naturalisations. 1 In the Netherlands and Germany, prior to the formalisation of the language and integration check, language skills were assessed by municipal officials. In the UK, language skills were initially tested by police officers. After a change in the procedure, applicants could go to a language school to take a language test, or have a legal practitioner vouch for their language skills.

148  Chapter Seven

Besides analysing the statistics relating to the numbers of naturalisations and the tests, this chapter will assess whether the tests have produced a different effect depending on age, gender, level of education and nationality, as was suggested in parliamentary debates.2 It will also be examined whether the exemption regulations provided for sufficiently prevent the knowledge requirements from forming barriers to acquiring citizenship for the mentally and physically handicapped.3 An analysis of the statistics will provide a good indication of the effects the new requirements for citizenship have produced, but can only partially reveal information about the ‘dark numbers’, i.e. those who do not or only to a certain extent show in the statistics because the tests made them give up their wish to become naturalised. Insight into the dark numbers will be provided in Chapters 8, 9 and 10, which analyse interviews with immigrants and stakeholders. 7.1. Lower Number of Naturalisations? This paragraph will analyse the number of naturalisations prior to and after the introduction of the tests, as well as numbers related to the citizenship tests. 7.1.1. The Absolute Number of Naturalisations The absolute numbers of naturalisations in the three countries under consideration are depicted below. Formalised language and integration requirements were introduced in 2003 (Netherlands), 2004 (UK), and 2007 (Germany). The figure above shows that in the Netherlands, the number of naturalisations in 2004, a year after the introduction of the naturalisation test, reduced by 50% compared to 2002, the year prior to the introduction of the test. In Germany, the numbers have more or less continuously decreased since the introduction of the 2000 Act. The numbers rose only in the UK. For the latter country, an image of the naturalisation rates, given below (figure 7.2), however shows a completely different picture, as these rates started to decline as of 2006. This could signal that

2 In the Netherlands, the low educated, women, people originating from countries using a different alphabet, and the elderly were singled out, whereas in Germany, only women (spouses) and the elderly received attention. In the UK, neither of these groups was explicitly singled out as categories for whom a formalised language requirement and a knowledge of society test would form barriers. Negative effects a formalised test might produce for “some people” were however referred to (MP Hughes (LD), House of Commons Standing Committee E, 30 April 2002). 3 In the Netherlands, illiterates and the handicapped were mentioned as groups for whom acquiring written language skills could pose problems.

Impact of the Citizenship Tests  149

200000 150000 100000 50000 0

1994

1996

1998 UK

2000

2002 Germany

2004

2006

2008

2010

2012

Netherlands

Sources: Netherlands: IND, Germany: EUDO citizenship statistics database (eudo-citizenship.eu), Statistisches Bundesamt, Fachserie 1 Reihe 2.1 (Bevölkerung under Erwerbstätigkeit – Einbürger­ ungen), several volumes, BAMF Working Paper 17, UK: Home Office Statistics.

Figure 7.1. Naturalisations in the Netherlands4 and Germany5 (1994–2011), and the UK (1994–2012). the effect of the formalisation of the language test and the introduction of the ‘knowledge of life’ requirement has been disguised by the number of immigrants present in the UK. The Netherlands An analysis of the absolute number of naturalisations in the Netherlands shows that, in 2004, the first full year the revised Act was in force (15,000 naturalisations), the number of naturalisations almost halved compared to 2002, the year prior to its entry into force (29,000 naturalisations). The decrease in the number 4 Only naturalisations of adults are depicted. The reason is that, prior to 2003, naturalisations of minors were separately reported, whilst they have been included in the reported numbers since 2003. When this difference in reporting is not taken into account, the entry into force of the revised DNA appears to have had only a modest effect on the number of naturalisations. The effect of the revision of the Act however appears to be a lot bigger if, as is shown in figure 7.1.1, only naturalisations of adults are taken into account. 5 For Germany, naturalisations of Aussiedler based on Article 6 of the Gesetz zur regelung von Fragen der Staatsangehörigkeit (StAnRegG, ‘Statusdeutsche’) are excluded. Account has been taken of naturalisations based on the following Articles: 85 and 85 Foreigners Act (old wording, applicable until 2000), 85 Foreigners Act (applicable until 2004), Articles 8 and 9 RuStAG, and Articles 8, 9 and 10(1), (2) and (3) and 40b Nationality Act, as well as ‘other’ naturalisations based on Articles 13 and 14 Nationality Act, the Gesetz zur Regelung von Fragen der Staatsangehörigkeit (StAnRegG), the Gesetz zur Verminderung von Staatenlosigkeit and the Gesetz über die Rechtsstellung heimatloser Ausländer im Bundesgebiet.

150  Chapter Seven

of naturalisations in the years 2003 and 2004 would have been much bigger if the majority of applications which were dealt with in those years had not concerned applications which had been filed prior to the coming into force of the revised Dutch Nationality Act (DNA) (IND 2009: 25). Even in 2005 many applications were dealt with that had been filed prior to the changes (IND 2011: 22). The extra high number of applications filed in 2002 and the first quarter of 2003 indicates that immigrants were anticipating the stricter provisions that were going to be introduced by filing an application before the coming into force of the revised DNA (Böcker, Groenendijk, De Hart 2005: 164). This could signal that both immigrants and their advisors considered the provisions in the new Act to be a barrier to acquiring nationality. This assumption is confirmed by the durable character of the decrease in the numbers: in 2010 (14,000 naturalisations), i.e. seven years after the introduction of the revised Act, the number of naturalisations was still 50% lower than in 2002. From figure 7.1 it appears that prior to the coming into force of the revised DNA the number of naturalisations had already decreased. What does this say about the negative effect of the revised DNA on the number of naturalisations? Is it less severe than assumed above? Probably not, since the decrease in the number of naturalisations prior to the introduction of the revised DNA has never been as high as after its entry into force. Moreover, the number of naturalisations after the introduction of the revised Act remained at the lower level permanently, also where the naturalisation rates, as depicted in figure 7.2 below, are concerned. This is a strong indication for the new naturalisation requirements introduced by the revised DNA constituting the main reason behind the decrease in the number of naturalisations from 2003. What is questionable is to what extent the naturalisation test has been responsible for the decrease. After all, the revised DNA also strengthened the residence requirement.6 Nevertheless, the period of residence required remained the same: five years. Furthermore, changes in the period of residence required will only temporarily affect the number of naturalisations, as with time, most naturalisation applicants will again automatically fulfil the requirement. As the decrease in the number of naturalisations in the Netherlands has a durable character, it is more probable that the naturalisation test and its successor, the integration examination, constitute the main reason behind the decrease. To get a more precise idea of the impact of the introduction of the formalised language and integration requirement on the number of naturalisations, numbers regarding the naturalisation test and the integration examination will be examined below (paragraphs 7.1.3. and 7.2). 6 Rather than ordinary residence, as of 1 April 2003 regular and continuous residence is required.

Impact of the Citizenship Tests  151

Germany In Germany, the number of naturalisations consistently rose until the year 2000, the year the reformed Nationality Act entered into force, when almost 189,000 persons naturalised. However, after that peak, the number of naturalisations decreased considerably. How can this evolution in the number of naturalisations since 1994 be explained? As for the rise until 2000, an explanation can be provided by the fact that, in 1990, a simplified naturalisation procedure for immigrants aged between 16 and 23 (Article 85 old) and for first-generation immigrants residing in Germany for a period of fifteen years or longer (Article 86 old) was introduced (paragraph 4.1).7 In 2000, new changes to German nationality law were introduced with a view to further liberalising access to German nationality.8 Immediately after the coming into force of the new Act, the liberalisation appeared to be successful: compared to 1999, the number of naturalisations in 2000 rose by 30%. As it turned out, this rise in the number of naturalisations however appeared to be only temporary: starting in 2001, the number of naturalisations, contrary to the goal of the reform and to expectations, decreased. A first explanation for the decrease since 2001 is the fact that the reduction in the residence requirement appears to have had only a limited effect: statistics show that the majority of the non-German population fulfils the residence requirement.9 Another explanation for the decrease since 2001 is that the 2000 law added new requirements, such as the language requirement and the declaration of loyalty, and reinforced existing requirements, such as the income requirement.10 Furthermore, the fees for naturalisation increased fivefold.11 Lastly, the 7 In 1993, naturalisation on the basis of these articles became a right. 8 See paragraph 4.2 for a detailed explanation of the liberalisations that took place in 2000. The most important liberalisation probably is the introduction of ius soli nationality acquisition for certain children born on German soil. For a limited period of one year, a possibility for children born before the coming into force of the new law, but who had not yet obtained the age of ten, to obtain German nationality was introduced provided for (Article 40b Staatsangehörigkeitsgesetz). In 2000, 40,000 children acquired German nationality this way, a number which quickly decreased afterwards. These naturalisations are also taken into account in figure 7.1. Acquisitions via ius soli (Article 4 German Nationality Act) are not. 9 At the end of 2008 almost three-quarters (72.2%) of Germany’s immigrant population fulfilled the residence requirement (Rühl 2009: 38). Furthermore, at the end of 2007, as many as 64% of all foreigners present in Germany resided there for a period exceeding eight years based on a residence permit not excluded by Article 10(1)(2) (Central Foreigners’ Register, quoted in BT-Drs. 16/13558, p. 12). 10 The new law required applicants for naturalisation to be employed both at the time of application and at the time of approval of the application. 11 Whereas applicants were required to pay 100 DM for their naturalisation application before 1 January 2000, after this date 500 DM were payable. Currently, immigrants pay € 255 for their naturalisation application.

152  Chapter Seven

so-called Inlandklausel, which allowed immigrants to re-apply for their former nationality after obtaining German citizenship, provided they had their habitual residence in Germany, was abolished (De Hart & Groenendijk 2007: 100).12 This amendment apparently led many of the immigrants who were required to renounce their original nationalities to decide not to apply for German citizenship, among whom were many Turkish nationals, whose naturalisation rate significantly decreased following the 2000 reform (see figure 7.6 below). The fact that the 2000 reform has not been successful for Germany’s largest immigrant group is one of the main explanations for the low number of naturalisations in total (Tränhardt 2008: 12).13 Seen in this light, besides explaining the decrease following the entry into force of the new Act, the introduction of the restrictive changes also offers an explanation for the peak in the number of naturalisations in 2000: in order to anticipate the changes made by the 2000 reform, immigrants filed their applications for naturalisation early, i.e. before 1 January 2000. Many of these applications would ultimately be dealt with or registered in 2000.14 Hence, as was the case for the Netherlands, this peak probably indicates that both immigrants and their advisors perceived the provisions of the new Act as barriers to naturalisation. And, as in the Netherlands, this assumption is confirmed by the fact that the decrease in the number of naturalisations has a permanent character: the numbers more or less continuously decreased until 2008, after which a modest rise occurred. Notwithstanding the decrease in the number of naturalisations since 2001, the 2007 Act further reinforced the requirements for naturalisation. In August 2007, the language requirement was formalised, meaning that proof of oral and written language skills at level B1 needed to be submitted. The Act furthermore introduced a knowledge of society test (Einbürgerungstest; applicable since 1 September 2008), strengthened the public order requirement, and abolished the rule that exempted immigrants aged between 16 and 23 from the income requirement. After a short-lived rise in the numbers of naturalisations in 2006, which might again be explained by immigrants filing their naturalisation applications early to avoid the new restrictions, the decreasing trend in the number of 12 As of 1 January 2000, the German Nationality Act provides that if naturalisation and reacquisition took place after the coming into force of the new law, German nationality is lost automatically (Article 25(1) Staatsangehörigkeitsgesetz). German nationality is also lost if naturalisation took place before 1 January 2000, but the foreign nationality was re-acquired after this date. This new practice has led to the automatic loss of German citizenship for 50,000 immigrants. 13 Alex Street mentions the introduction of birthright citizenship, which removed the intergenerational motive for naturalisation for many parents, as an alternative reason for the decline in German naturalisation rates after the 2000 reform (Street 2012: 21). 14 Federal statistics regarding the numbers of applications for naturalisation are not available.

Impact of the Citizenship Tests  153

naturalisations, unsurprisingly, continued in the years 2007 and 2008. In the years 2008 and 2009, for the first time since 1998, the numbers of naturalisations was below 100,000. In 2009, 2010 and 2011 the number of naturalisations however rose compared to the year before. In 2010, the number of naturalisations again passed 100,000. This rise could signal that the negative effects of the language and knowledge of society tests are starting to wear off. Naturalisations in the German Länder As we have seen in Chapter 4, the differences in naturalisation rates between the Länder constituted an important argument for amending the requirements for naturalisation: more uniformity in the application of the naturalisation requirements would reduce differences and prevent naturalisation tourism. To what extent has the introduction of the 2007 Act led to a more uniform interpretation of the requirements? Looking at the naturalisation percentages of the Länder prior to and after the introduction of the Act can provide insight into the answer to this question.15 In 2006, the naturalisation percentage was highest in Schleswig Holstein (3.18%) and lowest in Brandenburg (0.69%). After the introduction of the changes made by the 2007 Act, considerable differences however still exist: in 2011, the naturalisation rate of the Land with the highest naturalisation rate (2.4%; Hamburg) was more than twice as high as the rate in the Land with the lowest percentage (Sachsony; 1.02%). Judging from this difference, one of the central aims of the 2007 reform appears to have been only partially achieved. At the same time, however, the differences have become smaller: from a difference of 2.5 percentage points in 2006, to 1.4 percentage points in 2011. This diminishing difference can be explained by the fact that the 2007 reform has led to a general decrease in the number of naturalisations, which comes as no surprise given that the Länder that used to apply open naturalisation policies had to adapt their practices to the Länder that employed a restrictive policy. The United Kingdom From 1997 to 2008, the number of acquisitions of British citizenship via naturalisation almost constantly increased, notwithstanding the entry into force as of 2002 of stricter requirements: a citizenship ceremony was introduced, and, more importantly, the language requirement was formalised (July 2004), and a ‘knowledge of life in the UK’ requirement introduced (November 2005). Moreover, as of 2005, the fees were raised considerably on various occasions, something we also 15 The naturalisation percentages have been calculated on the basis of information obtained from Federal Statistical Office, Fachserie 1, Reihe 2 (Ausländische Bevölkerung), several volumes and Fachserie 1, Reihe 2.1 (Bevölkerung under Erwerbstätigkeit – Einbürgerungen), several volumes.

154  Chapter Seven

saw in Germany.16 These changes at first sight do not appear to have negatively affected the number of successful naturalisation applications. On the contrary: after the formalisation of the language requirement and the introduction of the ‘knowledge of life in the UK’ requirement, the numbers of naturalisations have only increased. In the year 2009, a peak was reached.17 The rise in the number of naturalisations would nevertheless have been lower if fewer applications had been filed in 2005, when 64% more applications were filed than in 2004 (Home Office 2006: 1). This rise has been ascribed to the introduction of the requirement to demonstrate knowledge of life and language, which took full effect in November 2005. Hence, as in the Netherlands and Germany, in the UK the knowledge requirements have exercised influence on the number of naturalisations: the rise in the number of applications prior to the introduction of the knowledge requirements shows that immigrants perceived these requirements as barriers to naturalisation. This barrier however does not appear to have negatively affected the absolute number of naturalisations, considering the more or less continuous rise in these numbers after 2005 and the peak of 2009. Whether the strengthened requirements have indeed not negatively affected the number of naturalisations, or whether this influence has been disguised by a rise in immigration and the number of potential applicants for naturalisation, will be analysed in the following paragraph, which discusses the relative number of naturalisations. 7.1.2. The Relative Number of Naturalisations Figure 7.2 below shows that the naturalisation rate (the number of naturalisations as a percentage of the total immigrant population) in Germany is (much) lower than the rates in the Netherlands and the UK. In the UK, the rate in 2009 (3.5%) was more than twice as high as the German rate in that year (1.4%). The difference became smaller in the years 2010 and 2011, even though the British rate in those years still was (almost) twice as high as the German rate. With a difference of 0.9 percentage points, the difference was however much smaller in 2008. Also compared to the Netherlands the German rate was half as low, at least in 2008 and 2009: whereas the Dutch rate in those years lay around 3%, the German rate was about 1.5%. In the years 2010 and 2011 the difference in the Dutch and German naturalisation rates however decreased. In those years, the Dutch rate was about 1 percentage point higher than the German rate. In all three countries 16 The fees were raised from £ 200 in 2005 to £ 575 in 2007. They then rose to £ 655 in 2008, £ 720 in 2009, £ 735 in April 2010, and £ 780 in October 2010 (Ryan 2010: 18, 34). Since April 2011, a single naturalisation application costs £ 836. 17 That is to say, a peak was reached in the period depicted in figure 7.1.

Impact of the Citizenship Tests  155

under consideration, the rates have been decreasing, albeit that the relative decrease is not as sharp as the absolute decrease. Figure 7.2 shows more or less comparable patterns as depicted in figure 7.1, at least where Germany and the Netherlands are concerned.18 For both countries the changes in the number of non-citizen immigrants present in the country hence cannot be blamed for the decrease in the absolute number of naturalisations, which therefore is apparently related to changes in the naturalisation requirements. As was the case for the absolute number of naturalisations, the relative number of naturalisations in the Netherlands was already decreasing prior to the restrictive changes made to the naturalisation requirements by the revised DNA. The decrease in the relative number of naturalisations since 2003 however again appears to be sharper than the decrease which took place prior to this year, and to have a durable character, indicating that changes made to the DNA are responsible for the decrease.19 Contrary to the Netherlands and Germany, the pattern followed by the absolute number of naturalisations in the UK is significantly different from the pattern of the relative number of naturalisations: whilst the absolute number of naturalisations shows a more or less continuous rise, the relative number of naturalisations has been decreasing since 2005, the year the ‘knowledge of life in the UK’ requirement entered into force, until 2009. In 2008, the general naturalisation percentage reached its lowest point since the year 2000. In that year, only 2.3% of Britain’s foreign population acquired British citizenship via naturalisation, compared to 3.7% in 2004 and 2005. This indicates that in the UK, the changes made to British nationality law since 2004, the introduction of the knowledge requirements being among the most significant, have possibly indeed had a negative effect on the number of naturalisations. The decrease in the relative number of naturalisations is probably also caused by the gradual but significant rise in the fees payable for naturalisation: whereas immigrants paid £ 200 in 2005, an

18 In Germany, the relative number of naturalisations was highest in 2000 (2.6%), after which it decreased until the years 2010 and 2011, when the rate slightly increased (from 1.44% in 2009 to 1.5% in 2010 and 1.54% in 2011). In the Netherlands, the naturalisation rate in 2003 (3.5%), the year the naturalisation test was introduced, was half as low as the rate in the year 2000 (7%), and continued to decrease until 2007 (3.3%), in which year the naturalisation percentage slightly increased compared to the year before (2006, 3%). The naturalisation rate in the years following 2007 however decreased again, until 2011, in which year the rate rose 0.2 percentage points compared to the years before. 19 Whereas the naturalisation percentage in 2002 (6.1%) was only 0.3% lower than the percentage in 2001 (6.4%), the percentage in 2004 (2.9%) more than halved compared to 2002. In 2009 (3.1%), the naturalisation percentage was still half as low as the percentage prior to the introduction of the revised DNA in 2003 (6.1% in 2002).

156  Chapter Seven 8% 7% 6% 5% 4% 3% 2% 1% 0% 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 Netherlands

Germany

UK

Sources: CBS, Federal Statistical Office, Home Office Statistics, EUROSTAT.20

Figure 7.2. Naturalisation percentages the Netherlands, Germany and the UK (2000–2011). application for naturalisation cost £ 836 in April 2011.21 An alternative explanation for the decrease in the relative number of naturalisations is however offered by the significant rise of immigration to the UK from new EU Member States after 2004. As these immigrants either have not been residing in the UK long enough to be eligible for naturalisation, or are not interested in UK citizenship because they are already EU citizens, this immigration has inflated the denominator of naturalisation rates.22 The year 2009 shows a remarkable recovery in the naturalisation rate. In that year, the naturalisation percentage reached the level of 3.6%, a rise of 1.3 percentage points compared to the year before. The rate however decreased again in 2010 and 2011, which shows that the rise in the naturalisation percentage was only temporary and the result of other variables, instead of a sign that the possible restrictive effect of the knowledge requirements and the fees started to wear off.

20 The German naturalisation percentage is based on the number of foreigners present in Germany according to the Ausländerzentralregister. 21 http://www.ukba.homeoffice.gov.uk/sitecontent/documents/aboutus/fees-table-spring-2011.pdf, site accessed 8 July 2011. The joint fee for spouses or civil partners who naturalise together in 2011 was £ 1,294. 22 I would like to thank Professor Rainer Bauböck for pointing this out to me.

Impact of the Citizenship Tests  157

To see to what extent the introduction of the ‘Life in the UK’ test and ESOL courses have been responsible for the decrease in the number of naturalisations, numbers relating to the course and test will be examined below (paragraph 7.1.3 and 7.2). 7.1.3. The Relative Influence of the Tests on the Number of Naturalisations As we have seen above, the (relative) number of naturalisations in the three countries under consideration decreased following the changes which were made to the naturalisation policy. This paragraph will analyse to what extent this decrease can be explained by the introduction of citizenship tests, or whether other variables should be held mainly responsible. In order to answer that question, for all three countries, pass rates in the language and knowledge of society tests will be analysed. Furthermore, for the Netherlands and the UK, we will look at the number of refusals per refusal ground. This information is not available for Germany. 7.1.3.1. The Netherlands Refusals As a rise in the number of refused naturalisation applications may explain the decrease in the number of naturalisations, it is useful to examine the refusals. Knowing how many applications were refused on which grounds however does not tell us to what extent these grounds actually constitute barriers to acquiring nationality, as the selection will generally take place before an application is made. In the years after the entry into force of the revised DNA, the relative number of denied applications rose. The percentage of denials was highest in 2004, when 17% of all applications filed were turned down.23 After 2004, the relative number of refusals started to decrease. In 2005, 11.6% of all applications were turned down, and in 2010, the relative number of denied applications appears to have returned to its pre-2003 level: with 6% of all applications denied, the level of denials is almost similar to the level in 2001 (5.7%).24 The revision of the DNA has hence only temporarily, i.e. in the years 2004 to 2009, led to an increase in the number of denied naturalisation applications. A rise in the number of refusals can hence not explain the decrease in the number of naturalisations post 2005. The temporary rise in the number of refusals is probably due to the lack of clarity or unfamiliarity with the new or revised requirements following the coming into force of the new Act. 23 In comparison, in 2002, 4.3% of all applications was turned down. Source: IND. 24 Source: IND 2007 and IND 2012.

158  Chapter Seven

The Tests Information on the number of immigrants who failed the test will provide insight into the extent to which the tests form a barrier to making an application for naturalisation. We will therefore look at the number of participants and pass rates in both the naturalisation test, which applied until 1 April 2007, and the integration examination, which has applied since. The Naturalisation Test In the period the naturalisation test was in force (April 2003 - April 2007), a total of 23,700 persons registered for the test. 14,300 candidates, i.e. 60%, made it through the complete test.25 Part of the explanation for the modest pass rate lies in the fact that not all persons who signed up for the test took part in it: of all those who registered, 81% actually participated in the test. This could be explained by the costs which were payable (€ 260), and by the absence of possibilities for preparation. In total, 19,314 persons participated in the test, of which 74% passed. The Integration Examination On 1 January 2007, the integration examination entered into force. Since then, passing this test has been obligatory for those with an integration obligation under the Integration Act. The new examination replaced the naturalisation test on 1 April 2007. Furthermore, since 1 January 2010, passing the test has been a condition for obtaining a permanent or independent residence permit. The level of the integration examination is similar to the level of the naturalisation test, namely A2 of the CEF, but the new examination is more extensive, and can be (much) more expensive (see paragraph 3.2). Despite its more demanding content, the pass rate in the integration examination from 2007 to the first half of 2011 is (almost) similar to that in the naturalisation test: 73%.26 Over the years, the pass rate has however decreased. Whereas the pass rate was 79% in the years 2007 to 2009, it dropped to 69% in 2011 and 65% in 2012.27 As an explanation for 25 Numbers obtained from IND. For more information regarding the naturalisation test, see paragraph 3.2. Possibilities to retake parts of the naturalisation test were offered until June 2007. 26 Sources: Significant 2010: 128, TK 2011–2012, 33000 VII, no. 2, p. 38 and http://www.rijksoverheid .nl/onderwerpen/inburgering/inburgering-in-cijfers/voortgang-inburgering/slagingspercentage inburgeringsexamen, site accessed 11 July 2011. 27 Sources: Significant 2010: 128, http://www.rijksoverheid.nl/onderwerpen/inburgering/inburger ing-in-cijfers/voortgang-inburgering/slagingspercentage-inburgeringsexamen (site accessed in 2011), an email provided to the author by Somajeh Gaeminia, freelance journalist, who was given the information on the number of candidates and persons who passed the examination in 2011 by a representative of the Ministry of Social Affairs and Employment (email dated 15 April 2013), and Voortgang Inburgering 2012, tot en met juli (https://abonneren.rijksoverheid.nl/media/00/66/040 531/1917/20120828_voortgang_inburgering_2012_tm_juli_def.pdf, site accessed 30 May 2013).

Impact of the Citizenship Tests  159

this decrease, respondents interviewed within the framework of the official evaluation of the Integration Act stated that in the years 2007 and 2008 more motivated or longer residing immigrants took the examination, the passing of which for them was relatively easy (Significant 2010).28 Among these candidates were many former asylum seekers whose status was regularised in 2007/2008 after they had been in the Netherlands for at least seven years. Many of these asylum seekers, 30,000 in total, had been offered integration courses shortly after their regularisation, and had been participating in Dutch society as irregular migrants after they had left the asylum seekers centres (Groenendijk 2011b: 344).29 Research showed that 7.950 of these former asylum seekers had passed the integration examination on 28 May 2012 (IND 2012b: 11). They have hence caused a one-time boost in the number of successful test candidates.30 It is expected that the pass rate will decrease further in the years to come when less motivated and less educated immigrants, who will need more time to complete a course in which they prepare for the examination, will be required to take part in the examination (Significant 2010: 129). The official policy evaluation of 2010 showed that 100,000 persons who between 2007 and 2009 had followed a course had not yet taken part in the examination (Significant 2010: 201). So far, the selection has hence been between the immigrants who were convinced of their chances of passing the examination and those who were not (Groenendijk 2011b: 344). As for the absolute number of successful candidates in the integration examination, this number is higher than that in the naturalisation test: 89,330 persons passed the integration examination between 2007 and the first seven months of 2012, compared to 14,300 passes in the naturalisation test (April 2003-June 2007).31 How can this difference be explained? First, the scope of potential candidates has widened: not only is passing the examination a condition for naturalisation, it is 28 Respondents were municipal officials and staff members in education institutions offering courses. 29 The official evaluation of the Integration Act mentions that 9,900 regularised immigrants were offered funding to follow an integration course, including an attempt at the integration examination (Significant 2010: 118). This number is probably higher, since some municipalities did not register the number of regularised asylum seekers correctly. 30 The requirement to present an authenticated birth certificate and a valid passport has for many regularised asylum seekers however proved to be a barrier to making a naturalisation application (TK 2011–2012, 19637, no. 1480). A ‘boost’ in the number of (applications for) naturalisation has hence not occurred (yet). 31 Information provided by IND and obtained from Significant 2010, TK 2011–2012, 33000 VII, no. 2, p. 38, an email provided by Somajeh Gaeminia, dd. 15 April 2013, on file with the author, and Voortgang Inburgering 2012, tot en met juli (https://abonneren.rijksoverheid.nl/media/00/66/040 531/1917/20120828_voortgang_inburgering_2012_tm_juli_def.pdf, site accessed 30 May 2013).

160  Chapter Seven

also obligatory for those with an integration obligation under the Integration Act, and, since 1 January 2010, required for permanent residence.32 The number of persons who have taken the test in order to obtain permanent residence will however be relatively low, since few people will have realised prior to 1 January 2010 that passing the test was a requirement for permanent residence. Furthermore, those obliged to pass the examination by virtue of their integration obligation will probably be inclined to skip the permanent residence stage and file for naturalisation when they become eligible, considering the benefits attached to holding Dutch nationality.33 Add to this the fact that the government, in a letter accompanying the integration certificate, informs those who have passed the test that they can apply for a Dutch passport, the centre-right coalition government’s announcement that the requirements for naturalisation will be raised, and skipping the permanent residence stage in favour of naturalisation seems rather self-evident.34 Another explanation for the higher absolute number of candidates in the integration examination than the naturalisation test could be the fact that funded possibilities to prepare for the examination have been offered. No such possibilities existed for the naturalisation test. The centre-right coalition Government led by Prime Minister Rutte (Conservative Liberals) has however decided to reduce the integration budget to zero in 2014, which means that immigrants, from 1 January 2013, are (again) required to pay for their own integration, which implies that the number of course participants and examination candidates will probably decrease.35

32 In the years 2007 to 2009, 127,000 persons were given notice that they would have to pass the integration examination within the next 3.5 years (Significant 2010: 26). In the first year the Act was in force, the number of immigrants who followed a course however remained low, because of the fact that the government did not pay for the courses and the tests. 33 This assumption is confirmed by two qualitative studies into the effects of the Integration Act, which pointed out that the Dutch passport is the main reason for candidates to take the examination (Strik, Luiten & Van Oers 2010a: 87, Significant 2010: 124). 34 Those who are required to give up their original nationality, but who are unwilling to do so, will however not opt for Dutch nationality. Fees for an application for a permanent residence permit (€ 150 in May 2013) are however much lower than the fees for an application for naturalisation (€ 810). A permanent residence permit however has to be renewed every five years. 35 The expectation is based on the fact that in the first year the Integration Act was in force, prior to the introduction of the ‘Deltaplan integration’, very few immigrants decided to follow a course and take the examination because they were required to pay for it themselves. A proposal of law to change the Integration Act, aiming to again place the financial responsibility for the integration on the immigrant, was accepted on 11 September 2012 (TK 2011–2012, 33086 nos. 1–3). The revised Integration Act entered into force on 1 January 2013. In 2013, municipalities will still receive funding to pay for the integration courses (EK 2011–2012, 33086, G, p. 4).

Impact of the Citizenship Tests  161

7.1.3.2. Germany To what extent did the formalised language requirement, applicable since August 2007, and the Einbürgerungstest, in force since September 2008, contribute to the (further) decrease in the number of naturalisations in Germany? The Influence of the Formalised Language Requirement As no data are published as to how applicants for naturalisation fulfil the language requirement, the number of persons who took part in integration courses within the framework of the Immigration Act (Zuwanderungsgesetz) and subsequently took the final test can be analysed. Those who passed the final test fulfilled the language requirement for naturalisation. Table 7.1 depicts the number of course participants, test candidates, and their pass rates for the years 2005 to 2009. From 2005 to 2009, a total of almost 163,000 persons reached level B1 in the test which is taken at the end of the integration course. 61% of all test candidates passed the test, compared to 51% of all course participants.36 This means that half Table 7.1. Participants in integration course, final test and pass rate 2005–2011.37

Year

Passed test Course (achieved participants Test candidates level B1)

Pass rate course Pass rate test participants candidates

2005 2006 2007 2008 2009 2010 2011 Total

31,478 76,401 67,052 73,557 70,968 83,818 73,647 476,921

39% 48% 44% 51% 66% 62% 68% 55%

17,482 50,952 43,853 61,025* 91,735* 103,875* 92,547* 461,469

12,151 36,599 29,544 37,438* 47,154* 51,791* 49,777* 264,454

69% 72% 67% 61% 51% 50% 54% 57%

Source: Integrationskursbilanz 2007 and 2008, Integrationskursgeschäftsstatistik 2009 and 2011. * Including participants who took the test for the second time. Those who took the test for a second time are only mentioned once as course participants.

36 In reality, the percentage of course participants among test candidates will be lower than 51%, since in the number of candidates who passed the test in 2008 and 2009, candidates who took the test for the second time are also included, whereas these candidates are only counted once as course participants. 37 Until 2008 the pass rates include both the language test and the ‘social orientation test’. Starting from 2009, the test parts have been separated. As of that year, only the numbers relating to the language test have been depicted.

162  Chapter Seven

of all those who participated in a language course did not reach the level required for naturalisation. Reaching level B1 is hence problematic, even for those who had the chance to follow a course. The language requirement therefore appears to form a barrier to naturalisation. Since 2006, and until 2011, the pass rate for test candidates decreased since: whilst 2006 almost three-quarters of all test candidates passed the test, only half of all test candidates were successful in 2010. The reason is that, until 1 January 2008, taking the test was optional. Research has shown that in the early years the test was in force, mainly more highly qualified immigrants, who were probably more confident that they could pass the test, had taken it (Rambøll 2006: 57). Hence, comparable to the Netherlands, a certain degree of self-selection has taken place among the immigrants. The pass rate among test candidates however rose to 54% in 2011. Einbürgerungstest Starting from its introduction on 1 September 2008 until 1 July 2011, about 174,000 persons took the Einbürgerungstest.38 The large majority of the participants, 98.4%, passed the test, which led the German government to draw the conclusion that the test could “hardly be interpreted as an important hurdle for naturalisation”.39 Compared to 2008, the relative number of naturalisations in 2009, the first full year the Einbürgerungstest was in force, increased slightly, and continued to increase in 2010. This might indeed imply that the introduction of the naturalisation test is not perceived as forming a barrier. The number of test candidates has however been decreasing: compared to 2009 (65,313 candidates), 17% fewer candidates took the test in 2010 (27,813 candidates). This development once again points to a certain degree of self-selection among the immigrants: those who felt confident that they would pass the test took it immediately after its entry into force, a number which is apparently decreasing. The percentage of test takers among the immigrants who naturalised is also decreasing. Whereas in 2009 about two-thirds of all those who naturalised took the test, only slightly more than half of all those who naturalised had passed the test in 2010.40 Compared to 2009, more immigrants hence relied on a diploma to be exempted from the test in 2010. This might signal that the German test is starting to function as a sieve to select the better educated immigrants from those who experience studying for and taking part in tests as a barrier to naturalisation. 38 BT-Drs. 16/11815, 17/941, 17/2223, 17/6041, 17/7701. 39 BT-Drs. 16/13707, p. 3. 40 As shown in figure 7.1, in 2009 (65,313 passes), 96,121 persons naturalised, which means that 66% did so after taking the test. In 2010 (27,813 passes), 52% of 101,570 naturalised immigrants took the test. Since not all those who took the test will have naturalised in the same year, these percentages are rough estimates.

Impact of the Citizenship Tests  163

To get support to prepare for the Einbürgerungstest, immigrants can follow a preparatory course. In 2009, such courses had however only been set up in a couple of Länder, because of limited demand (Michalowski 2010b: 196). Statistics have indeed shown that the vast majority of test candidates chose to take the test after preparing via self-study. From 1 September 2008 to 31 March 2009, 8% of all test candidates followed a preparatory course.41 The fact that most of the test candidates chose not to follow a course, but to prepare for the test via self-study, supports the assumption that the test is a means for selecting better educated immigrants, who have the skills to autonomously study for a test. 7.1.3.3. The UK For the UK, the influence of the introduction of the knowledge requirements on the overall number of naturalisations will be examined by an analysis of the number of refusals of naturalisation applications and data regarding the ‘Life in the UK test’ and the ESOL with citizenship courses. Refusals As the table below shows, prior to the year 2005, no applications were denied on the basis of insufficient knowledge of the language or society, even though a language requirement applied prior to November of that year. After the formalisation of the language requirement and the introduction of the ‘Life in the UK’ requirement, and with the exception of the year 2007, the number of applications denied due to insufficient knowledge of English and Life in the UK remained low: starting from 2005 and not taking account of 2007, less than 0.5% of all applications in the years 2002 to 2011 was denied because of nonfulfilment of the knowledge requirements.42 Applications that were denied because of non-fulfilment of the knowledge requirements were however also filed under ‘rejected applications’ (ninth row in table 7.2). In 2005 and 2006, the number of applications refused on these grounds was high. The rise in the number of rejected applications and applications refused on the basis of non-fulfilment of the knowledge requirements in 2005 and 2006 can, as in the Netherlands, probably be explained by unfamiliarity with the new requirements for citizenship. Nevertheless, as was the case in the Netherlands, because of the modest and only temporary rise in the number of refused applications, changes in the number of refusals cannot be held responsible for the decrease in the relative number of naturalisations. 41 BT-Drs. 16/13558. Their pass rate was 98%. 42 In 2007, the number of denied applications on this ground reached a peak with 2,365 denied applications. It is unclear what the causes for this peak have been.

164  Chapter Seven

Table 7.2: Refused applications for registration and naturalisation per ground for refusal, UK, 2002–2011. Year Reason for refusal

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

Incomplete 560 610 825 1,310 1,595 930 530 445 380 414 application Parent not a 530 775 1,100 1,670 2,160 2,535 1,305 1,690 786 627 British citizen Not of good 800 945 1,185 1,665 1,765 1,695 2,665 2,745 2168 2369 character 1,270 1,755 1,600 2,160 1,955 2,230 820 585 767 314 Delay in replying to enquiries from UKBA Residence 3,340 4,275 6,510 6,260 4,740 4,135 2,220 3,650 2,409 2,216 235 500 2,365 610 265 382 143 Insufficient knowledge of English and KOL Other 1,840 2,195 2,595 3,345 2,590 1,740 940 865 1,082 801 Rejected 3,260 6,105 255 40 130 200 151 applications43 Total 8,340 10,555 13,815 19,905 21,410 15,885 9,130 10,375 8,174 7,035 115,170 139,315 135,085 217,475 149,035 160,980 156,015 193,810 199,767 179,083 Total number of applications % refused 7 8 10 9 14 10 6 5 4 4

Source: Home Office, Immigration Statistics October-December 2012, British citizenship Tables cz.01 and cz.09.

Participants in test and course routes and pass rate in the ‘Life in the UK’ test To analyse the influence exercised by the knowledge requirements on the number of naturalisations, two questions need to be answered. The first question is to what extent the course route provides an effective alternative for those for whom passing the test is a barrier. The second question concerns the pass rate in the ‘Life in the UK’ test: knowing the number of persons who failed the test will allow us to determine to what extent the test forms a barrier to making an application for naturalisation. Course Route or Test Route? The number of persons who were granted citizenship after having followed the ESOL with citizenship courses is unknown, but an idea of the popularity of the 43 Applications not at the outset supported by the requisite evidence of entitlement to or qualification for British citizenship.

Impact of the Citizenship Tests  165

course route can be obtained by looking at the number of applicants for ILR who followed this route. Statistics show that the large majority of these, i.e. almost 90%, have followed the test route.44 This shows that the test route has proven to be a (much) more popular route than its alternative, which in turn makes one wonder to what extent the course route provides an effective alternative for those for whom the level of the test constitutes a barrier.45 Pass Rates in the Test From November 2005 to December 2009, the pass rate in the ‘Life in the UK’ test was 71%, a rate that would probably have been higher if immigrants with low language skills had chosen to follow the course route, instead of trying their luck in the test.46 However, as it is now, almost one in three test candi­ dates  failed the test.47 Hence, we can conclude that a significant proportion of applicants for naturalisation (and settlement) is being held back by actual test failures (see also Ryan 2008: 313). This shows that the ‘Life in the UK’ test exercises a selective effect, which is not counteracted by the existence of the course route. 7.1.4. Other Ways of Acquiring Citizenship Germany and the Netherlands for adults provide other ways of acquiring nationality than naturalisation. In the Netherlands, the revised Dutch Nationality Act, which introduced the naturalisation test, opened up extra possibilities for adults to opt for Dutch citizenship, in which case passing the test is not required.48 For Germany, there is the possibility to obtain German nationality via discretionary naturalisation. Possible rises in nationality acquisitions via these alternative routes may have reduced the negative effect of the strengthened requirements for ordinary naturalisation. Comparable alternative routes do not exist in the UK.

44 2 April 2007 - 15 June 2010. Source: UK Border Agency. Since nationals from EU or EEA states and Switzerland do not require ILR, and only nationalities where at least 100 persons obtained indefinite leave to remain (ILR) in the period in question have been included, the actual number of persons taking the test or following the course is higher. 45 Chapter 10 will show why the test route is more popular than the course route. 46 Source: Home Office statistics. The pass rate includes tests taken to obtain ILR. 47 See also ‘British Citizenship Test: One in three Immigrants fails’, BBC News 27 May 2010. 48 Those opting for citizenship are furthermore not required to renounce their original nationality. Moreover, the fees payable to opt for Dutch citizenship are lower than those payable for naturalisation: € 173 instead of € 810 for a single application (fees payable on 27 May 2013, according to http://www.ind.nl/Klant-informatie/leges/leges-nederlander-worden/Pages/default.aspx).

166  Chapter Seven

The Netherlands: Option With the coming into force of the revised Dutch Nationality Act in April 2003, the categories of immigrants who are attributed a right to opt for Dutch citizenship have been extended from two to eight (Article 6 paragraph 1 a to h DNA). An option right has for instance been granted to immigrants married to Dutch nationals and those aged 65 or over, provided they had been living in the Netherlands for at least fifteen years, as well as to former Dutch nationals and to immigrants born in the Netherlands.49 The number of options made increased considerably in the first years after the coming into force of the revised DNA. In the years 2003 to 2011, a peak was reached in 2007, when 7,779 options were registered, which is two and a half times as many as in 2002, the year prior to the entry into force of the revised Act (2,201 options).50 However, the rise in the number of options was not high enough to counterbalance the decrease in the number of naturalisations.51 Option is an attractive way to acquire Dutch nationality for those immigrants who have not followed education in the Netherlands, among whom, it is assumed, many can be found in the group of elderly immigrants, and those married to Dutch nationals. The option grounds for elderly immigrants and immigrants married to Dutch citizens gained popularity starting from 2006, which could imply that option has become more important for these categories as access to citizenship via naturalisation has become more difficult.52 However, after a peak was reached in 2007 (1,590 options, 19% of the total number of options made), the number of options by elderly immigrants decreased in the years thereafter, to 1,052 options (14% of total) in 2011.53 The number of options made by immigrants married to Dutch nationals reached a provisional peak in 2007 (2,282 options, 27% of total), but has remained high since (1,609 options in 2008 and 2,023 in 2009). A new peak was reached in 2011 (3,066 options, 41% of total). Nevertheless,

49 Children under shared custody of a non-Dutch parent and a Dutch national, and children recognised by Dutch nationals. 50 Source: CBS. 51 Compared to the number of options made in 2002, in 2004 almost 2,500 more options were made. But in comparison to 2002, in 2004 almost 14,000 fewer naturalisations were registered. Compared to 2002, the year 2006 also witnessed 14,000 fewer naturalisations. This difference cannot be compensated by the increase of 5,096 in the number of options made in that year. Lastly, the peak in the number of options made in 2007, in which year 5,578 more options were made than in 2002, only accounts for less than half of the decrease in the number of naturalisations (-12,605). 52 In 2004, a total of 1,331 elderly immigrants and immigrants married to Dutch nationals acquired Dutch citizenship via option. Their number rose to 4,118 in 2011. Source: IND. 53 Source: IND.

Impact of the Citizenship Tests  167

even though the newly introduced option ground for these immigrants is a bigger success than it is for the elderly immigrants, immigrants married to Dutch nationals do not use option on a large scale either. Germany: Discretionary Naturalisation For those who cannot fulfil the language and knowledge of society requirements for ordinary naturalisation, the possibility of discretionary naturalisation, provided in Article 8 of the Nationality Act, might be of help. In this case, the language requirement and the requirement to pass the Einbürgerungstest can, in some circumstances, be waived, or be applied more leniently.54 This applies for instance to those aged 60 or over if they have lived in Germany for at least twelve years, and to illiterates.55 An analysis of the number of naturalisations based on Article 8 however shows that naturalisations on the basis of this article went down by 70% between 2000 and 2011.56 This shows that illiterates and the elderly are mainly required to rely on the ‘ordinary’ naturalisation procedure of Article 10, and that naturalisations on the basis of Article 8 do not reduce the negative effect produced by the strengthened requirements for ordinary naturalisation. Elderly persons are however by law exempt from fulfilling the knowledge requirements for ordinary naturalisation if they submit a doctor’s certificate saying that their age hinders them from fulfilling these requirements.57 The possibility to exempt elderly people however appeared to be applied “as good as not” in practice (Lämmermann 2009: 295).

54 The requirements for naturalisation under Article 8 are that an immigrant has regular ordinary residence in Germany, is capable of earning a living and has a place to live. Persons applying for naturalisation under Article 8 may under certain circumstances be allowed to keep their original nationality. The income requirement is however more strictly applied compared to naturalisation on the basis of Article 10, since an application will also be refused if an applicant is not to blame for receiving social benefits. 55 The (binding) Verwaltungsvorschriften provide that those who have been living in Germany for at least twelve years and who are aged over 60 can be naturalised if they can ‘make themselves understood orally in daily life without considerable problems’, rather than requiring the uniform standard of level B1 to be achieved (No. 8.1.3.7 StAR-VwV vom 13. Dezember 2000). The temporary instructions (vorläufige Anwendungshinweise) from the Interior Ministry provide that those aged over 60 who have been living in Germany for at least twelve years, as well as illiterates can be exempted from proving language skills at level B1 (No. 8.1.2.1.3 Vorläufige Anwendungshinweise des BMI, 17. April 2009). The temporary instructions furthermore exempt the same categories of persons from proving knowledge of the juridical and social order and of the living conditions in Germany (No. 8.1.2.5). 56 In 2000, 15,440 persons acquired German nationality via Article 8, compared to 4,482 in 2011. Source: Federal Statistical Office Fachserie 1 Reihe 2.1 (2011), EUDO citizenship statistics database. 57 Article 10(6) German Nationality Act, number 10.6 of the temporary instructions of 17 April 2009.

168  Chapter Seven

7.1.5. Conclusions In all of the countries under consideration, the (absolute and/or relative) number of naturalisations decreased after 2000. In all cases, the introduction or strengthening of the language requirement, and the introduction of a knowledge of society requirement, has been, to a greater or smaller extent, responsible for the decrease. In all three countries under consideration, this restrictive effect exercised by the knowledge requirements appears inter alia from the rise in the number of applications immediately prior to the strengthening or introduction of the knowledge requirements. For Germany and the Netherlands, the restrictive effect furthermore becomes apparent from the lasting character of the decrease in naturalisations after the introduction of the requirements. The causal link between the decrease in numbers and the introduction of a citizenship test is probably most obvious in the Netherlands, where other requirements for naturalisation were only marginally strengthened during the same period. In the UK and Germany, on the other hand, the strengthening of other requirements, in the UK the significant rise in fees in particular, have almost certainly also played their part in reducing the (relative) number of naturalisations. 7.2. Gender, Age, Nationality, Level of Education and Handicaps: Differences in Effects? Did the knowledge requirements affect certain categories of immigrants more than others? We will answer that question by looking at the number of naturalisations, as well as pass rates in the tests, broken down by gender, age, and nationality. Furthermore, to see whether the formalisation of the integration requirements have produced negative effects for the handicapped, where information is available, it will be examined to what extent these categories can profit from the exemption possibilities which have been introduced. Information on the number of immigrants relying on a diploma to be exempt will provide insight into the question to what extent the knowledge requirements provide barriers to lower educated immigrants. 7.2.1. Gender Dutch parliamentarians expressed their fear that the formalisation of the language and integration test would have a negative effect on the number of naturalisations of women (see paragraph 3.1.2). Below, we will see whether this effect has indeed been produced by comparing the naturalisation rates (Netherlands and Germany) and absolute number of naturalisations (UK) for men and women.

Impact of the Citizenship Tests  169

Furthermore, for the Netherlands, the differences in pass rate in the naturalisation test and the integration examination between men and women will be compared. The Netherlands Figure 7.3 shows that in each year between 2001 and 2010, more women than men naturalised, except for the year 2009, when their share in the number of naturalisations was equal. The pattern followed by the number of naturalisations of men and women appears to be more or less equal, at least until 2009: a strong decrease in 2003, followed by a rise until 2007, after which the rates decreased again. In 2010, the naturalisation rate for women however rose, whereas the male rate continued to decrease. Figure 7.3 hence does not show that the naturalisation test and its successor have affected women more negatively than men. Looking at the results in the tests, the number of men and women registering for and passing the naturalisation test (applicable until 1 April 2007) was almost equal. 52% of those who registered were male, and they had a pass rate of 59%. The pass rate for the female immigrants who registered was 61%.58 In the integration examination, the number of female candidates however was much higher:

8%

6%

4%

2% 0%

2001

2003

2005 Men

2007 Women

2009

2011

Source: CBS.

Figure 7.3: Nationalisation percentages men and women in the Netherlands; 2001–2011. 58 Source: IND.

170  Chapter Seven

from 2007 to 2009, 60% of the candidates in the integration examination was female (Significant 2010). An explanation for the overrepresentation of women in the integration examination lies in the fact that the integration obligation in the Integration Act targets family migrants, who are more predominantly female. Women on the other hand had a 7% lower pass rate than men: 76% compared to 83% (Significant 2010). This shows that, despite the rise in the naturalisation rate in 2010, the examination has more negatively affected women than men, especially considering the fact that women, due to the integration obligation under the Integration Act, participate in courses more often. Germany The figure below shows that in Germany, as in the Netherlands, the naturalisation rate for women in the years 2000 to 2009 exceeded the naturalisation rate for men, except for the years 2008, 2010 and 2011, when the naturalisation rate for men and women was equal (1.4%, 1.2% and 1.3% respectively). With the exception of the year 2009, in which year the male naturalisation rate remained equal to the rate in 2008 (1.4%), whereas the female rate (1.5%) showed a small increase, the pattern followed by the naturalisation rates for men and women is similar. These figures show that the changes made to the naturalisation requirements in 2000 and 2007 appear not to have affected women more negatively than men.

3%

2%

1%

0%

2000

2002

2004 Men

2006

2008

2010

Women

Source: Statistisches Bundesamt, Fachserie 1 Reihe 2, Fachserie 1 Reihe 2.1 (Einbürger­ungen), several volumes.

Figure 7.4: Nationalisation percentages men and women in Germany; 2001–2011.

Impact of the Citizenship Tests  171 5% 4% 3% 2% 1% 0%

2003

2004

2005

2006 Men

2007

2008

2009

2010

2011

Women

Source: Home Office, Immigration Statistics October-December 2012, British citizenship table cz.05, Eurostat.

Figure 7.5: Nationalisation percentages men and women in the UK; 2003–2011. UK Also in the UK the pattern followed by the naturalisation rates for men and women is more or less similar, as shown in figure 7.5. Whereas the male naturalisation rate slightly exceeded the female rate in the years 2005 to 2007, the female rate was higher in the years 2009 and 2010.59 In 2011, the male and female rates are again equal. As was the case in the Netherlands and Germany, the changes made to the naturalisation requirements do not appear to have affected women more negatively than men. 7.2.2. Age The Netherlands Like women, elderly immigrants could, until 1 April 2003, profit from a more lenient application of the language and integration requirement.60 Since that date, these immigrants have also had to pass a formalised test. Has the introduction­of the formalised tests led to fewer naturalisations of elderly 59 For the years 2006 to 2008, no information on the number of immigrants broken down by gender was available. To estimate these numbers, the average rise in the number of immigrants between 2005 and 2009 has been taken as a point of departure. The numbers of naturalisations taken as a point of departure to calculate the naturalisation percentages also included registrations. However, as only grants of applications by adults have been used, the numbers of registrations included will be very low. 60 Manual for the application of the Dutch Nationality Act, June 2001, A1, p. 21.

172  Chapter Seven

immigrants?­To answer this question, the number of naturalisations and the pass rates in the tests broken down by age will be analysed. Naturalisations by Age In the years 2001 to 2011, and compared to immigrants in other age categories, immigrants aged 65 and over naturalised least often.61 Naturalisations by elderly immigrants have hence always been low, probably because the immigrant population is a relatively young population, and because of the fact that many immigrants aged 65 will already have naturalised prior to their turning 65, or by then have decided not to change their legal status. In the years 2001 to 2011, the share of elderly immigrants never exceeded 2%, except for the year 2003 when 2.1% of all naturalisations concerned naturalisations of immigrants aged 65 and over. As of 2003, the share of elderly immigrants in the total of immigrants who successfully applied for naturalisation started to decrease, and has since 2005 been below 1%.62 Furthermore, compared to 2002, 80% fewer immigrants aged 65 and older naturalised in 2006. Does this mean that the formalised language and integration requirement, applicable since 2003, has more negatively affected elderly immigrants than those in the younger age categories, where a less dramatic decrease took place? This is probably part of the explanation, but another, perhaps more important reason for the decrease seems to lie in the opening up of the possibility for elderly immigrants to opt for Dutch citizenship, referred to above (paragraph 7.1.4). In the years 2004 to 2009, almost 5,000 elderly immigrants acquired Dutch nationality via option, which largely makes up for the decrease in the absolute number of nationality acquisitions by elderly immigrants in those years.63 The introduction of the possibility to opt for Dutch nationality has thereby mobilised a group of elderly immigrants who, without this possibility, would perhaps have never obtained Dutch nationality. Pass Rate in the Tests by Age Immigrants in the highest age category have the lowest pass rate in both the naturalisation test and integration examination. In the naturalisation test, the average pass rate for persons aged 65 and over was only 40%, whereas in the integration examination, in the years 2007 to 2009, 60% of candidates aged 56 and over passed the test.64 Not only the pass rate, but also the number of candidates 61 Source: IND. 62 With the exception of 2008, when 1.3% of the immigrants who naturalised was aged 65 or over. 63 Compared to 2003, 215 fewer immigrants aged 65 and over naturalised in 2008. In the years 2005, 2006 and 2007, this difference was higher, but never higher than 330. 64 Source: IND and Significant 2010. In the division in age categories for the integration examination, immigrants aged 56 and over represent the highest age category, whereas in the naturalisation test, immigrants aged 65 and over represent the highest age category.

Impact of the Citizenship Tests  173

appeared to diminish with age: in 2003–2007, only 35 persons aged 65 and over registered for the naturalisation test, and in the years 2007–2009 merely 800 persons aged 56 or over took part in the integration examination.65 An explanation for the low number of immigrants aged 65 years or older registering for the tests might be offered by the possibility for immigrants aged 65 or over to opt for Dutch nationality. The elderly immigrants who cannot opt for Dutch nationality are however required to pass the integration examination if they want to obtain Dutch nationality. And as the above shows, the older the immigrant is, the less inclined she will be to take part in the test, and, once she decides to do so, the fewer her chances will be of passing. Germany In Germany, as in the Netherlands, immigrants aged 65 naturalised least often, at least in the years 2003 to 2011.66 Furthermore, their share in the total number of naturalisations decreased from 2.6% in 2004 to 1.7% in the year 2011. Elderly immigrants in Germany hence appear to have been negatively affected by the changes in the requirements for naturalisation, despite the fact that the law exempts those who, due to their age, cannot fulfil the knowledge requirements from passing the Einbürgerungstest and proving their language skills.67 This suggests that the possibility to be exempted on age grounds is not well known, or hardly applied in practice, something which has been confirmed in the literature (Lämmermann 2009: 295). Elderly immigrants who consider the knowledge requirements as barriers can file for discretionary naturalisation via Article 8, in which case the knowledge requirements can be waived, or applied more leniently. As paragraph 7.1.4 has already shown that naturalisations on the basis of Article 8 have been decreasing, we can conclude that, unlike the Netherlands, elderly immigrants in Germany can in the practice not profit from alternative ways of acquiring German nationality. UK As in the other countries under consideration, in the UK, immigrants in the highest age category naturalised least often. And similar to Germany and the Netherlands, the share of elderly immigrants in the total number of immigrants who naturalised has always been low, albeit that their share is higher than in the other countries under consideration. In the years 2001 to 2005, about 5% of the

65 In both the naturalisation test and the integration examination, immigrants aged between 18 and 35 obtained the highest pass rate, and constituted the largest group of candidates. 66 Source: Federal Statistical Office, Fachserie 1, Reihe 2.1, several volumes. 67 Article 10(6) German Nationality Act.

174  Chapter Seven

immigrants who successfully applied for naturalisation was aged 60 or over.68 This number decreased once the knowledge requirements became effective. In the years 2009 to 2011, only about 3% of the immigrants who naturalised was aged 60 or over.69 The introduction of the knowledge requirements has apparently caused a proportion of the elderly immigrants who might previously have naturalised give up their wish to become British citizens, despite the possibility for exemption from fulfilling the knowledge requirements.70 7.2.3. Nationality Below we will see whether the changes in the nationality laws in the countries under consideration have affected differently the two largest immigrant groups in the countries under consideration by an analysis of the naturalisation rates of these immigrant groups (Netherlands and Germany), and the pass rates in the tests per nationality (all three countries). The Netherlands Naturalisation Rates for Turkish and Moroccan Nationals In the Netherlands, the two largest immigrant groups are Turkish and Moroccan nationals. Together with the total naturalisation rate, the naturalisation rates for these groups are presented in the figure below for the years 1996 to 2011. The naturalisation pattern of Turks and Moroccans followed a somewhat different pattern than the total naturalisation rate. As of 1999, the naturalisation rate for immigrants with Moroccan nationality has been higher than the total naturalisation percentage, whereas the Turkish percentage has been below average since that year, with the exception of the years 2010 and 2011. The Moroccan naturalisation percentage rose until the entry into force of the revised DNA. Starting in 2003, it started to decrease and, contrary to the total naturalisation percentage, continued to decrease until 2009.71 In the years 2009 to 2011, the Moroccan naturalisation percentage however rose, to reach the level of 6.6% in 2011, which is 68 Source: Home Office. 69 2.9% in 2009, 3.2% in 2010, and 3.1% in 2011. 70 The nationality instructions for UK Border Agency Staff deciding applications for citizenship state that applicants aged 65 or over are exempted from fulfilling the knowledge requirements, whereas those aged between 60 and 65 can be exempted. Volume 1: the British Nationality Act 1981 – caseworking instructions, Chapter 18: naturalisation at discretion, p. 13. Accessible via http:// www.ind.homeoffice.gov.uk/sitecontent/documents/policyandlaw/nationalityinstructions/ni chapter18/ch18annexe?view=Binary, site visited 18 May 2010. 71 This is with the exception of the year 2005, which has shown a small increase in the naturalisation percentage. Since many of the applications treated in the years 2004 and 2005 were filed prior to the entry into force of the revised DNA, these years give a somewhat distorted image.

Impact of the Citizenship Tests  175

20% 18% 16% 14% 12% 10% 8% 6% 4% 2% 0%

1996

1998

2000 Turks

2002

2004 Moroccans

2006

2008

2010

Total

Source: CBS.

Figure 7.6. Naturalisation percentages in the Netherlands, for Turks, Moroccans, and all non-Dutch, 1996–2011. more than twice as high as the average naturalisation percentage, which decreased in the years 2009 and 2010. This could mean that, for the Moroccan immigrants, the deterrent effect of the revised Dutch Nationality Act is starting to wear off. The Turkish naturalisation percentage was highest in 1996 when almost one in five Turks in the Netherlands naturalised.72 After this peak, the relative number of naturalisations by Turkish nationals quickly decreased, and since 1999, and not taking account of 2010 and 2011, it has been below the total naturalisation percentage. The Turkish naturalisation percentage sharply decreased after the entry into force of the revised DNA. Unlike the total naturalisation percentage, but similar to the Moroccan percentage, the Turkish naturalisation rate continued to decrease until 2009. As for the years 1996 to 2011, the lowest percentage was reached in 2008, when 1.6% of all Turks present in the Netherlands naturalised. This means that, as was the case for Moroccans, the revised DNA has initially produced a more negative effect for Turks than for the total immigrant population, as the total naturalisation rate after 2003 rose following an initial drop. Contrary to the total naturalisation rate, but similar to the Moroccan rate, the Turkish naturalisation percentage rose in 2009, from 1.6% to 2.2%, and continued to rise in 2010 when 2.6% of all resident Turkish nationals naturalised. The rise continued in 2011, in which year the Turkish naturalisation rate was 3.2%. 72 That is to say that this percentage was highest in the years depicted in figure 7.6. In 1996, 19% of all Turkish nationals resident in the Netherlands naturalised.

176  Chapter Seven

As in the years 1996 and 1997 more than half of all naturalisations concerned naturalisations by Moroccan and Turkish nationals, the decrease in the naturalisation rates of these groups largely explains the decrease in the general naturalisation percentage. Whereas the decrease in the naturalisation percentage among these groups can partly be explained by the factors mentioned above, a certain saturation point can also be detected: in 2010, about 78% of all Moroccans living in the Netherlands had Dutch nationality, compared to 74% of all Turks.73 The saturation point appears to have been reached sooner by Turkish nationals living in the Netherlands than by Moroccan nationals. In 1998, 60% of all Turkish nationals had Dutch nationality, compared to 42% of all Moroccan nationals. Pass Rates in the Tests As a complete overview of pass rates by nationality has not been published, table 7.3 below was drawn up on the basis of a top ten of the nationalities that took part in the test and a top ten of the nationalities that passed the test.74 Neither of these rankings included EU or other Western countries. Whereas immigrants from these countries are less inclined to file an application for naturalisation, the absence of these nationalities in the published statistics also suggests that immigrants originating from Western countries can relatively often make use of an exemption on the basis of a diploma, such as the state examination Dutch as a second language, the level of which is higher, but which costs less.75 The pass rate was highest for persons holding Chinese nationality, followed by persons holding Afghan or Iraqi nationality.76 For all three nationalities, the pass rate was higher than the average pass rate of 74%. The pass rate was below average for all other nationalities for which pass rates could be calculated (i.e. for immigrants holding Moroccan, Turkish, Somalian and Sierra Leonean 73 Source: CBS. 74 Since the nationalities mentioned in both rankings did not match, a ranking of pass rates for only seven nationalities could be drawn up. Pass rates for nationalities that did not figure in either of the rankings could not be drawn up, and neither could pass rates for nationalities that were only mentioned in one of the rankings. 75 Furthermore, unlike the naturalisation test certificate, the certificate for the Dutch state examination will be acknowledged by educations institutions and employers because it is awarded on the basis of the Higher Education Act. 76 It could however very well be that the pass rate for other nationalities than those mentioned in table 7.3 was higher. It is for instance very likely that immigrants from Belgium or Surinam, who at the time the naturalisation test was in force could not rely on their Belgian and Surinamese diplomas to qualify for exemption, had a pass rate exceeding 79%. Since the table had to be drawn up on the basis of limited information, a full picture could however only be sketched for the nationalities listed.

Impact of the Citizenship Tests  177

Table 7.3: Number of candidates, passed candidates and pass rates for Dutch naturalisation test per nationality 1 April 2003–1 April 2007. Nationality

Candidates

Sierra Leonean Somalian Turkish Moroccan Total all nationalities Afghan Iraqi Other Chinese

410 574 1,494 2,621 19,114 1,428 1,163 10,632 792

Passes 122 386 1,049 1,870 14,303 1,069 875 8,301 631

Pass rate 30% 67% 70% 71% 74% 75% 75% 78% 79%

Source: IND.

nationalities).­The pass rate was lowest for persons having Sierra Leonean nationality. With a 30% success rate, two out of three persons holding Sierra Leonean nationality did not pass the test. Even though it cannot be exactly calculated, the pass rate was even lower for persons originating from Sudan, Egypt or Ghana. These nationalities figured in the top ten of the nationalities with the highest number of test candidates, but since fewer than 122 persons passed the complete test, they did not return in the top ten of the nationalities that passed the test most often. This means that, at most, the pass rate will have been 24% for persons having Egyptian or Sudanese nationality, and 27% for persons having Ghanaian nationality. With a pass rate of around 70%, roughly one in three persons with Moroccan or Turkish nationality failed the naturalisation test. This means that the test indeed functioned as a barrier for Turkish and Moroccan immigrants, thereby offering an explanation for the decrease in the naturalisation rates for these groups post-2003 (see figure 7.6 above). Has the replacement of the naturalisation test by the integration examination changed this conclusion? Again, a full list of pass rates per nationality has not been published. However, a list of nationalities with at least 200 candidates has. A selection of nationalities is presented below. Compared to the ranking presented for the naturalisation test in table 7.3, the ranking for the integration test represented in table 7.4 shows some remarkable differences and similarities. With a pass rate of 80% and 85% respectively, Afghan and Iraqi immigrants still score above average. As was the case in the naturalisation test, immigrants holding Moroccan, Somalian or Turkish nationality score below average in the integration examination. However, the pass rate for Moroccan and Somalian immigrants in the integration examination has risen

178  Chapter Seven

Table 7.4. Pass rates integration examination per country of origin, 2007–2009.*77 Country of origin

Exams taken

Turkey Somalia Cape Verdi Ethiopia China Morocco Vietnam Total Afghanistan Other Sierra Leone Iran Iraq former Yugoslavia former Soviet Union Indonesia Surinam Poland

5,100 860 210 330 1,100 4,000 270 33,000 1,500 4,700 530 820 2,600 1,400 2,100 700 260 250

Passes 3,900 620 160 240 900 2,900 200 26,000 1,200 3,900 440 690 2,200 1,200 1,900 630 240 230

Pass rate 63% 72% 73% 73% 74% 74% 74% 79% 80% 83% 83% 85% 85% 85% 90% 91% 91% 93%

Source: Significant 2010. * Only countries with 200 or more candidates are included. The category ‘other’ consists of all countries with less than 200 participants.

compared to the pass rate in the naturalisation test. This also applies to immigrants holding Sierra Leonean nationality: whereas their pass rate was well under the average in the naturalisation test, with 83% of all Sierra Leonean test candidates passing the integration examination, their pass rate is above average. This positive development should probably be ascribed to the introduction of courses, instead of to the replacement of the naturalisation test. Compared to their pass rates in the naturalisation test, the pass rates for Chinese and Turkish immigrants has decreased. One in four Chinese candidates failed the test, compared to roughly one in three Turkish candidates. This could

77 Account should be taken of the fact that it is not registered whether someone took the integration examination for the purpose of filing a naturalisation application, a permanent or independent residence permit, or because he or she was obliged to take the examination on the basis of his or her duty to integrate under the Integration Act.

Impact of the Citizenship Tests  179

signal that the new test poses a bigger hurdle for naturalisation for Chinese and Turkish immigrants than the naturalisation test did. Account should however be taken of the widened scope of test candidates: whereas only those interested in naturalisation would take the naturalisation test, the integration examination is also obligatory for those with an integration obligation under the Integration Act (mostly family migrants), and those interested in a permanent or independent residence permit. The group of candidates in the integration examination hence also consists of newcomers, who are likely to have fewer language skills. At the same time, however, these newcomers were assisted in their integration by virtue of the integration courses. What conclusions can be drawn from the differences in pass rate per country of origin? First, countries with a relatively high level of development (Poland, former Soviet Union, former Yugoslavia) appear to have a high pass rate (above 80%). The high pass rates for Surinam and Indonesia can be explained by the fact that these countries are former colonies. Moreover, in Surinam, Dutch is the official language.78 Immigrants from countries with a different alphabet have a significantly lower pass rate (China, Morocco, Vietnam). At the same time, however, the pass rate is high for persons coming from Afghanistan, Iraq or Iran, all countries which also have a different alphabet. This might imply that there is another intervening variable which comes into play. It could be for instance that the immigrants originating from Afghanistan, Iraq and Iran are more highly educated than the immigrants originating from China and Morocco. Germany Naturalisation Percentages In the figure below, the total naturalisation percentage in Germany for the years 1994 to 2011 has been depicted, as well as the naturalisation percentages for the two largest immigrant groups in Germany: Turkish nationals and persons from the former Yugoslavia. The low naturalisation percentages for Germany’s first and second largest minority groups explain the country’s low total naturalisation percentage, as well as the relatively low total naturalisation rate in Germany (1.5%), compared to the other countries under consideration (3%). Figure 7.7 shows that the pattern of the naturalisation rate for Turkish immigrants, Germany’s largest immigrant group, differs from the general pattern. Contrary to the naturalisation pattern for all immigrants in Germany, the relative number of acquisitions of Turkish immigrants reached a peak not in the year 2000, but in 1999. This peak can probably be explained by immigrants who made 78 As immigrants who obtained a secondary school diploma in Surinam are exempt, the group of candidates of Surinamese origin will mainly consist of immigrants who have not followed secondary education.

180  Chapter Seven

6% 5% 4% 3% 2% 1% 0%

1994

1996 Turks

1998

2000

2002

2004

2006

Citizens from the former Yugoslavia

2008

2010 Total

Source: Federal Statistical Office, Fachserie 1, Reihe 2 and 2.1, several volumes, Lange Reihe Ausländer nach Staatsangehörigkeiten 1994–1997.

Figure 7.7. Naturalisation percentage of the total immigrant population in Germany and naturalisation percentages of Turks and persons from the former Yugoslavia 1994–2010. their naturalisation applications early, i.e. prior to 1 January 2000, to avoid the stricter requirements. With almost 104,000 Turkish nationals acquiring German citizenship, the naturalisations of Turks in 1999 made up more than half of all naturalisations in Germany. The naturalisation percentage in that year was higher than five, meaning that in 1999, around one in twenty Turkish nationals naturalised. After the entry into force of the revised Act in 2000, the number of naturalisations of Turkish nationals decreased. Whereas since 1994 the naturalisation rate for Turkish immigrants had been above the national average, since 2005, it had more or less equalled the total naturalisation rate. The 2000 reform has hence not been successful for the largest immigrant group in Germany, which is one of the main explanations for the low number of naturalisations in total (Tränhardt 2008: 12). The relative number of naturalisations by Turkish nationals decreased further in 2007 and 2008.79 The changes brought about by the 2007 Act, such as the formalisation of the language requirement and the introduction of the knowledge of society requirement appear to have had their share in this decrease. Since 65% of all Turks present in Germany did not receive any education in Germany (Diehl & Blohm 2003: 149, Anil 2008: 1365), the majority of Turkish residents in Germany 79 The small increase in the number of naturalisations in 2006 might be explained by Turkish nationals anticipating the changes introduced in 2007.

Impact of the Citizenship Tests  181

cannot rely on a diploma to be exempted from the language and knowledge of society requirements. The naturalisation percentages in the years 2008 (1.4%) and 2009 (1.5%) are lower than the naturalisation percentage of Turks five years prior to the 2000 reform in 1995 (1.6%), and much lower than the rate in ‘peak year’ 2009 (5.2%). In 2010, the Turkish naturalisation rate however again equalled the level reached in 1995, to rise to 1.7% in 2011. This could signal that the effects produced by the changes brought about by the 2007 Act are starting to wear off. Germany’s second largest immigrant group, persons from the former Yugoslavia, has, except for the years 2004 and 2006, always had a lower naturalisation rate than the general immigrant population.80 A ‘peak’ was reached in the year 2001, when 20,800 citizens from the former Yugoslavia naturalised, equalling a naturalisation rate of 1.9%. After 2001, the number of naturalisations of persons from the former Yugoslavia decreased, but the percentage rose again after 2004. Another ‘peak’ was reached in 2006, when 1.8% of the citizens from former Yugoslavian states present in Germany naturalised. The rise in the number of naturalisations after 2004 could be explained by the fact that, after an initial deterrent effect, persons from the former Yugoslavia started to apply for naturalisation under the provisions of the revised Nationality Act. It could also be that, to a certain extent, former Yugoslavians were anticipating the changes introduced by the 2007 Act. In any case, after 2006, the number of naturalisations declined. With 8,900 successful applications, the relative number of former Yugoslavians who naturalised in 2009 was below 1%, to return to a level slightly above 1% in 2010 in 2011.81 The negative effect of the latest reform, including a formalised language test and the Einbürgerungstest, hence appears to also have affected immigrants from the former Yugoslavia. Pass Rates in the Einbürgerungstest Judging from the pass rates in the Einbürgerungstest, does the effect of this test differ per nationality? We have already seen (paragraph 7.1.3.2) that the average pass rate in the German Einbürgerungstest was high, i.e. 98.4%. From information published following questions asked in the Bundestag it becomes apparent that, for the nationalities with most test candidates (Iraqi, Turkish, Ukrainian, Polish, Iranian, Russian, Kosovan, Romanian and Moroccan), between 1 September 2008 to 31 March 2009, the pass rates were also high, i.e. higher than 96%.82 Between the country with the highest pass rate (Ukraine; 99.9%) and the country with the 80 An explanation could be sought in the hassle and high costs that come along with renouncing the original nationality. 81 The naturalisation percentage of citizens from the former Yugoslavia in 2009 was 0.98%. In 2010, the level was 1.12%, and in 2011, 1.28%of all former Yugoslavians present in Germany naturalised. 82 BT-Drs. 16/13558.

182  Chapter Seven

lowest pass rate (Turkey; 96.8%) lay a difference of only 3.1 percentage points. Later, the number of candidates of Turkish origin was published following new questions asked in parliament.83 In the years 2009 and 2010 and the first half of 2011, the large majority of Turkish test candidates (96% or more) passed the test. The total number of Turkish test candidates (9,684), 6% of the total in the period 1 January 2009 until 1 July 2011, is however quite low considering the size of the Turkish population in Germany, which means that Turkish nationals are not very much inclined to take part in the Einbürgerungstest. As most Turks present in Germany cannot rely on exemption grounds (see above), this suggests that the test is a barrier for Turkish immigrants. The latter assumption is confirmed by the share of Turkish immigrants who successfully filed for naturalisation among the total number of immigrants who naturalised, which decreased from 40% in 2003, to 26% in the years 2009 to 2011.84 The United Kingdom For the UK, naturalisation percentages per nationality could not be calculated. To see whether the effect of the knowledge requirements differs per nationality, the pass rates in the ‘Life in the UK’ test broken down by nationality can be analysed. An overview is presented in the table below. The pass rates per nationality appear to vary considerably. It is almost 100% for applicants holding the nationality of certain countries where English is the majority language (USA, Australia). However, the pass rate for the majority English speaking states in the Caribbean area is much lower (Barbados: 80.4%, Belize: 74.4%, St. Lucia: 68.2%, Jamaica: 67%). Speaking English is hence no guarantee for a high pass rate, which implies that the pass rate is probably influenced by the educational system in the country of origin, or the level of education of the immigrants coming to the UK. Countries with a different alphabet also have relatively low pass rates. Examples are Pakistan, China, Somalia, Thailand, Iraq and Afghanistan. The latter country appears at the very bottom of the list, with a success rate of less than 50%. The pass rate of Afghan and Iraqi nationals in the ‘Life in the UK’ test in this sense differs from the success rate of their compatriots in the Dutch integration examination, where, as we have seen, the pass rate for Afghans and Iraqis figures above the average. The reason for this difference in success is unclear, but it could be that the Afghan and Iraqi nationals coming to the Netherlands are more highly educated than those entering the UK.

83 BT-Drs. 17/941, BT-Drs. 17/6041, and BT-Drs. 17/7701. Questions asked by Die Linke parliamentarian Sevim Dağdelen. 84 Source: Federal Statistical Office, Fachserie 1, Reihe 2.1 (Einbürgerungen), several volumes.

Impact of the Citizenship Tests  183

Table 7.5. Pass rates ‘Life in the UK’ test total and per nationality (November 2005- December 2009).85 Nationality

Tests taken

Bangladesh Turkey Afghanistan Iraq Kosovo Thailand Somalia China Pakistan Jamaica St. Lucia Total Belize Philippines India Barbados Ghana Nigeria Kenya Romania Ukraine South Africa USA Australia

38,085 30,014 29,650 40,200 11,600 13,285 21,711 26,286 68,279 21,788 170 906,464 121 41,858 100,001 680 20,662 34,107 9,966 5,240 5,252 38,061 15,717 13,223

Passes 16,740 13,780 14,178 19,262 5,627 6,875 13,411 16,849 43,540 14,570 116 642,823 90 11,751 79,229 547 16,786 28,129 8,403 4,433 4,530 35,735 15,348 12,957

Failures

Pass rate

21,345 16,234 15,472 20,938 5,973 6,410 8,300 9,473 24,739 7,218 54 263,641 31 53,609 20,772 133 3,876 5,978 1,563 807 722 2,326 369 266

44.0% 45.9% 47.8% 47.9% 48.5% 51.8% 61.8% 64.1% 63.8% 66.9% 68.2% 70.9% 74.4% 78.1% 79.2% 80.4% 81.2% 82.5% 84.3% 84.6% 86.3% 93.9% 97.7% 98.0%

Source: Home Office.

With a pass rate of 45.9%, Turkish nationals figure at the bottom of the list. Hence, in all three countries, the tests are holding immigrants from Turkey back from applying for naturalisation (or permanent residence). Contrary to the Netherlands and Germany, those for whom the test constitutes too high a barrier will not need to give up their wish to become British citizens, since they can also fulfil the knowledge requirements by following the course 85 The data do not distinguish between applicants for indefinite leave and for naturalisation. The data are by test attempt, not by individual.

184  Chapter Seven

route.86 A low pass rate in the ‘Life in the UK’ test hence does not automatically mean that the knowledge requirements have more negatively affected certain nationalities. The course route however appears to be a much less used route towards citizenship (or ILR) than the test route (paragraph 7.1.3.3). Only around 13% of all immigrants who fulfilled the knowledge requirements to obtain ILR status did so by following a course (Ryan 2010: 31). Moreover, immigrants appear unlikely to rely on the course route, whether they have a high or low success rate in the test: despite a pass rate in the ‘Life in the UK’ test of less than 50%, less than one-third of all Turkish nationals who applied for ILR followed the course route, compared to 35% of all Bangladeshis (Ryan 2010: 31). In fact, only three nationalities (Afghan, Yemenite and Somali) have preferred the course route over the test route (Ryan 2010: 32). These figures show that the course route does in reality not appear to constitute a valid alternative for those for whom passing the test at level B1 after self-study is too high a barrier, something we already assumed from the information presented in paragraph 7.1.3.3. This means that firm conclusions can after all be drawn from the pass rates in the ‘Life in the UK’ test: fulfilling the knowledge requirements appears to constitute a higher barrier for certain nationalities than for others. 7.2.4. Level of Education The Netherlands and Germany exempt immigrants who have obtained certain diplomas from fulfilling the knowledge requirements. In the UK, a comparable exemption regulation is absent. Information on the number of immigrants that relied on a diploma to be exempt is available for the Netherlands. This information will provide insight into the question to what extent the knowledge requirements provide barriers for lower educated immigrants. After all, if the majority of the immigrants who naturalised relied on a diploma, the decrease in the number of naturalisations which has taken place since 2003 can be ascribed to lower educated immigrants no longer making applications. It has been estimated that in the Netherlands, between April 2003 and the end of 2006, only 25 to 30% of all applicants for naturalisation passed the naturalisation test (IND 2007: 70).87 The majority, i.e. 70 to 75% of all applicants, was either 86 As we have however seen, the government has announced plans to abolish the course route as of October 2013 (paragraph 5.3.2). 87 This estimate was not based on a spot check among filed applications, but by comparing the total number of persons who passed the test from 2003 to 2006 to the number of applications made in those years. A one-on-one comparison of the number of persons who passed the test and the number of applications made in a year is not possible, since there will always have been minimally a few months between the moment a candidate passed the complete naturalisation

Impact of the Citizenship Tests  185

exempt on the basis of a diploma or because of a medical impediment or illiteracy. As we will see, the latter exemption grounds in those years only accounted for a small percentage (