Research Handbook on Political Representation 9781788977098, 9781788977081

At a time when political representation can be said to be facing its ultimate crisis, this crucial work clarifies the te

217 15 2MB

English Pages [368] Year 2020

Report DMCA / Copyright

DOWNLOAD FILE

Research Handbook on Political Representation
 9781788977098, 9781788977081

Table of contents :
Front Matter
Copyright
Contents
Contributors
General introduction to the Research Handbook on Political Representation
SECTION I CONCEPTUAL AND THEORETICAL DISCUSSIONS
1. Political representation: concepts, theories and practices in historical perspective
2. The system of democratic representation and its normative principles
3. Democracy and representation
4. Policy responsiveness and democratic quality
5. Legitimacy and hegemony: two accounts of non-electoral representation
6. Challenges to political representation: participatory democracy, direct democracy and populism
SECTION II HISTORICAL AND CONTEMPORARY MODELS OF POLITICAL REPRESENTATION
7. From estate representation to the representation of the people and the nation in the Age of Revolutions
8. Individualistic representation in the liberal century - and beyond
9. Party-based representation: the paradox of democracy
10. Corporatism and representation
11. Technocratic representation
12. Representation in authoritarian regimes
SECTION III THE INSTITUTIONAL ASPECTS OF REPRESENTATION
13. Electoral systems and representation
14. Territorial and multilevel representation
15. Local jurisdiction size and political representation
16. Electoral quotas and group representation
17. The political representation of women: a feminist institutionalist perspective
18. Ethnic minorities and representation
SECTION IV REPRESENTATION AT WORK: OUTPUTS AND OUTCOMES
19. Types of democratic representative
20. Selecting representatives: methods and practices
21. Constituency service: incentives and consequences
22. Measuring representation: policy congruence
23. The fulfilment of election pledges
24. Interest groups and political representation
25. Responsiveness, the dynamic aspect of representation
26. What to expect when you are expecting: preferences for representation among voters and political elites
27. Representation and deliberation
Index

Citation preview

RESEARCH HANDBOOK ON POLITICAL REPRESENTATION

ELGAR HANDBOOKS IN POLITICAL SCIENCE Elgar Handbooks in Political Science provide an overview of recent research in all areas relating to the study of political science including comparative politics, international relations, political economy, political theory and research methods, ensuring a comprehensive and overarching guide to the field. The constituent volumes, edited by leading international scholars within the field, are high quality works of lasting significance, often interdisciplinary in approach. The Handbooks discuss both established and new research areas, expanding current debates within the field, as well as signposting how research may advance in the future. The series will form an essential reference point for all academics, researchers and students of political science. Titles in the series include: Handbook of Political Anthropology Edited by Harald Wydra and Bjørn Thomassen Handbook of Organised Crime and Politics Edited by Felia Allum and Stan Gilmour Research Handbook on Political Representation Edited by Maurizio Cotta and Federico Russo

Research Handbook on Political Representation Edited by

Maurizio Cotta Senior Professor, University of Siena, Italy

Federico Russo Associate Professor, University of Salento, Italy

ELGAR HANDBOOKS IN POLITICAL SCIENCE

Cheltenham, UK • Northampton, MA, USA

© Maurizio Cotta and Federico Russo 2020

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical or photocopying, recording, or otherwise without the prior permission of the publisher. Published by Edward Elgar Publishing Limited The Lypiatts 15 Lansdown Road Cheltenham Glos GL50 2JA UK Edward Elgar Publishing, Inc. William Pratt House 9 Dewey Court Northampton Massachusetts 01060 USA A catalogue record for this book is available from the British Library Library of Congress Control Number: 2020948598 This book is available electronically in the Social and Political Science subject collection http://dx.doi.org/10.4337/9781788977098

06

ISBN 978 1 78897 708 1 (cased) ISBN 978 1 78897 709 8 (eBook)

Contents

List of contributorsvii General introduction to the Research Handbook on Political Representation xi Maurizio Cotta and Federico Russo SECTION I

CONCEPTUAL AND THEORETICAL DISCUSSIONS

1

Political representation: concepts, theories and practices in historical perspective3 Federico Russo and Maurizio Cotta

2

The system of democratic representation and its normative principles Dario Castiglione

16

3

Democracy and representation Mark B. Brown

36

4

Policy responsiveness and democratic quality Leonardo Morlino

48

5

Legitimacy and hegemony: two accounts of non-electoral representation Eline Severs

58

6

Challenges to political representation: participatory democracy, direct democracy and populism Simon Tormey

SECTION II

70

HISTORICAL AND CONTEMPORARY MODELS OF POLITICAL REPRESENTATION

7

From estate representation to the representation of the people and the nation in the Age of Revolutions Pasi Ihalainen and Zachris Haaparinne

8

Individualistic representation in the liberal century – and beyond Heinrich Best and Lars Vogel

9

Party-based representation: the paradox of democracy Ian Budge

109

10

Corporatism and representation Alan Siaroff

121

11

Technocratic representation Daniele Caramani

135

12

Representation in authoritarian regimes Maurizio Cotta

148

v

84 98

vi  Research handbook on political representation SECTION III

THE INSTITUTIONAL ASPECTS OF REPRESENTATION

13

Electoral systems and representation Alessandro Chiaramonte

161

14

Territorial and multilevel representation Lori Thorlakson

171

15

Local jurisdiction size and political representation Bas Denters

183

16

Electoral quotas and group representation Mona Lena Krook

198

17

The political representation of women: a feminist institutionalist perspective Joni Lovenduski

210

18

Ethnic minorities and representation Miriam Hänni and Thomas Saalfeld

222

SECTION IV

REPRESENTATION AT WORK: OUTPUTS AND OUTCOMES

19

Types of democratic representative Luca Verzichelli

239

20

Selecting representatives: methods and practices Marco Lisi

253

21

Constituency service: incentives and consequences Zsófia Papp

266

22

Measuring representation: policy congruence Tom Louwerse and Rudy B. Andeweg

276

23

The fulfilment of election pledges Elin Naurin and Robert Thomson

289

24

Interest groups and political representation Laura Chaqués Bonafont

301

25

Responsiveness, the dynamic aspect of representation Federico Russo

314

26

What to expect when you are expecting: preferences for representation among voters and political elites Hanna Wass and Miroslav Nemčok

27

Representation and deliberation Jürg Steiner

326 338

Index346

Contributors

Rudy B. Andeweg is Professor of Political Science at Leiden University, the Netherlands. He has published on various aspects of political representation, parliamentary behaviour and coalition government. Heinrich Best is Professor Emeritus of Sociology at the University of Jena, Germany. He has published widely in the fields of state and nation building, European integration, elite studies and representative institutions. He has recently published the Palgrave Handbook of Political Elites, co-edited with John Higley et al. (2018). Mark B. Brown is a Professor in the Department of Political Science at California State University, Sacramento, USA. He is the author of Science in Democracy: Expertise, Institutions, and Representation (MIT Press, 2009), and various publications on the politics of expertise, political representation, environmental politics and related topics. Ian Budge is a political scientist who has pioneered the use of quantitative methods in studying party democracy across countries. He has been a professor at the European University Institute, Florence, Italy (1982–85), and in five other countries and is currently Emeritus Professor of the Department of Government, University of Essex. His achievements have been recognised by numerous research awards over his career. The citation for his European Award 2013 noted his ‘outstanding contribution to European Political Science … through international research projects … scholarly production and institutional service’. Daniele Caramani is Ernst B. Haas Chair in European Governance and Politics at the European University Institute, Florence, and Professor of Comparative Politics at the University of Zurich (on leave). He has recently published articles on representation and technocracy in the American Political Science Review and American Journal of Political Science, and co-edited a volume on The Technocratic Challenge to Democracy (Routledge, 2020). Dario Castiglione teaches Political Theory at the University of Exeter, UK, where he is the Director of the Centre for Political Thought. His research interests are in democratic theory, European constitutionalism, theories of sociability, and the history of modern political thought. Recent publications include the edited collection on Creating Political Presence: The New Politics of Democratic Representation (Chicago University Press, 2019), and From Maastricht to Brexit: Democracy, Constitutionalism and Citizenship in the EU (RLI-ECPR Press, 2019), co-authored with Richard Bellamy. Laura Chaqués Bonafont is Professor of Political Science at the University of Barcelona, Spain, and research fellow at the Institut Barcelona d'Estudis Internacionals (IBEI). She leads the research group ‘Quality of Democracy’ (www​.Q​-dem​.com) aimed at analysing agenda dynamics and interest groups in a comparative perspective, and participates actively in the comparative agendas project. She is also the author of several books and a large number of articles in academic journals. In 2014 she won the ICREA academia prize. vii

viii  Research handbook on political representation Alessandro Chiaramonte is Professor of Political Science at the University of Florence, Italy. He has published books and articles on elections, electoral systems and party systems. His works have appeared in international journals including West European Politics, Party Politics, Government and Opposition, Journal of Common Market Studies, South European Society and Politics, Journal of Elections, Public Opinion and Parties. Maurizio Cotta is Senior Professor of Political Science at the University of Siena, Italy. He has published on issues of comparative analysis of legislatures and executives, on political elites and the politics of the European Union. Among his recent works he has co-edited Technocratic Ministers and Political Leadership in European Democracies (2018) and The EU through Multiple Crises (2020). Bas Denters is Professor of Public Administration at the University of Twente, the Netherlands. He has published on issues of participatory and representative democracy in local government and on the issue of size and local democracy (Denters et al. Size and Local Democracy, Edward Elgar, 2014). Zachris Haaparinne is a doctoral student in European history at the University of Jyväskylä, Finland. His research focuses on petitions and representation in eighteenth-century Britain. Miriam Hänni is a senior researcher at the Swiss Federal Institute for Vocational Education and Training. In the field of representation she has primarily studied the descriptive and substantive representation of ethnic minorities and its effect on violent and non-violent protests. Pasi Ihalainen is Professor of Comparative European History at the Department of History and Ethnology, University of Jyväskylä, Finland. His research focuses on the comparative and transnational history of political discourse, nationalism, democracy and parliamentarism since the eighteenth century. His recent books include Parliament and Parliamentarism: A Comparative History of a European Concept, co-edited with Cornelia Ilie and Kari Palonen (2016), and The Springs of Democracy: National and Transnational Debates on Constitutional Reform in the British, German, Swedish and Finnish Parliaments, 1917–1919 (2017). Mona Lena Krook is Professor of Political Science and Chair of the Women and Politics PhD Program at Rutgers University, USA. Her research focuses primarily on dynamics of gender and political representation, but she has also written on the representation of young people in national parliaments in a series of reports for the Inter-Parliamentary Union. Her first book, Quotas for Women in Politics: Gender and Candidate Selection Reform Worldwide (Oxford University Press, 2009), received the Victoria Schuck Award from the American Political Science Association in 2010. Marco Lisi is Associate Professor in the Department of Political Studies, Nova University of Lisbon, Portugal and Researcher at IPRI-NOVA. His research interests focus on political parties, electoral behaviour, democratic theory, political representation and election campaigns. He has published several articles in national and international journals. He recently edited Party System Change, the European Crisis and the State of Democracy (Routledge, 2019). Tom Louwerse is Associate Professor at the Institute of Political Science, Leiden University, the Netherlands. His research focuses on political representation, legislative politics, government and opposition, and elections.

Contributors  ix Joni Lovenduski is Professor Emerita at Birkbeck College London, UK. Her published research focuses on gender, political representation and political institutions in Europe and the UK. She is chair of the board of The Political Quarterly. Leonardo Morlino is Emeritus Professor of Political Science at LUISS, Rome, Italy. His most recent book is Equality, Freedom, and Democracy: Europe After the Great Recession (Oxford University Press, 2020); he is also co-editor of the Handbook of Political Science (3 vols, Sage Publications, 2020). Elin Naurin is an Assistant Professor at the Department of Political Science, University of Gothenburg, Sweden. She studies representative democratic practices, such as politicians’ responsiveness to voters and parties’ making and breaking of election pledges. She also leads the Gothenburg Research Program on Pregnancy and Politics, where changes in political opinion and behaviour are tracked in large citizen panels to determine the effects of pregnancy and early parenthood. Miroslav Nemčok is a Finnish Cultural Foundation Postdoctoral Research Fellow at the University of Helsinki and a Research Specialist at the Masaryk University in the Czech Republic. His research focuses on electoral systems, satisfaction with democracy among citizens, and political participation. He has published in Party Politics, European Political Science Review and East European Politics and Societies, among others. Zsófia Papp is a Senior Research Fellow at the Centre for Social Sciences (Hungarian Academy of Sciences Centre of Excellence). She publishes on legislative behaviour, campaign personalisation, and electoral accountability. Her work is published in Parliamentary Affairs, European Union Politics, The International Journal of Press/Politics, Acta Politica and The Journal of Legislative Studies. She is currently serving as Associate Editor at Political Research Exchange, an Open Access journal of ECPR and Taylor & Francis. Federico Russo is Associate Professor of Political Science at the University of Salento, Italy. His research focuses on legislative behaviour, political representation and European integration. He co-edited the special section ‘Parliamentary Work, Re-Selection and Re-Election: In Search of the Accountability Link’ (Parliamentary Affairs, 2018) and the special issue ‘Policy Agendas in Italy’ (Italian Political Science Review, 2018). He currently serves as co-editor of Interdisciplinary Political Studies, an Open Access journal published by the University of Salento. Thomas Saalfeld is Professor of Political Science at the University of Bamberg, Germany. His research focuses on the substantive representation of citizens of immigrant origin in European legislatures, the political professionalization of legislators in advanced parliamentary democracies and cabinet duration. He has published extensively in peer-reviewed scholarly journals on these topics, including the British Journal of Political Science, the European Journal of Political Research, International Studies Quarterly and Parliamentary Affairs. He is an editor of The Oxford Handbook of Legislative Studies.

x  Research handbook on political representation Eline Severs is an Assistant Professor at the Department of Political Science, Vrije Universiteit Brussel, Belgium. Her research interests fall within the field of democratic theory. She is especially interested in democratic representation, the relationships between civil society and democratic governance, citizens’ conceptions of democracy, and what it means to include historically disadvantaged groups. Her recent work includes a special issue on ‘The Good Representative 2.0’ (with Suzanne Dovi, University of Arizona) in PS: Political Science and Politics (2018). Alan Siaroff is Professor of Political Science at the University of Lethbridge, Canada, and the author of Comparative European Party Systems: An Analysis of Parliamentary Elections Since 1945, 2nd edn, Comparing Political Regimes: A Thematic Introduction to Comparative Politics, 3rd edn, and various articles in both European and North American journals. Jürg Steiner is Professor Emeritus of Political Science at the University of North Carolina at Chapel Hill, USA and the University of Bern, Switzerland. Among his recent works is The Foundations of Deliberative Democracy (Cambridge University Press). Robert Thomson is Professor of Political Science at Monash University, Melbourne, Australia. His research focuses of comparisons of democratic representation, including work on election campaigns, and international governance, including work on the European Union. Lori Thorlakson is Professor in the Department of Political Science at the University of Alberta, Canada, where she was a Jean Monnet chair and the founding director of the European Union Centre of Excellence. She holds a PhD from the London School of Economics. She has previously taught at the University of Nottingham and from 2004–05 held a fellowship at the European University Institute. She is the author of the book Multi-Level Democracy (Oxford University Press, 2020). Simon Tormey is a Professor of Politics at the University of Bristol, UK. He is the author of many books and articles, most recently Populism (Oneworld, 2019). Luca Verzichelli is a Professor of Political Science at the University of Siena, where he teaches Italian Politics and Global Comparative Politics. Lars Vogel is Post-Doctoral Researcher at the Department of Political Science at Leipzig University, Germany. His research interests include political representation, elite studies, and political culture. He has recently co-edited the volume The Contested Status of Political Elites: At the Crossroads (Routledge 2018). Hanna Wass is an Adjunct Professor in Political Science at the Faculty of Social Sciences, the University of Helsinki, Finland. She is a work package leader in the research project Tackling the Biases and Bubbles in Participation (BIBU), funded by the Strategic Research Council. She is a member of the steering committee for the Finnish National Election Study (FNES) and a co-convenor of the ECPR standing group Public Opinion and Voting Behavior. Her work on political participation and representation has been published in several leading journals such as Public Opinion Quarterly, Political Research Quarterly, Electoral Studies, and the European Journal of Political Research. She is a co-author of the book Health and Political Engagement (Routledge, 2017).

General introduction to the Research Handbook on Political Representation Maurizio Cotta and Federico Russo

POLITICAL REPRESENTATION IN CONTEMPORARY DEMOCRACY AND BEYOND The contemporary model of democracy, one of the historically most relevant inventions in the field of ‘political technology’, originally conceived and developed during the last two centuries in Europe and in its external projections (such as the United States) and then diffused in different parts of the world, has at its centre the idea of representation and the crafting of institutions to translate this into practice. The representative qualifier has thus come to be strictly associated with the noun ‘democracy’ to define the new model of representative democracy. This development, which has significantly contributed to shaping the politics of a large part of the world, is both the result of passionate ideological and philosophical debates and of very concrete and often dramatic political struggles, as well as of often contested institutional choices. The two dimensions, ideational and practical, have been closely intertwined, though sometimes conflicting. Theories of democracy and representation have inspired political struggles and institutional choices, but they have also been used to criticize the latter as insufficient or misdirected. In turn, practical choices have sometimes been guided more by expediency and compromises among political actors than by full attention to principles. In the context of contemporary democracy, representation has thus obtained the position of the central and powerful metaphor used to address the oldest and most crucial problem of politics: how to normatively prescribe, practically organize and empirically describe the asymmetric and dangerous relationship between rulers and ruled. The solution offered by representation to this problem is seen as providing what is today (at least in principle) the predominant source of authorization and legitimation for the rulers (they can rule because they represent the sovereign people). It is also seen as the instrument that enables influence and control over those who detain power by those who do not wield direct power, and ensures that those who rule can be called to respond to the demands of those who are ruled, and in some way held to account. When discussing the concept of political representation and its ramifications, we must therefore remember they are the result of extensive theoretical, political and institutional elaborations and of the enormous quantity of practices which have taken place over two centuries and more, first in a limited number, then in a multitude of countries. Stretching over time and across countries, the ideas and realities of representation have been challenged and contaminated by a huge variety of socio-political realities. Large or small communities, territorially homogeneous or heterogeneous countries, egalitarian or highly unequal societies, the presence or absence of ethnic minorities, the politicization or non-politicization of gender differences, the rise of parties with their variable organizational forms, the activation of social and political movements mobilizing population sections: these and many other aspects have caused the concept of representation to be continuously reinterpreted and different institutional solutions crafted to make it practicable. It should not be a surprise that a sort of Pandora’s box opens up as soon as we start unpacking the concept. This is not to say, however, that the idea of representation did not exist before this period and that practices embodying it have not been used in the context of rather different political models. It is well known, in fact, that in Europe the modern concept and concrete forms of repxi

xii  Research handbook on political representation resentation were not developed ex nihilo, but were able to profit significantly from pre-existing experiences and theories. They did so, either by incrementally transforming and reinterpreting these traditions (the English experience is probably the best example) or, a contrario, by contesting and repudiating them as outdated and no longer acceptable and by inventing radically new formulas (the French revolutionary experience is here the best example). We are thus allowed to say that modern political representation is at the same time a truly new phenomenon, but it is also vitally linked to a significant chain of antecedents. Shifting our observation in another direction, we can say that even today the standard institutions and mechanisms of representation embodied in the contemporary form of liberal democracy do not exhaust the phenomenon. Different ideas and practices of representation also exist beyond the borders of that political model. Inspired by different principles, corporatist and technocratic forms of representation work sometimes in a complementary, sometimes in a conflicting way with the standard version. Moreover, different types of societal movements and aggregations do indeed produce elements of representation that do not fit easily into electoral representation. Finally, it must be remembered that more frequently than expected, non-democratic regimes adopt (and adapt) ‘representative’ institutions, mimicking those typical of democratic regimes with the purpose of enhancing their legitimacy, strengthening popular support and harnessing potential oppositions.

A RELATIVELY SIMPLE INSTITUTIONAL MODEL, BUT WITH MUCH LESS OBVIOUS MEANINGS In contemporary liberal democracies, representation is implemented through what is at first sight a relatively straightforward institutional configuration. This obviously refers to a result that was often far from easy and complicated to achieve. Popular elections of parliamentary assemblies endowed with law-making powers and in some cases of the head of state as well (often, but not always, also combining the role of head of the executive) are the typical institutional translations of representation at the national level, and more or less similar formats are reproduced at regional or municipal levels. In Europe an innovative supra-national application has also been developed with the direct election of the European Parliament (as well as with the indirect representation of member states through the Council of the European Union). As is explicitly declared by most democratic constitutions, members of parliament (and directly or indirectly, the members of the executive) are thus institutionally defined as the representatives, while the citizens/ the nation/ the people are alternatively indicated as the represented. Specific arrangements concerning electoral systems, configurations of parliaments and of parliament–executive relations may vary, but more or less within the frame of a relatively common institutional design. The coexistence also of representative institutions pertaining to different territorial levels – sub-national, national and in some cases supra-national – makes the institutional settings more complex, but does not alter the basic frame of reference. If the institutional set-up typical of democratic regimes has, so to speak, provided a normatively formalized version of representation, by which a relatively small set of people elected by a much larger set of other people has been seen as the way to achieve this goal, what this more precisely entails is much less clear and often contested. The ambiguous and polysemic character of the representation metaphor immediately emerges and is matched by a multiplicity of meanings and practices as soon as we open the box. The literal meaning implied by representation – making present what is otherwise absent – is the most obvious starting point for disentangling the concept, but at closer inspection is unable to define in a univocal and precise way the substance of the phenomenon. In fact, both at the normative level (what representation should be) and at the empirical level (what it is, in fact) a rather wide space of variation opens up. At least three sets of questions emerge: (a) who

General introduction  xiii is/are the representative/s; (b) who is/are the represented; (c) what kind of relational linkage unites and what distance divides the two terms of the representation equation? These questions have been the subjects on the one hand of intense theoretical and normative discussions and on the other of wide-ranging empirical analyses of their variable practical implementations. Are the represented the individual citizens, groups or categories of citizens, or a comprehensive and abstract collective entity (called the people, the nation, or sometimes the class)? When an individual, a social group or category, or an abstract collective entity is said to be represented, what is in fact at stake? Their interests, value orientations, specific demands, peculiar personal features, identities? All are in fact potentially involved. And if we also bring the temporal dimension into play, we can further distinguish between short-term and long-term interests, attitudes, and so on. On the other side of the representative relation, who do we expect the representative to be? The individual member of parliament, the party in parliament (or outside), a leader, or the parliament as a whole? Behind the answers given to these questions at the normative level it is not too difficult to detect rather different and potentially conflicting conceptions about the political body and its components, but also about democracy and its relationship with society. At the empirical level, the answers will reflect variations in the practical working of the political and social conditions which directly or indirectly impinge upon the representation processes. For instance, just to mention a few factors, changes in the extension of franchise, variations in electoral systems, the development (and decline) of mass organized parties, or at a more sociological level the decay of class society, the fragmentation of interests, and the mobilization of identity-based groups, can significantly alter the meaning of who is represented and who is representing. The research strategies will change accordingly. Questions about who is represented and who represents do not exhaust the representation debate. The nature of the relationship between the two terms of the equation is equally important both at the normative and at the empirical level. This element, while obviously influenced by the specific features of represented and representatives, deserves specific attention. As amply discussed in the literature on representation, a variety of aspects are at play here. There is first of all the aspect of authorization, that is, how (and within what limits) the representative is legitimized to act on behalf of the represented and to bind them by his or her actions. It may be seen as a purely formalistic fact, but we should not underestimate the symbolic value that can be attached to it: to be a representative of the people carries both prestige and responsibilities. The second aspect concerns the actions of the representatives and the criteria that (should) guide them and by which they can be evaluated. Acting in the interest of the represented is here the apparently easy answer. It is much less easy to specify what precisely defines this ‘acting in the interest’ and how the representatives should operate in order to absolve this duty. Does it entail special norms for guiding their actions (and for evaluating them)? Leaving aside the big problems arising from the concept of interest itself, the degree of freedom/constraint for the representative in interpreting and implementing the interest(s) of the represented is here the crucial issue which has animated the long and recurrently re-emerging debate on the delegate/trustee alternative. Both normatively and empirically the answer to this dilemma is strongly linked to the nature of the represented, but also to the prevalence of elitist or participatory interpretations of democracy. Is democracy a predominantly elite controlled process or is it rather driven by popular participation? Responsiveness and accountability are themes strictly linked to the solutions given to the previous questions. How important are they for the fulfilment of representation and how do they work in practice? But is representing necessarily and exclusively about interests (or, in other words, demands…) and the actions which (should) respond to them? Or is there something else at stake which has less to do with actions and more to do with identities (being rather than acting)? The very old idea of peer representation, which to a large extent inspired the format of pre-modern parliaments,

xiv  Research handbook on political representation has made a strong comeback in contemporary times, particularly, but not only, in response to the highly gender-biased state of representation. In this and other cases, the question of the degree to which representatives (should) reflect the identity, that is to say selected personal features, of the represented has gained prominence. Finally, in times of populistic politics, the personality traits of the individual leader may operate as a drastic shortcut in the representation equation. With regard to the representative relationship, a further point has to be raised. Representation is often discussed both from the normative and from the empirical point of view in a static perspective. The representative relationship (with all its potential shades discussed above) is often observed and evaluated in terms of the degree of congruence existing at a given point in time between the ‘positions’ (or identities) of the representatives and those of the represented, viewed as fixed. This, however, is a drastic simplification of representation which in democratic systems should be seen more as a process than as a static relationship. As signalled particularly by constructivist approaches, we should rather see the positions/identities of the two sides as constantly reshaped through the process of representation. This happens both in a top-down and in a bottom-up direction. In normal times, in particular, we may expect the top-down influence to be especially strong, as representative elites are in a better position to propose and suggest representative contents in simplified formulas to which the would-be represented can more easily react. On the other hand, however, we should not underestimate the extent to which, in times of accelerated change, representative elites can be under the unexpected pressure of new orientations and demands originating from below and be stimulated to redefine their offer.

TENSIONS BETWEEN NORMATIVE AND EMPIRICAL DIMENSIONS OF REPRESENTATION As we have already mentioned, the debate on representation entails both normative as well as empirical dimensions. The distinction between them, but also their important connections, must be underlined. On the one hand, it is quite clear that the concept of representation, particularly when linked to that of democracy (which redefines it as democratic representation), is strongly charged with normative content but at the same times acknowledges the constraints of reality. It proposes, in fact, an answer to the crucial problem of the relationship between rulers and ruled, which is highly demanding as it combines the idea of popular sovereignty (and all the expectations deriving from it) with the acceptance of the fact that governing is in the hands of a relatively restricted group of (elected) politicians. On the other hand, if we look at representation from the empirical side, we can observe a set of phenomena which are surely inspired by normative ideals but in their working are even more strongly conditioned by the ‘hard laws’ of politics. The limited information of citizens and the failures of collective action compared to the much greater resources of political elites and the weight of their self-interest obviously contribute to the factual experience of representation and to the structural imbalance in the relationship between represented and representatives. This means that expectations as to the working of representation generated by the norms are necessarily going to be disappointed to a greater or lesser degree. The quality of representation ordinarily supplied by representatives will easily attract the criticism of (both supportive and hostile) observers as being below the proclaimed normative standards. It will be simple to show that representatives fail to act in the interest of the represented, that they rather pursue their own goals, in particular that of preserving their seat, or that they are not men (and women) of the people but a privileged caste, and so on. In view of these failings, the voice of populist movements offering, as an alternative, direct identification with the leader or the party as a close community may have easy appeal for many citizens. A more realistic perspective, however, would suggest that ideal norms, rather than defining an empirical model that can be fully achieved in real life, should be seen as a ‘pull factor’ which can

General introduction  xv constantly help to make reality less distant from the ideal. They can also be used as a criterion of evaluation in the comparison between different political experiences. In this perspective, what we call representation may in real life offer a very imperfect representation of citizens and their interests but still appears to be the most effective instrument available to citizens for exerting a continuous influence over rulers.

THE HANDBOOK SECTIONS In the first section of the Handbook, the theoretical and normative debate is widely discussed. There is little doubt that the scholarly debate on what political representation is and what it does has never been so heated as in recent years. The broad consensus reached in the 1960s on the concept and practices of representation, as defined by the likes of Pitkin and Dahl, persisted almost unquestioned until the end of the last century. This is not to deny that within the Western world several political actors and thinkers criticized representation and its institutional form, questioning their democratic credentials; nevertheless, the theoretical language to discuss these issues was shared among all parts. This gradually changed at the turn of the century, when several pillars of the standard account were brought into question and, more or less at the same time, the difficult social and economic conditions in which democracies had to operate cast doubts on the capacity of political institutions to accommodate citizens’ needs, giving voice to old and new critiques. All in all, the chapters in this section highlight that while representation is still a contested concept, whose fundamental aspects can be questioned and redefined, no coherent theoretical or institutional alternatives to surpass it have yet been advanced. The second section is devoted on the one hand to a detailed discussion of the historical transformations of parliamentary representation, and on the other to exploring other models of representation which coexist with the former in contemporary democratic polities or which adapt or simulate this model in authoritarian regimes. With regard to the first aspect, the first chapters of this section discuss the changing meanings and institutional implementations of parliamentary representation from pre-modern experiences to the gradual developments linked to the process of democratization until the stage of party democracy. As to the second aspect, other chapters will analyse how corporatist and technocratic ideas offer the bases for other forms of representation which are partially complementary and partially antagonistic to the parliamentary one. The special case of non-democratic regimes and how they cope with their problems of legitimation, consensus building by exploiting some typical democratic institutions in a manipulative way, is also discussed. As representation in contemporary democratic systems has been implemented through a set of specific institutions, the third section provides an articulated discussion of variations in the design of these instruments and how they are related to different aspects and interpretations of the representative idea. Special attention is paid to electoral systems, to the territorial dimension of representation, to the relationship between size of jurisdictions and democracy, and to the problems of ensuring the representation of different types of minorities. The fourth section offers an overview of the most consolidated empirical programmes studying political representation, while also reserving some space for some relatively new research agendas. A first group of chapters focuses on individual representatives, reviewing existing typologies and the factors shaping parliamentarians’ behaviour, such as methods of selection and electoral systems. A second block of chapters shifts the focus from dyadic to aggregate forms of representation, looking at what interest groups, parties and governments do: in this vein the methods and findings of the rich body of literature on interest representation, policy congruence and government pledges are reviewed. Finally, the section also gives attention to less consolidated but promising fields of enquiry including studies concerning policy and non-policy responsiveness, citizens’ expectations about representation and the relationship between representation and deliberation.

SECTION I CONCEPTUAL AND THEORETICAL DISCUSSIONS

INTRODUCTION Federico Russo Section I explores and takes stock of the most important theoretical and conceptual debates that have arisen around the problem of political representation. This concept has a prominent place in contemporary political science literature simply because it also has a central place in democratic theory. Contemporary democracies are representative democracies by definition. However, the concept is not only central but also contested, controversial and complex. Recently, there has been a revival of theoretical studies on political representation reflecting two apparently contrasting trends: on the one hand, democratic theorists have rediscovered the centrality of the institutions of liberal democracy to pursue deliberation and participation, once seen as standing in contrast with representation. On the other hand, the theory of political representation has been innovated to surpass the standard model exclusively based on electoral representation. This ferment has not yet produced a consensus amongst scholars interested in the concept and practices of political representation, as it is reflected in the multiple standards utilized to assess the legitimacy and the quality of representation. In their first chapter, entitled ‘Political representation’, Russo and Cotta review the main turning points of the academic debate and provide an overview of the questions that remain to be the objects of an ongoing debate. Whilst the origin of modern representation can be traced back to the Roman legal tradition and to medieval ideas and practices, its connection with popular sovereignty emerged after the sunset of absolutism. After the Second World War, the neo-Schumpeterian model of democracy and the theoretical discussion of representation advanced by Hanna Pitkin, both based on responsiveness, cemented the union between the two concepts. However, starting from the end of the twentieth century, several assumptions of these models have been the objects of a renewed theoretical debate whose main points are summarized by this chapter. In the second chapter, entitled ‘The system of democratic representation and its normative principles’, Castiglione offers some reflections on the criteria that make representation legiti1

2  Research handbook on political representation mate. Before undertaking this endeavour, Castiglione clears the ground of two equivocations, namely, that it is possible and useful to reveal the core meaning of representation and that representation is an institutional mechanism. The chapter embraces a systemic view of (democratic) representation, seen as a social institution whose function is intermediating between social demands and governing capacity. Castiglione also discusses some of its core features and normative principles. In the ‘Democracy and representation’ chapter, Brown thoroughly examines the relationship between democracy and representation as it unfolded during the history of political thought, especially since the mid-twentieth century. The debate culminated with the emergence of the representative turn in democratic theory and the constructivist view of political representation, which challenged the standard model of representation based on the idea that existing constituencies, with fixed preferences, obtain representation through electoral politics. Whilst these perspectives help show some limitations of the standard model by downplaying the normative relevance of the opinions of citizens, they raise difficult questions about the relationship between representation and democracy. In his chapter, Brown, working within the constructivist view, elaborates on the criteria to evaluate the democratic credentials of representation. Despite the critiques coming from the constructivist camp, the concept of policy responsiveness is still central for empirical studies, which mainly work within the neo-Schumpeterian view of democracy. In the ‘Policy responsiveness and democratic quality’ chapter, Morlino, starting from an intuitive definition of responsiveness, reviews the theoretical and empirical problems related to its measurement. With regard to indicators, considering that measuring responsiveness has long been considered impossible, scholars have resorted to the measures of policy congruence or the satisfaction of citizens. Finally, the chapter spells out the connection between responsiveness and other related concepts, such as accountability, competition, participation and responsibility. The new conceptual language centred on claim making has broadened the applicability of the concept of representation well beyond the relationship between voters and their elected representatives. In the ‘Legitimacy and hegemony’ chapter, Severs analyses the literature on non-electoral representation, identifying two different perspectives with which it can be studied. On the one hand, some theorists, working within the transmission belt view of democracy, are interested in showing how non-electoral representation can contribute to the correction of some limitations of the traditional representative institutions. On the other hand, radical democratic theorists, denying the existence of objective foundations for assessing the legitimacy of democratic systems, consider non-electoral representation a preventive force against the hegemonic power of political authorities and other powerful actors. Severs further argues that two views are founded on distinct ontologies and normative concerns, leading to divergent theoretical and empirical agendas. However, pitting one against the other is not likely to pay dividends. All the chapters in this section show that political representation has been the frequent object of critiques and reconsiderations. In the ‘Challenges to political representation’ chapter, Tormey notes that by advocating the fuller forms of democracy and despite the divergent contexts in which they emerged, many of these critiques share the normative concern to address the concentration of power in political and economic elites. At the same time, the critiques of representative democracy are often sceptical about the role of political parties and thus downplay one of the factors accounting for the success of democratic systems in the last century. According to Tormey, the rise of populism should not be interpreted as a critique to representation but as the aspiration to a form of hyper representation. However, there are no signs that the future of democracy will be distinct from that of representation.

1. Political representation: concepts, theories and practices in historical perspective Federico Russo and Maurizio Cotta

INTRODUCTION The concept of representation is, without doubt, at the centre of what can be termed the most important political invention of the last two centuries, that is to say, liberal democracy. As such, it has animated vigorous theoretical and normative debates, has been used as a powerful flag in crucial political struggles (such as those for the extension of suffrage, or today for gender equality) and has inspired fundamental institutional arrangements at national, sub-national and also supra-national level. As a synthetic metaphor for delicate and complex political relations between rulers and the ruled it is a fundamentally polysemic and evocative rather than empirically analytic concept. It is not a surprise, therefore, that its meaning(s) have been highly debated throughout its history, which began well before the breakthrough of contemporary democracy. We will therefore start by briefly tracing ‘pre-democracy’ discussions on political representation, move to the issue of the relationship between representation and contemporary democracy and then devote our attention to how the debate on political representation has been influenced by the seminal work of Hanna Pitkin (1967). Next, we will discuss how her account has been discussed with regard to several dimensions, such as the direction of the relationship between represented and representatives, the proper object of representation, the normative criteria for evaluating representation and the context in which political representation takes place.

THE ORIGINS OF MODERN REPRESENTATION IN POLITICAL THOUGHT AND PRACTICES As is well known, the Roman legal tradition developed sophisticated views about representation and the powers and limitations of those standing for other people in judicial proceedings. The persisting cultural legacy of this tradition in Europe has provided, over the centuries, important conceptual elements which could also be applied in the political field. In particular, during the middle ages, when the distinctions between public and civil law were far from as neat as in the modern world, concepts deriving from that tradition were not only abundantly used in judicial proceedings but were easily transposed into other fields. In particular, they offered the bases for significant elaborations to be used for a diverse set of political and ecclesiastical practices where the need arose for people (ambassadors, delegates, etc.) to stand for cities, estates, communities or religious orders in different decisional bodies (councils, chapters, parliaments). Special attention was given to the issue of the plena potestas of representatives, that is to say to their ability to consent to decisions taken in their presence and 3

4  Research handbook on political representation to guarantee that such decisions would be binding for the communities they represented (Post 1943). Another important theme was that of the instructiones they received from those they represented, and which enabled them to be prepared when attending such meetings. Before the absolutist turn of monarchies, medieval legal traditions had thus developed ideas and practices which to some extent contemplated a two-way flow of representation, by which secular and religious authorities, needing the consent of subordinates for extraordinary decisions, could on the one hand summon representatives of communities and estates who had the power to stand for them and thus make the decisions, taken in council, binding for the represented (Marongiu and Woolf 1968). On the other hand, the represented could give instructions to their delegates to voice demands and defend local liberties and privileges. These ‘representational’ experiences had severe limitations, as their degree of institutionalization was often defective and as the top-down dimension was generally stronger than the bottom-up one. We know, however, that in some cases (Britain being the most notable one) they were able to survive until the dawn of modernity. The rise of absolutism in most European polities silenced these traditions to a large extent and gave a very different meaning to the concept of representation, as it was now the absolute sovereign (and his theorists) who seized and drastically transformed this concept. As expressed theoretically by Hobbes in the Leviathan (1651), and more practically interpreted by absolutist kings in real life, the sovereign was now seen as the representative of the ‘commonwealth’, that is to say, of the country seen as a unified body, while the pre-modern pluralism of society was progressively rejected. In fact, it was only through this ‘representational’ relationship between sovereign and country that the polity could be constructed as a unified entity and not reduced to a simple sum of individuals (Runciman 2009). This idea of representation, which also included a notion of authorization of the sovereign by the multitude, did not contemplate any role of control over him by corporate groups and even less by individuals. However, where traditional representative institutions managed to persist, as in Britain, the United Provinces of Netherlands and Sweden, and were undergoing an evolutionary transformation to face changing historical conditions (see Chapter 7 in this Handbook), or as in 1789 France where the monarchy, unable to face a critical situation, had revived them, or as in the United States where they were created in the building of a new polity, discussions about representation gained a new momentum both in intellectuals’ debates and in more down-to-earth political discussions. The characteristics of representative government and how the latter relates to democracy, the nature of the relationship between representative and represented, the role and freedom of the representative, what was to be the object of representation (individuals, interests, a pluralistic society or a wholistic entity): all these themes which then became the centre of discussions would continue to animate the debate on representation until today in spite of enormously changing social and political conditions. Montesquieu, in the eleventh book of his Esprit des Lois, while developing a comparative analysis of government types and using the English constitution as a model of free government, had already developed an elaborate version of representative government which was to have a fundamental influence on modern political conceptions (Montesquieu 1748). A representative assembly, endowed with law-making powers and capable of transmitting the views of the people, was openly presented as the instrument for ensuring in a large state, where direct democracy was considered impossible, the respect of the principle of self-government that free citizens have the right to enjoy.

Concepts, theories and practices in historical perspective  5 Very much along the lines of Montesquieu’s contribution, two crucial themes – a new view of society and of its internal articulation and the problem of power limitations – provided the background for lively political discussions in this period. The debate on representation and its institutional forms concerned both aspects seen as the instrument for giving voice on the one hand to the plural interests of society, and on the other hand for limiting absolute power. An important theme was also the problem of how to translate the plurality of voices into unified decisions for the common good. In Britain, the debate on the respective role of representatives and represented was epitomized in the famous electoral discourse of Edmund Burke to the electors of Bristol. In that speech he very explicitly touched on a point that was to have foundational value for the modern interpretation of representation, but at the same time would continue to be debated under different guises in the following centuries. The Burkean position did not deny the duty of the member of parliament to pay ‘high respect’ to the opinion of his constituency, as he would put it, implying by this a broadly conceived responsiveness, but affirmed clearly that a representative was not an ‘ambassador’ bound by ‘authoritative instructions, mandates… which [he] is bound blindly…to obey’, but a trustee endowed with the freedom (and responsibility) of deciding according to ‘his mature judgment’, as government is not ‘a matter of will’ but ‘government and legislation are matters of reason and judgment’ (Burke 1854 [1774], pp. 446–8). Central to this idea was the novel interpretation of parliaments and parliamentarians not as instruments for expressing particularistic grievances and demands arising from constituencies, but as governing and legislating bodies. In order that these could perform effectively, their members needed to be able to deliberate freely in order to serve the interests of the country seen as a whole (‘parliament is a deliberative assembly of one nation, with one interest’) (1854) [1774], p. 448). This debate inspired the prohibition of the imperative mandate written in many modern constitutions. We must remember, however, that in less principled and more pragmatic ways, the theme of the independence of MPs versus their duties to respect the mandates of the voters has recurrently re-emerged. It was, for instance, central to the empirical research conducted in the 1960s by Wahlke et al. (1962) on representative roles adopted by American legislators. More recently it returns in the discussion on electoral promises or pledges. The other grand theme was that of the relationship between representation and democracy, which became particularly intense at the turn of the nineteenth century but has continued since then. In the eighteenth century, democracy and representative government were typically presented as alternative systems and the different sides exposed their contrasting peculiarities and the reasons for their preferences. There were practical reasons, linked in particular to the size of the community as in the analysis of Montesquieu and of those building on this tradition. This discussion also implied, however, more explicitly normative reasons. Rousseau’s choice of the classic interpretation of (direct) democracy and his disparaging evaluation of representation hinged very much on the idea that the making of laws, as the highest expression of sovereignty (and of the volonté générale), could not be alienated by the people to representatives without renouncing liberty (Rousseau 1762). The sovereign people could at most alienate to representatives the governing power interpreted as an inferior function and not an embodiment of sovereignty. On the opposite side to Rousseau, Sieyès defended the merits of the (indirect) representative model precisely because it created a beneficial separation between the particularism of the

6  Research handbook on political representation individual voters and the required generality of the law-making function attributed to parliament. Representation was thus seen as the instrument for achieving the common interest of the nation (Sieyès 1789).

THE ‘MARRIAGE’ BETWEEN REPRESENTATION AND DEMOCRACY In the political practice of Europe, where representative institutions (elected parliaments), inspired by liberal constitutionalism and made possible in most countries by a compromise between monarchies and reformers, became the dominant political form, conservative and liberal ruling classes resisted for decades what had come to be called ‘democracy’, that is to say, universal male franchise, as a dangerous threat to the established political and social order. But this resistance had to give way, sometimes gradually but in other cases more abruptly, to pressures from below and strategic changes promoted from above, producing in real life the marriage between democracy and representation (Mill 1861), which at the end of the eighteenth century had seemed impossible. This marriage meant that previously excluded social classes could effectively advance representative claims and that these could be satisfied through important changes in the working if not in the form of representative institutions. Representation could no longer be a game played only inside rather restricted circles. The rise of mass parties produced fundamental transformations both in electoral practices and in the functioning of representative assemblies. On the other hand, democracy ceased to be seen as an alternative model, lost the original classic meaning of direct popular government and adopted the central mechanisms of liberal constitutionalism (electoral representation, division of powers, etc.). What happened in practice (but we must remember that in a significant number of cases this transformation did not succeed and from this failure new models of authoritarian and totalitarian regimes emerged) was also reflected in the new theorizing of democracy. From the extremely influential reading of Schumpeter to Dahl’s and Sartori’s more articulated formulations, the contemporary standard model of representative democracy emerged. Schumpeter, with his elitist interpretation of democracy focusing on the combined effects of pluralistic competition between opposed elites and the deciding role of elections, stripped representation of much of its normative and symbolic narrative (in particular by contesting the idea of a bottom-up process of transmission of popular will) and reformulated it as a rather blunt mechanism of selection by which top-down dynamics acquired a dominant role (Schumpeter 1943). This view provided an important inspiration for Downs’ Economic Theory of Democracy (1957) which, without using the word ‘representation’, introduced into scientific discourse on democracy a rational theory of the working of electoral representation, which saw the vote maximizing parties (the elites in Schumpeterian language) as the driving engine of the democratic process. Downs, however, compared with Schumpeter, paid much greater attention to the role of voters’ rational choices and to their influence in determining the positioning of parties. Not many years later, in his systematic discussion on the contemporary model of democracy (as opposed to the then significant challenge coming from the totalitarian model) Sartori explicitly and extensively placed the representative character as the centrepiece of modern liberal democracy (1962, ch. VI and pp. 252–66). The opposition between representative

Concepts, theories and practices in historical perspective  7 democracy and direct democracy was not based, for Sartori, only on the practical impossibility of the second, but also on the much more relevant combination of representation, political freedom and power limitations typical of the first. With regard to representation, Sartori was particularly keen on highlighting the proposing role of leaders, but also the counter influence of voters, enabled by the mechanism of ‘anticipated reactions’ (Friedrich 1941, pp. 589–91) by which representatives would be pushed to incorporate the orientations of the public (Sartori 1962, pp. 125–6). With a different approach and starting from a more ideal and normative view of democracy, Robert Dahl, particularly in his book Polyarchy, wrote that ‘the key characteristic of a democracy is the continuing responsiveness of the government to the preferences of its citizens…’ (Dahl 1971, pp. 1–2). Responsiveness (that is to say one of the aspects of the modern conception of representation), together with the institutional instruments needed to make it possible, which became the empirical attributes for assessing the democratic quality of a country, was thus given a central place in the discussion about democracy. Normative and empirical levels of analysis were in this way linked. Within this interpretation of democracy, which has become largely accepted in the scientific community, the centrality of representation, but also its internal paradoxes and limitations, has continued to stimulate a lively scientific debate on its multiple meanings and the problems it continues to raise.

THE CONTEMPORARY DEBATE: PITKIN’S HERITAGE The contemporary debate on representation has been profoundly shaped by Hanna Pitkin and her enormously influential treatment of the concept of political representation (1967) which soon became the standard reference for both normative theorists and empirical scholars. Building on J.L. Austin’s language theory, Pitkin found that the many different usages of the term ‘representation’ shared at least one core meaning: ‘the making present of something which is nevertheless not literally present’. In this sense, representation has an evidently paradoxical nature. In particular, Pitkin discussed different ways of looking at representation that had often appeared in political thought: formalistic representation, descriptive representation, symbolic representation. Each is seen as a picture of a complex object taken from a different angle: none of them is wrong, but taken in isolation they give an incomplete and misleading idea of the thing. Finally, she advances her own definition of substantive representation as ‘acting in the interest of the represented, in a manner responsive to them’ (Pitkin 1967, p. 209). Before analysing this definition more in depth, which is essential to understanding the current debate on political representation, it is useful to summarize the main points advanced by Pitkin in her review of the existing literature. The concepts of formalistic, descriptive and symbolic representation are often cited in the political science literature and we might be tempted to take them for granted. In fact, as often happens with classics, Pitkin’s book is more often cited than read, and some of its ideas have been progressively distorted. Formalistic views of representation focus on the institutional aspect of representation, but tell nothing of its substantive content. The representative relation is seen as a transaction between two parties and defined either in terms of authorization to act, or of accountability for the actions taken. Theorists who place emphasis on authorization see elections as acts giving officials the status of representatives. For those stressing accountability, a representative is one that will have to answer to another for what he or she does or, in other words, must be respon-

8  Research handbook on political representation sible to the represented. Putting together these diametrically opposed (though equally formalistic) views, the representative is someone who has been authorized to act and at the same time is to be held to account for his or her actions. However, nothing is said about the activities that a representative should carry out in order for representation to emerge. Similarly, no criteria are given to evaluate the quality of the representative. It is more interesting to discuss which characteristics should be represented, because Pitkin’s account of descriptive representation is often misleadingly reduced to the idea of social representativeness. By contrast, according to Pitkin, theorists of descriptive representation usually indicate that the representative assembly should mirror the opinions of the community; it is no coincidence that Downs’ spatial theory of party competition (Downs 1957) is included within this group. Pitkin does mention the social characteristics of the representatives, but she is very clear about one point: they are only important inasmuch as they have consequences for the opinions of the representatives. Symbolic representation has nothing to do with authorization, accountability or reflection of the popular will: however, it is a complement to all these factors. This doctrine reserves a central place to the perceptions of the represented: there is representation when the represented feel loyalty towards the representative. Focusing on the affective and emotional dimension of the representative relations has two main implications: first, ‘so long as people accept or believe, the political leader represents them by definition’ (Pitkin 1967, p. 102); second, it is worth noting that there are no requirements as to how this belief is formed. Leaders can build such feelings by manipulating the people, and shaping their identity and beliefs: as a consequence, symbolic representation need not be democratic but can be consistent with authoritarian doctrines. Finally, after reviewing different uses of the concept that give only partial ideas of what representation is and what representatives should do, Pitkin advances her own concept: ‘representing […] means acting in the interest of the represented, in a manner responsive to them’ (1967, p. 209). The focus is here on (substantive) actions rather than on the (formal) authorization to act. What should be represented are the interests of the represented, but representatives must minimize instances of conflict by taking into account the opinions of their constituents. When he or she acts in ways that are not in line with constituents’ opinions, the representative must at least have a good reason in terms of their substantive interest. Reacting positively to the demands of the constituents is perhaps Pitkin’s conception of responsiveness. However, Pitkin argues that representation is not a kind of person-to-person relationship between a representative and his or her constituency but an institutionalized arrangement involving multiple groups. The activities of single representatives contribute to the functioning of a more general system that can be considered representative insofar as it makes the people (or the constituency) present in governmental action.

TAKING STOCK OF THE CURRENT DEBATE As the previous discussion has shown, political representation is a complex and multidimensional concept. In order to clarify how contemporary authors contest or distance themselves from Pitkin’s definitions it is useful to schematically list four components that are specified in most accounts of political representation:1 1. 2. 3. 4.

the relationship between represented and representatives; the object of representation; the normative criteria to evaluate representation; the political context in which political representation takes place.

Concepts, theories and practices in historical perspective  9 Pitkin’s concept of representation is inextricably linked with democratic settings. As she later admitted, her book ‘more or less equated democracy with representation’ (2004, p. 336), treating their empirical equivalence as an axiom on the basis of what, at the time, appeared to be a truism: in modern societies democracy can only be representative (for an in-depth analysis of the relation between democracy and representation see Chapter 3 in this Handbook). This point of departure sets obvious constraints to the identity of the two sides involved in the representative relation, which were quite unproblematically identified with citizens and members of parliament. However, what counts is not so much the dyadic relationship between constituents and their own representative as the aggregate interactions taking place at the systemic level. With regard to the object of the representative relation, Pitkin specified that there is no representation unless representatives substantively act in the interest of the represented. However, this apparently straightforward view causes a paradox: is there legitimate representation when citizens do not agree with their parliamentarians’ actions? The solution proposed by Pitkin is rather vague: the representative must generally avoid this conflict, but when it sporadically emerges he or she must give an explanation of why citizens’ wishes are not in accord with their actual interests. In other words, representatives should be responsive to citizens’ requests but the proper object of representation is ‘interests’ rather than ‘opinions’. For several decades since the publishing of The Concept of Representation (1967) the theoretical debate on the topic, despite a few exceptions, languished. On the one hand, political theorists were more focused on different concepts, most notably political participation. On the other hand, empirical political scientists often paid lip service to the work of Pitkin without fully grasping its critical significance, and kept working within the neo-Schumpeterian paradigm (Urbinati and Warren 2008), as it was elaborated in the works of Sartori and Dahl. The Relationship between Represented and Representatives In Pitkin’s framework, the relation between citizens and elected politicians is unidirectional: representatives should be responsive to citizens, and not the other way around. This aspect fitted well with the procedural view of liberal democracy which constituted the prevailing view of the time. Pitkin’s critique notwithstanding, around the turn of the millennium the formalistic view of representation gained new currency in the empirical study of democratic representation after being translated into the language of agency theory (Kiewiet and McCubbins 1991; Strøm 2000; Strøm et al. 2003). Agency theory is especially concerned with situations in which one actor (the agent) acts on behalf of another (the principal). Representative democracy is thus seen as a chain of delegation of authority going from voters to elected representatives and eventually to bureaucracy, which is complemented by a corresponding chain of accountability running in the opposite direction. While the focus on delegation (authorization) and accountability is not especially new, the tools of agency theory have been used to analyse the mechanisms favouring or impeding the emergence of agency loss, offering an intuitive criterion to judge the quality of the representative relation. One limitation of conceptualizing representation as an agency relation rests in the inherently unidirectional nature of agency theory: when applied to democratic representation it suggests that citizens have endogenously formed preferences that are aggregated (for instance at the constituency level) and then represented. When applied only to the relationship between majority congressional parties and various institutions to which they delegate power, as originally done by Kiewiet and McCubbins (1991), this aspect is not problematic because parties are obviously capable of developing

10  Research handbook on political representation autonomous policy preferences. However, when broadening the scope of this framework and considering citizens as the ultimate principal, the assumption that their preferences are endogenously formed becomes problematic (Andeweg and Thomassen 2005; Disch 2011). Conceptualizing representation in the form of a principal–agent relation is a constitutive part of what Castiglione and Warren (2006) define as the standard version of political representation, but it captures only its formal and institutional structure: a more comprehensive picture needs to recognize that different opinions, interests and identities are formed in a deliberative space involving several actors. The recent constructivist turn in the study of political representation (Saward 2010 and 2017; Disch 2011) brings to the fore the constitutive nature of political representation. While in the principal–agent view the represented are assumed to be fixed, constructivists argue that representation is a dynamic process in which different types of constituencies (for instance territorial, functional or identity based) are constituted by the interaction between those who claim to be representatives and those who are depicted as represented. In other words, the activity of representing has ontological consequences, because it not only mirrors but actively shapes the identity and preferences of the represented. According to Saward, the performative instrument employed by would-be representatives is the representative claim, in which ‘a maker of representations (‘M’) puts forward a subject (‘S’) that stands for an object (‘O’), that is related to a referent (‘R’) and is offered to an audience (‘A’)’ (Saward 2010, p. 36). This framework can be applied to electoral and non-electoral contexts, in democratic and non-democratic settings. It must be noted that this framework is very general because it can be applied not only to cases where an actor is at the same time ‘maker’ and ‘subject’ of the representative claim, for instance when a politician puts herself forward as representative of a constituency, but also to cases in which the maker and the subject are different entities (for instance when a party leader puts forward his party as representative of the working class). It is also important to note the distinction between the object of representation and the referent. In the classic example, a member of parliament (maker) offers herself (subject) as standing for the interests of the constituency (object) to the members of that constituency (audience). In this case the referent consists of the citizens actually living in the constituency, but the object is the depiction of those citizens by the politician (for instance the hardworking people living in the city). On the one hand, this view can bridge a gap between the unidirectional views of representation and the results of empirical studies stressing that the preferences held by citizens are formed within the political process and with the active role of the political elites. On the other hand, as clearly put by Disch (2015), this account has potentially disruptive consequences for the normative attempt to distinguish legitimate and illegitimate instances of representation because, by equating a successful claim with one that resounds with a certain audience, it leaves no way to differentiate between the case in which someone is subjectively feeling represented and that in which someone’s interests are actually pursued. This problem resembles Pitkin’s scepticism toward symbolic representation: wherever we look at the subjective attitudes of citizens to recognize instances of representation, the problem of the citizens’ capacity to form independent judgements emerges.

Concepts, theories and practices in historical perspective  11 The Object of Representation The question of whether (class and sectoral) interests or citizens’ opinions should be represented was widely debated by political thinkers interested in the theory of representation (Birch 1972). The advent of mass politics and the development of party discipline apparently solved a problem, as political scientists saw in the theory of the electoral mandate a solution for avoiding an otherwise unsolvable problem. When the Committee on Political Parties of the American Political Science Association (1950), chaired by E.E. Schattschneider, formalized the responsible party model it also provided a pragmatic reply to the question of what should be represented: the policy positions presented by the parties in their electoral manifestos. Policy positions belong to the realm of opinions, but as long as voters are conceived as rational actors, their opinions should not differ too much from real interests. Pitkin (1967, p. 236) acknowledged the empirical role of parties and elections in the functioning of representative governments, but guarded against confounding the ideal purpose of representation with the institutions established for achieving it. For Pitkin, the essence of political representation remained as substantively acting for others, pursuing the interests of the represented without going against their wishes. The relative importance of interests and opinions cannot be determined once and for all, depending on the character of the issue at stake. Among political theorists – who look more at what representation should be than at what it actually is – the idea prevails that interests are the proper object of representation. For instance, when discussing the merits of descriptive representation, what Phillips (1998) calls politics of presence, Mansbridge (1999, p. 630) states that ‘The primary function of representative democracy is to represent the substantive interests of the represented […]’. However, emphasizing the distinction between interests and opinions can be misleading except in cases when there are good reasons to suspect that citizens are systematically manipulated. In more ordinary cases the contrast is artificial because in the words of Plotke (1997, p. 32), ‘in interest representation, representatives seek primarily to pursue their constituents’ interests as defined by their expressed preferences’. For empirical oriented scholars, interests indeed remain unobservable. For this reason opinions and wishes have been studied much more. Eulau and Karps (1977) lamented that in the work of Pitkin, the target of responsiveness was never explicitly specified, leaving political scientists without any obvious way to measure it. On what matters should the representative be responsive to her voters? By overgeneralizing the study of Miller and Stokes (1963) on constituency’s influence in the US Congress, many empirical studies equated representation with policy responsiveness: however, policy-making is only a part of the broader activities carried out by representatives.2 To overcome this limitation, Eulau and Karps enumerate four types of responsiveness which, taken together, give form to the representative relation. Beyond policy responsiveness they mention service responsiveness (granting individualized benefits to constituents), allocation responsiveness (attracting funds for the district), and symbolic responsiveness. By doing so they do not question that representation is acting on behalf of the represented, but they elucidate that representation cannot simply be assessed by looking at whether citizens get their preferred policies. The concept of policy responsiveness still has a central place in the debate on the quality of representation (for an overview see Chapter 5 in this Handbook), but it is only one component of a broader phenomenon.

12  Research handbook on political representation Normative Criteria to Evaluate Representation Agency relations are by definition dyadic, but Pitkin noted that representation should be conceived as a systemic property emerging from multiple interactions. Mansbridge (2003) acknowledges that political representation rests on different pillars, which she labels ‘promissory’, ‘anticipatory’, ‘gyroscopic’ and ‘surrogate’ forms of representation. These are best conceived as complementary mechanisms through which representation takes place, each of them requiring its own set of normative criteria against which it can be assessed. Promissory representation is just another name for the agency view: good representatives simply keep the promises made in the authorizing election. However, empirical research shows that elected politicians often try to anticipate the opinion of future voters even when this means breaking their old promises: these are instances of anticipatory representation. By admitting that voters can change their original opinion thanks to their interaction with representatives, other parties, interest groups or the media, this form of representation gives more space to deliberation. Moreover, the model makes less unrealistic assumptions as it conceives of the represented as having interests related to certain outcomes (for instance economic growth) rather than fixed policy preferences. Traditional accountability vanishes as a normative criterion, and is replaced by the quality of the deliberative relation between citizens and representatives, both conceived as being capable of autonomous judgement.3 What Mansbridge calls gyroscopic representation had been previously referred to using different terms, and basically amounts to the situation in which citizens choose a representative who shares their views and principles: ‘voters affect political outcomes not by affecting the behavior of the representative […] but by selecting and placing in the political system representatives whose behavior is to some degree predictable in advance based on their observable characteristic’ (2003, p. 521). In this case the normative criterion is the quality of deliberation at the time of the authorizing election or, to translate this into the language of agency theory, the quality of the screening process to prevent the problem of adverse selection. Finally, surrogate representation happens at the systemic level when citizens or groups are represented by someone who has not been elected by them. This mechanism is empirically very important since electoral systems are usually based on territorial representation while at present interests and opinions depend only marginally on the place where one lives and votes.4 Since there is no accountability relation between citizens and their ‘surrogate’ representatives the normative criterion to judge its legitimacy cannot be traditional accountability. Rather, at the systemic level, the normative question for surrogate representation is whether ‘each conflicting interest has proportional adversary representation in a legislative body (Weissberg 1978, esp. p. 542) and each important perspective has adequate deliberative representation’ (Mansbridge 2003, p. 524). The Political Context of Representation Since the marriage between representation and democracy (see above), empirically oriented scholars have tended to judge the legitimacy of political representation by making reference to democratic norms such as authorization, accountability and substantively acting for citizens. Following the so-called ‘representative turn’ in democratic theory, advocates of participatory forms of democracy, among whom liberal democracy was long seen as a form of thin democracy (Barber 1984), also rediscovered that representation is central to democracy, not least because it makes participation feasible (Plotke 1997; Urbinati 2006; Brito Vieira 2017). At the

Concepts, theories and practices in historical perspective  13 same time, political representation is also found in non-democratic contexts. For instance, the representative of an authoritarian country to the World Trade Organization cannot count on any democratic credential, but is universally recognized as standing for that country. Moving from this paradox, Rehfeld (2006) advanced a general theory of representation which aims to be simply descriptive (i.e. non-normative). At its core, his argument maintains that there is representation whenever a relevant group of people (the Audience) recognizes a particular actor (the Representative) claiming to represent someone else (the Represented) for the purpose of performing certain duties (the Function). Usually the Audience relies on certain rules of recognition to judge whether the claimant can be considered a representative, but these vary considerably in different contexts. In democratic contexts the Audience corresponds with the Represented and the rule to acquire the status of representative is being elected. Freeing political representation from the democratic setting, this account radically expanded the applicability of the concept which has been recently used for analysing several political phenomena once studied with different theoretical lenses (see Chapter 4 in this Handbook).

CONCLUSION The concept of political representation has been the object of several heated debates in the history of political thought and among contemporary scholars. Scholarly debates may have endogenous or exogenous causes, depending on whether they are triggered by theoretical innovations in related fields or by the evolution of social and political reality. What makes representation so central to the reflections of both political theorists and empirical political scientists is the central place assumed by the concept within the intellectual construction of liberal democracy. Therefore, it is not surprising that the debates on political representation can only be understood with reference to the political context in which they took place. Redefining representation – its ideal purpose as well as the criteria for evaluating its legitimacy – implies taking positions on the merits and limits of actual democratic systems. Following the marriage between representation and democracy, theorists building on the Schumpeterian view identified in the electoral competition between elites the means to ensure that citizens’ preferences are taken into consideration by a responsive government. The work of Pitkin (1967) did not challenge the idea that representation and democracy are fundamentally intertwined, but shifted the analytical focus onto the essence of representation (acting in the interest of the represented) rather than on the institutional settings built to pursue it. After three decades in which political theorists were more interested in participatory practices and empirical political scientists worked in the neo-Schumpeterian paradigm, perhaps based on different value-based considerations about the merits of existing liberal democracies, the reflection on the relation between representation and democracy gained new momentum (Plotke 1997). Since then, political theorists have reflected on various aspects of the representative relation, taking into consideration the insights of different strands of empirical research; for instance elaborating criteria to evaluate different forms of democratic representation or taking seriously the idea that interests, preferences and identities are formed within the representative relation and cannot be taken as pre-existing. Finally, a growing number of authors convincingly argue that representative relations can also be established without elections and even in non-democratic contexts. In turn, these theoretical insights will probably open further empirical streams of research in the years to come.

14  Research handbook on political representation

NOTES 1. We are indebted to Dovi (2018) for showing how making a list of key components of Political Representation helps to build a more systematic review. 2. Eulau and Karps (1977) wrote a dense critique of the empirical works reducing representation to the so-called policy congruence, a gross oversimplification that they attributed to the vulgarization of the seminal study Constituency Influence in Congress by Miller and Stokes (1963). The goal of the Miller–Stokes study was to understand to what extent and under what conditions the local constituency influenced the voting behaviour of congressmen on three different policy areas, and the congruence investigated was that between constituency attitudes, congressmen attitudes, congressmen perception of constituency attitudes and congressmen voting behaviour. While they never explicitly equated congruence with representation or responsiveness, many empirically oriented scholars replicating their study did. 3. This condition was already highlighted in the work of Pitkin (1967), who remarked that if citizens were completely amorphous the leader could simply manipulate their will through propaganda. In that case the similarity between their opinions would not count as representation but only as a distorted version of it, because the leader is certainly not acting for his or her constituents. 4. For a more radical critique of the concept of territorial constituency see Rehfeld (2005).

REFERENCES American Political Science Association (1950), ‘Toward a More Responsible Two-Party System: A Report of the Committee on Political Parties’ (Supplement to American Political Science Review, September), Washington and New York: Rinehart & Company. Andeweg, R.B. and J.J. Thomassen (2005), ‘Modes of Political Representation: Toward a New Typology’, Legislative Studies Quarterly, 30 (4), 507–28. Barber, B.R. (1984), Strong Democracy: Participatory Politics for a New Age, Berkeley, CA: University of California Press. Birch, A.H. (1972), Representation, London: Macmillan International Higher Education. Brito Vieira, M. (ed.) (2017), Reclaiming Representation: Contemporary Advances in the Theory of Political Representation. New York, NY: Routledge. Burke, E. (1854 [1774]), in E. Burke, The Works of the Right Honourable Edmund Burke. Volume I, London: Henry G. Bohn, pp. 446–8. Castiglione, D. and M.E. Warren (2006), ‘Rethinking Democratic Representation: Eight Theoretical Issues’, work presented at the Centre for the Study of Democratic Institutions, University of British Columbia, pp. 18–19. Dahl, R.A. (1956), A Preface to Democratic Theory, Chicago, IL: The University of Chicago Press. Dahl, R.A. (1971), Polyarchy: Participation and Opposition, New Haven, CT: Yale University Press. Disch, L. (2011), ‘Toward a Mobilization Conception of Democratic Representation’, American Political Science Review, 105 (1), 100–14. Disch, L. (2015), ‘The “Constructivist Turn” in Democratic Representation: A Normative Dead-End?’, Constellations, 22 (4), 487–99. Dovi, S. (2018), ‘Political Representation’, in The Stanford Encyclopedia of Philosophy, Stanford, CA: Stanford University Press. Downs, A. (1957), An Economic Theory of Democracy, New York, NY: Harper & Row. Eulau, H. and P.D. Karps (1977), ‘The Puzzle of Representation: Specifying Components of Responsiveness’, Legislative Studies Quarterly, 2 (3), 233. Friedrich, C.J. (1941), Constitutional Government and Democracy, 2nd edn, Boston, MA: Ginn. Hobbes, T. (1651), Leviathan, London. Kiewiet, D.R. and M.D. McCubbins (1991), The Logic of Delegation: Congressional Parties and the Appropriations Process, Chicago, IL: University of Chicago Press. Mansbridge, J. (1999), ‘Should Blacks Represent Blacks and Women Represent Women? A Contingent “Yes”’, The Journal of Politics, 61 (3), 628–57.

Concepts, theories and practices in historical perspective  15 Mansbridge, J. (2003), ‘Rethinking Representation’ The American Political Science Review, 97 (4), 515–28. Marongiu, A. and S.J. Woolf (1968), Medieval Parliaments: A Comparative Study, London: Eyre & Spottiswoode. Mill, J.S. (1861), Considerations on Representative Government, London: Parker, Son and Bourn. Miller, W.E. and D.E. Stokes (1963), ‘Constituency Influence in Congress’, American Political Science Review, 57 (1), 45–56. Montesquieu, C.L. de (1748), De l’Esprit des Lois, Geneva: Barrillot et Fils. Phillips, A. (1998), The Politics of Presence, Oxford: Oxford University Press. Pitkin, H.F. (1967), The Concept of Representation, Berkeley, CA: University of California Press. Pitkin, H.F. (2004), ‘Representation and Democracy: Uneasy Alliance’, Scandinavian Political Studies, 27 (3), 335–42. Plotke, D. (1997), ‘Representation is Democracy’, Constellations, 4 (1), 19–34. Post, G. (1943), ‘Plena Potestas and Consent in Medieval Assemblies: A Study in Romano-Canonical Procedure and the Rise of Representation, 1150–1325’, Traditio, 1, 355–408. Rehfeld, A. (2005), The Concept of Constituency: Political Representation, Democratic Legitimacy, and Institutional Design, Cambridge, UK and New York, NY, USA: Cambridge University Press. Rehfeld, A. (2006), ‘Towards a General Theory of Political Representation’, The Journal of Politics, 68 (1), 1–21. Rousseau, J.J. (1762), Du Contrat Social, Amsterdam: Marc Michel Rey. Runciman, D. (2009), ‘Hobbes’s Theory of Representation: Anti-democratic or Proto-democratic’, in I. Shapiro, S.C. Stokes, E.J. Wood and A.S. Kirshner (eds), Political Representation, Cambridge: Cambridge University Press, pp. 15–34. Sartori, G. (1962), Democratic Theory, Detroit: Wayne State University Press. Saward, M. (2010), The Representative Claim, Oxford, UK and New York, NY, USA: Oxford University Press. Saward, M. (2017), ‘Performative Representation’, in M. Brito Vieira (ed.), Reclaiming Representation: Contemporary Advances in the Theory of Political Representation, New York, NY: Routledge. Schumpeter, J.A. (1943), Capitalism, Socialism and Democracy, London: Allen & Unwin. Sieyès, E.J. (1789), Vues sur les Moyens d’Exécution dont les Représentants de la France pourront Disposer en 1789. Strøm, K. (2000), ‘Delegation and Accountability in Parliamentary Democracies’, European Journal of Political Research, 37 (3), 261–89. Strøm, K., W.C. Müller and T. Bergman (2003), Delegation and Accountability in Parliamentary Democracies, Oxford: Oxford University Press. Urbinati, N. (2006), Representative Democracy: Principles and Genealogy, Chicago, IL: University of Chicago Press. Urbinati, N. and M.E. Warren (2008), ‘The Concept of Representation in Contemporary Democratic Theory’, Annual Review of Political Science, 11 (1), 387–412. Wahlke, J.C., H. Eulau, W. Buchanan and L. Ferguson (1962), The Legislative System: Explorations in Legislative Behavior, New York, NY: Wiley. Weissberg, R. (1978), ‘Collective vs. Dyadic Representation in Congress’, American Political Science Review, 72 (2), 535–47.

2. The system of democratic representation and its normative principles Dario Castiglione

INTRODUCTION: APROPOS OF WHAT? There is no better time to write about political representation than the present. Its mechanisms, failings and transformations are under everyone’s eyes on the big stage of practical politics. Brexit, for instance, can be understood as a struggle over the ‘real’ meaning of people’s will and its representation. In the Brexit case, as in any political undertaking, we are confronted with the unavoidable question of how political decisions stand in relation to the interests (opinions, well-being and identities) of the people who are affected by them, and in whose name such decisions are allegedly made; in other words, whether, and how well those very people, either collectively or distributively, are represented in public decisions and acts. When put in this way, such a question leads to further interrogations regarding the meaning and mechanisms of ‘representation’ in politics; whether there are criteria that make representation both effective and legitimate. This chapter is concerned with such normative questions, but inevitably we first need to agree on what political representation involves from an evaluative perspective. Accordingly, this chapter is divided into two parts. The first sets out the assumptions underlying my normative analysis. These assumptions are bound up with aspects of the modern social imaginaries concerning democracy and legitimate state power (Taylor 2004, ch. 1). There is some circularity between factual and normative discourses. I consider this to be intrinsic to the way in which we understand and construct social facts, and in particular social institutions. Indeed, in the following, I propose to treat ‘democratic representation’ as a social institution. The second part of the chapter offers an interpretative framework within which to make sense of the normative principles of democratic representation.

A SYSTEMIC APPROACH: POLITICAL REPRESENTATION AS A SOCIAL INSTITUTION In trying to identify political representation, let me first clear the ground from a series of equivocations – two in particular: first, that in order to understand the nature of political representation we need to get at its core meaning(s); and second, that political representation is a mere institutional mechanism. The concept-focused approach is traditionally associated with the classic work of Hanna Pitkin (1967), which has been extremely influential in the Anglo-American literature. I hasten to say that the characterization of Pitkin’s contribution as exclusively engaging with conceptual clarification is a gross oversimplification. Conceptual analysis and distinctions are important, but insufficient either to characterize the phenomenon or to provide normative orientation. In the case of political representation this is so for at least two main related reasons. 16

Democratic representation and its normative principles  17 The first reason is the polysemic nature of ‘representation,’ something further complicated by the fractured history of the concept across a variety of intellectual fields and applications. Pitkin was fully aware of this, but at the start of her investigation she makes the reasonable assumption that, at least since the seventeenth century, political representation has had an ‘identifiable’ and ‘basic’ meaning captured by the word’s etymological origins: ‘re-presentation, a making present again’. She admits that many applications and interpretations of what she calls the presence/absence paradox are still possible, but describes such variety of theories as many ‘flash-bulb photographs’ giving pictures from different angles ‘of something there, in the middle in the dark, which all of them are photographing’ (pp. 10–11). The task of a political theory of representation is to explain how each of them offers a different angle on the ‘core’ concept. Arguably, the polysemic nature of representation is resistant to such an attempt. In his historical reconstruction of the philosophical and political uses of the Latin repraesentare, and of its vernacular derivates in many European languages, from medieval antiquity to the nineteenth century, Hasso Hofmann (1974 [2003]) challenges the idea of a ‘core’ meaning, suggesting instead three broad semantic fields. One, with its origins in theological disputes, later spilled over into issues of cognition and epistemology, concerns the relation between an original (Urbild) and an image-copy (Abbild). A second, of liturgical and ecclesiological significance, with juridical applications, concerns the idea and practices of vicariousness. Both sets of meanings can be reduced to the ‘core’ dialectic of presence/absence. But not the third semantic field identified by Hofmann, which he calls ‘identity representation’, and which conveys meanings of embodiment and personification in jurisprudential and later political discussions of corporate action and personality (Sintomer 2013; Mulieri 2016; Castiglione and Pollak 2019, pp. 22–5). The polysemic nature of representation connects with the other reason why conceptual analysis, particularly of an ordinary-language type, is inadequate. A few years after the publication of Pitkin’s book, Anthony Birch (1971, pp. 13–14) made the point that the search for a single or ‘core’ meaning is futile. One should rest content with establishing, in Wittgensteinian fashion, a series of usages that have a ‘family resemblance’ – because differences are equally as important as similarities. The Wittgensteinian point, which was not entirely lost on Pitkin (1967, p. 255, fn 20), is that semantics do not exhaust the meaning of concepts and linguistic utterances. Alongside a semantics of meanings, we need a pragmatics of their usages in order to understand what we do with them, and therefore what we try to do with representation. Semantics and pragmatics should not necessarily be seen as opposite projects. As Robert Brandon suggests, ‘concern with meaning and concern with use ought surely be understood as aspects of one, more comprehensive, picture of the discursive’ (2011, p. 165). The conjunction between the semantics of concepts and the pragmatics of political discourse as action makes semantic contents contingent, open-ended and therefore subject to change through use and new applications. But there is another Wittgensteinian point that follows from the idea of ‘family resemblances’. The intricate web of commonalities and differences suggested by such a metaphor accentuates the ‘unsystematic, unsurveyable variety of kinds of uses of expression and […] the dynamic character of linguistic practice’ (Brandon 2011, p. 163). Moreover, it shifts the focus of attention from the analysis of individual concepts to that of a more general vocabulary comprising a class of words and concepts that are deployed in order to identify and frame

18  Research handbook on political representation particular political problems in our attempt to resolve them. Political representation, and more specifically its democratic version, is one such vocabulary. The second equivocation, the one taking representation to be a mere institutional mechanism, is probably easier to dismiss, even though it has a common-sense ring to it. In the case of modern political representation, it is frequent to identify it either with the specific, one-to-one relation between a constituency and the agent whom they have selected to act or speak on their behalf, or with the electoral process itself, taken as a selection mechanism of the representative(s). The conflation between these institutions and political representation is partly due to the fact that they are considered as the engine of representative democracy, that is, of the dominant form of government in advanced constitutional states. There is something to be said in favour of such a metaphor. But the characterization of political representation provided by the mere-institution approach is either excessively reductionist, when it looks at representation as a one-to-one relation between principal and agent, or narrowly positivist, when it identifies representation with elections. Neither version is capable of giving a coherent account of how the institutions they take as paradigmatic perform a meaningful ‘representative’ function. As empirical research has amply illustrated, the simple principal/agent structure of the reductionist version offers contradictory explanations of how this relationship works. The idea of causal power underlying the mandate/responsiveness conception proves unworkable, unless it is put into a broader discursive and deliberative context operating both recursively and in relation to other agents (see Mansbridge 2003). Even then, it remains unsatisfactory. As for the positivist version, which collapses representation into the electoral process, it treats social institutions as merely instrumental. In our specific case, it fails to account for what ‘representation’ means in the electoral context, apart from taking this as a selection mechanism for positions of authority. This is precisely the way in which elitist theories conceive representative democracy, which they regard as a form of government where citizens act as ‘consumers’, who are given the opportunity to choose a governing party through open elections, but who are in no particular way represented by those they choose as rulers. In Joseph Schumpeter’s pellucid language, elections are an instrument for ‘competition for leadership’ and for ‘producing government’ (1943 [1976], pp. 269–73). From such a perspective, representation, along with the ‘will of the people’ and the ‘common good’, are largely fictitious ideas in the ordinary sense of the word. Ruling elites ‘represent’ the electors in no meaningful way, if not in the bland form of making decisions for their welfare and security, in the same way in which doctors do for their patients. Having excluded both the conceptual and the merely institutional approaches, is there some other way of defining political representation as an object of analysis and normative evaluation? I think there is, and an increasing body of literature fundamentally agrees on a more systemic way of considering political representation, even though it may not all describe it in the same way. There is one important qualification, however, that we should make, and this is that in most cases people are referring to democratic representation. Although nowadays the literature focuses more on analysing the forms and processes of political representation in their own right, without assuming that they are necessarily democratic, discussions of (modern) political representation, especially from a normative perspective, remain focused on its contribution to democratic legitimacy and its role as the engine of representative democracy. It is in this area that the vocabulary of political representation is largely deployed. It is therefore to the nature of the modern system of political representation in its democratic form that I now wish to turn. I do so by giving a systemic account of political representation as a ‘social institution’.

Democratic representation and its normative principles  19 In particular, I focus on three of its most notable features: its ‘institutional’, its ‘public’ and its ‘embedded’ nature. I shall discuss each in turn. Its Institutionalized Character On the institutional nature of political representation, Hanna Pitkin is once again a good starting point. Her subtle work of conceptual reconstruction has perhaps distracted followers, critics and commentators alike from noting the institutional and systemic aspects of her analysis. There is a certain progression in her argument: from formal to substantive concepts, and within the latter group from ‘standing for’ to ‘acting for’ concepts, which she considers to be the paradigmatic form of political representation. The reason for this is that she takes as a condition of democratic representation that both the represented and the representatives need to have a capacity for independent action and judgement, whose main object is the articulation and enactment of the interests of the represented. Formal and ‘standing for’ concepts exclude agency for the represented, and this may partly explain Pitkin’s rash dismissal of these concepts as politically (i.e. democratically) inadequate or perilous. The concluding chapter of the book, on ‘Political representation’, is therefore a fitting culmination to Pitkin’s analysis – the core of her normative argument. There she exposes her interest/responsiveness view: ‘representing here means acting in the interest of the represented, in a manner responsive to them’ (1967, p. 209). What is rarely noted, however, is the important qualifications she makes throughout the rest of the chapter, offering a conception of representing as ‘an extraordinary fragile and demanding human institution’ (1967, p. 155). The fragility and exacting nature of political representation is due to several factors besides the question of style (the how) at the centre of the long-standing controversy on whether representatives should act as delegates or trustees. There is a host of problems muddling the question of what representing means in politics: should electoral representation operate across issues or issue by issue? Is the constituency a single principal, or does it require ‘organization’ (p. 214)? How to reconcile partial with general interests, local with national, private with public? As Pitkin observes: the representative ‘works with other representatives in an institutionalized context at a specific task: the governing of a nation or a state’ (p. 215, emphasis added). This dual responsibility, as a representative of a motley constituency and as part of a governing assembly, makes it difficult, according to Pitkin, to find a clear-cut resolution to the mandate-independence controversy. The role of the representative is multifaceted: a ‘pleader’, a ‘judge’, an ‘agent’ and a ‘governor’; no single style may apply to the performance of all these different roles, since ‘the modern representative acts within an elaborate network of pressures, demands, and obligations’ (pp. 218–19). Given the many roles and variety of demands, Pitkin suggests that it is probably ‘a mistake to approach political representation too directly from the various individual-representation analogies’, so she is led to the conclusion that: Political representation is primarily a public, institutionalized arrangement involving many people and groups, and operating in the complex ways of large-scale arrangements. What makes it representation is not any single action by any one participant, but the overall structure and functioning of the system, the patterns emerging from the multiple activities of many people. (pp. 220–21, emphases added)

I have drawn attention to several points in this quote. I shall come back to ‘public’ and the ‘multiple activities of many people’. The other elements I wish to underline are those that

20  Research handbook on political representation I believe make political representation a ‘social institution’, requiring some level of concerted organization, and performing a recognized social function. The main elements are that political representation happens more as the result of a system of diverse interlocked actions, and not as a series of individual actions, isolated events, or single relationships; and that it involves a variety of variously organized (though not necessarily fully coordinated) arrangements across society. Its Public Nature The second feature, the ‘public’ nature of political representation, is also present in Pitkin’s quote, and I think it is closely related to its institutionalized character. It emerges more clearly from Pierre Bourdieu’s constructivist approach to political representation (1981 [1991], pp. 171–202), where, in agreement with neo-institutionalists (March and Olsen 1989, pp. 39–52), he discusses the role of meaning-giving in the way in which institutions (like the rules of a game) frame our overall experience and expectations. On this basis, Bourdieu makes a series of important points on how ‘political intention’ is constituted in relation to the rules of the political game: …more precisely, of the universe of the techniques of action and expression it offers at any given moment. In this and other cases, moving from the implicit to the explicit, from one’s subjective impression to objective expression, to public manifestation in the form of a discourse or public act, constitutes in itself an act of institution [acte de institution] and thereby represents a form of officialization and legitimation’. (1981 [1991], p. 173; original emphasis)

In Bourdieu’s characterization, there are two important elements of how intentionality works in a political system, or indeed in any other social system. On the one hand, the intentionality of a political act – in either words or deeds – acquires meaning within a set of established rules, customs and institutions. On the other, any appropriate articulation of political intentionality is a ‘constitutive’ action (acte de institution), capable of operating the passage from what is implicit, the agent’s intention, to what is explicit, the socially recognized meaning of the act, from the individual agents’ perceptions to their mutual understanding. This double passage, from intention to socially recognized meaning, from the subjective to the intersubjective gives to political actions a public dimension, making them public acts; hence, as Bourdieu points out, such a passage provides political intentions and political acts with a prima facie seal of officiality and legitimacy. I think it is important not to confuse political institutionalization with this public dimension for various reasons: first, because in contrast to the process of institutionalization, ‘becoming public’ highlights the aspects of meaning-giving and social recognition more than those of social organization. Second, because of the more normatively acceptable, consent-like, connotations that public recognition, rather than simple institutionalization, brings with it. Regarding something as an official and legitimate act implies some internalization of political intentionality, and not its simple acceptance as an external imposition. This does not exclude that political institutions can also be dominating. In the case of political representation, public institutionalization sanctions its social recognition as a meaningful and legitimate series of interconnected discourses and activities, which contribute to the governing of modern societies. But there is one further qualification that we should make. This is that publicness implies both being visible to the public, and being open to public discussion. This qualification is concerned more with the form, rather than with

Democratic representation and its normative principles  21 the constitutive function of being public. As Norberto Bobbio (1999, pp. 339–41, 352–63) remarks, visibility is an important requisite of the exercise of power in representative as well as in direct democracies. Both forms of government rely on institutions where deliberation and decisions are made in public arenas. He also cites Carl Schmitt, who argues that ‘representation can occur only in the public sphere. There is no representation that occurs in secret and between two people, and no representation that would be a “private matter”’ (Schmitt 1928 [2008], p. 242). In other words, it is in the very nature of representation to exhibit the exercise of power publicly. The visibility of power also comes with its being in principle open to public scrutiny and discussion. This theme is best illustrated by Jürgen Habermas’ reconstruction of the origins and transformation of the political role of the public sphere, which he considers as closely related to the development of a modern system of representation (1962; see also Manin 1997, and Urbinati 2006). I think it is not unreasonable to make such an assumption. Both public discussion and the social process of public opinion formation are contributing elements to the formation and maintenance of the democratic system of political representation. Its Embeddedness The final feature of political representation as a social institution, its ‘embeddedness’, follows directly from what we have just said about its strong link with opinion formation and public deliberation. In her description of political representation as a system, Pitkin talks of ‘multiple activities of many people’ (1967, p. 221). I wish to maintain that such activities should not be seen as exclusively located in the institutional machinery of political representation, but as taking place in a variety of contexts in which multiple agents contribute to the process of intermediation between, on the one hand, social needs, claims and demands, and, on the other, their political articulation. Clearly, there is an institutional and public core to the system of representation, and there are privileged instruments and practices through which it performs its function of intermediation; but there are many activities going on at both ends of this process, which are essential to guaranteeing its functionality as well as providing meaningful opportunities for representation and the cultivation of the required capabilities of both represented and representatives. From the perspective of the functioning of political representation as a social institution, representation does not merely happen in a few formally appointed places like parliaments, or solely through institutionalized mechanisms like elections. Echoing Sieyès, one could maintain that representation is everywhere in society (cited in Pasquino 1998, p. 39): social and political dealings in modern societies involve innumerable instances of representation. Even forms of direct social and political activism and mobilization rarely involve the whole relevant group, while often requiring some form of delegation of powers and decision making. Moreover, as the historical development of representative democracy shows, the process of political representation has been capable of incorporating some instances of self-government, and participatory and deliberative politics, through the use of referendums, citizens’ legislative initiatives, stakeholder consultation and, more recently, citizens’ forums. The process of political representation is often sensitive to changes in public opinion, as shown by the attention to opinion polls in inter-electoral periods, to the influence of specialized publics and professional opinion, and occasionally to popular pressure or what Simon Tormey (2015) has called the ‘politics of resonance’. Although electoral institutions still hold a key place in the politics of

22  Research handbook on political representation representation (Przeworski et al. 1999, p. 3), there has been a noticeable metamorphosis in some of its key elements, as Manin, for instance, has captured in his analysis of the passage from parliamentary to party, and eventually to ‘audience’ democracy (1997, ch. 6). The rise and eventual decline of mass parties is probably the most notable aspect of the profound transformation affecting the system of political representation throughout the twentieth century (Schattsneider 1942 [2004]; Mair 2013, chs 2 and 3), but the increasing attention to forms of descriptive representation (Phillips 1995; Lovenduski 2015), and more recently to informal representation (Montanaro 2018), and the quasi-representative role played by administrative or even legal institutions (Rosanvallon 2008; Keane 2012; Mansbridge 2019; Barroso 2016), all testify that in spite of the endurance of its institutional architecture, the general ecology of representative politics is in continuous evolution (Castiglione and Warren 2019, pp. 42–4). A Political Technology In sum, how can we characterize the system of political representation from a relevant normative perspective? I have suggested that this should be taken as a complex and variegated social institution whose main function is that of intermediation between social needs and demands, on the one hand, and the governing capacity, on the other. I have also emphasized three main features of political representation as a system of intermediation: its institutional character, which gives unitary scope to a variety of actions and processes by multiple agents, which are not always coordinated or in unison with each other. Its public character, which gives meaning and legitimate recognition to its function of intermediation. And its embeddedness both within a broader network of representative politics and in relation to other social processes that are relevant to the successful performance of the task of political intermediation in a democratic society. But before we move on to a discussion of the normative principles of democratic representation, I would like briefly to address one possible objection to my characterization of it as a social institution. It may in fact be suggested that it is too generic. There are two sides to this objection. One, because such a characterization is too broad. Representation could be found everywhere, thus making it into an amorphous series of practices of representative politics. This would undermine my contention that political representation is an ‘institution’ in some meaningful way. The other side of the objection is that by insisting on its general function of social intermediation, my characterization is not political enough, and therefore tends to emphasize a generic social functionalist perspective, implicitly undermining my argument that the ‘public’ nature of political representation has a constitutive (meaning-giving and legitimating) element to it. Moreover, treating political representation as a social institution may give the false impression that I am bringing back a transmission-belt conception of political representation as a mere process of reflection of already formed social demands into the political field. I reject both implications. The system of democratic representation that I am trying to identify is political in the double sense that it is a ‘political institution’, whose core organized practices and mechanisms operate within the political field, and that it has a ‘public function’, hence it is constitutive of, and constituted by, the political community. What I am suggesting, however, is that, as a political institution, representation also has a larger social function, which is entangled with other social processes (what I have described as its embeddedness); but that it remains distinctively political in both its scope and its modus operandi. Perhaps a more precise way of characterizing it would be as an institutionalized and publicly

Democratic representation and its normative principles  23 recognized series of ‘techniques’ (Bobbio 1999, Bourdieu 1981 [1991]) performing the role of political intermediation in modern societies. Since techniques may sound a bit reductive, mechanical and purely instrumental, one could talk of a political technology in its fuller sense: a body of knowledge, whose main scope is its practical application through the development of practices, techniques and human capacities in view of the fulfilment of a series of socially meaningful objectives of a distinctively political nature.

CAN POLITICAL REPRESENTATION BE DEMOCRATIC? In the second part of this chapter I explore the normative setting and principles of political representation from a democratic perspective. I do not claim that modern political representation is by definition democratic, nor that representative government is necessarily a form of democracy. I take both questions to be historically contingent. My main question is: can the political technology of representation be democratic? For reasons partly discussed and that will become clearer presently, I take the relevant context of modern political representation to be the historical affirmation of the widely held conviction that legitimate power must rest on a ‘popular’ foundation. Accordingly, in this second part I first discuss the way in which ideas and practices of representation apply to the symbiotic development of both modern statehood and new conceptions of democratic politics. The uneasy, but intriguing combination between these two processes puts in sharp relief a series of antinomies in the vocabulary and practices of political representation. I then move on to discuss the conditions that make the social intermediation provided by political representation democratic, and finally, the main principles of democratic intermediation. The Antinomies of Democratic Representation The persistent focus in theories of democratic representation on the transmission problem, how citizens’ interests are translated into governmental decisions, has perhaps obscured a more intractable problem for modern political representation: how to achieve unity of purpose through representation. Arguably, this unifying process takes place at two levels: through the establishment of unitary statehood (even in a federal regime), and through the development of a recognized governing capacity. Historically, the technology of political representation has contributed to both sides of this process. On the one hand, it has provided the vocabulary for the establishment of the ‘representative’ nature of the state and of its institutions; on the other, it has provided the means and mechanisms for the emergence of ‘representative’ forms of government. The combination of the vocabulary of political representation with that of sovereignty and legislation is telling, and I think it offers a kind of solution to the conundrum of representative democracy – whether it inclines more towards a democratic or an oligarchic form of government. Bernard Manin (1997) is probably right in suggesting that this tension is intrinsic to some of the institutional mechanisms that have historically characterized representative democracy: the practice of the free-mandate and recurrent elections open to all. But he sees this tension as the permanence of a social division, of the few and the many, making representative democracy similar to classical forms of mixed government. Of course, deep social differences and hierarchies are still present in modern democracy, but Manin’s interpre-

24  Research handbook on political representation tation may not account for the full story. He perhaps underestimates the equalizing ethos that comes with the democratization of modern societies and the universalistic culture of human and citizens’ rights. His reference to the classical category of mixed-government is interesting, but misleading, because it fails to recognize an important shift from classical to modern conceptions of politics in terms of the consolidation of a generalized conviction, in terms of ‘social imaginary’, that the legitimation of political power rests ultimately on a popular foundation. We can understand representative democracy as a unitary yet divided form of government, by considering the modern divarication between, on the one hand, discourses about the form of the state and its legitimation (which in classical time rested on a distinction of good versus bad forms of government), and on the other, discourses about the form of government, how it is organized and how power is exercised. The vocabulary of political representation plays a part in both these discourses. It is perhaps in such divarication, between legitimate statehood and effective governing processes, that one can locate an important tension, or antinomy, characterizing modern political representation: its role in creating unitary political agency from social differentiation. The classical text illustrating the emergence of the modern discourse of representation as unified political agency is Thomas Hobbes’ Leviathan (1651 [1996]). Pitkin (1967, ch. 2) characterizes Hobbes’ theory of representation as one of ‘authorization’. She finds his theory ‘partial, formal and empty of substance’ (p. 34), failing to address ‘the real problem of the creation of political consensus’ (p. 35). Pitkin’s interpretation is correct if we consider political representation exclusively as a technology for filtering interests and producing responsive governance. From such a perspective, Hobbes’ conception appears merely formal or at most symbolic. In Pitkin’s categorization, it lacks substance, for it does not specify the institutional and discursive processes through which the governing function tracks the representatives’ interests. In truth, Hobbes’ theory takes a different form from conceptions of representation as ‘reflection’, ‘substitution’ and vicariousness. It is a theory of ‘personation’, echoing more performative and theatrical practices (Brito Vieira 2009; 2017), through which a dispersed multitude is incorporated, or given body as a political community. It is more than symbolic, it is constitutive. The very act of representation gives unity to the represented, who up to that point were an amorphous collection of individuals. In Hobbes’ famous words: ‘for it is the Unity of the Representer, not the Unity of the Represented, that maketh the Person One’ (1651 [1996], p. 144; see also Brito Vieira and Runciman 2008, pp. 24–8; Loughlin 2003, pp. 58–61). The artificial personation of the political community is of course a ‘fiction’, but in the sense that this expression has acquired in the juristic tradition of persona ficta since Roman times (Thomas 1995). The recognition that certain entities have artificial personality has momentous consequences. Though ‘imagined’, the juristic-like invention of modern statehood has material and ideal effects on the relationships between natural persons. In the case of political representation, the personation of the state produces political unity (in Hobbes’ language: in the person of the ‘Sovereign’); establishes a public authority capable of political agency (the ‘Actor’); and makes the whole community (the ‘Author’) responsible for the acts of the ‘representer’. This last point is of crucial importance, because it ascribes (imputes) to the entire community the actions of the political authority. Hobbes was no democrat, but there is no need to embrace his absolutist version of sovereignty to recognize the more general relevance of his theory of personation for the legitimacy of the modern state. His theory of representation provides a certain unity for the exercise

Democratic representation and its normative principles  25 of political power (which, pace Hobbes, does not exclude the principle of the division and balance of power); a capacity for political agency; and the ultimate source of power in popular authority. Yet the legitimacy of the state as the ‘representer’ of the political community does not settle the problem of how power is exercised. In this Pitkin is right, the tension between social heterogeneity and political unity is not solved by fiat through personation. There is more work to do for representation as a political technology in the way in which the governing functions are arranged and exercised. At the level of governing, we see another aspect of the antinomy of diversity and unity in political representation, which manifests itself in the tension between particular and general interests. This was a central problem in the eighteenth-century French and American constitutional debates, overlapping with the well-known debates on actual versus virtual representation, and free versus imperative mandate. It has also continued in the nineteenth and twentieth centuries through discussions of corporatist and neo-corporatist theories of politics as well as pluralist and partisan theories of democracy (see Bobbio 1999, pp. 410–29). We can illustrate it briefly by looking at a number of twentieth-century authors such as Carl Schmitt ([1928] 2008), Gerhard Leibholz (1929 [1973]) and Eric Voegelin (1952), who, echoing Hobbes’ conception of personation, considered the main function of representation to be that of establishing political authority, whose actions and decisions are ‘legitimate for’, ‘binding on’ and ‘ascribed to’ the whole of the political community (see also Pitkin 1967, p. 53). Although this form of state representation takes place in all political regimes, these authors insisted that it applied to democracies too, otherwise these would dissolve in a multiplicity of conflicting interests and wills. In truth, their theories were directed against pluralist and parliamentarian versions of representative democracy; but, in general, their view of political representation showed that they regarded its main, if not exclusive, aim to be the unity of the ‘political form’. They did not think that this could be achieved through more mundane and empirical mechanisms and processes such as elections or the painstaking negotiation of, and mediation between conflicting interests. They understood political representation as a more ‘spiritual’ and ‘transcendental’ kind of process. Voegelin downplayed the more procedural and mediatory aspects of democratic representation as ‘elemental’, and therefore devoid of real value (1952, pp. 31–3). According to him, true political representation is ‘existential’. It consists in that process through which the actions of the representatives are ‘experienced by the members of the society as [having] […] obligatory force for themselves’ (1952, p. 37). Schmitt and Leibholz, for their part, argued that particular and privatistic interests have no place in the process of political representation. They rejected the very idea of vicarious representation, where the representative stands or acts for someone else by either following their instructions or attending to their actual preferences and opinions, as expression of their particular interests. They identified this type of ‘interest representation’ by using the German terms of Vertretung and Stellvertretung in opposition to Repräsentation (Leibholz 1929 [1973], ch. 1). As Schmitt ([1928] 2008) says, ‘that X steps in for the absent Y or for a few thousand such Ys is still not an instance of representation’ (p. 243). According to both Schmitt and Leibholz, Repräsentation only takes place where superior, spiritual entities are involved, and it belongs to the political rather than the private sphere. In their view, there is a clear separation between a private sphere, in which individuals and groups’ interests are considered through the categories of civil law, and political representation in the strict sense, which belongs to the domain of public law. When looked at from a democratic perspective, this understanding of political representation excludes that civil society organizations, private and

26  Research handbook on political representation corporative groups, or other particularistic instances can participate directly in the formation of the general will through public discussion. Their interests and opinions need to be represented (i.e. already transformed and incorporated) in and through the organic unity of the political form: through the state and its institutions, which are the true representatives of the entire people. This vision of representation through the corporate nature of the ‘organ’ or institution (the state, parliament, an assembly), rather than by the simple aggregation of interests and opinions in a deliberative forum, is one that has a complex genealogy, from pre-democratic theories of representation (like Burke’s and the English Whig tradition; see Sternberger, 1971; Birch 1971, pp. 37–40), to the French revolutionary debates, particularly Sieyès’ view that the National Assembly was the representative of the Nation (Baker, 1990; Violante 1981 [2012]). But as Böckenförde suggests, when talking of Carl Schmitt, in his theory: representation always relates to the political unity of the people, i.e. the state; it does not mean representation of the society vis-à-vis the state or representation of interests within the society. Moreover, the subject of representation is not the people in the state but, rather, the politically united and organized people which is the state itself. (1998, p. 49)

The antinomies of modern political representation – those between unity and diversity; between general and particular interests; between public and private domains; between virtual and actual representation; and between delegate and trusteeship relations – are all reflected in the two ways of considering the people’s role: either as the state or in the state. It is this dual process of democratic representation of the people that requires careful institutional articulation and considered evaluation. Arguably, the system of democratic representation and its principles should pay heed to the people both in its unity and in its diversity (see Brito Vieira and Runciman 2008, ch. 5). Democratic Intermediation The constitutive function that representation plays in the formation of the political community, and the way in which it gives legitimacy to the institutional machinery of public decision-making should have clarified its political nature as a system of social intermediation. There remains, however, to address the problem of the possibility of democratic intermediation, before coming to its principles. As we have seen, there are those who think that particular interests have no place in the representation process, and that this should only operate from a transcendental, exclusively public and institutional, perspective. Admittedly, some of the authors embracing such a view tend to reject political pluralism altogether, favouring more dirigiste and Bonapartist forms of democracy. But others, who subscribe to a more Rousseauian view of democracy, put equal emphasis on the transformative force of politics. Their understanding of representation inclines more towards a communitarian or teleological view, framing it as the capacity to ‘act for the common purposes of associations of people’, thus giving to the representatives a more educative and exhortative role as either guardians of the common good, or promoter of public-spiritedness (Schwartz 1988, p. 143). Moreover, it is possible to be sceptical about the role that the aggregation of already formed private interests and preferences plays in the representation process on more realist and constructivist grounds. Much empirical research on public opinion and electoral representation casts a sobering and realistic light on the existence of clearly pre-formed opinions and policy preferences in the electorate. Limited knowledge

Democratic representation and its normative principles  27 and lack of time makes it difficult for ordinary citizens to formulate clear policy choices, or even to express a considered judgement of the suitability for office of prospective representatives. Not only are citizens’ opinions and preferences often poorly formed and indeterminate, public opinion is also dynamically shaped by electoral communication and manipulation, or by changes in policies and in the policy environment (see Dalton and Klingemann 2010, chs 43, 45 and 46; Przeworski et al. 1999, chs 1 and 5). As Lisa Disch suggests, this empirical literature ‘turns the classical model of democratic legitimacy on its head: citizens are responding to politicians rather than the other way around’ (2019, p. 6). This observation is central to the recent constructivist literature on representation, which emphasizes the more performative aspects of political representation and at the same time suggests that there is a creative or entrepreneurial capacity, in the way in which representatives construct more than reflect the very identity and political orientation of the represented. Although the ‘constructivist turn’ in political representation offers a more nuanced way of looking at the functioning of representation, emphasizing its constitutive elements, it risks calling into question what I have previously referred to as the ‘public’ nature of the representative system, while raising afresh traditional concerns with regard to its democratic quality. In a series of influential contributions, Saward (2010, 2014, 2017 and 2019) has emphasized how what he calls the ‘representative claim’ is at the very centre of the act of representing. He stresses the inescapable performative element of representation both as a doing, and as a showing doing (Saward 2017, p. 76). Representation acquires its meaning as an ‘event’ in its own right, not just as reflection, or a substitution for, the absent presence of the represented. As Saward says: ‘a representative claim is a claim to represent or to know what represents the interests of someone or something. It invokes – consists of – claims that one stands for others by virtue of the role one can play’ (2010, p. 42, emphasis added). Although audience’s recognition is an important aspect of successful claim making, and in Saward’s view the constituency to which the claim refers can be empowered (not just silenced) by the claims made about them or on their behalf, there is the distinctive possibility that this approach re-evaluates more elitist and potentially manipulative (acceptance rather than consent) aspects of political representation. Moreover, by emphasizing the processual aspects almost to the exclusion of its institutionalized and publicly accepted features, this approach makes it more problematic establishing the validating criteria for democratic legitimacy. Admittedly, at least from the perspective that I am trying to defend here, the claim-making approach has a number of advantages. It opens up the institutional black box of representation, emphasizing the more discursive and interpretative aspects of the process of political representation. It suggests that representative claims can be made by a variety of actors and take place at many different stages of this process; in addition, it renders the distinctions between social and political domains, or between formal and informal processes, more porous. This allows the process of public recognition and identity constitution to be conceived in a more diffuse, dynamic and contested way, so that the social is not conceived as an entirely amorphous set of needs and relations, which are marshalled and colonized by the political. However, for the claim-making approach to allow for a more robust conception of the role played by the representative system in the intermediation between the social and the political, there are several qualifications to be made. One regards some ambiguity concerning the meaning of claim making in this context. In fact, when we say that someone is making a claim, we may be saying: (a) that someone is making a statement or an assertion, related to themself or someone/something else; (b) that someone is making a demand or asserting a right/

28  Research handbook on political representation status on their or someone/something else’s behalf. Obviously, as in many similar cases of polysemy, the distinction between these two semantic groups is neither precise nor internally stable. Nonetheless, it remains the fact that the distinction points to two different ways in which claims are made and articulated. Saward’s own use is almost exclusively of type (a), statement-like. At one point of his argument, he says: ‘To be sure, there are more things in the political world than claims – there are demands, for example’ (2010, p. 43, emphasis added). It may be that this sentence is an unintentional over-reach, but a careful analysis of most of the examples he cites confirms that the kind of claims he mostly refers to are of the statement-type – they are an invocation of representativeness, a claim to be representative. Even when they are presented as demand-type, their main function is still intended as a way of constructing the (self-)image and the (self-)understanding of the represented. In other words, they have an internal, identity-shaping function, rather than an external, politically purposive aim. Interestingly, another constructivist author, Ernesto Laclau, comes close to formulating something similar to the representative claim, but mainly of the demand-type. He suggests that the unit of analysis of political logic is neither separate individuals, nor already formed groups, nor ideologies, but the ‘category of social demand’ (2005, p. 73). As he says, ‘the notion of “demand” is ambiguous in English: it can mean a request, but it can also mean a claim (as in “demanding an explanation”).’ Laclau regards such ambiguity productive because it indicates a logic of progression from isolated requests, which can be met differentially within the political system, and which tend to confirm the status quo, to the accumulation of equivalent demands, which create new political subjectivities and produce new political logics: ‘the requests are turning into claims’ (2005, p. 74). I think that adding the demand dimension to the representative claims underscores the purposive and decision-type elements that characterize the construction of a more unified and generalizable kind of political agency that emerges from particular social demands. Moreover, this ‘material’ and action-like conception of claim-making suggests the possibility of a more interactive way of conceiving the process of representation, where identity-construction and demands-articulation claims are not simply elicited or evoked by the representatives, nor exclusively by those located within the institutional machinery of politics. They are instead the result of the interaction between multiple actors, operating at different levels and in different contexts. Lisa Disch’s conception of representation as mobilization captures this process well when she talks of the ‘emergence of a democratic constituency’ as that process that ‘rouses a constituency to action’ and ‘enables it to recognize itself in terms of a ‘generality’ […] that is neither given nor self-evident but must be narrated into being’ (2011, p. 108). My point is that the passage from the particular to the general, and from social demands to political agency – the formation of what Disch calls ‘a democratic constituency’ as the result of political representation – is neither a single nor a fully institutionalized event exclusively located in the political space. Alfio Mastropaolo (2018, pp. 322–5) has recently described this process of entanglement of the social and the political in the construction of democratic constituencies by referring to ‘texts’, ‘pre-texts’ and ‘con-texts’. Political representatives offer ‘texts’, in the form of claims and demands, which necessarily draw from ‘pre-texts’, consisting in a variety of statements, discourses, practices and experiences elaborated at various levels and in different social spaces, but which are rarely in a ‘raw’ or ‘unrefined’ state, since they have already passed through several forms of elaboration, conflict and mediation (2018, p. 323). The construction of political subjectivity neither starts nor ends in the act of formal representation,

Democratic representation and its normative principles  29 but it takes place within larger and multiple ‘con-texts’, as the result of ‘a much wider interweaving of power relations, cultural horizons, and all kinds of conditioning’ (2018, p. 324). In short, I think that the transcendence or transformation of the social into the political should not be seen in either metaphysical or spiritual terms, but as a concrete process of continuous negotiation, contestation, recognition and understanding. Although ultimately binding and productive of a sense of the public good, such a transformation process should not be presupposed on either a denial or the mortification of social diversity. On the contrary, its success is presupposed on the self-organizing capacities produced by and in civil society, which democratic politics must harness to the construction of the political community through the representative system. As Alessandro Pizzorno (1993) suggests, politics has something specific to offer to such a project. He talks of a specific political virtue, the ‘virtue of binding’ together: ‘political action mobilizes, constitutes new identities and binds individuals’ more or less strongly, more or less transiently, but it does so ‘through speech, the induction of trust, persuasion and conversion, and finally the decisions of authority’ (1993, p. 182–3). The binding force of politics should not be seen as the denial of the social, but as its complementary transformation. Principles of Political Representation I started this chapter by denying that the key to understanding political representation lies in semantic analysis or that, in the case of representation, we can get to some kind of essential, ‘core’ concept. But, in principle, there is nothing objectionable to using a particular concept of representation in order to make a normative argument on how it should work. When we come to the evaluation of political representation, many have pointed to the see-saw quality of re-presentation – the alternate, paradoxical character of a process that suggests a presence by its absence, and vice versa – as the paradigmatic way in which to think of it democratically. If democracy is some form of self-government (whether conceived in more individualist or collective terms), the ‘presence paradigm’ is inescapable. It therefore appears both natural and right to assess political representation in terms of the kind of descriptive, symbolic or substantive presence, to use Pitkin’s terms (1967, chs 4, 5 and 6), that it guarantees in governing. If the citizens are not, in some sense, present, if what political representation signals is their absence, the natural conclusion seems to be that representation cannot be democratic, and any pretention of this being otherwise is a sham. There is something important, from a democratic perspective, in the idea that ‘presence’ matters. But, as argued throughout this chapter, a systemic view of political representation cannot be reduced to the presence/absence paradox in any of its forms. The main limit of the paradox, as a normative standard, is that it tends to read the relationship between represented and representatives in identity terms, at least tendentially. What it misses is that the relationship of representation is one in which a ‘double’ identity (hence an inescapable difference) is established: the ‘identity’ of the representative consists at the same time to ‘represent the represented’, and to be the ‘representative of the represented’; and the same applies to the represented as represented. In other words, there is something irreducible in the relationship between what represents and what is represented. Such a conception, of how representation works, is philosophically complex, but it has obvious political consequences, which are not easy to translate into institutional forms and practices that guarantee some form of meaningful presence to the represented (i.e. the ability of self-governing, and the feeling of

30  Research handbook on political representation self-government), while accepting that this cannot take the form of an identity of views and actions with the representatives, who are those who ultimately make binding decisions. To put it another way, the passage between political judgement and wilful action that the system of representation tries to put in motion is never immediate, but always a mediated process. The logic of representation as difference and mediation offers some interesting analytical and normative ways of rethinking democratic representation, for the main precipitate of such a conception is that we need to think of the relationship of political representation as one involving the presence and contextual formation of two or more distinct political subjects. When looked at in these terms, the problem of democratic representation is transformed. It is no longer how successfully the representatives act for, stand for, or embody the represented, but how the process of political representation may aid (or impede) the establishment of meaningful practices of self-government. As a consequence, the forms and instruments of political representation must be judged on different criteria than presence, identity, responsiveness or congruence. The criteria I suggest are: (a) whether the system of political representation empowers citizens to have some voice and control over collective decisions; (b) whether it facilitates democratic inclusion on a fairly equal basis; (c) whether it allows for reflexive political judgement. As a conclusion to my argument, I shall briefly illustrate these three criteria in turn. Empowering citizens through the process of political representation means giving them the capacities and opportunities to exercise a certain control over their living conditions and welfare, in so far as these are dependent on the scope of public decision making and political cooperation. In a democracy, citizens’ capacity needs to be developed in two senses, as the capacity for individual autonomy, and as the capacity for collective self-rule (Warren 2019, pp. 48–9). Although the system of political representation is mainly functional to the latter, the two sets of capacities are closely connected, and are often cultivated by similar means. The ability, therefore, of exercising one’s role as a member of a political collectivity is also dependent on the ability of exercising autonomous choices in civil society and one’s own local community. Capacity itself is not enough; citizens must have opportunities for exercising some influence on collective self-rule. Traditionally, elections and the right to an equal vote have been regarded as the main, if not exclusive, democratic instrument for citizens to exercise such influence. But this is a poor view of how the system of representation works, or of the conditions according to which it may be said to empower the citizens both as autonomous individuals and as members of the political community. Elections, which are the traditional selection mechanism for political offices, are rather meaningless as the centrepiece of political representation, without taking into consideration several other factors. These comprise, first, the structure of organization of interests and identities through which both the political personnel and a general political orientation are formed and selected. Since the end of the nineteenth century, such a function was mainly performed by mass political parties and the party system as a whole. Particularly in the European experience, they have been instrumental as an opportunity structure for popular influence and for the development of ordinary citizens’ political capacities. Their current crisis as the main linkage between the state and civil society, and their transformation in governmental machinery, with a firmer foot in the state apparatus than in civil society (Mair 2013, ch. 3), shows the importance of the system of intermediation, more than the election mechanism, for citizens’ empowerment.

Democratic representation and its normative principles  31 The other factor to consider in order to assess democratic empowerment is the location of power itself. In many cases, elections, particularly parliamentary elections, are a blunt instrument for democratic empowerment also because there have been fundamental changes in the power map of modern societies. Without going into detail, one can mention the general shift of legislative power from parliaments to governments; the increasing role played by administrative powers in the application of general legislation; the proliferation of private powers and nominally independent regulatory agencies with quasi-legislative functions; and, of course, the growth of supranational collaboration and regulation that limits governments’ capacity for autonomous action, hence its ability to be receptive to public opinion pressures. This changing map of power shows the need for a more sophisticated system of social intermediation than the one traditionally associated with the electoral mechanism. It also shows that accountability through regular elections, though an important democratic instrument, plays a more symbolic role than a real one. Regular elections, universal suffrage and one-person-one-vote are also considered as the main institutional mechanisms through which political representation guarantees inclusion, the second of our democratic criteria. Once again, this over-evaluates the role of elections within the system of political representation. The right to vote is an important recognition of the status one has as member of the political community, and of one’s dignity as a person, as someone capable of self-direction. However, in modern, particularly multicultural and differentiated, society, inclusion may require not only confirmation of equal treatment in terms of rights and respect, but also identity recognition: of what we are, and who we feel we are. Such recognition extends to the things we identify with most, and which contribute to give sense and purpose to our life, including our social attachments, and, in some cases, that sense of identity that comes from personal or group suffering, or present and past grievances that we think need to be redressed as a matter of justice. Calls for a politics of presence and for more descriptive representation for women and minorities (Phillips 1995; Williams 1998) are the most obvious examples of the way in which inclusion may require special devices to energize the system of political representation in a way in which it mobilizes a sense of belonging to the democratic community. Arguments in favour of descriptive representation try to show that the introduction of identitarian elements in the selection process of representatives does not rest on a mimetic idea of representativeness, but it can also provide the conditions for better substantive representation, by putting in motion a new politics of trust. Trust and deliberation are another strategy for political inclusion through representation. Paradoxically, these are some of the central elements of Burke’s pre-democratic conception of representation. But Jane Mansbridge (2003, 2009) has recently revaluated this more trusteeship-based conception within a democratic context. She has proposed a sophisticated phenomenology of the relationship between represented and representatives, beyond the simple dichotomy between delegate and trustee, and suggested that what she calls the ‘selection model’, in opposition to a ‘sanction model’, is more successful in aligning voters’ expectations and legislators’ behaviour. This model also fits better with a more recursive and deliberative conception of representation across politics, administration and society (Mansbridge 2019). As she says, ‘the characters of the representative and the constituents are also formed in the representative-constituent relation’ (2009, p. 392), as their policy choice may be. Though this statement may be challenged empirically, it still retains an important normative value.

32  Research handbook on political representation Another inclusion strategy that is worth mentioning, but that mostly applies outside the circuit of institutional and electoral representation, is the one of self-appointed representatives. This group of representatives, mainly operating in civil society, speak for those who are affected by decisions, but who are excluded from the political process and lack the opportunity for their voice to be heard. From such a perspective, self-appointed representatives have an important role in democratic governance, where electoral constituencies fail to coincide with those interests affected by collective decisions. This form of representation poses problems in terms of democratic legitimacy according to criteria such as authorization and direct accountability. Nevertheless, its inclusionary function gains in legitimacy as the self-appointed representatives are recognized by either public authority or relevant publics. The third and final criterion that I propose to consider is that of reflexive political judgement. This criterion is more related to the way in which democratic forms of government aim to resolve conflicts and disagreements in a consensual way, avoiding as far as possible the use of force. The logic (or force) of numbers, via the majority principle, remains the ultimate way of settling issues. Nonetheless, public discussion and considered deliberation in either deputed settings, such as parliamentary assemblies, or during concentrated periods, such as electoral campaigns, has long been recognized as an integral part of the system of political representation. This discursive and deliberative aspect of representation extends more generally to the role played by the media and public opinion formation and sounding, or, as suggested by Habermas (1962), to the rise of the modern public sphere. Following Manin (1997), one may consider this criterion to consist of two institutionally embodied principles, the one guaranteeing freedom of opinion, and the other, the more deliberative principle, that puts opinions and decisions to the trial of discussion. This criterion has sometimes been regarded as having a decisive role in political representation, as in John Stuart Mill’s theory of political representation, not as an instrument of governing, but as a way of exercising control over power (1861). More recently, building in part on Habermas and other deliberative theories of democracy, Urbinati (2006) has eloquently argued for the central role that political representation plays in arriving at a considered public judgement. Like for the other two criteria, the satisfaction of the reflexive judgement criterion extends beyond the traditional electoral process and legislative institutions. We can consider two further aspects of such an extended conception of the deliberative function of representation. The first concerns the increased emphasis given to the role of non-electoral accountability, as the result of greater information and the capacity of civil society organizations to put governmental decisions under scrutiny. Manin (1997) considers this as the central aspect of a new metamorphosis of representative government, from a party-based form to an audience-based form. Various theories have been advanced (Rosanvallon 2008; Keane 2012; Green 2010; Dobson 2014), extending Manin’s analysis to the conclusion that democratic power is increasingly exercised through complex processes of accountability, by the more diffuse capacity of keeping an eye, carefully listening, scrutinizing and monitoring governmental and state action, or in some cases by attributing a representative function to the balancing powers of other non-majoritarian institutions such as constitutional courts and delegated agencies. The second aspect is that suggested by Lisa Disch, who further extends the reflexive function to the role played by social and political mobilization, and the importance of agonistic forms of politics (2011, pp. 111–13). She regards the encouragement of contestation and dissent as being intrinsic to the operation of political representation as mobilization.

Democratic representation and its normative principles  33 Although the proliferation of sites of discursive and agonistic forms of control and accountability may be regarded as a step forward in the democratization of the system of political representation by providing greater access to it, such a proliferation risks producing a cacophony of voices and the weakening of the system’s deliberative capacity. The explosion of what Mastropaolo (2018, pp. 327–30) has called ‘occasional’ representation, also favoured by new social media, may risk destabilizing what Pizzorno (1993) describes as the ‘virtue of binding’ that is one of the promises of politics, and that could be considered as the most prominent feature of a well-ordered system of political representation. On the criterion of reflexive judgement, as well as for empowerment and inclusion, the struggles for democratic representation go on.

ACKNOWLEDGEMENTS I would like to thank the editors of this volume for their help and encouragement in the preparation of this contribution and for their kind and boundless patience throughout.

REFERENCES Baker, Keith M. (1990), ‘Representation redefined’, in K.M. Baker, Inventing the French Revolution, Cambridge: Cambridge University Press, pp. 224–51. Barroso, Luís Roberto (2016), ‘Reason Without Vote: The Representative and Majoritarian Function of Constitutional Courts’, in T. Bustamante and B.G. Fernandes (eds), Democratizing Constitutional Law, Cham: Springer International Publishing, pp. 71–90. Birch, Anthony H. (1971), Representation, London: Pall Mall Press. Bobbio, Norberto (1999), Teoria Generale della Politica, Turin: Einaudi. Böckenförde, Ernst-Wolfgang (1998), ‘The Concept of the Political: A Key to Understanding Carl Schmitt’s Constitutional Theory’, in David Dyzenhaus (ed.), Law as Politics: Carl Schmitt’s Critique of Liberalism, Durham, NC: Duke University Press, pp. 37–55. Bourdieu, Pierre (1981 [1991]), ‘Political Representation: Elements for a Theory of the Political Field’, in P. Bourdieu, Language and Symbolic Power, Oxford: Polity Press. Brandon, Robert B. (2011), Perspectives on Pragmatism: Classical, Recent, and Contemporary, Cambridge, MA, USA and London, UK: Harvard University Press. Brito Vieira, Monica (2009), The Elements of Representation in Hobbes, Leiden: Brill. Brito Vieira, Monica (2017), ‘Performative Imaginaries: Pitkin versus Hobbes on Political Representation’, in Monica Brito Vieira (ed.), Reclaiming Representation: Contemporary Advances in the Theory of Political Representation, New York, NY: Routledge, pp. 25–49. Brito Vieira, Monica and David Runciman (2008), Representation, Cambridge and Malden, MA: Polity Press. Castiglione, Dario and Johannes Pollak (2019), ‘The Logic of Democratic Presence in Representation, in Dario Castiglione and Johannes Pollak (eds), Creating Political Presence: The New Politics of Democratic Representation, Chicago, IL: Chicago University Press, pp. 16–36. Castiglione, Dario and Mark E. Warren (2019), ‘Rethinking Democratic Representation: Eight Theoretical Issues and a Postscript’, in Lisa Disch, Mathijs van de Sande and Nadia Urbinati (eds), The Constructivist Turn in Political Representation, Edinburgh: Edinburgh University Press, pp. 21–47. Dalton, Russell J. and Hans-Dieter Klingemann (2010), Oxford Handbook of Political Behavior, Oxford: Oxford University Press. Disch, Lisa (2011), ‘Toward a Mobilization Conception of Democratic Representation’, American Political Science Review, 105 (1), 100–114.

34  Research handbook on political representation Disch, Lisa (2019), ‘Introduction: The End of Representative Politics?’, in Lisa Disch, Mathijs van de Sande and Nadia Urbinati (eds), The Constructivist Turn in Political Representation, Edinburgh: Edinburgh University Press, pp. 1–18. Dobson, Andrew (2014), Listening for Democracy: Recognition, Representation, Reconciliation, Oxford: Oxford University Press. Green, Jeffrey E. (2010), The Eyes of the People: Democracy in an Age of Spectatorship, Oxford: Oxford University Press. Habermas, Jürgen (1962), Strukturwandel der Öffentlichkeit: Untersuchungen zu einer Kategorie der Bürgerlichen Gesellschaft, Frankfurt am Main: Suhrkamp Verlag. Hobbes, Thomas (1651 [1996]), Leviathan, Richard Tuck (ed.), Cambridge: Cambridge University Press. Hofmann, Hasso (1974 [2003]), Repräsentation: Studien zur Wort- und Begriffsgeschichte von der Antike bis ins 19 Jahrhundert, Berlin: Duncker & Humblot. Keane, John (2012), ‘Monitory Democracy?’, in Sonia Alonso, John Keane and Wolfgang Merkel (eds), The Future of Representative Democracy, Cambridge: Cambridge University Press, pp. 212–35. Laclau, Ernesto (2005), On Populist Reason, London, UK and New York, NY, USA: Verso. Leibholz, Gerhard (1929 [1973]), Die Repräsentation in der Demokratie, Berlin: De Gruyter. Loughlin, Martin (2003), The Idea of Public Law, Oxford: Oxford University Press. Lovenduski, Joni (2015), Gendering Politics, Feminizing Political Science, London: Rowman and Littlefield International, ECPR Press. Mair, Peter (2013), Ruling the Void: The Hollowing of Western Democracy, London, UK and New York, NY, USA: Verso. Manin, Bernard (1997), The Principles of Representative Government, Cambridge: Cambridge University Press. Mansbridge, Jane (2003), ‘Rethinking Representation’, American Political Science Review, 97 (4), 515–27. Mansbridge, Jane (2009), ‘A “Selection Model” of Political Representation’, Journal of Political Philosophy, 17 (4), 369–98. Mansbridge, Jane (2019), ‘Recursive Representation’, in Dario Castiglione and Johannes Pollak (eds), Creating Political Presence: The New Politics of Democratic Representation, Chicago, IL: Chicago University Press, pp. 298–338. March, James G. and Johan P. Olsen (1989), Rediscovering Institutions: The Organizational Basis of Politics, New York, NY: The Free Press. Mastropaolo, Alfio (2018), ‘Le Acrobazie della Rappresentanza’, Comunicazione Politica, 3, 317–32. Mill, John Stuart (1861), Considerations on Representative Government, London: Parker, Son, and Bourn, West Strand. Montanaro, Laura (2018), Who Elected Oxfam?, Cambridge: Cambridge University Press. Mulieri, Alessandro (2016), ‘Hasso Hofmann and the Polysemy of Representation’, Redescriptions, 19 (2), 127–45. Pasquino, Pasquale (1998), Sieyès et l’Invention de la Constitution en France, Paris: Editions Odile Jacob. Phillips, Anne (1995), The Politics of Presence: The Political Representation of Gender, Ethnicity, and Race, Oxford: Oxford University Press. Pitkin, Hanna Fenickel (1967), The Concept of Representation, Berkeley, CA: University of California Press. Pizzorno, Alessandro (1993), Le Radici della Politica Assoluta: E Altri Saggi, Milan: Feltrinelli. Przeworski, Adam, Susan C. Stokes and Bernard Manin (1999), Democracy, Accountability and Representation, Cambridge: Cambridge University Press. Rosanvallon, Pierre (2008), Counter-Democracy: Politics in an Age of Distrust, Cambridge: Cambridge University Press. Saward, Michael (2010), The Representative Claim, New York, NY: Oxford University Press. Saward, Michael (2014), ‘Shape-Shifting Representation’, American Political Science Review, 108 (4), 723–36. Saward, Michael (2017), ‘Performative Representation’, in Monica Brito Vieira (ed.), Reclaiming Representation: Contemporary Advances in the Theory of Political Representation, New York, NY: Routledge, pp. 75–94.

Democratic representation and its normative principles  35 Saward, Michael (2019), Liminal Representation, in Dario Castiglione and Johannes Pollak (eds), Creating Political Presence: The New Politics of Democratic Representation, Chicago, IL: Chicago University Press, pp. 276–97. Schattsneider, E.E. (1942 [2004]), Party Government: American Government in Action, New Brunswick, NJ: Transaction Publishers. Schmitt, Carl (1928 [2008]), Constitutional Theory, trans. and ed. by Jeffrey Seitzer, Durham, NC: Duke University Press. Schumpeter, Joseph A. (1943 [1976]), Capitalism, Socialism, Democracy, London, UK and New York, NY, USA: Routledge. Schwartz, Nancy L. (1988), The Blue Guitar: Political Representation and Community, Chicago, IL, USA and London, UK: University of Chicago Press. Sintomer, Yves (2013), ‘The Meanings of Political Representation: Uses and Misuses of a Notion’, Raisons Politiques, 50 (2), 13–34. Sternberger, Dolf (1971), ‘A Controversy of the late Eighteenth Century Concerning Representation’, Social Research, 38 (3), 581–94. Taylor, Charles (2004), Modern Social Imaginaries, Durham, NC, USA and London, UK: Duke University Press. Thomas, Yan (1995), ‘Fictio Legis: L’Empire de la Fiction Romaine et ses Limites Médiévales’, Droits, Revue Française de Théorie Juridique, 21, 17–63. Tormey, Simon (2015), The End of Representative Politics, Cambridge and Malden, MA: Polity Press. Urbinati, Nadia (2006), Representative Democracy: Principles and Genealogy, Chicago, IL: University of Chicago Press. Violante, Piero (1981 [2012]), Lo Spazio della Rappresentanza: Francia 1788–1789, Rome: XL Edizioni. Voegelin, Eric (1952), The New Science of Politics, Chicago, IL: Chicago University Press. Warren, Mark E. (2019), ‘How Representation enables Democratic Citizenship’, in Dario Castiglione and Johannes Pollak (eds), Creating Political Presence: The New Politics of Democratic Representation, Chicago, IL: Chicago University Press, pp. 39–60. Williams, Melissa S. (1998), Voice, Trust, and Memory: Marginalized Groups and the Failings of Liberal Representation, Princeton, NJ: Princeton University Press.

3. Democracy and representation Mark B. Brown

The relationship between democracy and representation is a vast and vexing topic, especially today, when the claims of established democratic governments to represent their citizens seem to many people increasingly implausible. The global spread of ostensibly democratic regimes since the 1970s has been accompanied by growing social and economic inequality, a decline in voter turnout, an erosion of support for established political parties, and a general decline of trust in representative institutions (Hay 2007; Norris 1999; Tormey 2015). With the rise of transnational institutions, global social movements, and transboundary issues like climate change and migration, many citizens have grave doubts about the capacity of national states to represent them. Populist anger against elites has led to electoral upsets around the world. Protest movements such as Occupy (which started in 2011 with Occupy Wall Street) seem to be ‘expressions of a general rejection of the logic of representation’ (Sitrin and Azzellini 2014, p. 41). There has been a proliferation of ‘horizontal’ modes of organization and ‘prefigurative’ forms of politics by citizens who seek to live their future society today rather than elect representatives to build it tomorrow (Tormey 2015, pp. 9, 35). When protesters in the United States march in the streets chanting ‘This is what democracy looks like!’, they generally don’t mean political representation. In light of such developments, theorists of ‘post-democracy’ and the ‘post-political’ argue that democracy has become a mere façade (Crouch 2004; Wilson and Swyngedouw 2014). Simon Tormey (2015) says we have reached a ‘post-representative’ phase in the history of democracy. There has been a proliferation of books with titles such as How Democracies Die and How Democracy Ends (Levitsky and Ziblatt 2018; Runciman 2018). And yet, what we have been seeing is arguably not a rejection of either representation or democracy as such, but of liberal-democratic, elite-dominated, state-centric forms of representative democracy. As recent theorists of political representation have emphasized, political representation occurs in many places besides national legislatures, and it responds to issues that reach far beyond the boundaries of territorial states (Urbinati and Warren 2008, p. 390). In civil society organizations, labour unions, online forums, and a wide range of other non-state venues, people constantly make claims to stand for, speak for and act for others, often without formal authorization or accountability. Rather than a rejection of representation in favour of democracy, we may be seeing a variety of efforts to democratize representation. To bring such efforts to light, we need to avoid equating democracy with a particular set of political institutions, such as free elections or national legislatures, and instead think of democracy as a normative standard for evaluating political relationships of all kinds (Warren 1996). From this perspective, ‘The degree to which a representative relationship is “democratic” would depend upon the degree to which it enables individuals potentially affected by a collective decision to have opportunities and powers to affect the decision itself, in some way proportionally to his or her stake in the outcome’ (Castiglione and Warren 2019, p. 24). According to this standard, sometimes called the ‘all-affected principle’, democratizing rep36

Democracy and representation  37 resentation means empowering people to shape the political decisions that shape their lives, wherever those decisions may occur. This chapter examines selected aspects of the historical, conceptual, and political relationship between representation and democracy. The chapter first sketches six distinct approaches to understanding how these complex concepts relate to each other. The discussion then turns to two key challenges associated with efforts to democratize political representation: the representation of groups under conditions of structural inequality, and questions of democratic legitimacy.

FROM REPRESENTATIVE GOVERNMENT TO DEMOCRATIC REPRESENTATION Historically, the relation of representation and democracy has been understood in many different ways in Western political thought, and here I briefly summarize six of them. First, from antiquity until the early modern period, the two concepts were generally seen as having nothing to do with each other. Ancient Athens and other early self-governing communities had practices and institutions that people today would call representation, but they did not have a corresponding word or concept. The Latin word repraesentare was used in Ancient Rome, but until the thirteenth or fourteenth century it had nothing to do with politics (Brito Vieira and Runciman 2008, pp. 6–15; Pitkin 1989). Second, many thinkers have seen representation and democracy as distinct forms of politics, largely opposed to each other. Between the thirteenth and seventeenth centuries, the concept of representation gradually acquired a political meaning of ‘acting for’ someone else, but for a long time it had no link to democracy. Hobbes ([1651] 1991) provided the first modern theory of political representation, but he rejected democracy and republicanism. He argued that the ‘multitude’ in the state of nature authorizes a sovereign to represent them, and thereby acquires a collective identity as a unified people. But Hobbes insisted that security requires a sovereign that is unaccountable to the people. Rousseau also saw no way for representation to be democratic, but for different reasons. He famously insisted, ‘Sovereignty cannot be represented’, because it ‘consists essentially in the general will, and the will cannot be represented’ (Rousseau ([1762] 1978, p. 102). That is, nobody can express your will for you. For Rousseau, therefore, collective self-rule requires direct and unmediated expression of political will by the people themselves, an idea that later inspired participatory-democratic critiques of representation (Urbinati 2006, pp. 6–7, 60–62). Ironically, Rousseau also rejected democracy as utopian, and until the nineteenth century most political commentators saw democracy as an unstable and irresponsible form of government, forever associated with the execution of Socrates in ancient Athens (Wood 1992). Montesquieu, Sieyès, and other early liberal thinkers thus generally used the term representative government rather than representative democracy (Urbinati 2006, pp. 27, 73, 138–61). In Montesquieu’s influential scheme, democracy was the popular element in a system of mixed government, ideally held in check by aristocratic and monarchical elements (Montesquieu [1748] 1989, ch. 11). James Madison famously contrasted the tumultuous history of ‘pure democracy’ with the idea of a ‘republic’ in which ‘the scheme of representation takes place’ (Madison et al. [1788] 1987, p. 81). Most eighteenth-century political elites were interested in using representation to insulate government from democratic pressures, and especially from

38  Research handbook on political representation challenges to the interests of wealthy white men, rather than in how representation might help fulfil democratic ideals (Keane 2009, pp. 273–89; Brito Vieira and Runciman 2008, pp. 45–60; Manin 1997). The gradual expansion and protection of voting rights – a result of countless struggles from the early nineteenth century until well into the twentieth century, and continuing today in some countries, including the United States – led to a gradual change in the meaning of democracy. What Madison had called a republic became known as representative democracy or liberal democracy, leading to a third view of the relation between representation and democracy: democracy is representative, and representation is democratic, to the degree that government decisions correspond to public will and opinion. This basic view inspired a proliferation of political parties, interest groups, labour unions, and other political organizations that claimed to speak for both particular constituencies and the people as a whole. It motivated reformers who sought to close the gap between citizens and their representatives with provisions for government transparency, citizen initiatives and referendums, and popular election of party candidates. This view has been called the ‘folk theory’ of democracy, and it continues to play a prominent role in popular discourse (Achen and Bartels 2016, p. 21). It also underlies a lot of research on democracy in empirical political science, which has long sought to assess the ‘responsiveness’ of government, usually by measuring the correspondence between citizen preferences and government decisions (Sabl 2015, p. 346). Empirical researchers often cite Robert Dahl (1956; 1971) or Hanna Pitkin (1967) as sources for the criterion of responsiveness. But neither author actually views ‘responsiveness’ as the primary feature of representative democracy, especially not in the sense of measurable correspondence (Disch 2012; Sabl 2015). Indeed, democratic theorists have generally argued that representative democracy depends on a wide range of factors that go beyond direct responsiveness to citizen preferences, including social and political equality, informed citizenship, and concern for the public good (Dahl 1956, pp. 124–51; [1998] 2015, pp. 37–8; Pitkin 1967, pp. 232–40). To be sure, Pitkin wrote that representation requires ‘acting in the interest of the representative, in a manner responsive to them’. But she also stated, ‘The representative must act independently; his action must involve discretion and judgment’ (Pitkin 1967, p. 208, see also p. 233). Pitkin’s nuanced view of representation has been taken up in recent constructivist works on representation, which I will discuss in a moment, but when her book was published in 1967 most political theorists were more interested in participation than representation. They were seeking in part to defend democratic ideals of active citizenship against ‘minimalist’, ‘realist’, and ‘elitist’ views of political representation. The Cold War debates between participatory democrats and theorists of elite democracy relied on a fourth view of the relation between representation and democracy, upon which the two sides implicitly agreed: representative democracy is a contradiction in terms. For democratic elitists such as Joseph Schumpeter ([1942] 1950), the enormous scale and complexity of modern societies made it both impossible and undesirable for governments to be responsive to citizen preferences. Accordingly, these authors defined representation in aristocratic or technocratic terms, and they called for reducing democracy to a minimum. As Schumpeter put it, ‘Democracy means only that the people have the opportunity of accepting or refusing the men who are to rule them’ (p. 284–5). Participatory democrats criticized the elitist view of citizens as largely passive, but they implicitly accepted the elitist view of representation as anti-democratic. They thus defined participatory democracy in opposition to representative democracy. As Benjamin Barber made the point, ‘Representation is incompatible with freedom

Democracy and representation  39 because it delegates and thus alienates political will at the cost of genuine self-government and autonomy’ (Barber 1984, p. 145). Similarly, Sheldon Wolin (1996) argued that democracy, properly understood, appears only rarely in ‘fugitive’ moments of community-oriented popular mobilization. A more recent version of this perspective appears among those who define democracy, not as collective self-rule, but as the inherently disruptive, transgressive, anti-institutional ‘power of the demos’ (Devenney 2019, p. 235; Ranciere 2006, pp. 52–4). Either implicitly or explicitly, these authors suggest that representative democracy is an oxymoron (Barber 1984, p. xxii; see Urbinati 2006, pp. 4, 8, 57–8). In the 1990s, a fifth view of the relation between representation and democracy appeared in what has been called the ‘representative turn’ in democratic theory (Ankersmit 1996; Brito Vieira 2017; Disch 2015; Urbinati 2006; Urbinati and Warren 2008). Authors working in this vein rejected the widespread view, shared by theorists of participatory and elite democracy, that representation and democracy are inherently opposed to each other. Instead, they argued that nearly all political activity depends in some way on practices of representation. In David Plotke’s memorable formulation, ‘Representation is Democracy’ (Plotke 1997). For proponents of this approach, representative democracy is not a pragmatic concession to problems of scale in modern states. Rather, representative democracy is a distinct form of government, in many ways superior to direct democracy. Popular sovereignty involves not merely the expression of popular will, as for Rousseau, but a dynamic circular process that combines will and decision, on one hand, with deliberation and judgement, on the other (Urbinati 2006, pp. 52–9). In contrast to the voluntarist desire for immediacy associated with direct democracy, according to which representatives are supposed to mirror the will of their constituents, advocates of this view argue that representation creates a temporal and spatial gap between public opinion and government decision making (Kateb 1981; Urbinati 2006). Without such a gap, citizens would have no room for public deliberation, judgement and mobilization. From this perspective, representation actually makes democracy possible, and representative democracy is not an oxymoron but a tautology (Näsström 2006). Proponents of the representative turn have also challenged the focus on electoral politics common within both empirical political science and liberal political theory, often called the ‘standard account’ of representation (Disch 2015, p. 489; Urbinati and Warren 2008, p. 389). The standard account conceives of representation in juridical terms as a contract between a principal (a unified people) and its agents (the representatives). The boundaries of citizenship are defined by a territorial state, and the locus of representation is public officials in state institutions. In contrast to this view, theorists of the representative turn have argued that political representation is best understood as a dynamic relationship between representatives and constituents, dispersed through time and distributed across multiple venues (Mansbridge 2003; Urbinati and Warren 2008; Young 2000). Representatives do not replace their constituents, but instead depend on their active engagement in processes of representation (Plotke 1997, p. 19; Urbinati 2006, pp. 30–33). And many representatives are not popularly elected or located within state institutions (see also Chapter 5 in this Handbook). Claims to act or speak for others are made by non-governmental organizations, transnational social movements, and expert advisory bodies (Brown 2009). Potential constituents include not only voters in territorial states but also non-voters, undocumented immigrants, people in other countries, children, future generations, and non-humans (Brown 2018). This expansion of the range of possible constituents follows the extension of political problems beyond national boundaries: climate change, environmental pollution, migration, trade, financial markets, and so on. Work in this

40  Research handbook on political representation vein thus views political representation in systemic terms. It takes up a long-neglected feature of Pitkin’s account of representation, which noted that political representation is ‘not any single action by any one participant, but the over-all structure and functioning of the system, the patterns emerging from the multiple activities of many people’ (Pitkin 1967, pp. 221–2; Disch 2011, pp. 106–107; Mansbridge 2003, p. 515; Saward 2010, pp. 163–5; Young 2000, pp. 121–53). Finally, building on the ‘representative turn’, the ‘constructivist turn’ in democratic theory poses a more fundamental challenge to the standard account (Disch 2015). It goes beyond new ways of representing existing constituencies to examine how constituencies are created in the first place. It thus implicitly offers a sixth view of the relation between representation and democracy. The standard account defines ‘the people’ with regard to formal citizenship qualifications and the geographic boundaries of territorial states. But ‘the people’ has always been ‘a quintessentially political concept’, incessantly invoked by public officials, and yet inherently indefinite and amorphous (Canovan 2005, p. 140). From a constructivist perspective, representative claims should not aim for direct correspondence to fixed and pre-existing constituencies, but should instead be understood as partly constituting those same constituencies (Ankersmit 1996, pp. 45–51; Saward 2010; Disch 2012). Political representation contains an aesthetic and performative dimension, in the sense that representative claims shape both the representatives and those they represent. This does not mean that representatives create their constituencies from scratch. Citizens have various experiences, values, and attributes that exist before claims to represent them, and these factors shape and constrain the range of representative claims that will seem plausible to any given audience (Disch 2015, pp. 490, 493; Saward 2010, pp. 75, 80, 192). However, these pre-existing factors do not determine what counts as representation. As Lisa Disch puts it, ‘Absent representation, there may be a population but there cannot be a people, constituency or group’ (Disch 2019a, p. 9; Urbinati 2006, p. 37). Constructivist accounts of representation find empirical support in research that shows that citizens do not enter politics with fixed preferences but instead form preferences in response to the claims of their would-be representatives (Disch 2011). The constructivist perspective raises difficult questions about the relation between representation and democracy. If citizen preferences cannot provide a pre-existing standard for representative practices, what makes representation democratic? How can we determine when political representation should be democratized? Effectively addressing such questions requires maintaining a conceptual distinction between representation and democracy. If representative democracy is a tautology, it is so only as a normative ideal. And efforts to assess how well that ideal has been fulfilled in particular cases arguably depend on participation by the represented themselves. In this spirit, the remainder of this chapter briefly examines two key conundrums for efforts to democratize political representation: group representation and political legitimacy.

STRUCTURAL INEQUALITY AND GROUP REPRESENTATION Political representation has often been seen as a matter of speaking and acting for a socially homogeneous and unified people. However, many of the central concepts of liberal democracy are not politically neutral, but constructed to protect dominant interests. The idea of the social contract, for example, is closely enmeshed with capitalism, colonialism, slavery, and

Democracy and representation  41 the domination of women and racialized groups (Pateman and Mills 2007). From Hobbes to Rawls, the abstract ideal of government as based on freely contracting sovereign individuals has helped obscure the real conditions of oppression and exploitation under which the majority of people have lived. Similarly, the concept of citizenship as a matter of social status, rather than an activity of collective self-government, arose in the United States in direct and explicit contrast to those who lacked such status, especially women and enslaved African Americans (Shklar 1991; Olson 2004). From this perspective, one of the biggest challenges for efforts to democratize political representation today is economic and social inequality. Established democracies are characterized by large inequalities in wealth, health, education, housing, criminal justice, and many other social indicators. Many of these inequalities are structural, in the sense that they have been established and perpetuated by laws, policies, and institutions (Hayward 2009, pp. 112–13; Young 2000, pp. 92–9). And many structural inequalities are unjust, because they have been created by government policies that systematically favoured those socially defined as belonging to dominant groups, according to divisions of class, race, gender, sexual orientation, and so on. More broadly, structural inequalities are a result of capitalism, colonialism, patriarchy, white supremacy, and other systems of oppression and exploitation. (Of course, what counts as an unjust inequality often becomes a matter of dispute, subject to questions of democratic legitimacy, as I discuss in the next section.) Most of those who benefit from structural inequalities see themselves as having an interest in maintaining their superior social status. This means that efforts to make government laws and policies more closely reflect prevailing political interests, which according to the standard account is the task of political representation, will tend to perpetuate existing structural inequalities (Hayward 2009, pp. 121–2). One of the key conceptual tools for contesting structural inequalities has been descriptive representation: the idea that representatives should be ‘in their own persons and lives in some sense typical of those they represent’, with regard to politically significant demographic categories such as class, race, gender, sexual orientation, and so on (Mansbridge 1999, p. 629; Phillips 1995; Young 2000; Williams 1998). Such demographic categories have sometimes been understood in essentialist terms as a pre-existing foundation for representative claims, thereby denying the contingency and internal multiplicity of group identities. But from a constructivist perspective, social groups only acquire political identity and agency through being represented, whether by public officials, social movement leaders, or other political actors. Indeed, all political representation arguably has a descriptive dimension, in so far as it involves making claims not only for but also about the represented (Saward 2010). When politicians claim to speak for the ‘hard-working people’ in their districts, they construct their would-be constituents as such. The next section suggests that for group representation to be democratically legitimate, it should construct group members as, among other things, critical citizens with the capacity to evaluate and contest how they are represented (Castiglione and Warren 2019, pp. 36–7; James 2004, pp. 55–6; Young 2000, pp. 138–9). Group representation is clearly not always democratic, however, and some groups (e.g., white nationalists) are either internally anti-democratic, have external anti-democratic effects, or both. Descriptive representation can be usefully understood as serving the purposes of both advocacy and deliberation (Hayward 2009, pp. 114–18). With regard to advocacy, it is clear that not all members of any particular group have the same interests or policy preferences on all issues. Not all African Americans have the same views on affirmative action, for example. Nonetheless, members of disadvantaged groups are likely to share a general interest in elim-

42  Research handbook on political representation inating structural inequalities that harm their group. The lives and opportunities of African Americans, despite differing in all kinds of ways, have been affected by a long history of institutionalized racism, arguably creating a group interest in policies to promote racial justice. Descriptive representation is one way to increase the likelihood of effective advocacy on behalf of such group interests. With regard to deliberation, members of historically disadvantaged groups, while differing in their interests and policy preferences, often share group-related experiences and concerns (Young 2000). Women are more likely than men to have direct experience of sexual discrimination and harassment, for example. Women are thus more likely than men to raise the topic in discussion, and women are better positioned than men to fully understand the issue and how to address it. Women obviously differ from each other in all kinds of ways, and men may learn about sexual discrimination and seek to combat it. But if political discussions on relevant policies were to include only men, they would lack the experiential knowledge, personal testimony, and other epistemic resources that women can often bring to the table more effectively (Mansbridge 1999; Young 2000). Of course, neither of these approaches to descriptive representation is sufficient to eliminate structural inequalities and democratize political representation. Members of disadvantaged groups are often in the numerical minority in political institutions, which means that their interests will tend to be outweighed by the majority. And while deliberation may persuade some members of dominant groups to join efforts to eliminate structural inequalities, it would be naïve to expect most to do so (Hayward 2009, pp. 118–20). Therefore, in addition to attempting to shape people’s interests directly through deliberation, there is a need for laws and policies that indirectly facilitate the creation of new interests in eliminating structural inequalities. The challenge, as Clarissa Hayward argues, ‘is to restructure representative institutions so that they do not simply track the interests of all, but rather change political interests in ways that promote democratic inclusiveness and political equality’ (Hayward 2009, p. 132, original emphasis). What that entails is complex and case specific. It may involve redesigning electoral districts or election procedures to give greater influence to disadvantaged groups. It may require increased legal protections or government support for civil society organizations. The point here is that eliminating structural inequalities requires not merely greater participation in existing representative institutions but also new institutions. It is also worth noting that descriptive representation provides no guarantee that the interests or perspectives of subordinate groups will be adequately represented. Those who claim to speak for a group sometimes lack the support of a majority of group members, and disagreements often arise about who belongs to which groups. In some cases, members of subordinate groups may determine that they are represented more effectively by someone perceived as belonging to a dominant group in at least some respects (Brito Vieira and Runciman 2008, p. 114). From a constructivist perspective, effective group representation depends on developing vibrant relationships between group members and those who claim to represent them in a variety of institutional contexts (Dovi 2002, p. 735; Hayat 2019, pp. 153–7). As the next section explains, however, whether such relationships are sufficient to lend democratic legitimacy to a representative claim depends on the judgements of group members themselves.

Democracy and representation  43

DEMOCRATIC LEGITIMACY A second challenge for thinking about the democratization of political representation arises with regard to standards of political legitimacy. What makes political representation legitimate? What makes it democratically legitimate? Constructivist theories of political representation have two basic implications for questions of political legitimacy: they lead to an expansion in the forms and criteria of legitimacy, and they suggest an important shift in how political legitimacy is determined and by whom. At the most basic level, for political representation to be considered democratically legitimate, citizens must be persuaded that a portion of the people (the representatives) effectively act and stand for the entire people. As Pierre Rosanvallon (2011) argues, popular elections long served this function, but Western democracies have not been able to rely solely on elections as a source of legitimacy since the late nineteenth century. With the gradual adoption of universal suffrage, elites increasingly questioned the notion that majority rule alone could guarantee governmental virtue and competence. Many saw a remedy in the rise of bureaucracy and expert administration, which promised a substantive form of legitimacy to complement the procedural legitimacy provided by popular elections. At least since the 1980s, however, both elections and expertise have faced widespread public scepticism (Rosanvallon 2011, pp. 68–71). Elections remain crucial for providing the legal basis and positive legitimacy of democratic governments. But sceptical publics today demand more, and appeals to expert authority rarely work any more. In response to this situation, new forms of legitimacy have been emerging in contemporary democracies. Rosanvallon thus locates in contemporary democracies a ‘radical pluralization of the forms of legitimacy’ (2011, p. 8). He identifies three distinct kinds of legitimacy. First, expert advisory bodies, regulatory oversight committees, and other independent authorities generate legitimacy through impartiality. Such independent authorities once promised objective expertise and expected automatic public deference, but Rosanvallon argues that today they ‘must be subject to permanent open debate’ and seen as ‘truly political in nature’ (pp. 103, 107). Second, constitutional courts and other deliberative bodies examine legislative decisions and thus offer legitimacy through reflexivity. They facilitate a confrontation between two representations of the people: the people as the current electoral majority, and the people as an idea, captured in constitutional principles (pp. 140–41). Third, descriptive representation promises legitimacy through social proximity between representatives and constituents. Each of these forms of legitimacy is less direct, more dynamic, and more provisional than the legitimacy promised by either popular elections or supposedly objective expertise. Who should assess claims by would-be representatives to provide impartiality, reflexivity or proximity? Building on Rehfeld (2006), Saward (2010) argues that what determines whether someone counts as a representative is not formal authorization or accountability, but whether an audience accepts a person’s claim to represent it in some respect. If an audience accepts a representative claim, the person who made the claim is provisionally recognized as a representative. But if those who accept the claim are not also the represented, the claim is not democratically legitimate (Saward 2010, pp. 143–60). For example, a delegate to the United Nations might be accepted by other delegates as representing the citizens of his or her own country, but not by those citizens themselves, in which case the delegate is a representative but not a democratically legitimate representative. Representative claims are only democratically legitimate to the extent they are accepted by the represented themselves. Audiences emerge

44  Research handbook on political representation and change through the process of representation, so both the identity and legitimacy of representatives are often a matter of ongoing contestation. From a constructivist perspective, citizens not only evaluate representative claims but are also partly constituted by them, as discussed previously. Does the latter make the former impossible? It need not, as long as representative claims help to construct their audiences as critical citizens who are capable of making such evaluations. The democratic legitimacy of representation thus depends on representative practices that constitute and mobilize constituencies capable of evaluating representative claims (Disch 2012; Montanaro 2012). From this perspective, democratic representatives should be ‘responsive’ to their constituents, not by allowing their decisions to be entirely determined by pre-existing public opinion, but by facilitating and responding to citizen concerns and judgements (Fossen 2019, pp. 834–5). Finally, democratic legitimacy also depends on a social and economic context that ensures that all citizens, especially the historically disenfranchised, have the necessary resources to critically assess representative claims (Saward 2010, pp. 154–9). This means that the democratic legitimacy of political representation is best examined not primarily in terms of particular representative claims, but with regard to what Disch (2011) calls systemic reflexivity: the degree to which a process of representation ‘does more or less to mobilize both express and implicit objections from the represented’ (p. 111). A system of political representation is more democratically legitimate the more it mobilizes citizens to evaluate and contest a wide variety of competing representative claims. Indeed, from the perspective of political actors, the democratic legitimacy of representative claims is not an academic matter, but a resource for mobilizing constituencies, building popular movements, and winning elections (Disch 2019b, p. 179).

CONCLUSION In a 1989 lecture, Claude Lefort offered an early formulation of the view that representation and democracy depend on each other: ‘Representative democracy is not merely the system in which the representatives hold political authority in the place of citizens who have designated them; it is also the system that gives society a visibility’, in the sense that it depends on a network of associations that enable citizens to express their opinions and interests (Lefort 2019, pp. 106, 109, original italics). And conversely, ‘all that is fruitful in associations, in community actions, in the sometimes savage moments which revive the urgency of an active participation in public life – all of this acquires a general and lasting impact only if it is connected with political representation’ (p. 113). Echoing Lefort’s comments, the research discussed in this chapter suggests that the democratic ideal of collective self-rule, the creation of a world in which ordinary people have equal and effective opportunities to shape the political decisions that shape their lives, depends on diverse practices and institutions of representation. Given their myriad institutional failures and deeply entrenched systemic injustices, today’s established democracies are a long way from this conception of representative democracy. But efforts to remake political representation seem more promising than calls to reject it.

Democracy and representation  45

REFERENCES Achen, C.H. and L.R. Bartels (2016), Democracy for Realists: Why Elections do not Promote Responsive Government, Princeton, NJ: Princeton University Press. Ankersmit, F.R. (1996), Aesthetic Politics: Political Philosophy Beyond Fact and Value, Stanford, CA: Stanford University Press. Barber, B.R. (1984), Strong Democracy: Participatory Politics for a New Age, Berkeley, CA: University of California Press. Brito Vieira, M. (ed.) (2017), Reclaiming Representation: Contemporary Advances in the Theory of Political Representation, New York, NY, USA and London, UK: Routledge. Brito Vieira, M. and D. Runciman (2008), Representation, Cambridge: Polity. Brown, M.B. (2009), Science in Democracy: Expertise, Institutions, and Representation, Cambridge, MA: MIT Press. Brown, M.B. (2018), ‘Speaking for Nature: Hobbes, Latour, and the Democratic Representation of Nonhumans’, Science and Technology Studies, 31 (1), 31–51. Canovan, M. (2005), The People, Cambridge: Polity. Castiglione, D. and M.E. Warren (2019), ‘Rethinking Democratic Representation: Eight Theoretical Issues and a Postscript’, in L. Disch, M. van der Sande and N. Urbinati (eds), The Constructivist Turn in Political Representation, Edinburgh: Edinburgh University Press, pp. 21–47. Crouch, C. (2004), Post-Democracy, Cambridge: Polity. Dahl, R. (1956), A Preface to Democratic Theory, Chicago, IL: University of Chicago Press. Dahl, R. (1971), Polyarchy: Participation and Opposition, New Haven, CT: Yale University Press. Dahl, R. ([1998] 2015), On Democracy, 2nd edn, New Haven, CT: Yale University Press. Devenney, M. (2019), ‘The Improper Politics of Representation’, in L. Disch, M. van der Sande and N. Urbinati (eds), The Constructivist Turn in Political Representation, Edinburgh: Edinburgh University Press, pp. 224–38. Disch, L. (2011), ‘Toward a Mobilization Conception of Democratic Representation’, American Political Science Review, 105 (1), 100–114. Disch, L. (2012), ‘Democratic Representation and the Constituency Paradox’, Perspectives on Politics, 10 (3), 599–616. Disch, L. (2015), ‘The “Constructivist Turn” in Democratic Representation, A Normative Dead End?’, Constellations, 22 (4), 487–99. Disch, L. (2019a), ‘Introduction: The End of Representative Politics?’, in L. Disch, M. van der Sande and N. Urbinati (eds), The Constructivist Turn in Political Representation, Edinburgh: Edinburgh University Press, pp. 1–18. Disch, L. (2019b), ‘Radical Democracy: The Silent Partner in Political Representation’s Constructivist Turn’, in D. Castiglione and J. Pollak (eds), Creating Political Presence: The New Politics of Democratic Representation, Chicago, IL: University of Chicago Press, pp. 162–85. Dovi, S. (2002), ‘Preferable Descriptive Representatives: Will Just any Woman, Black or Latino Do?’, American Political Science Review, 96 (4), 729–43. Fossen, T. (2019). ‘Constructivism and the Logic of Political Representation’, American Political Science Review, 113 (3), 824–37. Hay, C. (2007), Why We Hate Politics, Cambridge: Polity. Hayat, S. (2019), ‘Varieties of Inclusive Representation’, in D. Castiglione and J. Pollak (eds), Creating Political Presence: The New Politics of Democratic Representation, Chicago, IL: University of Chicago Press, pp. 141–61. Hayward, C.R. (2009), ‘Making Interest: On Representation and Democratic Legitimacy’, in I. Shapiro, S.C. Stokes, E.J. Wood and A.S. Kirshner (eds), Political Representation, Cambridge: Cambridge University Press, pp. 111–35. Hobbes, T. ([1651] 1991), Leviathan, ed. by Richard Tuck, Cambridge: Cambridge University Press. James, M.R. (2004), Deliberative Democracy and the Plural Polity, Lawrence, KS: University Press of Kansas.

46  Research handbook on political representation Kateb, G. (1981), ‘The Moral Distinctiveness of Representative Democracy’, Ethics, 91 (3), 357–74. Keane, J. (2009), The Life and Death of Democracy, London: Simon and Schuster. Lefort, C. (2019), ‘Democracy and Representation’, in L. Disch, M. van der Sande and N. Urbinati (eds), The Constructivist Turn in Political Representation, Edinburgh: Edinburgh University Press, pp. 104–17. Levitsky, S. and D. Ziblatt (2018), How Democracies Die: What History Reveals about Our Future, New York, NY: Viking. Madison, J., A. Hamilton and J. Jay ([1788] 1987), The Federalist Papers, ed. by I. Kramnick, New York, NY: Penguin. Manin, B. (1997), The Principles of Representative Government, Cambridge: Cambridge University Press. Mansbridge, J. (1999), ‘Should Blacks Represent Blacks and Women Represent Women? A Contingent “Yes”’, Journal of Politics, 61 (3), 628–57. Mansbridge, J. (2003), ‘Rethinking Representation’, American Political Science Review, 97 (4), 515–28. Montanaro, L. (2012), ‘The Democratic Legitimacy of Self-Appointed Representatives’, Journal of Politics, 74 (4), 1094–107. Montesquieu, Charles de Secondat [1748] (1989), The Spirit of the Laws, A.M. Cohler, B.C. Miller and H.S. Stone (trans. and ed.), Cambridge: Cambridge University Press. Näsström, S. (2006), ‘Representative Democracy as Tautology: Ankersmit and Lefort on Representation’, European Journal of Political Theory, 5 (3), 321–42. Norris, P. (1999), Critical Citizens: Global Support for Democratic Government, Oxford: Oxford University Press. Olson, J. (2004), The Abolition of White Democracy, Minneapolis, MN: University of Minnesota Press. Pateman, C. and C.W. Mills (2007), Contract and Domination, Cambridge: Polity. Phillips, A. (1995), The Politics of Presence, Oxford: Clarendon Press. Pitkin, H. F. (1967), The Concept of Representation, Berkeley, CA: University of California Press. Pitkin, H. F. (1989), ‘Representation’, in T. Ball, J. Farr and R.L. Hanson (eds), Political Innovation and Conceptual Change, Cambridge: Cambridge University Press, pp. 132–54. Plotke, D. (1997), ‘Representation is Democracy’, Constellations, 4 (1), 19–34. Ranciere, J. (2006), Hatred of Democracy, S. Corcoran (trans.), New York, NY, USA and London, UK: Verso. Rehfeld, A. (2006), ‘Towards a General Theory of Political Representation’, Journal of Politics, 68 (1), 1–21. Rosanvallon, P. (2011), Democratic Legitimacy: Impartiality, Reflexivity, Proximity, A. Goldhammer (trans.), Princeton, NJ: Princeton University Press. Rousseau, J.-J. ([1762] 1978), On the Social Contract, with Geneva Manuscript and Political Economy, J.R. Masters (trans.), R.D. Masters (ed.), New York, NY: St. Martin’s Press. Runciman, D. (2018), How Democracy Ends, New York, NY: Basic Books. Sabl, A. (2015), ‘The Two Cultures of Democratic Theory: Responsiveness, Democratic Theory, and the Empirical-Normative Divide’, Perspectives on Politics, 13 (2), 345–65. Saward, M. (2010), The Representative Claim, Oxford: Oxford University Press. Schumpeter, J.A. ([1942] 1950), Capitalism, Socialism, and Democracy, 3rd edn, New York, NY: Harper and Row. Shklar, J.N. (1991), American Citizenship: The Quest for Inclusion, Cambridge, MA: Harvard University Press. Sitrin, M. and D. Azzellini (2014), They Can’t Represent Us! Reinventing Democracy from Greece to Occupy, London, UK and New York, NY, USA: Verso. Tormey, S. (2015), The End of Representative Politics, Cambridge: Polity Press. Urbinati, N. (2006), Representative Democracy: Principles and Genealogy, Chicago, IL: University of Chicago Press. Urbinati, N. and M.E. Warren (2008), ‘The Concept of Representation in Contemporary Democratic Theory’, Annual Review of Political Science, 11, 387–412.

Democracy and representation  47 Warren, M.E. (1996), ‘What Should We Expect from More Democracy? Radically Democratic Responses to Politics’, Political Theory, 24 (2), 241–70. Williams, M.S. (1998), Voice, Trust, and Memory: Marginalized Groups and the Failings of Liberal Representation, Princeton, NJ: Princeton University Press. Wilson, J. and E. Swyngedouw (eds) (2014), The Post-Political and its Discontents: Spaces of Depoliticisation, Spectres of Radical Politics, Edinburgh: Edinburgh University Press. Wolin, S. (1996), ‘Fugitive Democracy’, in S. Benhabib (ed.), Democracy and Difference: Contesting the Boundaries of the Political, Princeton, NJ: Princeton University Press, pp. 31–45. Wood, G.S. (1992), ‘Democracy and the American Revolution’, in J. Dunn (ed.), Democracy: The Unfinished Journey, 508 bc to ad 1993, Oxford: Oxford University Press, pp. 95–105. Young, I.M. (2000), Inclusion and Democracy, Oxford: Oxford University Press.

4. Policy responsiveness and democratic quality Leonardo Morlino

‘The representative system must … be responsive to public opinion.’ This is a key statement by Pitkin (1967, p. 224) that sets responsiveness at the core of representation. It captures the normative aspect of representation that again Pitkin and others after her call ‘acting for’. Thus, starting from this, we can take it for granted that political representation is at the core of (contemporary) democracy. Within the perspective of this Handbook, the next step is asking whether and how responsiveness is at the heart of political representation. This chapter suggests a reply to the question and assesses the role of it within representation, by also taking into consideration normative and empirical perspectives. Hence, this chapter starts by presenting the normative and empirical definitions with the related indicators, the core relationships with accountability, the connections with other procedural dimensions of democracy within the debate on the assessment of democratic quality, and the emerging problem about responsiveness vis-à-vis responsibility. A few concluding remarks recall the critical aspects of the analysis, the ones still open to this debate, and show the broad possibility of new directions of research.

FROM A NORMATIVE TO AN EMPIRICAL DEFINITION One of the most authoritative normative definitions of democracy focuses on responsiveness as the core feature of a democracy: ‘a political system one of the characteristics of which is the quality of being completely or almost completely responsive to all its citizens’ (Dahl 1971, p. 2) and, more precisely, to the preferences of them, ‘considered as political equals’ (ibid., p. 1). But classic thinkers, such as Mill, Madison or Tocqueville, emphasized how exclusive attention to the preferences of the people and, consequently, to responsiveness may imply a tyranny of the majority. Therefore, democracy would be deeply flawed and limited. Hence, when giving the definition above, Dahl also immediately adds that: (a) there is responsiveness if all citizens are supposed to have an equal opportunity of formulating their preferences, to manifest those preferences to the other citizens, to have the preferences weighed equally; and (b) this happens if there are the liberal institutional guarantees related to the traditional set of civil and political freedoms, free and fair elections and alternative sources of information included. In a nutshell, responsiveness is at the core of democracy if complemented by the freedoms that have been empirically developed, especially in Western democracies. This reasoning recalls an essential observation by Sartori (1987, p. 7) on ‘[what] democracy is cannot be separated from what democracy should be’ (emphasis added). In fact, the two assumptions made by Dahl are normative statements, which would imply that democracy can only be liberal. Such an inextricable link brings at least two related analytic consequences. First, responsiveness, which we accept as a critical normative characteristic of representative democracy, is also empirically highly salient. Second, from an analytic perspective, if we accept the centrality of responsiveness, the challenge is proposing a normative notion that can 48

Policy responsiveness and democratic quality  49 be translated into an object of empirical research. In this way, we can better understand and partially overcome that kind of divide between normative and empirical scholars as discussed by Sabl (2015) when examining responsiveness. In this vein, the simplest common-sense definition of responsiveness is ‘the capacity of the government of satisfying the governed by executing its policies in a way that corresponds to their demands’ (see Morlino 2012, p. 208). From an empirical perspective, this implies that: (a) there is an effective government; (b) we can identify the demands of citizens; (c) there are the resources for implementing the consequent policies; and (d) there is the willingness of elected authorities to respond to citizens by satisfying their demands. For each of these features, there are empirical problems in detecting them. And if those difficulties are solved, there can be additional impediments and constraints related to achieving a high level of responsiveness. First, the effectiveness of government implies the existence and proper working of the rule of law. Briefly put, legal order, judicial independence, functioning bureaucracy, personal and institutional integrity or absence of corruption, civilian control of police and military are the usual dimensions that characterize the rule of law (see, among others, Bingham 2010). As is immediately evident, we are considering highly demanding aspects that can be differently measured. During the first two decades of the twenty-first century especially, international institutions such as the World Bank, foundations such as Freedom House or International IDEA, and research groups such as those who built data sets such as Varieties of Democracies or the Quality of Government, developed different, relevant measures, all considered helpful in research despite limitations and flaws. Second, if we put aside the first assumption by Dahl on preference formation, identifying the demands is often difficult and problematic. On the one hand, in complex contemporary democracies even the educated, informed and politically engaged citizens may have difficulties in defining their own needs, apart from the most vital ones such as being out of poverty and living a life with dignity. In different situations such as, for example, climate and environment issues, where the citizens would need some expertise to identify and decide what their wants and desires are, that kind of identification is impossible and easy to manipulate by political leaders. More emphatically, Pitkin has no hesitation in stating: ‘the represented has no will on most issues’ and, consequently, ‘the duty of representative is to do what is best for them’ (1967, p. 163). On the other hand, building on the theory of voting and the consequences of it, and taking into full account the development of the rational choice from Arrow (1963) onwards, William Riker (1982, esp. ch. 10) concluded that identifying the demands is a virtually impossible task. In fact, from an empirical perspective only a homogeneous distribution of preferences can allow an acceptable, easier identification of demands. In contemporary democracies, characterized by social and political high fragmentation and more recently by polarization, this is a virtually impossible result. The conclusion by Riker, which is correct from a strictly theoretical point of view, is so devastating that a way of finding an empirically viable solution is necessary. With obvious limitations, this can be envisaged by looking at the party proposals, their electoral results and the government that is formed. Moreover, as also recalled by Powell (2005), some institutional solutions can make elected authorities more responsive, even if the electoral majorities are artificially achieved through majoritarian electoral laws and also constitutional arrangements. Let us recall that during the last decades the enormous diffusion of polls and mass survey research, especially in advanced democracies, has been trying to perform the task of identi-

50  Research handbook on political representation fying the wants and preferences of citizens, but on several occasions in different countries the percentage of no response has been high or very high (even beyond 40 per cent). Hence, although the statement by Pitkin and the conclusions by Riker are too radical and eventually not empirically entirely correct, their observations are also empirically relevant. Third, let us assume that a government is effective and we are able to identify the demands of citizens: what if there is a profound, protracted economic crisis or if decision-makers have to cope with massive social or natural phenomena, such as waves of immigration or natural disasters? If there is climate change with related consequences? In other words, what if there are no or poor resources and other dramatic events change the possible decision-making agenda setting other priorities? On the one hand, the need and wants of citizens and non-citizens are still there; but, on the other hand, they cannot be met because of the scarcity of resources or the need to cope with unexpected events. No doubt this may be a contingent situation, but it is one which is more and more recurrent in some countries and at the same time bears a deeply delegitimating effect. Fourth, the willingness of elected authorities to respond to citizens by satisfying their demands cannot be taken for granted. The solemn declarations during the electoral campaign and even the party manifestos or party programmes, when they exist, may be inconsistent with the decisions subsequently made for objective and subjective reasons (see also Naurin and Thomson, Chapter 23 in this Handbook on ‘the fulfilment of election pledges’). In fact, on the one hand, the social and/or economic situations may change after the elections, and the elected authorities have to change their previous willingness on some issue into an unwillingness or an adaptation and reshaping of it. On the other hand, from a subjective point of view, at least four different situations can be identified. The incumbent politicians committed themselves before the elections to conflicting decisions that require a subsequent choice in terms of priorities. The government is a coalition, with different parties, whose leaders committed themselves to alternative solutions on the same issue or to different priorities. Third, the pressures of organized groups are able to change the availability and perspectives of incumbent leadership in different ways, financial contributions included. And, last but not least, elected leaders do not always seek to respond to the perceptions and positions of the citizens. On some occasions they try to influence citizens’ perceptions and understandings of what their needs are. That is, they can change the perception and priorities of citizens, also profiting from the complexity of problems and from the changes in economic and social situations that occur over the course of a legislature, which is usually a period of four or five years. Moreover, to better understand this fourth key feature, two additional aspects are evident. In the suggested definition, the response is always in terms of policy. Hence, responsiveness is necessarily ‘policy responsiveness’. However, even more relevant, there is an explicit analytic shift. That is, we assess responsiveness not by itself, as we assume this is empirically impossible, but with regard to its result: citizen satisfaction. A strong assumption is implicit but rarely expressed: there is effective responsiveness if the citizen is eventually satisfied. With this assumption, it is evident that because of the necessity of building a detectable, empirically simplified concept, the output (a decision is adopted and implemented) is merged with the outcome (a decision or set of decisions give results that are positively assessed by citizens who are for this reason satisfied). This is the reason why a few authors have preferred to define responsiveness in terms of congruence, that is, correspondence or consistency between the positions of party leaders on policy issues and voter positions. Thus, for example, Arnold and Franklin (2012, p. 1218),

Policy responsiveness and democratic quality  51 among other scholars, recall the previous literature on the ‘responsible party model’ and analyse the congruence by making reference to the policies that ‘enacted by governing political parties closely reflect the wishes of those who elected those parties to government office’. As is evident, in contrast to the previous perspective on voter or citizen satisfaction, here the focus is mainly on the output, not on the results (outcome) (see also Louwerse and Andeweg, Chapter 22 in this Handbook). As a way of summing up and clarifying this short definitional discussion, we can recall Powell’s chain of responsiveness (see Powell 2005). Three linkages characterize it: linkage I (structuring choices) connects citizens’ preferences to their voting behaviour; linkage II (institutional aggregation) connects election outcomes and selection of policy-makers who are committed to do what citizens want; and linkage III (policy-making) connects policy-makers committed to citizen desires and implemented public policies with the related outcomes. As already stated, despite the clarity of linkages, all of these have empirical problems and detection limits.

INDICATORS With these problems and constraints in mind, how has responsiveness been analysed empirically? If we stay at the core of research in the field, there are two main streams, one that is focused on congruence and the other on citizen satisfaction. If we start from the first stream, Lijphart (1999, pp. 287–8) referred to the congruence in terms of government–voter proximity. Following Huber and Powell (1994), such proximity is measured by ‘government distance’, which is basically the distance between the government’s position on the left–right scale and the position of the median voter. The expected conclusion by Lijphart is that consensual democracies are more responsive than majoritarian ones (see, however, Louwerse and Andeweg, Chapter 22 in this Handbook, for an in-depth operationalization of policy congruence). With all simplifications, indicated above, the other, recurrent way of measuring responsiveness is through citizen satisfaction. Starting from a common question (‘how satisfied are you with the way in which democracy functions in your country?’), which has been commonly used for many years in several mass surveys in Western Europe, Latin America, Eastern Europe, and other countries around the world, we assume that a high percentage of satisfied people means that the government is effectively responsive. In other words, the better the citizen satisfaction, the stronger the responsiveness. The simplification implicit in the analysis of this phenomenon has already been recalled above. Here, another limitation can be made explicit: such a responsiveness is a perceived one, and within contemporary democracies where manipulation of the media and technological developments are very relevant, this is not only simplifying but it may also imply a responsiveness that actually does not exist, or at least not at the (high) level indicated by the polls. This measure can be complemented, and consequently strengthened by measures of confidence for government and institutions, which are always policy oriented. These data can be obtained through mass survey research that is also recurrent in several countries. But when reflecting on all the indicators and measures we have been considering, we can easily see that we are actually analysing the perceived legitimacy under the implicit assumption that the higher the legitimacy, the higher also the responsiveness for a sort of empirical, although non-analytic, overlapping. There is nothing wrong in this

52  Research handbook on political representation research direction, but it is, however, something to be fully aware of. Let it be added that no better indicators have been elaborated and adopted in these years.

ANALYTIC AND EMPIRICAL CONNECTIONS WITH ACCOUNTABILITY For clarity’s sake, until now we have been discussing responsiveness by itself. But to better understand this mechanism that is so important in a democracy, we do need to connect it with another dimension that is at the core of every representative democracy. This is accountability (see also Chapters 21 and 25 in this Handbook) and from both a normative perspective and an empirical one the two dimensions are inextricably linked to justify and maintain a democratic regime. From a normative point of view, accountability with its three main features (information, justification and punishment/compensation) (see, for example, Schedler 1999, p. 17) is necessary in order to establish a chain of representation whereby the identified governmental actor/s, who are responsible for being or not being responsive, are appropriately rewarded or punished. Thus, only then can the intertwining of accountability and responsiveness assure representation. From an empirical perspective, the effectiveness of that intertwining is deeply affected by all the aspects we have already mentioned in the section on the definition. Now, we add them to all other constraints stemming from the three features of accountability discussed above. All three of these elements require the existence of a public dimension characterized by pluralism and independence and the real participation of a range of individual and collective actors. Within this perspective, at least three aspects should be emphasized. The first one is depicted by Papadopoulos (2020, p.701) when stating that: contemporary … governments are … characterised by high levels of complexity and even ‘messiness’. The process of governing is often interactive, entailing collaboration among various interdependent state and non-public agents. Political decisions are formulated or implemented through bargaining or deliberation in polymorphous networks, usually involving politicians, administrators, interest representatives, stakeholders, and experts. In policy networks, multiple principals with distinct preferences cohabit with multiple agents, while the latter are embedded in intricate accountability webs with many other forums endowed with unequal resources and power.

Second, as suggested by Mansbridge (2003), in addition to the traditional ‘promissory representation’, the one that from his point of view (elections) Fiorina had labelled ‘retrospective voting’, other forms of representation were spelled out in these years. They include the ‘anticipatory representation’, which refers to the representative who tries to please future voters; the ‘gyroscopic representation’ when ‘voters select representatives who can be expected to act in ways the voter approves without external incentives … the representatives act only for “internal” reasons’; and the ‘surrogate representation’, namely a ‘representation by a representative with whom one has no electoral relationship – that is, a representative in another district … Individual and interest groups … often turn to surrogate representatives to help advance their substantive interests, including their ideal-regarding interests’ (Mansbridge 2003, pp. 517, 520, 522). All these new forms of representation circumvent accountability, but for the gyroscopic and the surrogate ones in particular they are also ways to assure better responsiveness.

Policy responsiveness and democratic quality  53 Thus, the new forms of representation decouple accountability and responsiveness, violating in this way a key tenet of normative theory. We cannot ignore that these are democratic developments that deeply affect the way democracy works. Third, when looking at the empirical research, especially in the USA, the double interrelated dimensions we are considering here work more effectively if the so-called ‘public opinion’, meant as the aggregated responses of individuals that are detected in all usual opinion polls carried out in several countries, ‘expresses a coherent mood or view on a particular policy question (or bundle of policy or political questions) in a way that is recognizable by political elites’. If this is not so, first, ‘politicians and policy entrepreneurs often have substantial room to maneuver policy in detailed ways that are not visible to the public’; second, ‘the combination of contradictory public views on many key policy issues and the capacity of political elites to shape or direct citizens’ views significantly reduces the independent causal impact of public opinion’; and, third, ‘although policy will tend to follow public opinion more often than not, there is sufficiently wide variation in the extent of responsiveness across different issues and at different points in time to warrant increased scholarly attention to examining the institutional and political sources of variation’ (Manza and Lomax Cook 2002, pp. 657–8).

QUALITY OF DEMOCRACY: THE OTHER PROCEDURAL DIMENSIONS The closer attention to the connection between accountability and responsiveness should not lead us to overlook the highly relevant relations with the other critical qualities of a democratic regime. Here, taking for granted the salience of the rule of law in terms of effective implementation of existing rules and of levels of integrity/corruption (see also above), we should especially highlight the relations with two additional qualities. They are competition and participation. When considering the first dimension, the probability of a higher, more adequate responsiveness can be assured by more or less organized political actors, especially party leaders and parties, which are in competition. This is particularly evident when analysing electoral competition, but it is a mechanism that also works during the interval between one election and another when there is an active political opposition. The activity of the opposition may continuously influence the effective policies of the incumbent authorities through the pressures of public opinion that the opposition is able to activate. This observation also brings to the fore the key role that participation, whether it be institutionalized or not, may play to get a more effective responsiveness. As we may expect, the different forms of partisan participation are deeply influenced by how party leaders shape the various issues to respond to. However, it is more interesting to stress that especially on all issues related to socio-economic equality there is an important role of non-conventional participation, expressed through social movements and protest, in bringing about better responsiveness through the reactions and related decisions of the government (see Morlino 2012, ch. 8, section 2). To these more general results, based on a data set covering all existing democracies, Hutter and Vliegenthart (2018, esp. p. 366) added at least two other relevant findings from a data set on a few European countries (more precisely: France, the Netherlands, Spain and Switzerland). First, parties respond to protest covered in the news, and this is especially so for leftist parties that are more sensitive to societal demands rather than for rightist parties.

54  Research handbook on political representation Second, party response seems driven by party competition. In other words, parties (and party leaders) seem to be more responsive if other parties have already reacted to the issue supported by protest. In this vein, parties in opposition are more sensitive to protest than incumbent parties, and participation and competition are intertwined. This short discussion shows how close empirical connections are among the different democratic dimensions. Also, because of the ways in which it has been developing historically and adapting to the new changing social, economic and technological contexts, democracy works as a system: the key democratic aspects influence each other. This is so evident that research on those dimensions has been singling out a mechanism of mutual convergence: ‘a democracy … acquires its own inner logic that strongly connects procedural aspects and substantives ones with regards to contents and results’ (Morlino 2012, p. 248). We can take for granted that there are contingencies, constraints and problems that can weaken these relations. However, especially in the long run, the working of the mechanism is evident in the sense of mutually improving or worsening all qualities. If because of some specific event or the action of a leader, a quality is deeply affected for a while, the other qualities will also be affected for better or worse. When analysing such a mechanism more in depth, however, an additional aspect emerges that is worth highlighting here. That is, in terms of explained variance and significance, it is responsiveness that is more weakly correlated to the other dimensions of quality, also mentioned above (see Morlino 2012). However, this can be easily understood when reflecting on the fact that in comparison with the other qualities, responsiveness is a more volatile dimension as it is much more affected by a number of external factors, economic crises included, than the other dimensions. Thus, the focus on the mechanism of convergence helps a specific feature of responsiveness to be better understood: it is the dimension of democracy, which is more reactive to the change of the surrounding reality, and consequently the one that keeps the connections between institutions and citizens.

RESPONSIVENESS AND RESPONSIBILITY: A NEW THORNY QUESTION? From a normative point of view the alternative between an exclusive attention to responsiveness, which can be traced back to Rousseau and his focus on the volonté générale – what we could call today public opinion – and the attention to the liberal aspects of democracy on the limits of power, the protection of individual rights, or also the attention to the socio-economic equalities and solidarity, is still present as an object of discussion and possible different choices. But from an empirical perspective, the position taken by Dahl and other scholars has become the dominant one. Responsiveness is linked to other qualities in the way Powell was suggesting with the three linkages. And the subsequent empirical research on the quality of democracy showed it. But since the beginning of the twenty-first century and above all since the Great Recession, the economic crisis that started in 2008 with protracted consequences in the entire subsequent decade, there is a new difficult question that is again relevant from both a normative perspective and an empirical one: even if we accept that responsiveness is not the only democratic criterion, can governments make decisions that are against that responsiveness? The issue is not whether to be responsive or unresponsive. The crux of the matter is whether an incumbent elected authority can consciously go against their citizens. Mair (2009) made this point when, in discussing this issue, he drew attention to the responsibility

Policy responsiveness and democratic quality  55 toward third actors, usually other democracies linked by treaties and consequent obligations, especially within the European Union. From a normative theoretical point of view, what is the way out, if there is one? From an empirical point of view, what have the actual reactions of incumbent leaders been to this dilemma? Starting with the first question, let us recall that the term ‘responsibility’ has traditionally been a synonym of accountability. This is already evident in traditional political thought and above all in the early political science of post-World War Two (see, for example, Pennock 1952). In this vein, there is no third notion possible beyond responsiveness and accountability. Thus, theoretically, there are only two possible ways out. The first one was mentioned above and is shown by Mair, who in front of a new phenomenon related to the development of the European Union and the obligations accepted by the member states with the related limitations of sovereignty, interprets responsibility as external responsibility. And this is a category that was overlooked by a democratic theory focused on the nation state only with the citizens as the principals of democracy. The other solution is to add explicitly a time dimension to responsiveness, up to now evidently missing. More precisely, when considering responsiveness, we should take into account both a short-term process and a long-term one. Consequently, what Mair and others consider responsibility by coining a new additional notion, different from accountability, in a more traditional way can still be considered responsiveness in the medium or long term. In other words, to comply today with the obligations a country undertook toward third parties is also a way to protect the interests – not the immediate ones – of its citizens within a positive assessment of belonging to the European Union. From an empirical perspective, research points to a different conclusion vis-à-vis the one anticipated by normative theory. As shown by Laffan (2014) and other authors, some years after the beginning of the Great Recession, the member states deeply hit by the economic crisis accepted a new regulatory regime with at least three new agreements (the so-called Six-Packs, Two-Packs and the Fiscal Compact) that imposed additional constraints to national governments in budgetary and fiscal matters. In this perspective, the incumbent authorities considered that it was in the immediate interest of their country to sign these agreements and accept those constraints, although they could expect different positions by the citizens. Of course, other factors, especially the pressures of the market on the national public debt, contributed to accepting those new regulations. But the tension between the responsiveness and the responsibility has not been a salient aspect in the minds of decision-makers. The policies that are consequent to those agreements are a different matter that has been deepening the proor anti-European Union divide, as shown by the 2019 European elections. The negative impact of the economic crisis on responsiveness is also shown by research that stresses the salience of classic economic factors, such as public debt, unemployment, GDP and inflation, in explaining the limits to responsiveness (see Morlino and Quaranta 2014, but also Traber et al. 2018). Finally, accepting the distinction between responsiveness and responsibility, Linde and Peters (2018, p. 11) show how ‘responsiveness and responsibility do not need to be traded off per se. They can rather be related in a more positive way in which responsiveness facilitates responsibility’. On the grounds of the results of the empirical analysis of the European Social Survey: governments that are being perceived as responsive to short-term demands from citizens generate something like a ‘responsiveness capital’ that eventually allows governments to make important but non-responsive decisions and that these decisions are more likely to be accepted by the citizens. In this way, responsive government actions create leeway for governments to act in a responsible way.

56  Research handbook on political representation Thus, eventually responsiveness and responsibility can be mutually reinforcing rather than in conflict, and ‘responsive government seems to be a precondition for citizens’ acceptance of responsible government’ (Linde and Peters 2018).

CONCLUDING REMARKS When drawing the lines of this analysis, a scholar cannot help being puzzled by the normative and empirical problems that such an important, core aspect of democracy still has. This awkward situation is even more relevant today when responsiveness is achieving a higher and higher salience. For contemporary democracies in the first decades of this century, this is so because they have to cope with urgent problems brought by poor or no economic growth, immigration phenomena, higher socio-economic inequality, climate change and so on. The political consequences include political dissatisfaction at the citizen level, the electoral success of neo-populist parties and political polarization. And all this brings more and more to the fore the issue of responsiveness. Moreover, within such a difficult context, policy responsiveness has to confront two different innovations that we singled out in our analysis and would like to emphasize here again. For most European countries a theoretical notion and a related reality that was conceived and implemented within the boundaries of the nation state or a multinational federal state, has now to be reconsidered within the new supra-national reality that the European Union embodies. A previous section mentioned the different ways to cope with this, but the reality is that there will be additional developments in connection to the direction the European Union will follow. Moreover, from a different point of view, the technological advancements with the internet and social networks profoundly affect the perception of implemented policies and consequently of responsiveness, leaving much scope for manipulation. With this in mind, the reflection and empirical research on responsiveness should start again from these two key aspects to try to keep or even to deepen it in future democracies. A last, critical remark is in order here. The entire analysis in this chapter and the related scholarly literature has ignored the topic of responsiveness in non-democratic regimes. Again, especially during the last four decades and in the expected future, this issue will become more and more relevant in the traditional regimes that have been slowly transforming into hybrid regimes. Examples of these are Morocco or Jordan, to mention only two relatively well-known cases, and in the new electoral authoritarianisms, such as Russia or even Turkey, where some sort of opposition is tolerated and pursuing of consensus is among the goals of the incumbent, elected authorities, although compounded by suppression, vote manipulation and fraud, which make elections in those countries flawed and merely ceremonial events without effective choice. Let us add, however, to attenuate this lacuna that the key aspects and mechanisms of responsiveness that lie at the core of democracies are also relevant in the traditional, hybrid and authoritarian regimes with all the additional difficulties, problems and limitations due to the non-guaranteeing of fundamental rights and, consequently, a highly defective formation of public opinion (see also Chapter 12 in this Handbook, ‘Representation in authoritarian regimes’).

Policy responsiveness and democratic quality  57

REFERENCES Arnold, C. and M.N. Franklin (2012), ‘Introduction: Issue Congruence and Political Responsiveness’, West European Politics, 35 (6), 1217–25. Arrow, Kenneth (1963), Social Choice and Individual Values, New Haven, CT: Yale University Press. Bingham, Tom (2010), The Rule of Law, London: Allen Lane. Dahl, Robert A. (1971), Poliarchy: Participation and Opposition, New Haven, CT: Yale University Press. Huber, J.D. and B.H. Powell (1994), ‘Congruence Between Citizens and Policymakers in Two Visions of Liberal Democracy’, World Politics, 46 (3), 291–326. Hutter, S. and R. Vliegenthart (2018), ‘Who Responds to Protest? Protest Politics and Party Responsiveness in Western Europe’, Party Politics, 24 (4), 358–69. Laffan, B. (2014), ‘Testing Times: The Growing Primacy of Responsibility in the Euro Area’, West European Politics, 37 (2), 270–87. Lijphart, Arendt (1999), Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries, New Haven, CT: Yale University Press. Linde, J. and Y. Peters (2018), ‘Responsiveness, Support, and Responsibility: How Democratic Responsiveness Facilitates Responsible Government’, Party Politics, 20 (10), 1–14. Mair, P. (2009), ‘Representative vs. Responsible Government’, MPIfG Working Paper 09/8, Cologne, Max Planck Institute for the Study of Societies. Mansbridge, J. (2003), ‘Rethinking Representation’, American Political Science Review, 97 (4), 515–28. Manza, J. and F. Lomax Cook (2002), ‘A Democratic Polity? Three Views of Policy Responsiveness to Public Opinion in the United States’, American Politics Research, 30 (6), 630–67. Morlino, Leonardo (2012), Changes for Democracy: Actors, Structures and Processes, Oxford: Oxford University Press. Morlino, L. and M. Quaranta (2014), ‘The Non-Procedural Determinants of Responsiveness’, West European Politics, 37 (2), 331–60. Papadopoulos, Yannis (2020), ‘Political Accountability’, in Dirk Berg-Schlosser, Bertrand Badie and Leonardo Morlino (eds), Handbook of Political Science, Thousand Oaks, CA, USA and London, UK: Sage Publications. Pennock, R.J. (1952), ‘Responsiveness, Responsibility, and Majority Rule’, American Political Science Review, 46 (3), 790–807. Pitkin, Hanna F. (1967), The Concept of Representation, Berkeley and Los Angeles, CA: University of California Press. Powell, B.H. (2005), ‘The Chain of Responsiveness’, in Larry Diamond and Leonardo Morlino (eds), Assessing the Quality of Democracy: Theory and Empirical Analysis, Baltimore, MD: Johns Hopkins University Press. Riker, William H. (1982), Liberalism Against Populism: A Confrontation Between the Theory of Democracy and the Theory of Social Choice, San Francisco, CA: W.H. Freeman. Sabl, A. (2015), ‘The Two Cultures of Democratic Theory: Responsiveness, Democratic Quality, and the Empirical-Normative Divide’, Perspectives on Politics, 13 (2), 345–65. Sartori, Giovanni (1987), Theory of Democracy Revisited, New York, NY: Chatham House Publishers. Schedler, Andreas (1999), ‘Conceptualizing Accountability’, in A. Schedler, L. Diamond and M. Plattner (eds), The Self-Restraining State: Power and Accountability in New Democracies, Boulder, CO: Lynne Rienner. Traber, D., N. Giger and S. Häusermann (2018), ‘How Economic Crises Affect Political Representation: Declining Party‒Voter Congruence in Times of Constrained Government’, West European Politics, 41 (5), 1100–24.

5. Legitimacy and hegemony: two accounts of non-electoral representation Eline Severs

One of the core innovations of recent work on political representation is the conceptual separation between what representation is and what it should be about (Saward 2010; Rehfeld 2006). This innovation stands in sharp contrast to traditional accounts which firmly located political representation in assembled bodies and explained it as the outcome of democratic elections. This framework blurred the distinction between descriptive and normative conceptions of representation and, as such, heavily restricted theorizing on political representation. Because representation was conceived as the result of elections, everything a representative did after getting elected was considered representation. Similarly, the quality of representation was heavily associated with the exercise of democratic control, which was measured by the degree of congruence between constituents’ preferences and representatives’ actions. Since the mid-1990s, scholars (e.g., Plotke 1997; Ankersmit 2002) became increasingly critical of this ‘transmission belt’ conception of political representation. Primarily concerned with democratic control, research within this tradition focused on how representatives link to their electoral constituents and routinely treated the latter as exogenous to their representation, thereby downplaying the very contribution which representation makes to defining constituents and their interests. Scholars’ unequivocal focus on democratic institutions, furthermore, occluded attention from the representative work of unelected actors, such as advocacy groups, celebrities and civil society organizations. The insight that political representation is not defined by nor confined to democratic institutions promoted closer attention to what representation is (a selective and invariably contestable depiction of reality) and what it does (re-affirm, challenge or otherwise contribute to generalized ways of understanding the political). Reflecting this ‘constructivist turn’ in the literature, representation is now increasingly conceived as a communicative exchange between actors who formulate claims to represent others and audiences who accept, reject or otherwise engage with these claims (Saward 2010). This account dissociates the representative character of claims from their mimetic qualities, or how well they replicate a given reality. Instead, the representative character of claims depends on their performance before relevant audiences, and their ability to elicit the latter’s recognition. Whether a claim is accepted as ‘standing for’ or ‘being about’ those represented depends crucially on how well a claim taps into audiences’ familiar frameworks or habitual understandings of those represented and of the interests that are typically associated with them (Severs 2012; Dutoya and Hayat 2016). This new conceptual language radically expanded the scope of representation studies. Previously, the claims advanced by non-electoral actors (such as civil society actors, grassroots movements, and citizen-spokespersons) were primarily studied by social movement researchers and communication scientists and were ordinarily conceived as forms of agenda-setting, participation or contestation. However, and as a result of the broad applicability of the ‘claim-making’ framework, a growing range of studies now applies a representation 58

Legitimacy and hegemony  59 lens to the study of INGOs (Rubenstein 2014), demonstrators (Judge 2013), self-appointed representatives (Warren 2008; Montanaro 2012; Vieira 2015), digital platforms (Knops and Severs 2019), social movements (Weldon 2011; Strolovitch 2006), lobbyists and interest groups (Kröger 2018). This chapter takes stock of the extant literature and seeks to provide further direction to the study of non-electoral representation. To this end, I identify and discuss two accounts of non-electoral representation; each prioritizing different normative concerns (for a similar account of diverging normative concerns, see Disch 2019) and raising different challenges for empirical research. The first account conceives of non-electoral representation primarily as a correction to the shortcomings of electoral representation. It does not radically challenge the legitimacy of elections or of the institutions and model of government it produces. Conceived in terms of advocacy, non-electoral representation primarily serves as a correction to the ‘transmission belt’ of electoral representation: it provides citizens a political presence in between electoral moments and, by voicing new or previously overlooked interests, complements and potentially corrects electoral processes of interest aggregation. Because they conceive of non-electoral representation as a potential corrective to the limitations of electoral representation, accounts of this nature have primarily invested in specifying the conditions under which non-electoral representatives may be considered democratically legitimate themselves, and may, in this sense, be seen to contribute to processes of interest aggregation. The second account situates non-electoral representation more firmly within radical democratic theory and challenges the existence of objective foundations of democracy. It argues that there are no objective benchmarks for democratic legitimacy and, instead, turns attention to the political debate as to what is legitimate (Lefort 1988) and the power relations that underpin such debate in the context of contemporary representative democracies. Accounts of this nature treat non-electoral representation as a preventive force against hegemony, and task it with the job of multiplying and challenging governmental, or otherwise powerful, claims to represent the popular will. Its quality, then, depends crucially on its capacity for de-stabilizing political authorities and de-centring power. Both accounts lead to fundamentally different research agenda on non-electoral representation and provide different starting points to scholarly discussion on the merits of applying a representation lens to political practices situated outside parliament (Rubenstein 2014).

REPAIRING THE ‘TRANSMISSION BELT’ OF REPRESENTATION From their inception, theories of representative democracy have faced the challenge of explaining how the democratic principle of self-government could be made compatible with the elitist features of elections. Theorists of the revolutionary tradition, such as Jean-Jacques Rousseau, had introduced the concept of popular sovereignty as a means to delegitimize the absolute order of the Ancien Régime. With the concept of general will, Rousseau captured the democratic belief that all citizens are equal and that no individual can impose her views on the rest. This belief implied that the only legitimate source of authority formed the people as a whole and that no part of it could lay a claim to sovereignty. Because Rousseau considered the general will to be indivisible by nature (if it is not general, it is only a particular will), it could only be delegated (in the sense of ratification by others) but not represented, for rep-

60  Research handbook on political representation resentation would break up the unity of will and would subordinate the will of the people to that of a part (Garsten 2009, pp. 92–3). Against this backdrop, eighteenth-century theorists as diverse as James Madison and Edmund Burke defended the indirectness of representative democracy as an alternative to the desirable yet unattainable ideal of direct democracy. They emphasized the masses’ low level of political sophistication, and pleaded for a separate sphere of deliberation (generated by the election of capable elites) that would allow for a peaceful arbitration of interest conflicts and that could protect the general interest from both the tyrannical passions of the majority and the particular interests of factions (Enroth 2006, p. 461; Urbinati 2005, p. 196). Marrying self-government with the elitist features of elections, Madison and Burke defended representative government as a constitutional compromise that, while limiting the people’s sovereignty, nonetheless guarantees – via the principle of universal suffrage – citizens’ individual rights and political equality (Dahl 1956, p. 4). The principle of regular and free elections allows, in this regard, the majority of the people to self-exclude or opt out of systematic involvement in decision-making without making such exclusion permanent or irreparable: at all times, citizens retain their right to judge their representatives and to organize themselves in an attempt to exert influence on their representatives (Judge 1999, p. 9). Although juxtaposing itself to the model of direct democracy, constitutionalist theories of representative democracy implied, as Nadia Urbinati (2005, p. 196) succinctly stated, an ‘elitist transmutation rather than a rejection of’ Rousseau’s conception of the general will (see also Chapter 3 in this Handbook). While defending elections as a method for conferring (and retracting) consent to decision-making, constitutionalist theorists accepted the conceptual opposition between constituent power (the source of power that can authorize the creation of a political order) and constituted power (the sovereign political body that embodies that power) (Jennings 2011, p. 40). Their belief that the general interest could only be ‘preserved under the condition that its authoritative interpreters were severed from external influences’ (Urbinati 2005, p. 196) echoed, furthermore, Rousseau’s contention that while power may be divided, will cannot. Lack of unity would subordinate the whole to the part, thus constituting good cause for citizens’ disobedience to government. This transmutation of Rousseau’s principles of popular sovereignty had far-reaching implications for the development of representation theory. By accepting the conceptual opposition between constituent and constituted power, scholars located the origin of legitimate authority with the people and, in doing so, cast suspicion on all forms of constituted power. This conceptual opposition helps explain why representation scholars traditionally emphasized linkage. Congruence between citizens’ preferences and representatives’ actions presents an alignment of constituent and constituted powers and, in this sense, attenuates the division of labour brought about by elections: while it is still the representatives who act and make decisions, congruence allows for conceiving of the people as giving, through its representatives, laws to itself. Citizen contestation, in contrast, evidences a rupture between constituted and constituent power which, depending on its severity and longevity, threatens the legitimacy of constituted power. Many accounts of non-electoral representation follow in this tradition. They treat non-electoral claims as a rupture or de-alignment between constituent and constituted power and explain such rupture by the limitations of electoral representation. Scholars here generally refer to the low information which votes carry and how this produces flawed interpretations of the people’s will. Elections, furthermore, define constituents primarily in

Legitimacy and hegemony  61 terms of their geographical location and, thus, prioritize territoriality as a determining value of citizens’ political interests. Feminist scholars (Phillips 1995; Young 2002) were among the first to criticize the relations of accountability structured by elections. They argued that territorially-defined constituencies make the representation of interests that are associated to alternative identities (such as those founded in gender, class, age or race relations) secondary at best, and emphasized the importance of descriptive representation as a (partial) corrective. More recently, ecologist thinkers have begun to criticize the short time span of electoral cycles and the political myopia it produces. To enhance the sustainability of political decisions and promote inter-generational justice, scholars (e.g., González-Ricoy and Gosseries 2016) advocate a series of future-focused institutions, such as youth quotas, and special franchise rights for environmental groups that can help elect future-oriented representatives into parliament. While being critical of its limitations, accounts of this kind do not typically renounce elections, nor the institutions and model of government which they produce. Instead, they conceive of non-electoral representation as potential correctives to the ‘transmission belt’ model of electoral representation: it provides citizens a political presence in between electoral moments (Castiglione and Pollak 2019) and, by voicing new or previously overlooked interests, complements and potentially corrects electoral processes of interest aggregation. Because they primarily treat non-electoral representation as a correction to forms of under-representation and misrepresentation, accounts of this nature have primarily sought to clarify the conditions under which non-electoral representatives may be considered democratically legitimate themselves and may, in this sense, contribute to processes of interest aggregation.

THE GOOD REPRESENTATIVE 2.0: NEW ACCOUNTS OF LEGITIMACY Clearly, scholarly debates on what defines a democratically legitimate representative predate the ‘constructivist turn’ in the literature. It would be impossible for me to offer a full account of scholarly discussions on representation ethics (i.e. the standards we apply to the behaviour of individual representatives). I therefore limit myself to discussing the contemporary literature on non-electoral representation and making some commentaries on the diverging conceptions of legitimacy that arise within that literature. First, discussions on what makes a democratic representative are increasingly separated from attempts at defining concrete sets of interests that warrant representation. Such definitions are considered inherently problematic: they risk reproducing essentialist conceptions of those represented and overlook how political interests are co-constituted through representation. Instead, and as Severs and Dovi (2018, p. 311) state, the identification of normative standards is increasingly linked to clarifying the nature of the power relationship between those represented and their representatives. Central to scholars’ renewed attention to power (inequalities) is the insight that a representative claim is never simply an act of naming and defining the political. It is never simply an aesthetic act. Depending on the status of the claimant, representative claims have the capacity to produce ontological effects (making what is represented real) and, by doing so, affect the lives of those represented. Claims may, for instance, activate and empower those represented. But claims may also silence those represented (Saward 2010). Individuals, moreover, cannot simply opt out of, or escape, the consequences of being depicted in a particular way. Political

62  Research handbook on political representation representation lumps individuals into groups and, as with other acts of naming (such as stereotyping), individual acts of resistance do not exempt citizens from the treatment that results from having one’s group represented in a particular way (Severs 2019). Reflecting this renewed attention to power inequalities, scholars increasingly define representatives’ democratic legitimacy in relational terms, as arising from a constant and careful balancing between the autonomy of representatives (who are the ones that act) and the agentic qualities of those represented (who are likely to disagree, from time to time, with representatives’ interpretations of their interests). This notion of power balance or, in Hanna Pitkin’s (1967, pp. 145–6) terms, relative equivalence between those represented and their representatives is conceptualized in various ways. Some scholars (for instance Montanaro 2012) look for functional equivalents to electoral mechanisms of authorization and accountability (such as group membership and exit) for they expect these formal conditions to induce representatives to act with consideration for the preferences of those represented (should they hold any). A second set of scholars turns to representatives’ behaviour for evidence of a power balance. While Mansbridge (2003) emphasizes representatives’ willingness to listen to those represented and give an account of their actions, Dovi (2018) underscores representatives’ active commitment to citizen empowerment (in the sense of promoting citizens’ critical judgement of the claims made in their name). Other accounts (Rubenstein 2014; Severs and Dovi 2018) define democratic legitimacy by the absence of forms of power misuse, such as privileging self-interest, essentializing social groups, or continuing relationships of (representational) dependence when citizens can speak for themselves. A third set of scholars (for instance Maia 2012; Mulieri 2013) applies a more systemic approach to defining the legitimacy of non-electoral representatives. They no longer focus on dyadic relationships between those represented and their representatives but, instead, invoke the general institutional architecture of representative systems to guarantee power-sharing. In these accounts, the legitimacy of representatives is not defined by their individual characteristics or actions but by relations of horizontal accountability that allow representatives to hold one another in check. Other institutional conditions, such as the absence of dogma, warrants to individual political freedoms, the interlocking character of representation venues (allowing claims to move from margin to centre) and possibilities for appeal, are believed to empower citizens. What is interesting about this literature is that not only does it account for the constitutive qualities of political representation but also, and in a productive manner, it reconnects scholarly debates to traditional principal–agent accounts of representation. Contemporary, constructivist accounts of political representation routinely invoke metaphors, such as that of the artist, to describe the activities of political representatives. Yet the emphasis on representation’s aesthetic qualities (as constituting the political) should not distract us from the coercive aspects of representation. Or, from the insight, once central to principal–agent accounts, that those represented are bound by their representatives’ actions. Another helpful insight from traditional principal–agent accounts is the fact that representative relationships often serve a particular goal, such as the procurement of a right, or the solution of a problem. This signals that accounts that define good representatives solely by the absence of power misuse may fall short of actually capturing what is at stake in representative relationships. Especially in contexts where constituents have formed and self-identify as an interest group, good representation may depend as much on the promotion of particular interests as on the absence of power misuse.

Legitimacy and hegemony  63 Clearly, there are big differences in scholars’ accounts of democratic legitimacy. Yet, despite these differences, they face similar challenges when it comes to empirical operationalization. Scholars have distanced themselves from traditional approaches that relied on a delineated and predefined set of political interests for measuring the quality of representation. Yet, to measure the quality of representative relationships, empirical scholars still need to determine which individuals are potentially affected by representative claims, whether they are equally affected or whether some should hold greater influence over their representatives. Empirical researchers, at this stage, find themselves confronted by a normative dilemma. When they identify a particular set of people as most relevant for measuring the legitimacy of a representative, they risk privileging a particular account of the political. Alternatively, when they refrain from casting normative judgement, scholars risk reaffirming the status quo and those who benefit from it (Severs 2019; Näsström 2011). While systemic accounts have the advantage of not needing to identify who is (potentially) affected by a representative claim, they produce a different kind of challenge for empirical scholars; namely the specification and identification of the institutional arrangements that guarantee those represented a stake in representation processes. At present, with criteria as abstract as ‘the absence of dogma’ and ‘warrants to political rights’, there is a genuine risk of misunderstanding the power inequalities that underpin actual democracies. In The Semi-Sovereign People (1960) Schattschneider formulated a stringent critique on proceduralist approaches to democracy. What is needed for inclusion, he argued, is not simply competition between representative actors or fair opportunities for political mobilization. Instead, measures for inclusion should be centered on contextualized accounts or analyses of the brute, socio-economic circumstances and other systems of domination (related but not limited to gender, race/ethnicity, sexuality, and age) that negatively affect the opportunities of some people. The absence of marginalized groups in political debates does not originate in political disinterest or apathy but instead, as shown by a large body of empirical research, from the suppression of the policy issues and alternatives that reflect their needs (Severs 2019). This signals the need for more contextualized understandings of the particular power relations and dynamics at play in contemporary democracies. Contextualization also holds out a solution for empirical scholars who struggle with identifying who is affected by representative claims and what would count as evidence of a relationship of relative power equivalence. Clearly, any type of operationalization involves scholarly intervention, thus warranting justification. Yet such interventions do not, I argue, run counter to the constructivist underpinnings of contemporary accounts. Instead, they should be understood as contextualized applications of general normative principles, such as inclusion. While attention to context-specific mechanisms of inclusion and exclusion limits studies’ scope for generalization, it provides in-depth knowledge of the operation of power in a particular representative system, making indispensable contributions to representation theory.

CONTESTING HEGEMONY, DE-CENTRING POWER A second account of non-electoral representation is more closely linked to radical democratic theory and draws inspiration from the medieval origins of political representation. In medieval times, Ankersmit (2002, p. 24) argued, representation did not function as an instrument for democratic rule but as a form of embodiment: the presence of the three estates (the clergy, the

64  Research handbook on political representation nobility and the commoners) at the royal court imputed unity on the inhabitants of the realm, which in turn allowed the King to speak for and bind ‘the people’. The original function of representation, thus, consisted in generating a sense of unity or a master narrative (Disch 2019) that could be used to justify state coercion. This insight promoted a break with traditional accounts that problematize the indirectness of representative democracy and treat it as a second-best option to direct democracy. More specifically, the language of embodiment points to the co-originality of constituent and constituted power, and to the constitutive qualities of political representation (Geenens et al. 2015). ‘The people, as a totality, taken in the singular’ (Rosanvallon 2008, p. 206) only acquires its shape and value after and through representation. Hence, constituent power should not be understood as preceding representation in any way. Instead, constituent and constituted power should be conceived as ‘completely imbricating each other, and as ontologically incapable of separate existence’ (Jennings 2011, p. 41). This constructivist turn has important implications for democratic theory: because the people, as a constituent power, can never be completely outside some kind of constituted form (be it government, tradition or contextual repertoires of communal life), it cannot be treated as an objective benchmark for democratic representation. It is logically impossible, as Sunstein (1991, p. 8) argued, to measure the legitimacy of representatives’ actions by making reference to the very groups and interests they have helped to produce. In the absence of objective foundations, every claim to sovereignty remains contestable, and the locus of popular power can at best be situated within the process of representation itself. Or, as Lefort (1988, p. 39; italics in original) stated, democracy is ‘founded upon the legitimacy of a debate as to what is legitimate and what is illegitimate – a debate which is necessarily without any guarantor and without any end’. This account stands at odds with the transmission belt account outlined earlier. The latter chronologically separated an extra-institutional constituent power (the plurality of individual citizens) and constituted power (the symbolic unity produced by government) and advocated a circular movement between both powers (Urbinati 2005, p. 177). While allowing for temporary disaccords, government’s legitimacy hinges, in this account, on the alignment of both powers. For radical democratic theorists, however, the legitimacy of government originates from the representation process itself, and the extent to which it prevents artificial closures of the debate over what constitutes legitimate authority (Disch 2011). According to radical democratic theorists, the spatial-temporal gap opened by representation does not require closure: it holds democratic value in that it uncovers the contrived, and therefore, contestable character of any claim to represent the people (Garsten 2009, p. 94; Disch 2011, p. 11). Closure, in the sense of establishing congruence between citizens’ preferences and political decisions, is neither necessary nor sufficient for democratic representation. Congruence, in itself, does not inform us of the circumstances by which it came about, and whether it reflects power misuse or genuine democratic consensus. As Hayward (2009, p. 125) remarks, radical democratic theorists ‘no longer conceive of conflict as an empirical reality that should be overcome or pacified’. Instead, they attribute conflict a normative function and treat it as evidence of the vitality of democratic systems, or its capacity for democratic self-actualization. In keeping with this view, they attribute an alternative democratic function to non-electoral representation. Its primary objective no longer centres on advocacy or on repairing the transmission belt of electoral democracy. Instead, non-electoral representation is endowed with a ‘revelatory function’ (Disch 2019, p. 168). By

Legitimacy and hegemony  65 giving expression to alternative imaginaries, or understandings of the social, it exposes the contrived nature of dominant imaginaries, norms and beliefs, and, more importantly, has the capacity to unmask the historical power struggles through which imaginaries have acquired dominance (see also Chapter 6 in this Handbook). This is important. For contemporary norms, beliefs and institutions do not owe their status to their legitimacy (in the sense of being broadly supported by those affected) but to the suppression of ‘historically available and politically articulated alternatives’ (Marchart 2014, p. 275 quoted by Disch 2019, p. 168). Historically converging processes of selection, control and standardization not only elevated particular interpretations of socio-political reality to a standard, but equally contributed to stabilizing the individual observer, in the sense of fixing the issues and problems she focuses on (Ezrahi 2012, pp. 287–8). Non-electoral claims, to the extent in which they help articulate political alternatives, have the potential to unfix our interpretations of the political, potentially de-stabilizing the institutions and authorities that are privileged by dominant imaginaries. Such de-stabilization is needed to shield democratic debates from artificial closures.

TO MAKE A COUNTER-SOVEREIGN In keeping with theories of radical democracy, non-electoral claims are often conceived as an expression of sovereignty in the negative, or the manifestation of a counter-sovereign (Rosanvallon 2008). The democratic potential of non-electoral representation should not, however, be reduced to individuals’ capacity to object to the claims made on their behalf. What is at stake is not some macho-like contest over who manages to hold out and object. Instead, what is at stake is a representative system’s capacity for self-actualization and transformation. To this end, non-electoral representation should not only challenge the structures and institutions (such as norms and rules) that mask, perpetuate and legitimize dominant imaginaries but should equally articulate credible alternatives. Non-electoral representatives, however, face many challenges. As Claude Lefort (1988) argued, the indeterminacy of democracy yields a permanent quest for provisional closures. Only such closures provide a sense of unity sufficiently stable for subordinating citizens to the outcomes of collective decision-making. Because challenges to political unity risk destabilizing the political order, they are likely to meet resistance. The more so as the status of non-electoral representatives differs radically from that of elected representatives who can invoke democratic elections as the basis of their legitimacy (Grant 2014, p. 584). As a method for aggregating individual preferences, elections guarantee that individual preferences have, in the abstract, an equal chance to affect the outcomes of decision-making (Williams 1998). This account, while overlooking the somewhat arbitrary character of voting and the low information votes carry, has become one of the founding myths of representative democracy. Non-electoral representatives lack comparable credentials to speak for the people and, as a result, their claims risk being cast to the side as mere partisan claims. This weakens their ability to challenge prevailing or monistic conceptions of the people effectively (Severs et al. 2015; Urbinati 2014). To date, scholars have attributed little attention to the ways in which non-electoral representatives seek to bolster and justify their challenger claims. Scholarly debate has mostly centred on the difference between the so-called ‘authorization-based’ claims of electoral rep-

66  Research handbook on political representation resentatives and the ‘authenticity-based’ claims of non-electoral representatives, emphasizing the latter’s adoption of a stylistic register that is closer to citizens’ own ways of understanding, relating to and discussing the political (see for instance Saward 2009). Authenticity-based justifications are not, however, exclusive to non-electoral actors and the simple fact of being elected does not shield politicians from critique and the need to give account of their actions. What sets electoral representatives apart from other representatives is not authorization per se but rather their ability to rely on elections as a democratic mechanism for constituting ‘the people’ (Knops and Severs 2019, p. 5613). Empirical researchers may take inspiration from Rosanvallon’s (2011, pp. 117–18) discussion of the alternative sources of legitimacy that may bolster the credibility of non-electoral representatives when they claim to speak for ‘the people’. A first source of legitimacy, impartiality, originates in representatives’ detachment from particularities. According to Knops and Severs (2019, p. 5613), claims to impartiality could be expressed either by demonstrating the claim-makers’ lack of self-interest in a matter or by showcasing their factual independence from the authorities of surveillance and regulation that are associated with the state. Reflexivity, then again, roots the legitimacy of non-electoral representatives in their ability to showcase the limitations of electoral democracy ‘which assumes the electoral majority to be the will of the whole social body’ (Rosanvallon 2011, p. 117). The credibility of such claims depends on demonstrating the biases inherent to contemporary democracies, unmasking the underlying power structures that (re-)produce them, and claiming recognition for previously under-represented or misrepresented groups. A third source of legitimacy implies a descent in generality, whereby non-electoral representatives appeal to citizens’ lived experiences and demonstrate a mismatch between decision-making and citizens’ everyday needs and knowledge of the political. The credibility of such claims depends on representatives’ proximity to the people they claim to represent, and the recognition they receive from them. These sources of legitimacy are not exclusive to non-electoral representation. Neither are they sufficient to account for the plurality of representative practices. They, however, underscore the crucial insight that the potential of non-electoral claims for disrupting dominant imaginaries of the political hinges on their capacity to convey a meaningful or persuasive alternative. Studying how non-electoral actors justify their political imaginaries and how these justifications are received by relevant audiences (such as citizens and other representatives more closely associated with the state) may help us differentiate between dominant imaginaries and those that have acquired hegemonical status, in the sense of suppressing the articulation and circulation of rivalry accounts of the political. Here, scholars may also want to pay greater attention to representatives’ performances, and the emotional or evocative language they resort to. Emotions, as Edelman (1972) showed, underpin our ability to understand the world. They are also crucial to the solidification of hegemony. Political imaginaries owe their status of hegemonical to their capacity of creating the illusion of objectivity, thereby concealing their own creativity. They cover up, not seldom resorting to rhetoric, theatricality or other grand emotions, their part in selecting and delineating referents to citizens’ shared understandings of ‘what is real’ (Ezrahi 2012, p. 4). Imagination, as opposed to fact-giving, is what causes us to experience the results of historical processes of meaning-making as natural, and to suppress facts, to accept and even rationalize evidence that contradicts or otherwise challenges dominant understandings of our socio-political environments (cognitive dissonance).

Legitimacy and hegemony  67

CONCLUSION AND DISCUSSION The aims of this chapter have been two-fold: to take stock of the burgeoning literature on non-electoral or extra-parliamentary forms of representation, and to provide direction to future studies. To this end, I identified and discussed two accounts of non-electoral representation. The first treats non-electoral representation as a potential correction to the ‘transmission belt’ model of electoral representation and does not radically challenge the legitimacy of electoral institutions. The second account rejects a foundationalist approach and, instead, defines the legitimacy of representative government by the open-ended character of public debates over what constitutes legitimacy. It primarily attributes non-electoral representation the task of de-stabilizing governmental claims to represent the people, protecting democratic debate from hegemonic powers and artificial closures. In my discussion I have deliberately avoided references to moderate or strong constructivism (cf. Fossen 2019). This is in part because this dichotomy tends to produce hierarchical relations within the field of representation studies which I consider hardly conducive to innovation and cross-fertilization. In part, it is also because the dichotomy occludes attention from the insight that there exists a world outside of representation. The indeterminacy of democratic constituents, however correct, should not lead to a fetishization of the inchoate (Deveaux 1999, p. 14). Just as processes of state formation have rendered the impact of state membership real, historical processes of representation (consisting of a reiteration of power relations) structurally privilege some groups and their experiences of the political while disadvantaging others (Severs 2019). The labels strong versus moderate constructivism are potentially misleading. The central dividing line between strands in the representation literature centre less on the extent to which scholars believe that political reality is constructed. Our political worlds are always both constructed and real. Instead, the central differences between research traditions are the normative concerns they prioritize: respectively, the inclusiveness of decision-making and the political imaginaries that affect how citizens understand and engage with their socio-political environments. While both lead to fundamentally different research agendas, there is no need for pitting one against the other. Instead, innovation is likely to come from attempts at making both accounts speak more closely to each other.

ACKNOWLEDGEMENTS This chapter benefited greatly from discussion at the ‘Representation and Power’ panel at the APSA Annual Meeting in Washington DC (August 2019). I am grateful to all participants but especially to Mark Warren and Suzanne Dovi for their generous and insightful comments. My thanks also go to the editors of this Handbook, Maurizio Cotta and Federico Russo, for their careful reading of this chapter and their suggestions on improving its central argument. The title of the section ‘The good representative 2.0: new accounts of legitimacy’ references the title of Severs and Dovi’s (2018) article ‘The good representative 2.0: why we need to return to the ethics of political representation’, published in PS: Political Science and Politics.

68  Research handbook on political representation

REFERENCES Ankersmit, Frank R. (2002), Political Representation, Stanford, CA: Stanford University Press. Castiglione, Dario and Johannes Pollak (eds) (2019), Creating Political Presence: The New Politics of Democratic Representation, Chicago, IL, USA and London, UK: Chicago University Press. Dahl, Robert A. (1956), A Preface to Democratic Theory, Chicago, IL: University of Chicago Press. Deveaux, Monique (1999), ‘Agonism and Pluralism’, Philosophy and Social Criticism, 25 (4), 1–22. Disch, Lisa (2011), ‘Toward a Mobilization Conception of Democratic Representation’, American Political Science Review, 105 (1), 100–114. Disch, Lisa (2019), ‘Radical Democracy: The Silent Partner in Political Representation’s Constructivist Turn’, in Dario Castiglione and Johannes Pollak (eds), Creating Political Presence: The New Politics of Democratic Representation, Chicago, IL, USA and London, UK: Chicago University Press, pp. 162–85. Dovi, Suzanne (2018), ‘Good Representatives Foster Autonomy’, PS: Political Science and Politics, 51 (2), 323–6. Dutoya, Virginie and Samuel Hayat (2016), ‘Making Representative Claims: The Social Construction of Political Representation’, Presses de Sciences Po, Revue Française de Science Politique, 1 (66), 7–25. Edelman, Murray (1972), Politics as Symbolic Action: Mass Arousal and Quiescence, Chicago, IL: Markham Publishing Company. Enroth, Henrik (2006), ‘Beyond Unity in Plurality: Rethinking the Pluralist Legacy’, Contemporary Political Theory, 9 (4), 458–76. Ezrahi, Yaron (2012), Imagined Democracies: Necessary Political Fictions, New York, NY: Cambridge University Press. Fossen, Thomas (2019), ‘Constructivism and the Logic of Political Representation’, American Political Science Review, 113 (3), 824–37. Garsten, Bryan (2009), ‘Representative Government and Popular Sovereignty’, in Ian Shapiro, Susan C. Stokes, Elisabeth Jean Wood and Alexander S. Kirschner (eds), Political Representation, Cambridge: Cambridge University Press, pp. 90–110. Geenens, Raf, Thomas Decreus, Femmy Thewissen, Antoon Braeckman and Marta Resmini (2015), ‘The “Co-originality” of Constituent Power and Representation’, Constellations, 22 (4), 514–22. González-Ricoy, Iñigo and Axel Gosseries (eds) (2016), Institutions for Future Generations, Oxford: Oxford University Press. Grant, John (2014), ‘“Becoming One”: Visions of Political Unity from the Ancients to the Postmoderns’, Constellations, 21 (4), 575–88. Hayward, Clarissa Rile (2009), ‘Making Interest: On Representation and Democratic Legitimacy’, in Ian Shapiro, Susan C. Stokes, Elisabeth Jean Wood and Alexander S. Kirschner (eds), Political Representation, Cambridge: Cambridge University Press, pp. 111–38. Jennings, Ronald C. (2011), ‘Sovereignty and Modernity: A Genealogy of Agamben’s Critique of Sovereignty’, Anthropological Theory, 11 (1), 23–61. Judge, David (1999), Representation: Theory and Practice in Britain, London, UK and New York, NY, USA: Routledge. Judge, D. (2013), ‘“Word from the Street”: When Non-electoral Representative Claims Meet Electoral Representation in the United Kingdom’, British Politics, 8 (4), 388–409. Knops, Louise and Eline Severs (2019), ‘Self-appointed Representatives on Facebook: The Case of the Belgian Citizen’s Platform for Refugee Support’, International Journal of Communication, 13 5610–28. Kröger, Sandra (2018), ‘How Limited Representativeness Weakens Throughput Legitimacy in the EU: The Example of Interest Groups’, Public Administration, 97 (4), 770–83. Lefort, Claude (1988), Democracy and Political Theory, Cambridge: Polity Press. Maia, Rousiley C.M. (2012), ‘Non-electoral Representation: Expanding Discursive Domains’, Representation, 48 (4), 429–43. Mansbridge, Jane (2003), ‘Rethinking Representation’, American Political Science Review, 97 (4), 515–28. Marchart, Olivier (2014), ‘Institution and Dislocation: Philosophical Roots of Laclau’s Discourse Theory of Space and Antagonism’, Distinktion: Journal of Social Theory, 15 (3), 271–82.

Legitimacy and hegemony  69 Montanaro, Laura (2012), ‘The Democratic Legitimacy of Self-appointed Representatives’, Journal of Politics, 74 (4), 1094–107. Mulieri, Alessandro (2013), ‘Beyond Electoral Democracy: A Critical Assessment of Constructivist Representation in the Global Arena’, Representation, 49 (4), 515–27. Näsström, Sofia (2011), ‘The Challenge of the All-affected People’, Political Studies, 59 (1), 116–34. Phillips, Anne (1995), The Politics of Presence: The Political Representation of Gender, Ethnicity, and Race, Oxford: Oxford University Press. Pitkin, Hanna Fenichel (1967), The Concept of Representation, Berkeley, CA: University of California Press. Plotke, David (1997), ‘Representation is Democracy’, Constellations, 4 (1), 19–34. Rehfeld, Andrew (2006), ‘Towards a General Theory of Political Representation’, Journal of Politics, 68 (1), 1–21. Rosanvallon, Pierre (2008), Counter-Democracy: Politics in an Age of Distrust, Cambridge: Cambridge University Press. Rosanvallon, Pierre (2011), ‘The Metamorphoses of Democratic Legitimacy: Impartiality, Reflexivity, Proximity’, Constellations, 18 (2), 114–23. Rubenstein, Jennifer C. (2014), ‘The Misuse of Power, Not Bad Representation: Why it is beside the Point that No One Elected Oxfam’, The Journal of Political Philosophy, 22 (2), 204–30. Saward, Michael (2009), ‘Authorisation and Authenticity: Representation and the Unelected’, Journal of Political Philosophy, 17 (1), 1–22. Saward, Michael (2010), The Representative Claim, New York, NY: Oxford University Press. Schattschneider, Elmer E. (1960), The Semi-Sovereign People: A Realist’s View of Democracy in America, New York, NY: Holt, Rinehart and Winston. Severs, Eline (2012), ‘Substantive Representation Through a Claims-making Lens: A Strategy for the Identification and Analysis of Substantive Claims’, Representation, 48 (2), 169–81. Severs, Eline (2019), ‘Theorising Representation Fairness: Why we need to Account for Social Groups’, in Mihnea Tanasescu and Claire Dupont (eds), The Edges of Political Representation: Mapping, Critiquing and Pushing the Boundaries, Colchester and London: ECPR Press and Rowman & Littlefield International, pp. 81–102. Severs, Eline and Suzanne Dovi (2018), ‘The Good Representative 2.0: Why We Need to Return to the Ethics of Political Representation’, PS: Political Science and Politics, 51 (2), 309–13. Severs, Eline, Karen Celis and Petra Meier (2015), ‘The Indirectness of Political Representation: A Blessing or a Concern? A Study of the Conceptions of Members of the Flemish Regional Parliament’, Parliamentary Affairs, 68 (3), 616–37. Strolovitch, Daria Z. (2006), ‘Do Interest Groups Represent the Disadvantaged? Advocacy at the Intersections of Race, Class and Gender’, The Journal of Politics, 68 (4), 894–910. Sunstein, Cass R. (1991), ‘Preferences and Politics’, Philosophy and Public Affairs, 20 (1), 3–34. Urbinati, Nadia (2005), ‘Continuity and Rupture: The Power of Judgment in Democratic Representation’, Constellations, 12 (2), 194–222. Urbinati, Nadia (2014), Democracy Disfigured: Opinion, Truth, and the People, Cambridge, MA: Harvard University Press. Vieira, Monica Brito (2015), ‘Founders and Re-founders: Struggles of Self-authorised Representation’, Constellations, 22 (4), 500–513. Warren, Mark E. (2008), ‘Citizen Representatives’, in Mark E. Warren and Hilary Pearse (eds), Designing Deliberative Democracy: The British Columbia Citizens’ Assembly, Cambridge: Cambridge University Press, pp. 50–69. Weldon, S. Lauren (2011), When Protest Makes Policy: How Social Movements Represent Disadvantaged Groups, Ann Arbor, MI: The University of Michigan Press. Williams, Melissa S. (1998), Voice, Trust, and Memory: Marginalized Groups and the Failings of Liberal Representations, New York, NY: Princeton University Press. Young, Iris Marion (2002), Inclusion and Democracy, Oxford: Oxford University Press.

6. Challenges to political representation: participatory democracy, direct democracy and populism Simon Tormey

Representation, according to authoritative commentators such as Hannah Pitkin, is absolutely intrinsic to how we think about politics, and how we enact politics in democratic settings. As she puts it: ‘In modern times, almost everyone wants to be governed by representatives […] every political group or cause wants representation; every government claims to represent’ (Pitkin 1967, p. 2). Notwithstanding what for many decades stood as a truism about the nature of democratic politics, confidence in the practice as well as the process and institutions associated with political representation appears to be at a low ebb. Citizens have become less inclined to vote in many advanced democracies. We are reluctant to join political parties, which are in turn the main transmission belt between those represented, citizens, and those who represent, politicians. Trust in our representatives is in steep decline, and interest in mainstream and electoral politics as measured in terms of media output has been in freefall since the 1960s (Tormey 2015). In view of this chronic condition, often referred to as the crisis of democracy, it is unsurprising that political theorists as well as activists have suggested that we need to rethink our attachment to political representation as the paradigmatic form in which our engagement with politics takes place, and also by which our societies are governed (Manin 1997; Urbinati 2006; Keane 2009). This chapter discusses several kinds of challenge to the representative paradigm including participatory democracy, direct democracy and populism. Whilst populism does not offer an alternative model of democracy, it is common to paint it as a threat to democracy, and can thus be considered as a challenge to democracy on the terms being discussed here.

THE CRITIQUE OF REPRESENTATION Amongst the many criticisms that we hear of representative democracy is the accusation that representation by definition presupposes a division of the community into representatives, the politicians and political elites, and those who are represented, the citizens. Representatives exercise power through political institutions whilst everyone else is subject to that power, and therefore rendered passive in relation to it (Manin 1997). Jean-Jacques Rousseau articulated this position clearly when he noted that ‘the People of England’ is ‘free only during the election of members of Parliament. As soon as they are elected, slavery overtakes it, and it is nothing’ (Rousseau 1762 [1985]). In his view, and the views of many of those who followed him who articulated a critique of early modern democratic practices, democracy is supposed

70

Challenges to political representation  71 to be exercised by the whole community, not just some part of it. It is supposed to be directly participatory as opposed to representative. From this point onward, democratic theory was divided between those who sought to defend representative politics and those who looked for a grander, more involving model that would overcome the passivity that the representative model supposedly enacted. The political thought of the nineteenth and twentieth centuries is full of various iterations of democracy that allegedly deepen and broaden engagement to overcome this distinction consecrated by representation and by the practice of electing politicians. Even where representation was accepted as an inevitable and necessary feature of the practice of politics, there was – and still is – deep interest in the mechanics of representation. A key distinction that has its roots in arguments heard in the English Civil War concerns the nature of the difference between representation and delegation. One way of overcoming the alienation of political power to representatives was to force them to articulate in close terms the views and interests of those represented. This delegate model implied a more active role for those being represented, including in the most radical iterations of the kind found, for example, in the work of Marx and Lenin the right of immediate recall of representatives and also a commitment to ensuring that in terms of pay and conditions representatives were no better off than the average working person (Lenin 1918 [1977]; Marx 1871 [1996]). Suggestions were also forthcoming for adapting political institutions in ways that guaranteed or encouraged wider participation by those not directly elected. Many theorists have tinkered with the idea of creating other kinds of institution or assembly to thicken the participation of social groups, trade unions, indigenous peoples, and those who might otherwise find themselves underrepresented or not represented at all (Brito Viera and Runciman 2008). The argument here is that if we encourage the participation of groups, collective interests and different identities then this will provide for a fuller representation of the different kinds of subject to be found in modern settings. This concern to give voice to positions and identities that might not otherwise come to the fore in our political life is particularly prominent in the call for citizens’ assemblies. Countries such as Iceland and Ireland have sought recently to augment or supplement political life through the creation of broad-based assemblies composed of citizens to debate contentious topics. This has had considerable success in Ireland where citizens’ assemblies were created to debate the matter of abortion and also gay rights. The positive and progressive adoption of new measures to promote women’s rights and the rights of sexual minorities has been ascribed to the public activity of citizens’ assemblies in giving space to the fears, hopes and expectations around these topics, taking them out of the partisan arena of parliament, and permitting the free flow of comment and debate. So it is not difficult to point to examples of how citizens have been encouraged to participate in new and imaginative ways in the political process. On the other hand, it should also not be ignored that many commentators and political scientists query the neat separation of representative from other forms of democracy found in the work of critics such as Rousseau. The image of a citizenry mute in the face of their own democratic institutions ignores, so it is argued, the key innovation promoted by representative democracy in the nineteenth century, which is the political party (Duverger 1959; Sartori 2005). Whilst it is certainly true that political representation preceded the emergence of political parties, it is often argued that the basis for stable and successful democratic institutions rests on the creation of parties to represent

72  Research handbook on political representation the various needs, interests and ideologies found in contemporary societies (see Chapter 9 in this Handbook). This is particularly so in the case of social democratic and socialist parties which were created towards the end of the nineteenth century to give voice to those who have previously either been denied the vote or denied the means of articulating a voice through existing political structures. The birth of these parties gave rise to mass political organizations, which could call on the participation and loyalty of many millions of citizens. This in turn gave rise to mass organizations of the centre and right which similarly encouraged citizens to become active in their own right. In this fashion the political party became the basis for political participation in modern democracies. If you had an animus to vent or a dream to pursue, you did it in and through the political party. Parties became an absolutely crucial transmission belt between citizens and democratic political institutions. They offered community, resources, solidarity and a sense of participating in the governance of the system. It’s on this basis that party scholars argue that the distinction between representative and participatory democracy is an illusory one under modern conditions. Political parties are a mechanism for participating. So, it is argued, are social movements and the array of other kinds of organization found in contemporary democracies. It is, from this point of view, inaccurate to say that our democracies are not participatory when so many citizens participate in all manner of organizations, from single issue movements to mass political parties. Part of the success of contemporary representative institutions is that they admit and indeed encourage a free flow of participation, albeit beyond or outside the governing institutions themselves (Dalton 2002; Lawson 2010). So what then of the critique offered by those arguing for participatory, as opposed to representative democracy? Given the undeniably participatory nature of representative democracy, does it make sense to maintain the distinction? The point that critics return to is the efficacy of these non-parliamentary, non-mainstream forms of participation. Why, if civil society is so vibrant and participation so evident, is it that they seem so powerless to alter the direction of policy in ways that challenge entrenched interests and, in particular, the power of wealth and privilege? This is the kind of critique we see articulated by the likes of C. Wright Mills and Noam Chomsky, and then in a more theoretical mode in the work of Benjamin Barber (1984) and Carole Pateman (1970). The common denominator here is a scepticism concerning the claims of democratic pluralism to ensure a fair and just political process based on party competition on the one hand, and a democratic civil society on the other (Mills 1956; Chomsky 1997). They, and many others, claim that such is the centrality of money to politics that the wealthy always win. The elites can buy representatives through showering them with the resources that they need to mount their campaigns and electoral offices. Such is the power of business in terms of lobbying Congress and parliaments that the voices of less well endowed groups are sidelined. A properly participatory democracy, or what Barber terms strong democracy, is therefore at odds with party-based electoral democracy which favours the wealthy. It has to be grounded in forms of representation and institutions which are by design more inclusive of minority voices, as well as the mainstream. Hence the insistence that we actively pursue ways of making our democracies more responsive to the needs, interests and voices found in contemporary society (Wolin 2017).

Challenges to political representation  73

BEYOND THE CRITIQUE OF PLURALISM: DIRECT DEMOCRACY Haunting the discussion of the shortcomings of representative democracy is the ideal of democracy as rule by the people. It is inspired by the example of Athens and with the practice whereby all citizens were expected to participate in the political life of the polis, to perform key tasks regarding the governance of the community, as well as to come to its defence when threatened by external forces. This model has long been held up by critics of representative democracy as a truer or more authentic expression of what democracy means. Athens and by extension the idea of democracy as something that all citizens should be engaged with – and indeed themselves enacting – became a kind of gold standard for the work of those who were unimpressed by the apparent elitism of representative politics. Rousseau was deeply influenced by classical Athens and seemed to have had in mind a directly democratic model to underpin the formation of a general will to displace the will of all, which he associated with representative democracy generally and electoral politics in particular. However, even he was prone to lapsing into representative modes of thinking as per the example of the Lawgiver who emerges as a key figure at the back end of The Social Contract. Marx and then Lenin seem also to have been inspired by the directly democratic nature of the Paris commune of 1871. The commune enacted a kind of democracy from below in the sense that many citizens were either directly engaged in decision-making, or able to maintain a tight grip on delegates through practices such as the rotation and recall of officials, ensuring that delegates were paid no more than ordinary workers, and more generally fostering a sense that the functions of the political apparatus were fully integrated into civil society. This was to ensure that there was no separation or alienation of the state from society, which was the great fear of Democrats as they witnessed the emergence of bureaucracy and administration over the course of the nineteenth and twentieth century. As it turned out, this effort to create an ecology of reciprocity and mutuality between society and the state came unstuck with the hegemony of the Communist Party in its self-anointed role as vanguard of the revolution. As the Russian revolution and indeed most other socialist revolutions demonstrated, the party quickly took on the task of governing in the name of the people and for the people. Direct democracy quickly became a proxy for the party representing the masses. Notwithstanding the failure of various Marxist experiments to enact a more radical form of democracy, other initiatives of a directly democratic kind were established over the nineteenth and twentieth centuries. The most notable of these is the practice of citizen assemblies and referendums in Switzerland. Again stemming from a collective suspicion about the centralizing tendency of national bureaucracies, the various cantons that compose the Swiss Federation built on local practices of delegation and direct recall to establish directly democratic practices that subsist today as an example of how societies can maintain a firm grip on the decision-making apparatus at both state and federal level. Many decisions that in other states are ceded to parliamentary representatives are still subject to referendums, which are held on a very regular basis across the country. This is not to say that such practices do not have their own critics, both within the country and without. Critics have argued that it is an anachronism to demand a referendum on every item of public interest. Others argue that direct democracy produces a slow-moving and rather cumbersome system of decision-making that is not well adapted to the fast-flowing needs of a modern economy. On the other hand, the evidence seems to suggest that direct democracy has greatly aided the sense of citizen engagement in public life. It has helped keep down the efflorescence of bodies, quangos and committees that

74  Research handbook on political representation otherwise provide a buffer between citizens and their representatives. More generally it has led to high levels of satisfaction relative to other European countries as far as citizens’ well-being is concerned. Elsewhere, the idea that citizens should be encouraged to participate directly in government underpins a number of recent initiatives from the development of assemblies and consensus decision-making, as found in movements such as #15M and Occupy, to Internet-based initiatives such as the Pirates and Liquid Democracy. With regard to the former, the citizen insurrections in Spain in 2011 manifested both a widespread disaffection with the practices of representative democracy and a desire to create and sustain forms of participation that in the terms of the activists characterized real democracy. What this slogan meant is of course open to considerable interpretation. It is worth noting here that the lasting impact of #15M was not the creation of new democratic institutions or processes, but rather new political parties such as Podemos and citizen platforms such as Barcelona en Comu, both of which remain resolutely within the frame of representative politics (Tormey and Feenstra 2015). This re-emphasizes a point made above: trying to make a hard distinction between forms of democracy that are representative and those that are participatory is more demanding than many critics appear to imagine. The demand for greater democracy is, as commentators such as Laclau have argued, one that requires articulation. It needs someone to take on the task of representing this position to others. One cannot in this sense escape the logic of representation by demanding participation. The desire for greater participation is one that itself requires articulation often, paradoxically, through the representative process itself (Laclau 1996). With regard to Internet-based initiatives, we now find the demand for direct democracy being taken up by radical movements and parties, most notably the Pirates, who have enjoyed considerable success electorally in countries such as Germany and Iceland. Originally created with the idea of pushing back against what was felt to be the overly intrusive influence of the state regarding regulating the Internet, the Pirates quickly developed a broader political message to promote a libertarian and directly democratic system of governance. They argue that where once we needed representatives to undertake the business of decision-making, the Internet now makes it possible for all of us to be engaged in the process of governance. Nearly all citizens have access to the Internet. With the political will we could design platforms that enable all of us to take part in decision-making. So much for the theory, what about the practice? It is interesting to note in this regard that a number of cities around the world are now experimenting with techniques to encourage participation and reduce the sense of alienation felt by citizens from their own representative systems. Barcelona and Madrid, to take two high-profile examples, are both experimenting with online platforms to encourage citizen engagement. Both start from a similar method. This is to encourage citizens to come forward with proposals which are then ranked by other citizens to create a list of priorities for the City Hall. A budget is then allocated for each priority, and citizens can then trace online how these proposals and projects unfold. It is early days for this form of citizen engagement. The results are rather mixed, notwithstanding the rapturous reception such initiatives have received, not least by the UN, which awarded Madrid its citizens prize for its online platform. So far the amount of take-up has been limited, with the percentage of citizens engaged in these online platforms representing a small fraction of the overall population of the city. This reminds us that direct democracy takes time, and in the context of the contemporary metropolis many citizens are hard pressed to juggle domestic commitments, several jobs, long commutes and other pressing demands. Sitting at a computer

Challenges to political representation  75 or tapping away at an online platform debating municipal priorities is arguably a luxury that only a few can afford. Another frequently discussed model in terms of bridging the divide between representatives and citizens is deliberative democracy (Dryzek 2002). It also takes aim at representation as a form of political alienation, but it does so with a different goal in mind, which is to generate policies and outcomes on the basis of considered collective reflection. Representation, and particularly representation as expressed through the contemporary party system is, so it is held, partisan, antagonistic and designed to produce losers as well as winners. Deliberative democrats reject the view that democracy has to be majoritarian. They believe that through carefully constructing processes of deliberation designed to produce consensus, we can eliminate or at least sharply reduce the winner takes all approach that turns so many citizens off politics and produces the crisis discussed above. Several of today’s protest movements similarly pin their faith on the healing power of deliberative processes. Extinction Rebellion, for example, has made the creation of a citizens’ assembly to address issues around sustainability one of its three key demands. Others have also called for citizens’ assemblies to help the UK find a solution to the apparently intractable problem of Brexit. So how does deliberative democracy work? The key it seems is the initial orientation of participants in the discussion. If we see decision-making in zero-sum terms, that is as a process that will inevitably produce winners and losers, then our orientation will be to ensure that we come out on the winning side. It will emphasize the centrality of rhetoric, gamesmanship and the black arts of persuasion as means of ensuring we get what we want. If on the other hand we see decision-making as a process whose objective is to create a consensus, then we will approach the discussion in a very different way. It is less about getting what I want, and more about establishing what everyone can live with. This means opening oneself up to the possibility of change, negotiation and compromise for the greatest good, which is achieving a goal that everyone can live with. Deliberative democrats often stress the key role of information sharing, of reducing our libidinal investment in certain fixed positions, and in sharing our hopes as well as fears in terms of particular outcomes. In short, deliberative democrats encourage us to be participants, rather than contestants in a shared process. On the other hand, this is not to say that this process is easy, or indeed desirable from the point of view of establishing a democratic practice. Chantal Mouffe for example has criticized deliberative democracy for depoliticizing politics (Mouffe 1993). Her view is that politics has an inevitably conflictual aspect to it. There will always be differences between individuals and groups. We should not downplay or ignore fundamental differences because the quest for consensus requires it. This will lead, she argues, to the disavowal of fundamental differences which then often resurface in violent and disruptive ways. Better in her view to work on translating antagonistic positions into agonistic ones, or ones that do not threaten the integrity of the decision-making process, but which also recognize the fundamental differences that she argues are constitutive of social life. In sum then, it’s important in our discussion not to overlook the fact that direct democratic practices are an important supplement to regular representative politics, and also in some places important and well embedded practices in themselves. Most democratic systems embrace referendums as a means of empowering citizens’ voices directly. Many others use devices such as citizens’ assemblies in order to inform decision-making on particularly contentious matters, or more generally to foster a sense that participation goes further than periodic elections. Switzerland has, however, gone the furthest in this direction. Here referendums

76  Research handbook on political representation are a very regular event in many cantons, with the result that representatives often complain that too much power has been placed in the hands of communities, and too little in the hands of the central state. As mentioned, cities such as Barcelona and Madrid enact direct democratic processes through online and digital platforms. Whilst the success of such initiatives is uncertain, the clamour is growing for further strengthening and deepening citizen engagement with local and central politics as manifested in many activist movements across the developed and developing world.

POPULISM AS A CHALLENGE TO REPRESENTATIVE DEMOCRACY In contradistinction to participatory and deliberative approaches, we also need to consider the challenge posed by populism, in turn one of the most significant features of the contemporary political landscape (Moffitt 2016; Müller 2016; Mudde and Kaltwasser 2017). Populism is the political expression of discontent with political elites, which is often in turn translated as discontent with the system of political representation. There is some justification for the assumption that populism can be read on these terms, but there are also some problems with this approach as well. It is certainly true that, for many citizens, rejecting current political elites in favour of outsider movements or leaders equates to a condemnation of how politics is conducted as well as who is conducting it, our representatives. We often hear the expression that politics is broken or that democracy is, as already mentioned, in crisis. All this suggests that it is indeed the system that requires fundamental rethinking. On the other hand, what is also apparent is that for the most part populism remains within the frame of representative politics, but it does so with one particular idiosyncrasy. This is the claim that the party or the leader represents the people versus the elites or the establishment. So Trump is elected with a promise to challenge Washington with the implication that he will not be prey for the lobbying groups, special interests and government departments who, it is alleged, really run the system. Much of the drama of the Brexit vote was provided by the leave campaign which promised Britons that if they voted to leave the EU it would be tantamount to taking back control from those villainous elites in Brussels, and indeed in the Commons. Similar appeals can be heard across Europe, as for example in the emergence of the Five Star Movement and the Lega in Italy, in the discourse of Marine Le Pen in France, and indeed in the developing world as well, where figures such as Rudy Duterte and Thaksin Shinawatra claim to understand and know the people much better than the elites. Such is the dynamic of contemporary politics. The populist claim seems therefore to make a real – or imagined – break from the discourse of democratic politics, and from its pluralist underpinnings. It challenges the conception of representation as necessarily partial, particular and related to the interests and identities of a particular part of the population, as opposed to the whole. It also breaks from the inheritance of party-based representative politics of a kind familiar in the advanced democracies. It is perhaps for this reason that populism is often characterized as a challenge to political representation (Müller 2016). Populism is, however, better characterized as a challenge to how political representation operates in competitive, party-based systems. Whereas, so it is argued, representation from the nineteenth century onwards has been articulated on the basis of partial interests, needs and

Challenges to political representation  77 desires, so now we have actors who claim to represent the people as an undifferentiated or homogeneous mass. For commentators such as Jan-Werner Müller (2016), this is what makes populism a threat to democracy. Democracy is a system built on a recognition of pluralism. Pluralism has expressed itself historically in and through political parties who compete for power on the basis of representing particular needs and interests, such as those of the working class, or of Catholics, or of business or whatever. In claiming to represent the people as a whole, populists at the same time disavow these particular needs and identities, and thus the inevitable or constitutive pluralism of contemporary society. They reduce us, the citizens, to a mass which can then be represented by a single individual or a single party, in turn obviating the need for party competition or indeed elections. Again the ongoing saga of Brexit illustrates this point well. Feeling hampered by the actions of representatives, Boris Johnson prorogues or suspends Parliament on the basis that his task is to enact the will of the people. So this is a slippery slope, and a dangerous one as far as critics of populism are concerned. Making a fetish of the people, assuming that one can know the will of the people, contains a particular kind of logic which is antithetical to the pluralist underpinnings of representative democracy. Left unchecked, such a logic quickly translates into authoritarian practices such as the suspension of Parliament or criticism of the judiciary. It then provides a justification for the overturning of civil society in favour of autocracy. For others, however, such claims are overblown, mistaking the rhetorical basis of democratic discourse with a largely imagined threat. As Laclau explains, it is intrinsic to the discourse of democracy that we invoke the people (Laclau 2005). The people is both the subject of democracy, and the object of a properly democratic discourse. Appealing to the people is not in this sense an authoritarian gesture, even incipiently. Rather, we appeal to the people because ultimately the people are sovereign and have the power to make or break political actors. It’s for this reason that we often hear politicians claiming to understand what ordinary people want or need. We often hear politicians talking about how the people feel about a particular policy, a particular lack, a particular issue. All politicians in this sense claim to represent the people. The question is not therefore whether it is legitimate to speak for the people or not; clearly it is. The issue is whether people see or hear themselves represented in these statements. If they do, if they feel that their needs or interests are being spoken for, then they will feel that they are being represented. If they don’t, then they will keep searching for someone, or some party or some movement that gets closer to their own ideas about how to address contemporary issues. So, assuming Laclau is right about the nature of democratic discourse, the question is then what the difference is between populism and other kinds of politics. If for any kind of articulation to be successful under democratic conditions it requires, at a certain level, an appeal to the people, what is it that constitutes the particular character of populism? As we alluded to at the start of this section, the particular quality that differentiates populism from other kinds of discourse is the clear distinction between the people on the one hand and the elites on the other. Populism introduces a particular kind of antagonism into democratic politics that might not otherwise be evident. It suggests that the only way that the people can be represented is by an outsider, by someone who is not him or herself of the elite. This in turn suggests that political representation has in a sense broken down. People have lost faith in their representatives, and perhaps by extension in the system of representation itself. The latter position is of course a more radical form of populism than the former. The former suggests that by voting for an outsider, this will repair the broken link between representatives and the represented, the people. This is the kind of populism we associate with the election of figures

78  Research handbook on political representation such as Trump. The latter kind of populism contains a more radical suggestion. This is that the system is broken and needs to be replaced by something else. What that something else is depends on the form of political analysis put forward by the relevant political actors. Those on the right tend to default to market solutions over those proposed by government agencies, bureaucrats or experts. Those on the left usually propose some thickening or deepening of democracy so that the needs and interests of citizens come to the fore, over those of the wealthy, the bankers, La Casta. Right-wing solutions tend in this sense to be anti-statist and often aligned with what we would otherwise term neoliberalism. Left-wing versions tend to be statist, but also desiring of introducing democratic deliberation into areas such as the economy or the welfare state to thicken the sense of ownership over areas that might otherwise be felt to be separate from civil society. So does all this represent a threat to representative democracy? Does populism undermine representation? On the basis of what we have heard so far, this would have to be doubted. Populism is better characterized as a form of hyper-representation in the sense that populism feeds off both the desire to be represented, with a critique of existing forms of representation. To put the same matter differently, populists do not seek to supplant or displace representation, so much as to render it fuller or more authentic. This is to overcome the antagonisms and contradictions marking social life in the name of wholeness. It’s this quality that paradoxically makes populism an intrinsic feature of democracy, but also potentially a threat to it. The demand for greater representation is after all one that is historically intrinsic to the expansion of democracy. Without the demand for representation, democracy falls into stasis. However, populism also provides a critique of representation, one that may represent a threat to the forms of representation that characterize contemporary democracy. A perhaps better way of characterizing populism is to note that it is particularly evident during periods of profound crisis. Indeed, it is a manifestation of crisis. Its appearance tells us that citizens have become disaffected, not only with a particular class of politicians, but perhaps also with politics as usual. That crisis might have its origin in different domains. It might be the consequence of economic crisis, of recession and austerity. It might have its origins in a perceived threat to cultural identity or integrity. It might have its origins in the feeling that politicians have become corrupt, inefficient or detached from the people they are supposed to represent. It’s for this reason that populism is not a singular phenomenon. It is a symptom of a crisis at the heart of contemporary politics, and of the system of representation that has characterized how democracy functions for two centuries.

CONCLUSION So what have we learned about the challenge to political representation? Two points stand out. The first is that, notwithstanding the representative basis of democratic systems since the dawn of democracy, in the early modern period critics from a wide variety of political viewpoints – including anarchism, socialism, communism and varieties of republican thought – evince a discontent with it. The sense that representation cannot be the basis for the full flourishing of the democratic order persists in the thought and practice for many political movements and

Challenges to political representation  79 initiatives today, from Occupy, through the Pirate parties to the various experiments in deliberative and participatory democracy discussed above. What is equally clear, however, is that the notion that it is a straightforward matter to displace representation with some other logic of political organization is fraught with difficulty. This is not to say that we cannot improve on systems of representation. Far from it. The desire to render our democracies more democratic, more engaging and inclusive is one shared by many across the political spectrum. It is to say that the dichotomy frequently drawn between representation and participation is not always helpful. As we have had occasion to note, participation is intrinsic to the operation of representative systems. Representation without participation would indeed make for a hollow kind of democracy. But one of the reasons why representative democracy subsists is that it permits and indeed encourages participation. This can be in the form of involvement in and engagement with political parties. It might be in the form of social movements, protest or other forms of mobilization. It might be in the form of engaging in online discussions or debates. All of these are forms of participation. More than this, without such forms of participation our democracies would be dull and lifeless. If all democracy means is, as Schumpeter put it, the rotation of elites, then we would have every reason to feel that participation had been sidelined. But this is not what it means. It is not what democracy has ever meant. Democracy is, as Lincoln reminds us, a threefold articulation between those who represent, those who are represented, and the people as a whole which authorizes democracy – democracy of, by and for the People. Remove one of those articulations and we change the basis of democracy itself. We don’t improve democracy. We don’t enhance it; we diminish it. The challenge to political representation is to take on board valid criticisms of the kind that insist on broadening and deepening representation, on ensuring that representatives represent. But seeking to supplant political representation altogether cannot be the goal of those motivated to improve or ameliorate democracy. The people cannot itself exercise power. We cannot collapse the state into civil society. Or rather we can, but the result would not be democracy, but anarchy – and anarchy is not democracy. Democracy requires a separation of those who exercise power and those to whom they are subject, the represented. From this point of view it is no exaggeration to note that democracy requires representation. Representation is intrinsic to democracy. If we want to improve or build on our present institutions and structures, we will do so from within the logic of political representation, and not from outside or beyond it.

REFERENCES Barber, Benjamin R. (1984), Strong Democracy: Participatory Politics for a New Age, Berkeley, CA: University of California Press. Brito Viera, Monica and David Runciman (2008), Representation, Cambridge: Polity. Chomsky, Noam (1997), Democracy in a Neoliberal Order: Doctrines and Reality, Cape Town: University of Cape Town. Dalton, Russel J. (2002), Citizen Politics: Public Opinion and Political Parties in Advanced Industrial Democracies, London: Chatham House Publishers. Dryzek, John S. (2002), Deliberative Democracy and Beyond: Liberals, Critics, Contestations, Oxford: Oxford University Press. Duverger, Maurice (1959), Political Parties, London: Methuen. Keane, John (2009), The Life and Death of Democracy, London, UK and New York, NY, USA: Simon & Schuster.

80  Research handbook on political representation Laclau, Ernesto (1996), Emancipation(s), London: Verso. Laclau, Ernesto (2005), On Populist Reason, London: Verso. Lawson, Kay (2010), Political Parties and Democracy, Santa Barbara, CA: Praeger. Lenin, Vladimir I. (1918), ‘The State and the Revolution’, reprinted in Vladimir I. Lenin [1977], Selected Works in Three Volumes, Moscow: Progress, pp. 238–327. Manin, Bernard (1997), The Principles of Representative Government, New York, NY: Cambridge University Press. Marx, Karl (1871), The Civil War in France, reprinted in T. Carver (ed.) [1996], Marx: Later Political Writings, Cambridge: Cambridge University Press. Mills, C. Wright (1956), The Power Elite, New York, NY: Oxford University Press. Moffitt, Benjamin (2016), The Global Rise of Populism: Performance, Political Style, and Representation, Stanford, CT: Stanford University Press. Mouffe, Chantal (1993), The Return of the Political, London: Verso. Mudde, Cas and Cristóbal Rovira Kaltwasser (2017), Populism: a Very Short Introduction, Oxford: Oxford University Press. Müller, Jan-Werner (2016), What is Populism?, Philadelphia, PA: University of Pennsylvania Press. Pateman, Carole (1970), Participation and Democratic Theory, Cambridge: Cambridge University Press. Pitkin, Hanna Fenichel (1967), The Concept of Representation, Berkeley, CA: University of California Press. Rousseau, Jean-Jacques (1762), The Social Contract, reprinted in J.-J. Rousseau [1985], The Social Contract, London: Penguin. Sartori, Giovanni (2005), Parties and Party Systems: A Framework for Analysis, London: ECPR Press. Tormey, Simon (2015), The End of Representative Politics, Cambridge: Polity. Tormey, S. and R.A. Feenstra (2015), ‘Reinventing the Political Party in Spain: The Case of 15M and the Spanish Mobilisations’, Policy Studies, 36 (6), 590–606. Urbinati, Nadia (2006), Representative Democracy: Principles and Genealogy, Chicago, IL: University of Chicago Press. Wolin, Sheldon S. (2017), Democracy Incorporated: Managed Democracy and the Specter of Inverted Totalitarianism, Princeton, NJ: Princeton University Press.

SECTION II HISTORICAL AND CONTEMPORARY MODELS OF POLITICAL REPRESENTATION

INTRODUCTION Maurizio Cotta While political representation is a defining feature of contemporary liberal democracy, its phenomenology neither begins nor ends with this specific political experience. Forms of political representation existed well before what we see as the standard model came into existence; in its turn this standard model is far from being a fixed entity and has, with time, undergone important transformations. If pluralistic and competitive elections are the basic institutional channel of representation in democratic regimes, a wide range of non-democratic regimes make use of the same institutions in a non-competitive (or semi-competitive) setting. Finally, both in democratic and non-democratic settings, corporatism and technocracy provide alternatives to electoral representation. A wider survey of different historical and contemporary models and experiences is therefore necessary in order to place democratic representation in an adequate perspective and to fully understand the potential space of application of the concept of political representation, as well as some of the dilemmas that are involved in its practical implementation. In the first chapter of this section, ‘From estate representation to the representation of the people and the nation in the Age of Revolutions’, Ihalainen and Haaparinne highlight how, well before liberal democratic systems could affirm themselves, the political history of Europe from the Middle Ages to modern times had a rich and varied experience of representative institutions. Most of them disappeared under the hammer of monarchical absolutism, but others managed to survive until the dawn of the new era. In connection with these experiences, and particularly in the course of the eighteenth century, a spirited discussion developed on the meaning of representation, the figure and role of the representative and the nature of what was to be represented. Even if the foundations and guiding principles of pre-modern representation differed significantly from contemporary ones, these experiences of the past have left signif81

82  Research handbook on political representation icant institutional traces and contributed in a far from marginal way to preparing the ground for the modern liberal and democratic versions of representation. Their relevance is thus not purely historical. In fact tensions between the concepts of representation, democracy and popular sovereignty in the Age of Revolutions contributed to the contestability of representation in later times. When we turn, then, to the developmental processes which over the last two centuries have produced the contemporary form of democracy, we find that the principles and institutions of representation have undergone significant transformations and innovations. From the period of limited franchise to that of mass democracy, the narratives and practices of representation have experienced important changes involving the qualities of the representatives and the perimeter and the characteristics of the represented as well as the substance of the linkage between the two sides of this relationship. In the chapter ‘Individualistic representation in the liberal century – and beyond’, Best and Vogel explore the meaning of representation in the age of the notables and how the progressive displacement of amateur politicians by party professionals, and today by new forms of personalization of politics and by the web age populists, has deeply changed conceptions and practices of representation. The role of the organized party and of party competition in the process of representation, and how this has reshaped contemporary democracy and its performances is detailed analytically by Budge in the chapter titled ‘Party-based representation: the paradox of democracy’. If democratic representation is based on the electoral right of individual citizens and essentially leaves to the parties the role of aggregating individual preferences into collective choices, the role of organized socio-economic interests in advanced industrial countries cannot be overlooked. Particularly in the golden age of industrialization, the representation of the big conflicting societal groups – workers and entrepreneurs – and of their basic interests has been a key problem in all political systems. The possibility of envisaging special mechanisms of representation fostering social cohesion and consensus has stimulated the birth of corporatist theories of representation and the implementation of institutional experiments based on the direct role of the organizations expressed by these groups. As illustrated by Siaroff in the chapter ‘Corporatism and representation’, under special conditions, socio-economic organized interests gain a central role in the process of representation and decision, and contribute to establishing a corporatist system which largely sidesteps, with its method of trilateral bargaining and consensus decision-making, the conflict-oriented model of democracy. In highly developed societies, with their increasingly complex problems, the relationship between technocracy and democratic representation has gained prominent interest. Democratic governments make great use of technical experts, and both at national and supranational level important decisions are allocated to institutions of governance, inspired by the principle of competence, and which enjoy a large degree of autonomy. Apparently, this does not pose a direct challenge to the democratic model of representation and its political legitimacy. The role of technocrats and technocratic institutions is generally depicted as auxiliary or subordinate, and is restricted within defined borders. Their language is that of competence and expertise, which is quite distant from that of popular representation. However, as discussed by Caramani in the chapter on ‘Technocratic representation’, as the defence and promotion of certain public interests attributed to technocratic institutions entail a reference to the interests of the country/the people, the issue of representation is far from being irrelevant even in these forms of governance. Skirmishes between technocrats and elected representatives are an increasingly frequent event in political life. Particularly in crisis situations and when divi-

Section II  83 sions among democratic politicians make the reaching of decisions more difficult, the appeal of technocrats and technocratic institutions, with their image of neutrality, competence and attention to long-term interests, becomes significant. Technocratic representation may, then, challenge democratic representation. While the main contemporary discourse on political representation and the invention of the institutional mechanisms for implementing it have developed essentially within the context of democratic regimes, we cannot disregard that non-democratic regimes of different types have not neglected the opportunity of exploiting the appeal of this idea for their legitimation and preservation and have developed instruments which pay tribute to this idea without endangering their survival. The great totalitarian experiments of the twentieth century challenged liberal democracy, proposing themselves as more perfect answers to the will of the people. To do so they proposed models of representation which were radically alternative to that based on free and competitive elections. These models profoundly redefined the representative actor (the single party, the vanguard of the people, the totalitarian leader), the represented (the class or the nation as an organic unit) and the relationship between the two sides of the equation. In more recent years, as discussed by Cotta in the chapter ‘Representation in authoritarian regimes’, the expanding phenomenon of hybrid regimes, often denoted as electoral or competitive authoritarianism, deserves special attention for its pragmatic use of non-competitive or semi-competitive elections and legislatures. Without the ambition of proposing a strong alternative model, these regimes exploit the representative institutions of democracy to increase their bases of consent and promote their internal cohesion while thwarting in different ways the potential of contestation. When representation becomes a predominantly top-down process and consensus-building gains first place, the question arises whether representation can still retain some influence on political reality or whether it is reduced only to an instrument of ideological manipulation.

7. From estate representation to the representation of the people and the nation in the Age of Revolutions Pasi Ihalainen and Zachris Haaparinne

In Western countries since the Age of Revolutions, the legitimacy of political rule has typically been constructed with the concepts of popular sovereignty and parliamentary representation. Inherent tensions between the ideals of popular sovereignty, democracy, political representation and parliamentary government have never been fully resolved, however. It is important to understand the trajectories of tension that revolutionary applications of the ideal of popular sovereignty left in modern systems of representation. This chapter focuses on parliaments as the conventional institutional solution for representative government. Instead of functionalist approaches defining, evaluating or measuring ‘representation’ analytically or theoretically (Morgan 1988; Urbinati 2006; Congleton 2010) or focusing on the ‘mechanisms’ of representation in state building (Jansson 2007; Albareda and Sánchez 2019), we have adopted an empirical, source-based and language-sensitive approach, analysing competing meanings assigned to popular sovereignty and representation in their historical contexts. We aim to grasp what past political actors themselves meant when they talked about representation. We use digitized parliamentary sources to locate competing conceptualizations of representation. Our examples originate from the British parliament, Swedish-Finnish four-estate diet and French revolutionary assemblies. As a special case we analyse British conceptualizations of representation in petitions. We also review related literature on the German lands, the North American Colonies and the Netherlands. Our analysis focuses on the transition from the conventional early modern representation of the interests of estates by delegates (managed by monarchical rulers) to the ideal of the representation of the people or the nation by independent representatives. We explore how tensions between the representation of privileged groups and the more individualistic representation of the people manifested themselves conceptually in the Age of Revolutions. We start from the domestic political crises following the Seven Years’ War in the 1760s and conclude with the Restoration after the fall of Napoleon around 1815, when new representative institutions were created in several countries. This period also leads us to address conflicting meanings assigned to democracy and representation – a divide that still manifests itself in political battles today.

FROM FUNCTIONALIST TO LINGUISTICALLY SENSITIVE HISTORIES OF REPRESENTATION Democratic representation did not exist before the nineteenth century; in early modern Europe representation was based on estates with differing privileges. In much research, comparison 84

From estate representation to the representation of the people  85 between representative systems continues to stand for separate national narratives presented side by side. Jansson, for instance, examines representative systems on two sides of the Atlantic from the sixteenth to the eighteenth centuries, concluding that the ‘tension between monarchy and early representative institutions […] provides its own dynamic for understanding early modern state building’ (Jansson 2007, p. 2). Albareda and Sánchez have been interested in the mechanisms of representation at diverse levels of estate societies. They have explored diversities of representation and interactions between social groups that either contributed to or resisted monarchical sovereignty in the building of the modern state characterized by constant wars and growing bureaucracies. Continental systems of representation were typically multi-layered and diversified structures used to defend local autonomy and the privileges of each corporation (Albareda and Sánchez 2019, pp. 1–2). Such functionalist approaches to the history of representation focus on macro-level societal processes and their social historical aspects, not on the discursive formation of the concept of representation. Transformations in the German concept of representation have been explored most comprehensively by Barbara Stollberg-Rilinger (1999). In her view, the fiction that the estates were identical with and representative of the entity of subjects was based on late medieval theories of corporations. Such representation was intended to produce binding decisions serving the interests of monarchs, not to legitimize the estates as a political elite. In the course of the eighteenth century, a new concept of citizen introduced by natural law theorists and supported by influences from American, British and French debates challenged this understanding by viewing the estates as a party in a social contract made by the people. The dilemma of appropriate ‘national representation’ consequently emerged. The Revolutionary Wars temporarily increased the powers of some German estate assemblies, and differences between privileged and non-privileged groups decreased, but it was far from simple to modernize estates that had concentrated on preserving corporate privileges. Distinct breaks with the early modern concept followed, introduced by reformist bureaucrats who saw the estates as a means to articulate the economic interests of various classes. Some wished to retain old estate privileges, denying the role of the estates as representatives of the people, while others wanted to open the estates up to new forms of participation in the name of supposedly ancient traditions of popular representation. After the Restoration, French revolutionary innovations were rejected and the estates were generally defined as representatives of particular interests, though no longer in the traditional sense. There was thus no direct continuity between old estates and emerging representative bodies despite terminological similarities and historical-political narratives created to bridge them (Stollberg-Rilinger 1999, pp. 298–304; Gehrke 2005, pp. 6–7). This point constitutes a starting hypothesis for our review of other national cases below.

REPRESENTATION IN LATE EIGHTEENTH-CENTURY BRITAIN Continental experiences of representation differed from Anglo-American ones, which have often been interpreted through teleological narratives on the rise of popular sovereignty and the representation of the people. It has been typical to draw a line from the Commonwealthmen of the 1640s to the Continental Congress and modern representative democracy (Morgan 1988). According to George Yerby, the English revolutionaries ‘could set a value on the services of parliament which heralded a change in the political position of the representative assembly’: the people were ready to use parliament through their representatives (Yerby 2008, p. 1).

86  Research handbook on political representation British early modern traditions of representation became a model for continental reformists challenging deficiencies in their native political systems. Seaward and Ihalainen (2016, pp. 32–3) have highlighted considerable continuity in concepts that defined the British parliament but also an accelerated change in their meanings from the late eighteenth century onwards. Representation became one of the concepts through which an increasingly modern understanding of parliamentary government was defined. While the legitimacy of the English parliament had built on the notion of representation since the thirteenth century, when representatives were required to arrive in parliament with full power from their community (Morgan 1988, p. 39), the Civil War and the republican experiment clarified the implications. The House of Commons, and not the monarch only, became understood as ‘representative of the people’. Yet, particularly once party disputes arose from the 1680s, the electors continued to doubt whether ‘the voice of the people’ was properly represented in parliament. Wealthy men paid for parliamentary seats, the ministry arranged seats for office holders, very limited groups could vote in most constituencies, MPs did not necessarily feel responsible to their voters, and the lack of parliamentary publicity made it impossible for the public to control the representatives. It was justifiable to ask how well the constituencies actually were represented (Seaward and Ihalainen 2016, pp. 33–4, 39–40, 42–3; Dickinson 2007, pp. 20–22). Representation often stood for representing information or the opinion of a community to the monarch or parliament with expectations that a decision favourable to the community would follow. Alternatively, representation could stand for the process of electing representatives that reflected the views of the community (Seaward and Ihalainen 2016, pp. 39–40). Parliament was defined as ‘the grand Inquest of the Nation, they are to represent the grievances of the People to their Sovereign; and the People are always to choose proper representatives for that Purpose’ (House of Commons/House of Lords (HC/HL), 1734/35, pp. 421, 674). The opposition considered that Robert Walpole’s government of 1721–42 was disrupting this relationship, as the Commons was becoming ‘a Representative of an Administration, or of one single Minister, but could no longer be a true Representative of the People’ (HL, Protest by Several Lords, 29 March 1732, p. 1059). Calls for parliamentary reform and more frequent elections were loud during the American Crisis in the 1760s. The ministry reacted by presenting ‘the people’ as a collective political agent independent of the monarch and represented by parliament, while pamphleteers placed ‘national representation’ and ‘sovereignty’ in parliament. The ministerial side insisted that it was the elected MPs rather than the population at large or even the voters that constituted ‘the People’ (Ihalainen 2010, pp. 476–7, 486). A period of intense argumentation for parliamentary reform followed, leading politicians such as William Pitt the Younger and Charles James Fox to call for the restoration of the democratic element of a constitution that suffered from corrupt representation, emphasizing the people’s trust in their representatives in the Commons as opposed to the monarchy and ministry (Ihalainen 2010, p. 486). Such rhetoric reinforced the principle of representation of the people as an essential element in the parliamentary system, yet it did not lead to reform until 1832. A typical response to reformists was that ‘interests’ rather than communities needed to be represented, that ‘a virtual representation’ assured that the interests of the entire realm were heard or that increasing parliamentary publicity compensated for the lack of representation (Seaward and Ihalainen 2016, pp. 43–4). According to Charles Jenkinson, publicity created

From estate representation to the representation of the people  87 a link between the members and the people and increased accountability in a way that made a French-style universal suffrage unnecessary (HC, 26 May 1797, p. 689). The British form of parliamentary representation, like most representative systems, was based on the free mandate. Although chosen by particular communities, electors could not dictate how their MPs should vote. It differed from the so-called imperative mandate of representation in which representatives ought to follow the advice of their constituents (Dickinson 2007, p. 30). In other words, the will of the people only existed, in the procedural sense, during elections. Once elected, the MPs became independent of electors’ interests. Opposing calls to dissolve parliament in 1770, the author of An Address to the People of England recognized electors’ right to criticize representatives’ decisions, but maintained that, ‘at the same time, they, to whom the people have entrusted their rights, are to act for themselves, and to be governed by their own consciences’. The author claimed that the MPs should recognize electors’ desires and opinions, but also act on their own – ‘otherwise they were mere machines and lifeless instruments, meant only to echo the voices of others’ (An Address to the People of England 1770, p. 10). Although often referred to as representatives of the people, elected agents in the House of Commons remained, primarily, members of parliament – chosen by the people, but, formally, independent of the people. Contemporaries used two specific arguments to legitimize the independence of the Commons and the MPs. First, parliament consisted of the ablest individuals (Dickinson 2007, pp. 23–7). Charles Jenkinson, a Lord of Treasury, castigating mass petitions in 1770, questioned the abilities of people ‘out-of-doors’: common people were easily misinformed and misled by opposition agents and, therefore, could not ‘be judges of the motives that lead to those decisions [of the House of Commons]’. Jenkinson claimed that ‘were the authority of this House to depend upon the popular opinion, scarce any laws would be regarded: all laws, in their immediate effects, are restraint and inconvenience, and the multitude never consider remote advantages’. Hence, he concluded, ‘to found […] the authority of this House upon the popular voice, is vain and idle’ (HC, 9 January 1770, pp. 690–91). William De Gray, the Attorney General, even encouraged representatives to ‘preserve the independence of our own body, as involving the liberty of our people, and defend it against the people themselves misguided and inflamed by faction and self-interest’ (HC, 25 January 1770, pp. 796–7). As the ablest individuals, members of parliament could resist the temptations of passion and self-interest. Hence they, rather than the easily misled members of the public, ought to judge the advantages and disadvantages of legislation. Secondly, the nature of parliament ensured that, in principle, decision-making was based on the common good. Once elected to parliament, MPs became representatives of the nation instead of particular constituencies (Dickinson 2007, pp. 28–30). Edmund Burke, for instance, declared to Bristol electors in 1774 that ‘Parliament is not a Congress of Ambassadors from different and hostile interests; which interests each must maintain, as an Agent and Advocate, against other Agents and Advocates’. Instead of serving ‘local Purposes’ and ‘local Prejudices’, ‘Parliament is a deliberative Assembly of one Nation, with one Interest, that of the whole’ (Burke 1775, pp. 28–9). Consequently, representatives could not be obligated to follow constituents’ instructions. Instead of courting popular opinion, members of parliament ought to guard the common interest by protecting the constitution. Charles James Fox, then an ardent opponent of the opposition out-of-doors, claimed in 1771 that representatives ‘have higher obligations to justice, than to our constituents; we are […] bound to promote [people’s] true interests in preference to the dearest desires of their hearts’. Fox claimed that the consti-

88  Research handbook on political representation tution made representatives ‘the sole arbiters of those interests, notwithstanding the imaginary infallibility of the people’ (HC, 25 March 1771, pp. 145–6). Both arguments referred to the superior character of the MPs, though the second one also made use of the unique nature of parliament as an institution.

CHALLENGES TO FREE MANDATE IN BRITAIN Even if the free mandate remained the dominant principle of British parliamentary representation throughout the eighteenth century, it was not absolute. Several agents used implicit means to challenge it and to oblige representatives to act as desired. Petitioners employed three discursive strategies to challenge representatives’ independence (Haaparinne 2021). First, petitioners used representative claims to persuade representatives to observe their recommendations. They often emphasized their status as officials, as members of corporations, assizes and “quarter sessions”1 and as representatives of local communities who, if discontented, could complicate the MP’s re-election. After the Middlesex election dispute surrounding the numerous elections and expulsions of the radical John Wilkes to and from parliament, constituents reintroduced mass petitions emphasizing the scale of subscriptions that had been used in the late seventeenth century (Knights 1993; 2009, pp. 45–6). Mass petitions became an important feature of British political culture in the late eighteenth and early nineteenth centuries: campaigners against the slave trade used mass petitions in 1788 (Knights 2009, p. 45), and later on so did the Chartists. Representative claims were used to convince MPs that instead of speaking for themselves, the petitioners represented the sentiments of the nation. Secondly, petitioners sought to persuade representatives using ideals and counter-ideals. They could claim that their objectives were essential to the common interest, even on explicitly local issues. Petitioners could also characterize themselves as ‘determined to act the Part of Free-born Englishmen’ (London Evening Post, 4 December 1756), anticipating that representatives shared their definition of Englishness and hence their objectives. Petitioners’ opponents, by contrast, were depicted as servants of the faction (General Evening Post, 4–6 January 1770), prioritizing the interest of a corrupt administration over that of the nation. Ideals and counter-ideals functioned as implicit signals of petitioners’ desires and expectations. Ideal representatives, of course, shared petitioners’ senses of right and wrong. Thirdly, petitioners’ choices of verb mattered. Most petitioners used humble verbs, desiring and beseeching representatives to act. Others used more authoritative verbs. The mayor, magistrates and inhabitants of Wareham claimed that Walpole’s tax reform would demolish consent-based politics and deprive electors of the right to choose their representatives. Hence the petitioners ‘intreat[ed] and require[d]’ their MPs to oppose the excise scheme because ‘we that have chosen you are zealously against it’ (London Magazine, June 1733). Both of the borough’s representatives voted in favour of excises. Nathaniel Gould lost his seat in 1734 and Thomas Towers moved to represent another constituency. The constituents of Wareham elected two opposition-minded representatives to replace them. An analysis of petitions reveals that the free mandate was less authoritative than is often assumed. Although direct challenges to it remained infrequent, political agents implicitly challenged the independence of MPs. The discourses of the petitions did not constitute any consistent doctrine: though endeavouring to influence their opinions, most petitioners recognized representatives’ independence. Even uses of authoritative language were ad hoc rather

From estate representation to the representation of the people  89 than explicit defences of the imperative mandate. Yet petitions demonstrate how the public outside the political elite participated in defining the nature of representation.

PECULIARITIES OF REPRESENTATION IN THE NORTH AMERICAN COLONIES Britain also encountered a more explicit challenge to the legitimacy of its representative government. The colonies in North America, most of which had long traditions of political autonomy and local representative institutions (Greene 2007, pp. 172–4, 176–83), began to question the boundaries of the sovereignty of parliament in London. Tensions between the Colonies and the mother country started to mount after the passing of the Sugar Act of 1764 and the Stamp Act of 1765. Whereas the British parliament maintained its sovereign right to impose taxes on the Colonies, colonists and the British opposition insisted that such taxes could only be enacted by the consent of the colonial assemblies. Defenders of the Colonies generally used references to the British constitutional canon to resist the imposition of British taxes on American colonists. According to the canon, subjects could not be taxed without their consent, and as the Colonies did not elect representatives to the Commons, parliament could not enact taxes on them. British opposition MPs sometimes used similar arguments to delegitimize taxes on the American Colonies. Lord Camden, for instance, argued that ‘taxation and representation are inseparably united; God hath joined them, no British parliament can separate them’. Whoever confiscates private property without consent, he continued, ‘commits a robbery’ and ‘destroys the distinction between liberty and slavery’ (HL, 10 February 1766, p. 178). The proponents of colonial taxation by and subordination to London disputed such assertions by claiming that the inhabitants of the American Colonies were in fact represented in the British parliament. Even if they lacked the right to send representatives to the Commons, they were represented through an instrument referred to as ‘virtual representation’. George Grenville, former prime minister, claimed in 1766 that parliament had always used its power to tax people, entities and regions without actual representation in the Commons. Such a right was exercised over the East India Company, merchants of London and ‘many great manufacturing towns’ and several regions ‘before they sent any representatives to parliament’ (HC, 14 January 1766, pp. 101–103). Lord Lyttelton recognized that ‘[n]o subject is bound by any law to which he is not actually or virtually consenting’ but noted that ‘[i]f the Colonies are subjects of Great Britain, they are represented and consent to all statutes’ (HL, 10 February 1766, p. 167). Such reasoning was based on a collective conception of representation: once elected to parliament, representatives did not merely represent their constituencies, but the nation as a whole. Hence, according to the proponents of taxes, the MPs also represented places such as Manchester, Birmingham and the Thirteen Colonies even if these did not elect any members; the representatives considered only the common good of the nation and thus the interest of the virtually represented subjects and entities. After gaining independence, the former Colonies preferred the opposite of the British form of centralized government. The United States came to favour local assemblies and decentralized rule, both geographically and institutionally.

90  Research handbook on political representation

SWEDISH ESTATE REPRESENTATION In the meantime, the Swedish Diet, especially in the so-called late Age of Liberty (1766–72), has sometimes been seen as a herald of modern popular sovereignty and representative democracy (for historiography, see Ihalainen 2010, ch. 3 and pp. 477–84; Ihalainen and Sundin 2011; Ihalainen 2015). The Swedish version of representative government was based on the electoral rights of an exceptionally broad proportion of the population, the free peasant estate covering, in principle, a quarter of the male population alongside the noble, clerical and burgher estates. The rule by the estates was extensive when the diet was in session. By the late 1760s, the concepts of ‘the people’ (folket) and nation were used to legitimize political demands by the competing estates and parties known as the Hats and Caps. Even if the Riksdag could be presented in the free press as representatives of the people, it by no means yet constituted a ‘representative democracy’. Disputes over estate privileges led to politicization of the concept of the people. The noble estate had traditionally viewed itself as the people who made use of the supreme power of the nation. Some of its Hat opposition radicalized the concept of the people in 1769, extending it to cover the lower orders, to challenge the government. However, as the peasant estate demanded more political influence, most noblemen rejected such ‘democratic’ participation. In 1771−72, the Cap opposition of the peasant, burgher and clerical estates likewise employed the concept of the people to challenge the Hat ministry and increase their privileges. The Hats of the nobility responded by appealing to the king to save the established order. All members of the estates typically prioritized their own representation and privileges over general popular representation. The popular origin of political power in the past was recognized by many contemporary politicians, the supreme power of the estates by most and the relevance of popular opinion in decision-making by some. Yet the basic assumption remained that political power had been surrendered by the people to the estates, who constituted the people rather than simply representing them. The members of the diet defined themselves as ‘plenipotentiaries’ (fullmäktige), emphasizing their independence from the electors and avoiding references to responsibility to the people. References to ‘representatives of the nation’ occasionally appeared, yet it was more common to talk about ‘the representation of an estate’ in the Council of the Realm (Swedish Diet (SD), RAP 1769–70, 1:35; all translations are ours). Noble members represented their entire family, clerical members a group of the clergy or ‘class of civil servants’, burghers ‘all the inhabitants of their town’ and peasants ‘their brothers at home’ or ‘the Swedish commoners’ (SD, RAP 1771–72, 3:76–7; BgP 1809, 415; BnP 1809–10, 3:448). After a royal coup in 1772 the regime became increasingly autocratic. King Gustavus III referred to the people to legitimize his power against the nobility, emphasizing the common interests of the monarch and the three lower estates as (re)presenting (föreställa) ‘the entire people of Sweden’ (SD, RAP 1778–79, pp. 48, 89). Just before the outbreak of the French Revolution, the Swedish king broadened his prerogative further and increased the rights of the lower estates. The nobility responded by emphasizing ‘the knowledge, free will and consent of the Estates of the Realm which historically represents the Swedish people, and with which the king has in all times taken counsel and consulted’ (SD, RAP 1789, 1:406). The king, for his part, encouraged the peasants to claim that their estate expressed ‘the voice of the nation’ (P.Z. Ahlman, SD, BnP 1789, pp. 6–8), reinforcing the myth of a uniquely free Swedish or Finnish

From estate representation to the representation of the people  91 peasant. In reaction to regicides in Sweden and France, however, the interests of the monarch and his subjects were presented as identical. Napoleon destroyed this fiction by persuading Alexander I of Russia to force Sweden to join his blockade against Britain by conquering Finland in 1808. The war led to a change of Swedish monarch and to the adoption of a constitution that updated conceptualizations of representation. Discourse on representation grew exponentially in all estates. ‘The improvement of national representation’, ‘the election of plenipotentiaries to the Diet’ (SD, BgP 1809, p. 179; BnP 1809–10, 1:312) and the creation of a five-estate assembly or a uni- or bicameral legislative body were discussed (SD, RAP 1809, 1:190), emphasizing the political rights, liberties and duties of citizens, not those of the estates alone. The revolutionary notion of popular sovereignty was rejected but the Swedish nation was viewed as an active political agent rewriting its mixed constitution through its representatives (Sundin 2006, pp. 75–83; SD, BgP 1809, 6:129–30). The estates were defined as ‘the representatives of the Swedish people’ (SD, BgP 1809, 8:494–5), and some noblemen boasted about the title of ‘representative of the people’ (folk-representant, SD, RAP 1809, 5:259). These redefinitions, echoing those made in the German-speaking lands, reflected parallel difficulties in updating early modern principles of representation. The object of estate representation could be the said estates but also ‘the nation’, ‘the country’, ‘the native country’ (fosterlandet), ‘the public’ or ‘the realm’ (SD, BgP 1809, 2:124, 551, 553; 1809–10, 1:49; BnP 1809–10, 1:49, 10:197; PrP 1815, 1:143). Critical debate on ‘the general right of representation’ (SD, BgP 1809, 3:693) as ‘the right of a citizen’ (SD, BgP 1809, 8:807), ‘the system of representation’ and ‘the transformation of representation’ (SD, RAP 1809, 2:157, 4474) still concerned a very small group of landowners and entrepreneurs as a potential new ‘body of representation’ (SD, BnP 1809–10, 1:112). The traditional conclusion was that ‘the entire Swedish people’ and ‘every Swedish man’ was ‘represented through the four Estates of the Realm’ even if there was no ‘direct contribution to representation’ (SD, RAP, 1800, 2:920, 1110–11). All innovations in representation and the structure of the diet were postponed, the peasant estate representing ‘the mass of the people’, the burghers ‘a population of 60 000’, the clerics 14 000 and the noblemen 9500 persons, the latter including both sexes (SD, KU 1810/1812/1815, p. 324). The estates wished to retain their immemorial rights. Yet any conflict between members of the government and the representatives of the people was seen as unthinkable (SD, RAP 1815, 4:631, 641–3, 5:218), and the representatives recognized their responsibility to the nation (SD, RAP 1815, 5:484). Such updates were recorded even by members of the clerical estate who viewed the Swedish system as ‘representative government’ by 1815 (ett representatift styrelse-sätt, SD, PrP 1815, 3:1335). Conceptions of representation were gradually changing even on the peripheries of Europe, as reflected in the ways in which Alexander I, when summoning the Finnish Estates in 1809, updated the language of politics. Educated in Enlightenment ideals as he was, the Russian Czar reinforced his role as a constitutional monarch of his new Finnish grand duchy, referring to the estates as the representatives of the people and flattering the Finns as a brave and loyal people ‘[p]lacé désormais au rang des nations, sous l’empire de ses loix’ (Alexander I, FD, PrP 1809, p. 514). In reality, 54 years of monarchical rule without representation followed. While both Swedish and Finnish political elites continued to hold the inherited principle of the representation of the people, the Finns were bound to articulate it more forcefully when opposing Russian rejections of the same.

92  Research handbook on political representation

RADICALLY REDEFINED REPRESENTATION DURING THE FRENCH REVOLUTION Representation was one of the central terms of the French Revolution. Theories and practices of representation had previously been based on corporate bodies and unequal representation in provincial estates. The monarchy had made modest attempts to reform representation by delegating prerogative to provincial estates but a major transformation was seen only when, on 17 June 1789, the Third Estate declared that sovereignty belonged solely to the nation and that they, as the National Assembly, represented the will of the people constituted by individuals (Jones 1995). Emmanuel-Joseph Sieyès had famously argued for a representative government based on national sovereignty and equality among citizens and, like most contemporaries, preferred representation to ‘real democracy’ (Rosanvallon 1995; Wright 2002; Dunn 2005; Urbinati 2006). The representative institution would speak on behalf of the nation, and all citizens participate in legislative work through their representatives (Baker 1990; Wright 2002). By summer 1789, even conservative and initially moderate members of the National Assembly saw it as their duty to ‘secourir le peuple, ce peuple qui souffre, ce peuple que nous avons l’honneur de représenter, et l’obligation de défendre’ (Archives parlementaires [AP], Gérard de Lally-Tollendal, 5 July 1789, p. 195) as they possessed ‘la puissance législative de la nation, représentée par la collection de ses députés’ (AP, Bertrand Barrère, 6 July 1789, p. 205). According to Sieyès, ‘la nation française’ was: toujours toute entière légitimement représentée par la pluralité de ses députés, … les mandats impératifs, …ne peuvent jamais ni arrêter son activité, ni altérer la liberté, ni atténuer la force de ses statuts, ni enfin restreindre les limites des lieux soumis à sa puissance législative, laquelle s’étend essentiellement sur toutes les parties de la nation et des possessions françaises. (AP, 8 July 1789, p. 207)

The principle of representation of the nation, people or citizens was extended to different levels of the revolutionary political system, and the French Constitution was claimed to provide ‘un modèle de la représentation la plus exacte’ for all other European nations (AP, 15 January 1790, p. 495). It is noteworthy that, despite the status of the revolutionary assemblies as representatives of national sovereignty, debate continued on representation derived from inheritance, the question being whether the royal representation of national dignity, the French people and France in relation to other nations was still valid (AP, Pierre Louis Rœderer, 10 August 1791, pp. 323, 325; Antoine Barnave, 331). Most revolutionaries, in line with Sieyès, prioritized parliamentary representation over direct democracy as the means to express national sovereignty (Rosanvallon 1995), while others underscored the original sovereignty of the people. As the revolution became increasingly radical with the Revolutionary Wars and the founding of the republic, the assemblies adopted the identity of the defenders of ‘les droits sacrés de l’humanité et la représentation nationale’ (AP, President Armand Gensonné, 20 March 1793, p. 344). As ‘la puissance la plus sacrée que le peuple souverain puisse former’ the members saw themselves as deliberating ‘des destinées du genre humain’ threatened by counter-revolutionaries (AP, 30 May 1793, p. 630), viewing the goals of French revolutionary national representation as universal. Yet disagreements on the best ways of obtaining ‘national opinion’ through representation and on the proper representatives of the people deepened (AP, Bertrand Barrère, 21 March 1792, p. 425; 29 April 1793, p. 585), challenging the revolutionary ideal of the unity of the sovereign nation and its representatives (Gauchet 1995). Radical calls

From estate representation to the representation of the people  93 for the most ‘democratic’ representation possible rose in response to concerns that the people tended to lose their sovereignty with representation (AP, Jean Joseph Victor Genissieu, 14 June 1793, p. 518). As some revolutionaries believed that the representatives would turn into a new aristocracy out of popular control (Markoff 1999), they prioritized the active participation of the people. During the Jacobin Republic of 1793, associations between the revolution, republic and democracy rather than representative government gained popularity. In February 1794 – during the Terror – Maximilien Robespierre combined the concepts of representation and democratic rule into ‘representative democracy’, which has generally been seen as a turning point in the history of democracy (Dunn 2005; Hobson 2008). This conceptual innovation identified the people as the source of power through representation instead of direct popular rule. It was made possible by Robespierre’s understanding of both democracy and representative government as originating from the revolutionary principle of the sovereignty of the people (Rosanvallon 1995). In the British parliament, such radicalization of representation led to defences of the status quo. The early phase of the revolution was still open to visions for rethinking the British political system. In March 1790, Henry Flood made ‘the representation of the people’ the key idea of his motion for parliamentary reform. Flood wished to see a stronger ‘popular form of government’, replacing virtual with actual representation so that the majority of the people (men) would be entitled to vote. In his view, representation enabled ‘a great people’ to ‘possess an efficient influence in their own legislature, without being legislators themselves’. It would turn the Commons into an institution no longer ‘under another influence than that of the people’ (HC, 4 March 1790, pp. 453–64). A new discourse cycle started in 1792 with Thomas Paine’s suggestion that combining representation and democracy (in the American way) would produce a constitution embracing and confederating all the various interests of the population (Paine 1792, p. 33). Prime Minister Pitt responded by recognizing that the ‘legitimate authority’ of ‘a proper representative assembly’ was based on the people feeling the identity of interests between themselves and the MPs. Yet the British political order already enabled both representation of the people (if not the nation at large) as well as ‘the true spirit of proper democracy’ (HC, 30 April 1792, pp. 1309–12). The Association of the Friends of the People soon demanded the use of the concept of representation ‘in its fair and obvious sense’ rather than in its current unconstitutional one. The reformist Charles Grey welcomed the creation of ‘a free and fair representation of the people’ instead of the current Commons (Ihalainen 2010, p. 399), but the majority of the House turned increasingly suspicious of reform after the opening of hostilities with Revolutionary France. Rhetorical redefinitions of the established system were the usual response in the late 1790s, John Thelwall calling the Commons ‘representative democracy’ (The Tribune, No. 25, pp. 217–18). The realization of representation continued to be debated but there were limited prospects in post-revolutionary Britain for extending representation as in France. In the Batavian Republic, formed on the ruins of the Dutch Republic after a French intervention in 1795, representation was radically redefined in line with the revolutionary regime. The ideas of popular sovereignty, representation and democracy were supported by American influences and native debates. Rutger Jan Schimmelpenninck, for instance, had argued in 1785 for popular sovereignty delegated to representatives, and some of his fellow authors associated popular sovereignty with (representative) democracy. The ideas of popular sovereignty and representation were amalgamated with French revolutionary ones into ‘popular government

94  Research handbook on political representation by representation’ (Rutjes 2012, pp. 8–9, 11, 27, 71–4, 77, 79–80, 118). The experiment did not lead to lasting representative democracy, as the notion of purely popular representation was rejected in connection with the Restoration in the Netherlands. Yet elements of a representative system were included in the monarchical government, just as in France and Southern Germany (Prutsch 2013, pp. 206, 208, 212–13).

CONCLUSION This chapter has compared contemporaries’ perceptions of representation in Britain and its American Colonies, Sweden, Finland, Russia, France, the German lands and the Netherlands, considering transnational interaction as well. Our conceptual historical approach has foregrounded the coexistence of competing understandings of representation in late eighteenth-century representative assemblies. We have highlighted both patterns of continuity of early modern conceptions, and the rise of more modern ones during the critical and turbulent period between 1760 and 1815. Not only the German but also the British, Swedish and Finnish cases exemplify the durability of early modern conceptions of representation in the revolutionary period. In Britain, representation had been a highly politicized concept in constructing and challenging parliamentary legitimacy as early as the seventeenth century. It became increasingly so under Robert Walpole’s government and with popular challenges to parliament, the American troubles and calls for reform from the 1760s onwards. The majority of the British parliament maintained an interpretation of representation supportive of parliamentary sovereignty under transnational revolutionary pressure, turning into a counter-force to French forms of popular representation and making mere rhetorical adjustments to maintain established order. Petitions and reform proposals exemplify that rather than being uniform and stable, representation was essentially a process of constant negotiation. Constituents and representatives may not have been equal partners, but neither was eighteenth-century representation as one-sided as is often assumed. Britain’s former colonies in North America, although revolutionary in their conduct, claimed to perfect the British system of representation, emphasizing their republican interpretation of a mixed constitution and checks and balances. In international comparison, the Swedish case reminds one of the German and to some extent the British one. Early modern ideas of representing the competing interests of the estates – abused by an autocratic monarchy in the late eighteenth century – continued to flourish in Sweden with just minor adjustments towards recognizing popular sovereignty under transnational pressures that had culminated in the French Revolution. Owing to the very gradual nature of change, the national narrative has sometimes overemphasized the modernity of eighteenth-century forms of representation in Sweden. In Revolutionary France, representation became a major tool for realizing popular sovereignty but remained an object of dispute between those who preferred revolutionary assemblies to direct democracy and those who believed that the latter form of representation endangered popular sovereignty. This tension supported adoption of the concept of representative democracy. The radicalized revolutionary concept of representation evidently had a transnational impact in all the studied countries, including Britain and Sweden (and Germany, and even Finland and Russia). Confrontations between representation and democracy were also seen in the Netherlands where pre-revolutionary transnational debates on popular sovereignty, representation and democracy had created opportunities for their positive reception. No

From estate representation to the representation of the people  95 permanent representative democracy was created by the French revolutionaries or their sister republics, however. The post-Napoleonic regimes typically aimed at restoring traditional forms of representation, only to find themselves updating forms of interest representation. The amalgamation of democracy and representation nevertheless remained part of the legacy of the French Revolution. Ever since, the inherent tensions between representation and democracy in the realization of popular sovereignty have contributed to the contestability of representation. Representation remains a central rhetorical concept, especially in parliaments: parliamentary government can be both legitimized and delegitimized with references to representation.

NOTE 1.

Quarter sessions were traditional local courts in the kingdom of England.

BIBLIOGRAPHY Primary Sources Archives Parlementaires (AP), French Revolution Digital Archive, https://​frda​.stanford​.edu/​en/​ap. Burke, Edmund (1775), Mr. Edmund Burke’s speeches at his arrival at Bristol, and at the conclusion of the poll, 2nd edn, London. Finnish Diet (FD) (1899), Prästeståndets protokoll vid Borgå Landtdag år 1809 (PrP), ed. Elis Lagerblad, Helsinki: Svenska Litteratursällskapet. General Evening Post. House of Commons (HC) and House of Lords (HL), Parliamentary History, William Cobbett (ed.), London 1806–20. The Parliamentary Register; or, History of the Proceedings and Debates of the House of Commons (PR), London 1775–1800 (Eighteenth Century Collections Online). Independent freeholder of Middlesex (1770), An Address to the People of England, on the inexpediency of dissolving the present Parliament, 1st edn, London. London Evening Post. London Magazine. Paine, Thomas (1792), Rights of Man. Part the Second, 2nd edn, London. Swedish Diet (SD), https://​riksdagstryck​.kb​.se/​standsriksdagen​.html, Bondeståndets protokoll (Minutes of the peasant estate, BnP). Swedish Diet (SD), https://​riksdagstryck​.kb​.se/​standsriksdagen​.html, Borgarståndets protokoll (Minutes of the burgher estate, BgP). Swedish Diet (SD), https://​riksdagstryck​.kb​.se/​standsriksdagen​.html, Konstitutionsutskottets protokoll (Minutes of the Constitutional Committee, KU). Swedish Diet (SD), https://​riksdagstryck​.kb​.se/​standsriksdagen​.html, Prästeståndets protokoll (Minutes of the clerical estate, PrP). Swedish Diet (SD), https://​riksdagstryck​.kb​.se/​standsriksdagen​.html, Sveriges ridderskaps och adels protokoll (Minutes of the Swedish knights and noblemen, RAP). The Tribune.

96  Research handbook on political representation Literature Albareda, Joaquim and Manuel Herrero Sánchez (eds) (2019), Political Representation in the Ancien Régime, New York, NY, USA and London, UK: Routledge. Baker, Keith Michael (1990), Inventing the French Revolution, Cambridge: Cambridge University Press. Congleton, Roger D. (2010), Perfecting Parliament: Constitutional Reform, Liberalism, and the Rise of Western Democracy, Cambridge: Cambridge University Press. Dickinson, H.T. (2007), ‘The Representation of the People in Eighteenth-Century Britain’, in Maija Jansson (ed.), Realities of Representation: State Building in Early Modern Europe and European America, New York, NY, USA and London, UK: Palgrave Macmillan, pp. 19–44. Dunn, John (2005), Setting the People Free: The Story of Democracy, London: Princeton University Press. Gauchet, Michel (1995), La Révolution des pouvoirs: La souveraineté, le peuple et la représentation 1789–1799, Paris: Gallimard. Gehrke, Roland (2005), ‘Zwischen altständischer Ordnung und monarchischem Konstitutionalismus: Begriffsklärungen und Fragestellungen’, in Roland Gehrke (ed.), Aufbrüche in die Moderne: Frühparlamentarismus zwischen altständischer Ordnung und monarchischem Konstitutionalismus 1750–1850, Cologne: Böhlau, pp. 1–11. Greene, Jack P. (2007), ‘Traditions of Consensual Governance in the Construction of State Authority in the Early Modern Empires in America’, in Maija Jansson (ed.), Realities of Representation: State Building in Early Modern Europe and European America, New York, NY, USA and London, UK: Palgrave Macmillan, pp. 171–201. Haaparinne, Zachris (2021), Petitions and Disputes on Political Representation in Britain, 1721–1776, PhD Dissertation, University of Jyväskylä. Hobson, C. (2008), ‘Revolution, Representation and the Foundations of Modern Democracy’, European Journal of Political Theory, 7 (4), 449–71. Ihalainen, Pasi (2010), Agents of the People: Democracy and Popular Sovereignty in British and Swedish Parliamentary and Public Debates, 1734–1800, Leiden, the Netherlands and Boston, MA, USA: Brill Academic Publishers. Ihalainen, P. (2015), ‘The 18th-Century Traditions of Representation in a New Age of Revolution: History Politics in the Swedish and Finnish Parliaments, 1917–1919’, Scandinavian Journal of History, 40 (1), 70–96. Ihalainen, Pasi and Anders Sundin (2011), ‘Continuity and Change in the Language of Politics at the Swedish Diet, 1769–1810’, in Pasi Ihalainen, Michael Bregnsbo, Karin Sennefelt and Patrik Winton (eds), Scandinavia in the Age of Revolution: Nordic Political Cultures, 1740–1820, Farnham: Ashgate Publishing Limited, pp. 169–92. Jansson, Maija (2007), ‘Introduction’, in Maija Jansson (ed.), Realities of Representation: State Building in Early Modern Europe and European America, New York, NY, USA and London, UK: Palgrave Macmillan. Jones, P.M. (1995), Reform and Revolution in France, 1774–1791: The Politics of Transition, 1774–1791, Cambridge: Cambridge University Press. Knights, M. (1993), ‘London’s Monster Petition of 1680’, The Historical Journal, 36 (1), 39–67. Knights, Mark (2009), ‘Participation and Representation before Democracy: Petitions and Addresses in Premodern Britain’, in Ian Shapiro et al. (eds), Political Representation, Cambridge: Cambridge University Press, pp. 35–57. Markoff, J. (1999), ‘Where and When was Democracy Invented?’, Comparative Studies in Society and History, 41 (4), 660–90. Morgan, Edmund S. (1988), Inventing the People: The Rise of Popular Sovereignty in England and America, New York, NY: W.W. Norton & Company. Prutsch, Markus J. (2013), Making Sense of Constitutional Monarchism in Post-Napoleonic France and Germany, Basingstoke: Palgrave Macmillan. Rosanvallon, P. (1995), ‘The History of the Word “Democracy” in France’, Journal of Democracy, 6 (4), 140–54. Rutjes, Mart (2012), Door gelijkheid gegrepen: democratie, burgerschap en staat in Nederland 1795–1801, Amsterdam: Vantilt.

From estate representation to the representation of the people  97 Seaward, Paul and Pasi Ihalainen (2016), ‘Key Concepts for the British Parliament, 1640–1800’, in Pasi Ihalainen, Cornelia Ilie and Kari Palonen (eds), Parliament and Parliamentarism: A Comparative History of a European Concept, New York, NY: Berghahn Books, pp. 32–48. Stollberg-Rilinger, Barbara (1999), Vormünder des Volkes? Konzepte landständischer Repräsentation in der Spätphase Des Alten Reiches, Berlin: Duncker & Humblot. Sundin, Anders (2006), 1809: Statskuppen och regeringsformens tillkomst som tolkningsprocess, Stockholm: Elander Gotab. Urbinati, Nadia (2006), Representative Democracy: Principles and Genealogy, Chicago, IL: University of Chicago Press. Wright, Johnson Kent (2002), ‘The Idea of Republican Constitution in Old Régime France’, in Martin van Gelderen and Quentin Skinner (eds), Republicanism: A Shared European Heritage, Vol. 1, Cambridge: Cambridge University Press, pp. 289–305. Yerby, George (2008), People and Parliament: Representative Rights and the English Revolution, Basingstoke: Palgrave Macmillan.

8. Individualistic representation in the liberal century – and beyond Heinrich Best and Lars Vogel

THE CONCEPT OF INDIVIDUALISTIC REPRESENTATION Individualistic representation can be defined as the freedom of representatives to make their own decisions in office without interference by others. Here, the authority in the relationship between constituents and representatives is clearly allocated to the latter, based on their assumed moral standing, intellectual competence and the quality of their leadership. Quality matters because in individualistic representation representatives are not just ‘mirrors’ or ‘agents’ of their constituents, but fully empowered actors. The concept of individualistic representation is a normative rationalization of institutional practices developed in the process of state building in the late eighteenth and early nineteenth centuries which were related to social change that freed the upper echelons of society from corporative restrictions. In this historical situation free representation by independent representatives became a political tool to integrate dynamically changing and increasingly differentiated societies. In his typology of authority and imperative coordination (Typen der Herrschaft) Max Weber sees representation as a fundamental form of organization of authority in corporate groups (2012, pp. 416–23). It is characterized by the underlying ‘primary fact that the action of certain members of a group, the “representatives” is binding on the others or is looked upon as legitimate so that its result must be accepted by them’ (ibid. p. 416). Under the wide umbrella of this definition, Weber distinguishes between various emanations of representation that differ mainly according to the source of the authority of representatives and the definition of their representative roles. The most archaic of these variants of representation is ‘appropriated representation’, which is ‘found in all kinds of patriarchal and charismatic groups’ and typically claimed by the heads of these groups such as chieftains, monarchs and sheiks (ibid.). Weber also identifies a closely related variant of appropriated representation on the basis of ‘socially independent grouping’. Here claims of representing ‘others’ beyond the representatives themselves are based on privilege, whereby ‘others’ comprise those who are not members of the privileged class from which the representatives are drawn. Weber includes privileged estates, like those of the late Middle Ages, in this category. The British House of Lords can be considered as a living relic of representation appropriated by privilege (ibid.). Whereas the beneficiaries of an appropriated mandate can exercise their function freely, the third type of ‘instructed representation’ combines the choice of representatives by election, rotation or lot with an imperative mandate and their submission to instruction. Here the representatives’ mandate is not received as an awarded privilege or as an inherited property but as given through a selective procedure involving those to be represented (Weber 2012, p. 417). Because ‘instructed representation’ is subject to the consent of those represented, ‘this type of “representative” is, in effect, an agent of those he represents.’ Max Weber refers to the 98

Individualistic representation in the liberal century – and beyond  99 representatives of the communes in the French ancien régime and to the Russian soviets as examples of ‘instructed representation’. In the case of ‘free representation’, Weber’s fourth and last type of representation, the representative is generally elected and ‘not bound by instruction, but in a position to make his own decisions. He is obligated only to express his own genuine conviction, and not to promote the interests of those who have elected him’ (ibid.). Under the auspices of ‘free representation’ the representative ‘by virtue of his election exercises authority over the electors and is not merely their agent.’ He is, however, bound by ‘the obligation to conform to abstract norms, political or ethical’ (ibid.). Weber assigns this type of representation to ‘modern parliamentary representation’ (ibid.). In the context of this chapter it denotes individualistic representation by elected representatives who exercise their mandate without having to submit formally to instructions, being bound only by their adherence to universal norms and their dedication to the common good. Although Weber does not give an outline of the genealogy of types of representation, he affirms that ‘appropriated representation’ and ‘free representation’ emerged in a sequential order: ‘“appropriated representation” is archaic and found in groups, “free representation” underlies the representative bodies of modern political organization of parliaments’(ibid., p. 419). It is obvious that under the premises of constitutional continuity, archaic forms of representation can survive into modern times, the most significant case being Great Britain with its Monarchy and House of Lords. But even these institutions are actually subordinate to the House of Commons, which is based on the principles of free or individualistic representation.

INDIVIDUALISTIC REPRESENTATION: CONCEPT AND PRACTICE IN THE LIBERAL CENTURY The principles of free or individualistic representation evolved over a centuries-long process and were very often born out of a situation where instructions for members of corporative representations were missing or incomplete while immediate action was needed. This was regularly the case in times of heated conflicts between representative bodies and monarchs, particularly about issues of taxation. In these conflicts, an issue which technically concerned only a socially limited group of taxpayers could become of constitutional relevance, transcending the established rules and practices of authority (Bryant 2014, pp. 173–235). In such situations, no instructions were available or could be obtained. Representatives were summoned to act independently of external intervention on the basis of universal norms that transcended the narrow interests of some class or region. They redefined themselves as proxies of abstract and superordinate entities such as ‘Commonwealth’, ‘Nation’ or ‘the People’. The English House of Commons in the seventeenth century, the American Constitutional Convention of 1787 and the French Assembly of the Third Estate of 1789 (the latter soon recast into the Assemblée nationale constituante), exemplify this process. The French and the American cases particularly epitomize individualistic representation in that they explicitly abolished terms of instructed or appropriated representation (Palmer 1959; Lefebvre 1973; Tocqueville 1998). The emergence of individualistic representation defined as free representation by formally independent representatives is obviously linked to the development of the modern state and its need to create centralized political institutions that fulfil three functions: to integrate complex and inherently conflictual elite structures; to create reliable links between the levels of central

100  Research handbook on political representation political institutions and the general population; and to come to rational decisions which are binding for and accepted as legitimate by the polity as a whole. In particular the challenges of integration and of efficient decision-making could and can be achieved only by actors who have leeway to compromise and the authority to decide. Both capacities formed the core of the ideologemes of individualistic representation, which were given their classical formulation in Edmund Burke’s Speech to the Electors of Bristol after he was declared by the Sheriffs to be elected as one of the two Representatives of Bristol on 3 November 1774 (1774 [1999], pp. 5–15). It marked the advent of what can be called the ‘liberal century’, which extended to the 1870s when mass parties and collective representation started to dominate politics. In his remarks concerning the highly controversial ‘topic of instructions’, Burke distanced himself from the coercive authority of such instructions, which had been supported by his rival for office, Henry Cruger. The latter had claimed that ‘it has ever been my opinion that the electors have a right to instruct their members. For my part I shall always think it my duty in Parliament to be guided by your counsels and instructions’ (Broke 2019). Burke’s riposte went to the core of the issue (1774 [1999], p. 12): Parliament is not a congress of ambassadors from different and hostile interests, which interests each must maintain as an agent and advocate against other agents and advocates, but Parliament is a deliberative assembly of one nation with one interest, where not local purposes, not local prejudices ought to guide, but the general good, resulting from the general reason of the whole. You choose a member indeed, but when you have chosen him, he is not a member of Bristol, but he is a member of Parliament. If the local constituent should have an interest or should form a hasty opinion evidently opposite to the real good of the rest of the community, the member for that place ought to be as far as any other from any endeavour to give it effect.

Burke’s concept of representatives as ‘trustees’, that is, independent advocates of the common good, only bound by reason and submission to universal norms, established members of parliament as a representative elite, a segment of the leading stratum of the polity which was supposedly, within constitutional limitations, entitled to make binding decisions for society and polity as a whole (Best and Vogel 2018). In the following decades, Burke’s propositions were fully developed, codified, but also modified and refuted in a debate where political and theoretical positions merged inseparably. In the Federalist Papers, free representation becomes a cornerstone of a republican form of government which hedges against the risks of the ‘tyranny of a majority’ in a plebiscitarian democracy and forms an integrative core for the emerging union of federal states (Palmer 1959). The establishment of an Electoral College in the process of US American presidential elections is a lasting relic of these concerns. However, these precautions also show how ephemeral the construct of a ‘free representation’ was under the pressures of mass democracy and party politics. In France, the first post-revolutionary constitution of 1791 established a legislative assembly elected on the basis of a two-tier indirect voting system with increasing censitary eligibility requirements at each level and with those bound to the personal service of others, such as domestic servants, being excluded from becoming active citizens (Rudé 1988). Deputies enjoyed the parliamentary privilege of being legally inviolable while their mandates were protected against premature dissolution of the assembly. These precautions against an onslaught of uncontrolled interventions of popular or royal caprice were, however, done away with by the Constitution de l’an I (1793) which introduced universal and direct male suffrage and an ‘appeal to the people’

Individualistic representation in the liberal century – and beyond  101 for important laws. The French National Convention of 1792, formally elected on the basis of universal male suffrage, allowed and sometimes provoked the disruption of parliamentary sessions by the populace and turned the principle of inviolability on its head by having large parts of the actual opposition executed (Rudé 1988). In many respects, the French constitution of 1793 and the parliamentary practice of the Convention were the nightmares come true of the authors of the Federalist Papers because they replaced the principles of individualistic representation by the principles of popular sovereignty and introduced an element of direct popular control into the running of public affairs. It was, however, indicative of a growing elitist spirit of the members of the revolutionary Convention that the constitution of 1793 was never enforced. Instead, we see for decades to come, across the whole of Europe, a suppression and retreat of the ideas of popular sovereignty and a restoration of the monarchical principle. However, the directed parliamentarism of the Napoleonic era, where elections under highly restrictive censitarian electoral laws were combined with governmental control over parliamentary proceedings, was also not fully conforming to the principles of individualistic representation. It nevertheless became the model for the continental European states of the post-Napoleonic era, in so far as they introduced parliamentary representation at all. In many European states, the reconsolidation of monarchies in the Napoleonic and post-Napoleonic era saw an introduction of parliamentary representations which were at least partly modelled after the ‘individualistic’ principles outlined in Burke’s speech to his voters. Usually based on highly censitary suffrages and candidate selection, they resulted in a representation which was socially dominated by the juste milieu of the propertied and educated classes (Bentley 1996). The archetypal representative was the dignitary, who was anchored in the local power structure, with independent means and some higher education (Beck 1974). The mix of these elements of social distinction varied between countries: for example in France one finds a stronger element of wealth in the qualification for office, whereas in Germany (where, after 1815, parliaments were introduced in the territorially fragmented central and southern parts of the German Confederation) it was primarily education combined with a high-ranking position in the state administration that qualified one for office (Best 2001). In many cases, the chambers of post-Napoleonic parliaments were interspersed with members of the lower echelons of the nobility or with second sons of the aristocracy. These recruitment patterns endured well into the nineteenth century – for example, in the 1870s, more than 40 per cent of the members of the German Reichstag, which had been elected on the basis of universal male suffrage, had a noble title (Best et al. 2001). Social exclusivity of legislative recruitment in combination with the elitist ideology of individualistic representation led to homogeneity and – notwithstanding persistent infighting – coherence of the representative elites of the nineteenth century. They became formidable opponents of the protagonists of the monarchical principle and formed a bulwark against the onslaught of revolutions, at least as far as those attempting to establish regimes based on popular sovereignty are concerned (Best 2001). The ‘chambers’ of the nineteenth century, which were modelled on the House of Commons, were the pivots of the constitutional monarchies and used their power of the purse, most times successfully, as leverage against authoritarian and absolutistic ambitions of monarchs and their coteries. Therefore, as far as the nineteenth century can be seen to qualify as a liberal century, it was due to representative elites formed under the premises of individualistic representation.

102  Research handbook on political representation This institutional setting was, however, fraught with internal contradictions and shortcomings. The social exclusivity of parliamentary recruitment, as far as it was enforced by electoral laws, narrowed the pool of contenders who might otherwise have been well qualified to promote the common good and to direct the course of state affairs. It was also in doubt as to whether having independent financial means could be equated with freedom from special financial interests or from career ambitions. Free representation was definitely not free from special interests – an impairment which had already become visible in the founding process of the United States of America (Beard 1913). Parliamentary rules, which tried to prevent the corruption of parliamentary procedure by special interests or patronage, were usually not comprehensive enough to prevent these evils from happening. According to Burke (1999, p. 12), however, individualistic representation was based on the assumption that it was the moral duty of a representative to ‘sacrifice his repose, his pleasures, his satisfactions […] and, above all, ever, in all cases to prefer [his constituents’] interests to his own.’ The compliance with this duty was, however, hard to control, particularly if individual interests coincided with some collective interest, for example those of the social class or the segment of the elite from which the representation was drawn. In view of this reality, free representation and the independence of representatives appears to be fictitious, but a necessary fiction if one wanted to uphold the normative foundations of individualistic representation. In structural terms, individualistic representation in the liberal century was based on the private means of candidates and incumbents. On the one hand, the censitarian electoral system linked the right to vote to property and affluence. On the other hand, parliamentary mandates were not remunerated, so that affluence was a necessary condition for eligibility. The movements for electoral reforms and the democratic parties of the nineteenth century strived to overcome this advantage for the nobility and bourgeoisie by reforms to abolish the censitarian electoral system and by funding elected representatives through party means or by employing them in the party press. Not until parliamentary mandates became – as a rule – remunerated by the state at the beginning of the twentieth century, were legislators able to live off politics without private means or party sponsorship (Best and Cotta 2000). In normative terms, exclusiveness was compatible with individualistic representation in the liberal century, but inclusiveness is the contemporary normative cornerstone of democratic representation. In the liberal century, the autonomy of representatives was legitimized by the need for reasoned deliberation of the common good and the goal of effecting a homogeneous national interest. Representation was not based on particular interests, but on deliberation by a minority of reasonable men who were able to overcome their individual interests, whereby reason was usually linked to property ownership. The common good supposedly transcends and thereby includes all the different segments of society, implying that their actual presence in parliament is not necessary. Nevertheless, the proponents of this virtual representation were aware that a free mandate may tempt representatives to realize their private interests. Elections were thus not seen as aimed at bringing diverse societal interests into parliament but rather as something with which to threaten representatives vis-à-vis the potential loss of their mandate. Even John Stuart Mill, who supported the idea that representation should proceed from particular interests, presumed that these interests can be moderated and suggested their accommodation within the common good (Viera and Runciman 2008, p. 50).

Individualistic representation in the liberal century – and beyond  103

FROM INDIVIDUALISTIC TO COLLECTIVE REPRESENTATION One answer to the dilemma of interest aggregation was the formation of parties. Max Weber saw them historically as a complement to ‘modern parliamentary representation’ which cedes its ‘freedom’ regularly to the ‘voluntary intervention of parties’ (1947, p. 417). Parliamentary parties were initially established to bring together like-minded members of parliament and to form relatively stable blocs of votes, often in support of or in opposition to the government, but also to promote policies based on common idealistic commitments and material interests (ibid., pp. 407–15). In contrast, individualistic representation gives no immediate answer as to how to implement desirable public policies and to translate them into powerful political action. Burke’s (1999, p. 12) idea of a ‘lonely’ representative who is guided only by his ‘unbiased opinion, his mature judgement, his enlightened conscience’ distances him from his voters, but also from ‘any man, or any set of men’. This is, however, a highly unrealistic concept of representation, which obscures the more unsavoury realities of parliamentary practice of his day, with its horse-trading, log-rolling, cue-taking, patronage and outright corruption. Indeed, Burke himself was no saint with regard to some of these things. Once in a while, a powerful speech may have sufficed to gain a majority, but what to do if a good orator was not available to support good policies? Submitting parliamentary decisions to the shifting whims, moods and emotions of the floor would have undermined any claim of parliaments and representative elites for political power. If parliamentary parties were the answer to this challenge, they produced, however, another massive and inherently unsolvable dilemma: with parties came party discipline and the whip, submission to party leaders and commitment to a party programme, all hardly compatible with the ideals of free and individualistic representation (Blondel and Cotta 2000). The dilemma intensified when parties established themselves as intermediary organizations between parliaments and voters. With the arrival of party members and party officials, additional actors appeared who exerted influence and pressure to turn free representation into a modern form of instructed representation and to convert trustees into agents. These pressures were amplified with the advent of mass democracy, when the control of electoral policies and candidate selection shifted from caucuses of local power holders to national party machines (ibid.). The emergence of working class parties was a particular driving force of this development: the party functionary or the (nominal) worker turned member of parliament is the reverse image of the individualistic representative, whereas practising free representation may breach a sanctified party line, sometimes even being seen by fellow party members as treason to the party cause. The pressure on representatives to submit to the demands of party discipline was even stronger because free representation was not very popular amongst voters either and had to be protected by legal statutes establishing parliamentary privilege. Burke himself witnessed this unpopularity when he acted contrary to the wishes of his constituents on issues related to the trade of Ireland (Canavan 1999). As a direct consequence, he decided not to stand a second time for the highly contested seat of Bristol and to accept a safe seat from his parliamentary patron, Lord Rockingham. After that, voters were no longer of concern for Burke because Lord Rockingham’s tenants voted as he instructed. The age of mass parties eventually transformed party organizations into pivotal collective actors of ‘collective representation’ (Weissberg 1978) who compete to maximize their electoral support among voters (Schumpeter 1959). In so doing, they address specific material or idealistic interests linked to class or religion through the mechanism of cleavage representation

104  Research handbook on political representation by mass parties (Best 2007b), or they maximize their votes by targeting the median voter and transforming themselves into catch-all parties (Dalton et al. 2011). Party affiliation thus became the decisive indication for the ideological profile and policy-positions of the representatives, and voters made their choice based on their partisanship (Bowler et al. 1999). At the level of electorates, the extensions of suffrage and the formation of mass parties eroded the deference towards traditional authorities based on social standing and personal qualities. Professional competences and qualifications, which are instrumental for the exertion of representational roles, characteristics of symbolic significance such as skin colour, gender, religion, ethnic or working-class background, and the relations contenders control through their linkage positions held in organizations and networks, became more important assets for legislative recruitment and careers (Best 2007a). The increasing competitiveness of parliamentary elections and candidate selection through the extension of suffrage and the introduction of parliamentary pay curtailed the autonomy of representatives and increased their dependency on the goodwill of voters and on the support of party machines (Cotta and Best 2007). The ‘professional politician’ who lives ‘for and from’ politics emerged from these changes as the guiding figure for the personnel of modern mass democracies in the twentieth century (Weber 1958). This process in which he and, increasingly, she replaced the dignitary was, however, slow, and stretched over a century (Best and Cotta 2000). Even under the condition of universal or near to universal male suffrage, as in the Paris and Frankfurt National Assemblies of 1848/1849, most of their members would have qualified for a mandate under censitarian regimes of the pre- (and post-) revolutionary era (Best 2001). However, careers preceding recruitment to stand for election to parliaments are still controlled by parties and the structural composition of legislators indicates advantages for those whose occupation is compatible with running for office (Best and Cotta 2000). Wealth and property are still important and have even gained additional importance for access to legislative mandates since the beginning of the twenty-first century (Fox and Lawless 2011; Jacobson and Carson 2019). Although contemporary individualistic representation may be less exclusive in terms of private means, political and non-political criteria in recruitment still produce exclusion and disproportionately favour socially privileged representatives (Best and Cotta 2000). Although parties are collectives, individuals remained important actors. This importance was obvious in parties with charismatic leaders and was especially pronounced in parties that emerged from inner parliamentary caucuses of representatives. Also, at the top of working-class parties, which formed as mass organizations outside parliaments, leadership was characterized by oligarchic tendencies (Michels 1915). Further, the influence of parties was conditional upon institutional and historical context (Aldrich 1995). National political leaders were more central in presidential systems, whereas local or regional legislators were more so in majoritarian electoral systems. Research has further identified dimensions of representation beyond ideology or policy in which the individual behaviour of legislators towards their represented has been decisive, for instance, in directing resources to their own constituency (Eulau and Karps 1978). The transformation of mass-parties to catch-all parties by the middle of the twentieth century shifted the focus of election campaigns and voters’ choice away from ideological considerations towards the qualities and characteristics of their national leaders. Since the end of the twentieth century, the relation between individualistic and party-representation seems to have changed further. Parties are decreasingly characterized as mass organizations linked to social groups and structured by competing ideologies (Mair

Individualistic representation in the liberal century – and beyond  105 2008; Drummond 2006). Their core functions of aggregating and articulating the preferences of the population into the political realm are challenged by growing social heterogeneity and by the variety of new civil society actors that competitively claim to represent social and political groups (Saward 2010). This diversification is fuelled by increasing support for new institutional forms of participation and deliberation intended to give more weight to the principals (i.e. voters) vis-à-vis their agents (i.e. parties) or to abolish the principal–agent relation completely (Dalton 2014). In consequence, the role of parties changes: they are no longer the central mobilizing agencies for citizens and the choice of the voters is decreasingly influenced by party identification. In other words, the representational function of parties is waning. At the same time, while ceasing to provide the main linkage mechanism for representation, their centrality in organizing the functioning of parliaments and governments seems to remain untouched: they form governments and negotiate coalitions and recruit and allocate political personnel to ministries and committees (Mair 2008). This shift of parties from agencies of representation to agencies of institutional organization is accompanied by the personalization of politics. This term denotes the growing importance of individuals in politics at the expense of collective organizations and institutions (Karvonen 2010; McAllister 2007; Caprara and Vecchione 2017; Rahaṭ and Ḳenig 2018; Cross et al. 2018). Personalization may occur in different arenas, for instance in parties, legislatures, governments, the electorate or the media and has garnered an institutional, media and behavioural dimension (Rahat and Sheafer 2007). Personalization may further occur at the level of political elites or leaders (centralized personalization) or at the broader level of professional politicians (decentralized personalization) (Wauters et al. 2019). While personalization describes the shift towards individuals, personalism is the current degree of individuals’ influence compared to that of organizations and institutions (Cross et al. 2018). Accordingly, personalization focuses on the individual behaviour of legislators vis-à-vis their parties and their voters and may thus indicate a revival of individualistic representation.

THE PERSISTENT SIGNIFICANCE OF INDIVIDUALISTIC REPRESENTATION Wealth and education, the necessary de jure criteria for becoming a member of parliament in the era of limited suffrage, nevertheless remain important de facto criteria for becoming a representative today (Cotta and Best 2007). ‘Free representation’, although de facto under massive pressure from various groups of actors in the environment of members of parliament, has been de jure institutionalized as a cornerstone of representative democracies. Under this perspective, the DNA of individualistic representation in the liberal century is still present in modern democracies and a requirement for their performance and liberty. We therefore ask whether elements of individualistic representation are still present in party-based representation and if so, what these are, and whether they are regaining importance in the process of personalization. We also ask how these contemporary forms differ from their predecessors in the liberal century. We highlight that various types of representation which comprise in their normative core the representation by individuals formally conducting a free mandate with independence from others, still shape the role-conceptions of representatives. However, the contemporary mode of individualistic representation is distinct from the liberal mode, since it is embedded in party-representation, is less based on private means, is

106  Research handbook on political representation normatively linked to the democratic principles of equality and inclusiveness, and lacks significant references to a generalized common good. Individualistic representation in the liberal century did not initially consider parties, whereas modern individual representation remains inextricably linked. Burke’s primary concern was the conflict between local interests and the (national) common good, but in contemporary representation the focus is rather on the relation between party and individual representatives. Thus, individualistic parliamentary representation may come at the expense of parties. This is especially the case if the unity and coherence of parties is undermined by the diversity between individual representatives (Harden and Carsey 2012). Further, the emerging focus on individuals provides opportunities and incentives for representatives on a national platform to base their political support less on their party but rather on their own assets, achievements and policy-stance. The conflict becomes most obvious by defections from the party line in roll-call votes, but the coherence of parties is already endangered, if their functionaries at the top or at the grassroots propose policies that are too fuzzy to serve as basis for voters’ choices. Parties may, therefore, strategically pursue personalization, because they benefit from individualistic representation by putting their personnel in the focus of competition to substitute vanishing ideological differences. Individuals emphasize their competence and pragmatic leadership skills instead of ideological goals or policy-proposals (Pruysers et al. 2018). The conjunction between exclusion, deliberation and the common good was condensed in the Burkean figure of the trustee, who – together with the delegate as its antipode – formed the core of the most influential normative typology of (self-)conceptions of the role of representatives. Contemporary debates on representation, however, subsequently modified this dichotomy. Pitkin (1967) underlined that the role of a representative is not exclusively that of trustee or delegate but that both represent legitimate expectations of their behaviour, since representation is ‘acting in the interest of the represented, in a manner responsive to them [so that] there is no conflict, or, if it occurs, an explanation is called for’ (p. 209). Empirically, research into the role-conceptions of parliamentary representatives differentiated Burke’s dichotomy into two dimensions of representation: while focus indicates the object of representation, either national, regional or own party, style indicates whether representatives primarily attempt to realize the preferences of their respective focus or their own convictions (Eulau et al. 1959) or both (the ‘politico’) (Wahlke et al. 1962). This modification shows that nowadays autonomous deliberation for the sake of the common good is restricted by emphasizing that representation requires the inclusion of particular societal interests and party policy proposals into decision making. The trustee concept remains influential, however, in that autonomy is needed to achieve compromise-based decisions despite the heterogeneous interests of the represented. To fulfil their functions as decision-makers in a representative democracy, representatives must have the opportunity to interact and, in particular, to compromise with relative freedom in a setting of antagonistic cooperation (Best and Vogel 2018).

Individualistic representation in the liberal century – and beyond  107

REFERENCES Aldrich, John H. (1995), Why Parties? The Origin and Transformation of Political Parties in America, Chicago, IL: University of Chicago Press. Beard, C.A. (1913), An Economic Interpretation of the Constitution of the United States, New York, NY: Macmillan. Beck, T.D. (1974), French Legislators 1800–1834: A Study in Quantitative History, Berkeley and Los Angeles, CA, USA and London, UK: University of California Press. Bentley, Michael (1996), Politics without Democracy 1815–1914, London: HarperCollins. Best, H. (2001), ‘Structures of Parliamentary Representation in the Revolutions of 1848’, in Dieter Dowe et al. (eds), Europe in 1848: Revolution and Reform, New York, NY, USA and Oxford, UK: Berghahn Books, pp. 475–506. Best, H. (2007a), ‘New Challenges, New Elites? Changes in the Recruitment and Career Patterns of European Representative Elites’, Comparative Sociology, 6 (1–2), 85–113. Best, Heinrich (2007b), ‘Cleavage Representation in European Parliamentary History’, in Maurizio Cotta and Heinrich Best (eds), Democratic Representation in Europe, Oxford: Oxford University Press. Best, Heinrich and Lars Vogel (2018), ‘Representative Elites’, in Heinrich Best, and John Higley et al. (eds), The Palgrave Handbook of Political Elites, London: Palgrave Macmillan, pp. 339–63. Best, Heinrich and Maurizio Cotta (eds) (2000), Parliamentary Representatives in Europe 1848–2000, Oxford: Oxford University Press. Best, H., V. Cromwell, C. Hausmann and M. Rush (2001), ‘The Transformation of Legislative Elites: The Cases of Britain and Germany since the 1860s’, The Journal of Legislative Studies, 7 (3), 65–92. Blondel, Jean and Maurizio Cotta (eds) (2000), The Nature of Party Government: A Comparative European Perspective, Basingstoke: Palgrave. Bowler, Shaun, David M. Farrell and Richard S. Katz (1999), ‘Party Cohesion, Party Discipline, and Parliaments’, in Shaun Bowler, David M. Farrell and Richard S. Katz (eds), Party Discipline and Parliamentary Government, Columbus, OH: The Ohio State University Press, pp. 23–48. Broke, J. (2019), ‘Cruger Henry’, in History of Parliament Research, available at http://​ historyofparliamentonline​.org/​volume/​1754​-1790/​member/​cryer​-henry​-1739​-1827. Bryant, Chris (2014), Parliament: The Biography, Vol. 1, Ancestral Voices, London: Doubleday. Burke, Edmund (1999 [1774]), ‘Mr Edmund Burke’s Speech to the Electors of Bristol’, in Francis Canavan (ed.), Edmund Burke, Miscellaneous Writings, Indianapolis: Liberty Fund. Canavan, F. (1999), ‘Introduction to Edmund Burke’s Speech to the Electors of Bristol’, in Francis Canavan (ed.), Edmund Burke, Miscellaneous Writings, Indianapolis: Liberty Fund, pp. 3–14. Caprara, Gian Vittorio and Michele Vecchione (2019), Personalizing Politics and Realizing Democracy, Oxford: Oxford University Press. Cotta, Maurizio and Heinrich Best (eds) (2007), Democratic Representation in Europe: Diversity, Change, and Convergence, Oxford: Oxford University Press. Cross, W.P., Richard R. Katz and Scott Pruysers (eds) (2018), The Personalization of Democratic Politics and the Challenge for Political Parties, Colchester: ECPR Press. Dalton, Russel J. (2014), Citizen Politics: Public Opinion and Political Parties in Advanced Industrial Democracies, 6th edn, Los Angeles, CA: Sage, CQ Press. Dalton, Russel J., David M. Farrell and Ian McAllister (2011), ‘The Dynamics of Political Representation’, in Martin Rosema, Bas Denters and Kees Aarts (eds), How Democracy Works: Political Representation and Policy Congruence in Modern Societies, Amsterdam: Pallas Publications, pp. 21–38. Drummond, A.J. (2006), ‘Electoral Volatility and Party Decline in Western Democracies: 1970–1995’, Political Studies, 54 (3), 628–47. Eulau, Heinz and Paul D. Karps (1978), ‘The Puzzle of Representation: Specifying Components of Responsiveness’, in Heinz Eulau and John C. Wahlke (eds), The Politics of Representation: Continuities in Theory and Research, Beverly Hills, CA: Sage, pp. 55–72. Eulau, H., J.C. Wahlke, W. Buchanan, and L.C. Ferguson (1959), ‘The Role of the Representative: Some Empirical Observations on the Theory of Edmund Burke’, American Political Science Review, 53 (3), 742–56. Fox, R.L. and J.L. Lawless (2011), ‘Gaining and Losing Interest in Running for Public Office: The Concept of Dynamic Political Ambition’, Journal of Politics, 73 (2), 443–62.

108  Research handbook on political representation Harden, J.J. and T.M. Carsey (2012), ‘Balancing Constituency Representation and Party Responsiveness in the US Senate: The Conditioning Effect of State Ideological Heterogeneity’, Public Choice, 150 (1–2), 137–54. Jacobson, Gary C. and Jamie L. Carson (2019), The Politics of Congressional Elections, 10th edn, Lanham, MD: Rowman & Littlefield Publishers. Karvonen, Lauri (2010), The Personalisation of Politics: A Study of Parliamentary Democracies, Colchester: ECPR Press. Lefebvre, Georges (1973), The Coming of the French Revolution 1789, Princeton, NJ: Princeton University Press. Mair, P. (2008), ‘The Challenge to Party Government’, West European Politics, 31 (1–2), 211–34. McAllister, Ian (2007), ‘The Personalization of Politics’, in Russel J. Dalton and Hans-Dieter Klingemann (eds), Oxford Handbook of Political Behavior, Oxford, UK and New York, NY, USA: Oxford University Press, pp. 571–88. Michels, Robeert (1915 [1911]), Political Parties: A Sociological Study of the Oligarchical Tendencies of Modern Democracies, New York: Free Press. Palmer, Robert R. (1959), The Age of the Democratic Revolution: A Political History of Europe and America, 1760–1800, Princeton, NJ: Princeton University Press. Pitkin, Hanna F. (1967), The Concept of Representation, Berkeley, CA: Berkeley University Press. Pruysers, Scott, W.P. Cross and Richard S. Katz (2018), ‘Personalism, Personalization and Party Politics’, in W.P. Cross, Richard S. Katz and Scott Pruysers (eds), The Personalization of Democratic Politics and the Challenge for Political Parties, Colchester: ECPR Press, pp. 1–18. Rahat, Gideon and O. Ḳenig (2018), From Party Politics to Personalized Politics? Party Change and Political Personalization in Democracies, Oxford: Oxford University Press. Rahat, G. and T. Sheafer (2007), ‘The Personalization of Politics: Israel, 1949–2003’, Political Communication, 24 (1), 65–80. Rudé, George (1988), The French Revolution: Its Causes, its History and its Legacy after 200 Years, London: Weidenfeld and Nicolson. Saward, Michael (2010), The Representative Claim, Oxford: Oxford University Press. Schumpeter, John A. (1959 [1940]), Capitalism, Socialism and Democracy, London: George Allen and Unwin. Tocqueville, Alexis de (1998), The Old Regime and Revolution: The Complete Text, Vol. 1, Chicago, IL: Chicago University Press. Viera, Monica B. and David Runciman (2008), Representation, Cambridge, UK and Malden, MA, USA: Polity Press. Wahlke, John C., Heinz Eulau, W. Buchanan and L.C. Ferguson (eds) (1962), The Legislative System: Explorations in Legislative Behavior, New York, NY: Wiley. Wauters, B., N. Bouteca and B. de Vet (2019), ‘Personalization of Parliamentary Behaviour: Conceptualization and Empirical Evidence from Belgium (1995–2014)’, Party Politics, doi: 10.1177/1354068819855713. Weber, Max (1958 [1919]), ‘Politics as a Vocation’, in H.H. Gerth and Charles W. Mills (eds), From Max Weber: Essays in Sociology, Oxford, UK and New York, NY, USA: Oxford University Press. Weber, Max (2012 [1947]), The Theory of Social and Economic Organization, ed. by Talcott Parsons, Mansfield Centre, CT: Martino Publishing. Weissberg, R. (1978), ‘Collective vs. Dyadic Representation in Congress’, American Political Science Review, 72 (2), 535–47.

9. Party-based representation: the paradox of democracy Ian Budge

OVERVIEW: PARADOXES OF DEMOCRACY Democracy is distinguished from other types of political regime by seeking to bring public policy into line with popular preferences for it. Doing so obviously requires more complicated arrangements than where decision-making is confined to an elite or technical experts. A first complication is that modern democracies emerged from constitutional arrangements which sought to make policy at one remove from the people, by allowing them only to elect representatives better qualified by education and experience than they to make the best policy choices for them (elected Parliaments and/or Presidents with their advisers). The rise of mass political parties organizing blocs of candidates to compete in elections on a common policy programme (a platform or manifesto), which they were pledged to carry through in government, subverted such indirect arrangements. Government was now in the hands of a partisan majority or coalition, committed to doing what the people had shown they wanted (which they assumed to coincide with their own ideas). A second paradox thus emerged in that parties based on competing segments of the population – workers, bourgeoisie, peasants, opposing religious groups, linguistic minorities – were supposed to implement the wishes of the State population as a whole. Unresponsive parties focused on their own ideology and the sectional interests of their core supporters, and thus dedicated to promoting their own special interests, were somehow supposed to respond to the General Will and implement it. Scepticism about whether they could do this fuelled – and still fuels – anti-party sentiment, facilitating the rise of anti-party parties (populists) which claim to transcend older divisions and aim at ousting all the older parties to produce policies which truly reflect the popular will. This is a process which has gone on ever since modern democracies got going, from Jacksonian Democrats in 1830s America through the Know Nothings of the 1880s to Trump today (with parallels in most other democracies). Populism has raised particular worries at the present time, to the extent that democracy itself is widely felt to be under threat (Albright 2018, pp. 77–85; Crewe and Sanders 2020, pp. 1–71). One purpose of this chapter is to put such fears into historical perspective, by showing how often past diagnoses of party and democratic ills have got them wrong, shifting from one diagnosis to an opposed one over a few years (see Table 9.3 below). A more fundamental question is how parties with strongly contrasting ideologies and support bases ever manage to deliver public policies acceptable for electorates, which evidence shows to have generally moderate and centrist views (Powell 2000, pp. 167–8; Budge 2019, pp. 53–4). The question is rendered particularly acute by the fact that parties are elected to government on the basis of many non-policy factors (competence, candidate appeal, current scandals, crises etc.). Parties choose to regard general election success as a ‘mandate’ to 109

110  Research handbook on political representation pursue their particular policy targets. However, for many electors these are just a peripheral part of the party package. General elections by their open-ended nature rule out direct policy voting by citizens as a guaranteed way of getting public policy to conform to majority preferences, as democracy requires. How can party-based democracies square this circle? The answer lies in viewing their characteristic institution, the general election, not as a series of separate events, but as a continuing process over time. As one crisis after another impacts on votes, governments (even coalition governments) alternate between different parties with opposing policy targets. New governments generally try to change policy in their preferred direction, often the opposite one to their predecessors. Barriers to actually implementing their policy targets, however, render real change slow, so targets are rarely met in practice before one party is replaced by another in government, often by its direct rival. Their attempts to reverse their predecessors’ policies pull actually implemented policy back through the centre where the electoral majority wishes it to be. However, the majority or plurality party cannot go too far in its preferred direction before another election weakens it or puts it out of office, when its rival comes in and reverses policy again. It is thus regularly recurring general elections and party alternation in government rather than direct policy voting which match popular centrist preferences to the policy actually implemented. Other conditions of course have to be present for these mechanisms to work. These are explored below (in the section ‘Matching policy to preferences’). A final paradox of democracy is that parties, besides forming and implementing their own particular policies, also play an important role in shaping the popular preferences they are supposed to respond to (Budge 2019, pp. 26–8 and 185–204). Electors find difficulty in translating their own aspirations for personal and family stability, general security, good education and so on into public policies (as do experts and civil servants). Parties and party governments simplify things for electors by presenting them with proposals they can either support or reject – or ignore, in favour of other party-related shortcuts to voting decisions such as their competence or past record. All this seems to add up to a circular process in which established parties, supposed to represent the population, simply impose their own policy preferences and agenda on it while claiming that these are the popular will. This in fact is the basis of the populist critique and of their promise of really listening to and following through on popular grievances (which of course also happen to coincide with their own policies!). The latitude which democracies offer for new parties to emerge nevertheless offers a corrective to such a potential circularity. It offers electors (or more precisely, previously overlooked groupings among them) a chance to broaden the policy agenda to any problems that concern them but are passed over by existing parties. Party formation is a long-term process usually ignored in election and vote analyses, which take the existing party alternatives as given and as constraining voting choices. Through new party formation, however, electors can always break into the process and establish a broader range of alternatives which more accurately reflect those they want to choose between. New party formation is what saves party democracy from the accusation of being a self-reflecting process which fails to respond to true popular preferences. Where no new parties emerge, one can take it that there is broad popular support for the policies being pursued by established parties. The democratic freedom to form new parties also allows us to infer that the centrist position of the majority vis-à-vis the party positions on offer (Powell

Party-based representation: the paradox of democracy  111 2000, pp. 167–8; Budge 2019, pp. 53–4 and 121–3), is a genuine expression of preferences and not one manufactured by keeping some off the table. Parties, both old and new, provide the central dynamics for all this and so provide the main focus for the discussion below. Given the crucial role of new party emergence in guaranteeing a genuine expression of popular policy preferences, it starts off by considering the developmental theory associated with Stein Rokkan (1975). This then broadens into a consideration of the main party families and their ideologies and identifying policies. This leads us back to the initial question of how much inward-looking parties can truly respond to popular preferences. As just pointed out, popular preferences are far from spontaneous and autonomous of parties (even though in the long-term they can extend the range of party choices available). What we are really dealing with is a process of interactive preference formation, best described in the thermostatic theory of Soroka and Wlezien (2010) (see the section ‘Matching policy to preferences’ below). Saved by the possible entry of new parties from being a self-reflecting process in which established parties simply see their pre-existing policy mirrored in the electoral preference, this leads on to a consideration of the populist movements of the present time and the role they currently play. Are they an existential threat? Or is it business-as-usual in systems which have absorbed many new parties in the past? This then raises the question of how democratic processes typically operate, which draws on the alternation and thermostatic theories mentioned above. Since parties are crucial to these, the discussion ends by considering whether they are in long-term decline or constantly renewing themselves – and concludes that reports of their death are greatly exaggerated. Since ‘who says democracy says party’, this is an encouraging conclusion.

SUBVERTING REPRESENTATIVE DEMOCRACY: THE RISE OF MASS PARTIES Modern States have all developed the institutions – army, bureaucracy, law courts, police – without which they could not defend and administer their territory. The most basic distinction between States, however, is whether public policy is decided exclusively by the military – bureaucratic elite, alone and apart from the rest of the population – or whether other institutions have been added to express and impose popular preferences on them. The most important of these are regular competitive elections, freely formed parties, an independent media, parliaments determined by election results and governments responsive to such parliaments. Central to such institutions are the political parties, since they operate at all these levels and link them up – framing choices for voters, focusing candidates on a common programme in elections, organizing them to support it afterwards and forming governments to carry it through. It is no exaggeration therefore to say that modern democracies are better described as party democracies than as representative democracies, since elected representatives do as their party tells them to. Mass parties entered democratic politics from the mid-nineteenth century onwards, as new Nationalist ideologies presented the people (or the dominant majority within the State) as a nation bound together by a unique language, culture and destiny which could be realized only through their own government, and State institutions. From this perspective it was difficult to argue against their freedom to organize themselves politically through mass parties, participat-

112  Research handbook on political representation ing in general elections and parliaments, which gave them a share in national policy-making. Taking advantage of such freedoms, mass parties were organized by social groups (workers, farmers, religious groups, cultural minorities) which felt themselves politically excluded under the existing set-up. By asserting their claim to direct State policy in line with the views of their supporters, and then competing in free elections for control of policy-making, mass parties rendered the population proactive in making the political decisions which affected them, rather than having to simply submit to policies made by their rulers. The mass party transformed the political situation in three ways. In contrast to earlier factions which opposed each other only at elite level – often covertly, and exclusively within the traditional institutions – mass parties created state-wide organizations which grouped supporters in local branches. Such branches collected money to finance party activities – giving the party a financial base independent of State subsidies. They also rallied and organized supporters in elections contested between parties, in order to get the maximum number of party candidates elected to the legislature and thus empowered to push party policy there. In many ways they were a social invention comparable to the steam engine in technology, and transformed interactions in the political sphere just as much as factories did in the social sphere.

PARTY FAMILIES: IDEOLOGICAL POLICY-CARRIERS This general account is focused and summarized in Table 9.1. Table 9.1

The developmental-family theory of parties and their policy targets

1. All significant socioeconomic and opinion groupings in society find an existing party to promote their concerns or develop a new one to do so. 2. Parties develop these core concerns into an overall ideology showing how they advance the general as well as their group interest. 3. On this basis parties can be grouped cross-nationally into ideological ‘families’. 4. Party election programmes can thus always be distinguished in terms of the core policy concerns they uniquely and continually emphasize in line with other members of their international family abroad. 5. Party policy is thus generally more extreme than the majority policy preference, reflecting the views of its core supporters and their policy concerns.

Source:

Table adapted from Ian Budge, Politics: A Unified Introduction to How Democracy Works (2019, p. 77).

Proposition 1 summarizes the process we have just been describing, where all significant groupings in a democracy either find or create a party to promote their interests in its programme, and through the efforts of its members get it accepted and implemented. Otherwise they will lose out – or even, in the case of linguistic minorities seeking to have schooling in their own language – disappear altogether. The second proposition specifies the main strategy parties employ to compete effectively in general elections – that is, developing a general ideology to show how their own concerns also advance the general interest of society. For example, respecting the rights of the ethnic or religious minority represented by the party can be presented as contributing to more general satisfaction, support and stability of existing States and to a wider and more varied culture within them. Since most democracies contain groupings divided by class, ethnicity, religion or residence (urban versus rural), the parties that form on each of these social bases are broadly

Party-based representation: the paradox of democracy  113 similar: they borrow ideas from each other, and can thus be grouped cross-nationally into families with similar names, ideology and support bases. This is noted in Proposition 3 of the Developmental-Family theory. A consequence of this – often overlooked in political discussions and analyses – is that parties, embedded in a particular ideological family, cannot just abandon it in pursuit of election votes. To do so would endanger both its long-term support and very identity. They have in their programmes and proposals always to stick to their core concerns, even though they may emphasize other issues more at any one election because of their immediate importance. It follows from Proposition 4, which makes this point, that parties can almost always be placed in a family in terms of the issues which they (in common with other family members in other countries) uniquely emphasize in their programmes. Another consequence of such an emphasis is noted in Proposition 5. This is that party preferences on ‘its’ issues will always be more extreme than those of the overall majority in a democracy which is partly formed by supporters of opposed parties. The median or ‘average’ citizen is thus likely to find herself between the parties, round about the centre of any section of a policy space marked out by their positions. This is a long-term consequence of the way parties develop and their ties with particular groupings in the population, built into the Alternation Theory of Party Representation presented in the ‘Matching policy to preferences’ section. A more immediate prediction from the Developmental-Family theory which can be used to check its plausibility against actual evidence, is Proposition 4. That specifically expects party programmes to be distinguishable in terms of their particular issue concerns – which in turn should permit clear identification of their family affiliation. As expected, 80 per cent of individual parties across Europe can indeed be successfully assigned to ideological families and groupings on the basis of the characteristic issue concerns emphasized in their election programmes (Klingemann et al. 2006, pp. 28–50). This view of party development and continuing ‘family’ concerns as generating a relatively extreme policy position in respect to other parties and the general population, creates the ‘Paradox of Democracy’, which forms a realistic starting point for the alternation account of democratic representation outlined in the section ‘Matching policy to preferences’ below. A further contribution which developmental theory makes to explaining democratic representation is stated in Proposition 1. All significant opinion groupings in a democracy have their concerns catered for in some (or several) parties’ programmes for action. The implication is that the centrist position taken up by the popular majority vis-à-vis the parties is a genuine reflection of their policy preferences, and does not result artificially from excluding some of them from political debate. This is obviously an important consideration in assessing the quality of democratic representation. Table 9.2 gives a broad overview of the major party families with their characteristic issue concerns as emphasized in their election programmes. The class-based parties in the bottom half of the table can be generally characterized in terms of a Left–Right continuum, opposing welfare concerns, government intervention and peaceful internationalism on the left to free markets and internal and international security on the right. Christian Democrats mix these appeals, so are often classified as Centrist, between Left and Right. However, what distinguishes them most clearly are their religious concerns for traditional morality and the stable social framework this engenders. Minority ethnic parties, like the Christians, arose alongside class-based parties in the nineteenth century. With their concerns focused on decentralization of powers or even political

114  Research handbook on political representation independence for their own region, they are even harder than religious parties to classify in Left–Right terms, as are Agrarians representing the interests of farmers, small and large. Green ecological parties are more recent entrants to electoral politics, but more inclined to ally themselves with the Left in demanding government intervention against climate change. However, it is their ecological focus which uniquely distinguishes them. Sociocultural parties with their more tightly focused single issue concerns are more easily characterized on their basis than the bigger parties which define their position in broad Left– Right terms and have to spread their electoral appeal more widely. Nevertheless, these can still be placed quite accurately in their broader party grouping. Table 9.2

Distinguishing issues which each party family uniquely emphasises in their manifestos: and the extent to which they can be clearly identified by these

Party-Families

Characteristic Issues

Predictive Efficiency

Sociocultural Parties Ecologists

Environmental Protection

0.938

Agrarian

Agriculture

0.898

Ethnic

Decentralization

0.835

Traditional Morality including respect for

0.854

Religious Parties Christian Parties

law and order Mainstream (Left–Right) Parties Left Socialists

Peace and Detente

0.482

Social Democrats

Welfare State: Expansion

0.516

Liberals

Market Economy

0.638

Left Conservatives

Market Economy

0.631

Right Conservatives

Military Strength

0.511

State Nationalists

Military Strength

0.543

Notes: The table reports those issues which each party uniquely stresses. The third column reports on how well such emphases enable individual national parties to be placed in the appropriate family, on a scale from 0.000 to 1.000. Source: Table adapted from Table 3.7 in Klingemann and Budge (2013).

All these parties push their own policy concerns when they are in government, even as a junior coalition partner. So the alternation and thermostatic models of policy representation apply to single issues as well as to general Left–Right ones (as described in the section ‘Matching policy to preferences’ below). A final point arising from this discussion of party concerns and origins is that all the parties discussed, apart from those originating from established elites, businesspeople and landed proprietors (i.e. Liberals and Conservatives) were regarded as illegitimate and potentially revolutionary when they first appeared (as many minority ethnic parties still are). Democracies have on the whole been able to give all of them space and integrate them into the system over time. This cautions us against regarding current populist parties, whether of Left or Right, as existential threats to democracy. Rather they form part of the continuous process of party creation, giving groups that feel excluded an opportunity to extend the policy alternatives to be voted on – just as the older parties did in their day.

Party-based representation: the paradox of democracy  115

ELECTORS, PARTIES, GOVERNMENTS: INTERACTIVE PREFERENCE FORMATION A weakness of the representational literature is that popular policy preferences are often taken as springing directly out of electors’ political cogitations into their voting choices or replies to opinion polls. Figure 9.1 shows instead how difficult it is for citizens to translate their private and family concerns about personal security and goods and service delivery into public policy. Should the government intervene in a foreign war? Or spend more money on Medicare? How does this relate to personal security and goods and service delivery? Citizens are neither diplomats nor doctors, so how can they decide between policy alternatives about which even experts are unclear? Under democracy the People decide. But how can they be helped to speak? It is here that parties and the governments formed by them come in, by offering policies and proposals which the citizens can decide between. Figure 9.1 specifies how.

Figure 9.1

How electors’ preferences are bounded and defined by political parties in general elections (3) and between elections (7)

The boxes in the figure represent the various choice situations that voters and electors find themselves in. Electors’ private preferences (box 1) are relatively unconstrained politically as they are ‘framed’ by their personal and family needs and ambitions (mainly for good public service delivery and personal security). Immediately they are expressed in the public domain

116  Research handbook on political representation (box 2) however, they have to be translated into collective political alternatives. Electors who go on to vote in general elections then have to relate their preferred courses of action to the policy priorities parties commit themselves to (among other things) (box 3). These priorities are defined by party policy concerns (box 4) and their history (box 5). Electoral preferences – private and public – have to be sharply distinguished from voters’ election choices between parties’ policy alternatives. The aggregated votes that parties receive in an election (box 6) are publicly declared and recorded. They form a record of the last general election, which takes on a life of its own independent of electors’ current preferences, giving government parties a ‘mandate’ for effecting their own policy targets and priorities (box 7). These are operationalized as specific proposals for change in current policies (e.g. for cutting welfare payments by 10 per cent). This then provides a specific context within which electors can formulate a preference for or (more usually) against the specific proposal, (box 2), leading to general discussion and often protests and criticisms, which result in it being modified. While the government parties lose support, they continue to be associated with the proposal, which goes into their record (box 5) and affects evaluations of them and voting in future elections. In the next section we will show how this overview leads on to the government alternation and thermostatic theories, which explain how public policy is brought into line with popular preferences for it in the long run. The important point to note here is that parties are essential in making voters and electors think about policy in the first place (if they do think about it – many other factors enter into voting decisions, as the figure illustrates). They also define the possible courses of action to be chosen on each issue (cutting or increasing expenditure for example). The danger of circularity in the process (preferences being imposed by parties on electors rather than public policy reflecting their ‘real’ preferences) is avoided by electors’ freedom to create new parties, which then offer a wider frame for voting decisions.

MATCHING POLICY TO PREFERENCES – THERMOSTATIC REACTIONS AND GOVERNMENT ALTERNATION The open-ended nature of general elections – allowing in many other influences on voting than policy, means that governments are formed by relatively unresponsive parties with policy agendas that are more extreme than most electors want. How then can they be brought into line with each other (and guaranteed to be in line) – as democracy requires? A first step in answering this question is distinguishing between policy targets and the actual policies governments are implementing at any one time. The two are often conflated as though a new government can change everything at a stroke. A moment’s reflection, however, shows that new policies can actually only be effected well into its first term and only partially after that. Legally binding contracts and entitlements have to be respected, new bureaucracies created, and legislative and judicial processes carried through. In most areas governments can only hope to move policy towards their target for it at a slow rate of 10–15 per cent of the gap each year (Budge et al. 2012, pp. 153–8). This gap is reinforced by the thermostatic reactions denoted in Figure 9.1 by the arrows linking box 7 (government policy proposals) to box 2 (elector policy preferences). Such proposals usually set off hostile popular reactions, for two reasons. If service cuts are proposed, electors usually feel too little is being provided and oppose the cuts. If spending is to be increased, electors are liable to feel that too much is being provided and should be spent

Party-based representation: the paradox of democracy  117 elsewhere or used to reduce taxes (Soroka and Wlezien 2010). Hostile reactions are reinforced by the fact that those electors losing by the change will react more strongly against it than potential beneficiaries will support it, since immediate losses are felt more strongly than distant benefits. Such negative ‘thermostatic’ reactions have two effects: first adding to the other factors slowing policy change down, and second, bringing actual implemented policy closer to popular preferences as the government is forced to compromise. The effects are even stronger where government proposals are put to a referendum or initiative. The second overarching process bringing implemented policies closer to centrist popular preferences are the changes in party governments brought about by the unpredictable effects of general elections. As new governments slowly attempt to modify policies actually being implemented in line with their own targets, they bring their implemented policies (initially anyway) closer to the centre where most electors locate themselves. Policy changes then carry on towards the more extreme position the new government prefers. Before actually implemented policy moves too far to the extremes, however, an opposed government is likely to be elected and reverse the changes. Over time therefore, actual policy tends to oscillate around the centre where the majority electoral preference lies. Figure 9.2 illustrates this process schematically for Left–Right divisions over policy (but it could equally apply to single issues such as financing church schools). The majority or plurality parties which dominate most governments have relatively extreme targets which they rarely have enough time to effect fully (periods when they are in government are shown by thicker sections of their policy lines). By trying to do so however, they keep implemented policy oscillating around the position of the median elector, which marks out the preferred position of the citizen majority. Of course such oscillation will only occur if the two major parties’ policy positions fall on each side of (‘bracket’) the median majority elector position. This is likely to be the case in most democracies as electors’ preferences will naturally spread out between the positions they provide as a frame. Alternation thus shifts the focus of representational theory from congruence between preferences and (declared) government policy in each election, viewed as a discrete event, to how the electoral process as a whole brings about a dynamic correspondence between the public policies actually implemented and preferences over time. The prerequisites for doing so are: (1) regular alternation of the major parties in government with relatively rigid, ideologically fixed policy targets; (2) a slow rate of policy change; (3) party bracketing of the median citizen policy preference. These conditions do not unrealistically require parties to trade policy positions for votes – creating possible confusion and mistrust by doing so. Nor do they unrealistically ignore the open-ended nature of general elections, where many factors other than policy affect voting choices and outcomes. All that is required of elections is that they regularly alternate parties with different policy positions and priorities in government. How far is this ‘alternation theory’, plausible in itself, upheld on existing evidence? On the basis of data available for nine democracies (the US and eight European countries), actual graphs similar to Figure 9.2 have been created. Seven of these broadly match the patterns shown in that figure. That is, plurality or majority parties bracket the median elector, governments alternate and policy expenditures change slowly. As expected, where such conditions are present, the policy expenditure line is close to median elector preferences and comes closer over time. The two remaining democracies upheld theoretical expectations in another way –

118  Research handbook on political representation bracketing of the median elector preference by plurality parties was absent and, as expected, preference–policy correspondence was much less (Budge et al. 2012, pp. 162 and 230–41). Political parties are of course carriers of many specific policy priorities and positions as well as general Left–Right ones (Table 9.2). However, government alternation applies in exactly the same way, oscillating implemented policies around the centrist position taken up by most electors (Budge 2019, p. 65).

Figure 9.2

Relating public policy to citizens’ left–right preferences for it over time

Notes: The figure shows parties ‘bracketing’ the median preference over time (the horizontal dimension), in the sense that their policy positions consistently fall right or left of the median policy preference. As the parties alternate in government, actually implemented policy veers slowly to the left or right. Actual policy remains, as a result, generally close to the median position.

LONG-TERM PARTY DECLINE? As this discussion has shown, ideologically rooted parties with contrasting policy proposals are central to the whole process of democratic representation, whether ‘thermostatically’ or by government alternation. The question of whether they (or at least their ideological differences) are disappearing becomes, therefore, of central concern to the future of democracy itself. With the proliferation of new parties in recent years – Greens, State and minority nationalists, Left Socialists – the question of a general party weakening might seem academic. However, general party decline is often conflated in the literature with the weakening of established parties like the Social Democrats, so new parties (‘populists’) are often seen as threats to the stability of the central party system and its representational functions rather than as invigorating it with new blood. A further tendency of the general party literature is to base diagnoses of what is happening to political parties on single dramatic events, generalizing this to all parties and their policy differences. Such diagnoses – the rise of Rightist or Leftist ‘authoritarian’ populism or the End of Ideology for example – then dominate discussion for a few years before being displaced by another, often totally opposed, diagnosis. Table 9.3 graphically illustrates this process by listing popular diagnoses of party health over the post-war period, showing how one generally

Party-based representation: the paradox of democracy  119 accepted diagnosis is often replaced by an opposed one in the course of a few years. Their common characteristic is to be so vaguely formulated as to seem to cover everything that is happening or going to happen to parties at that point without really being put to any test. Discussion is then left in the air when another general diagnosis becomes popular. Thus, the idea that Left-wing parties were abandoning their ideology while their ‘catch-all’ rivals were collecting votes wherever they could, did not hold up when checked against what they actually said in party platforms and manifestos. Quantitative analyses showed them all aligned along Left–Right lines as before (Budge et al. [1987] 2008, pp. 392–9). The medley of often conflicting diagnoses of the state of parties shown in Table 9.3, from a Social Democratic consensus to a Neo-Liberal one 20 years later, or from the triumph of democracy ‘ending history’ in the 1990s, to the threat to democracy posed by Rightist Populism in the last few years, discredits current predictions of party weakness and decline. Instead what we see is party democracy bumping along its usual rocky road of unexpected crises and events, but continuing to function much as normal, with a certain ability not just to absorb but to respond to new parties and issues – and to citizen preferences, more or less as it has done throughout the post-war period. We should not be complacent about how parties are doing at present but we should not be too alarmist either. The best contribution political science can make to political debate is to understand democratic processes better through serious theorizing and analysis, as provided by the developmental, thermostatic and alternation explanations reviewed above, rather than the facile diagnoses and generalizations shown in Table 9.3. Table 9.3

Popular diagnoses of parties’ current and future condition 1950–2020

Diagnosis

Period

Thesis

‘Social Democratic Consensus’

1950–1970

Politics is only about managing the welfare state, not challenging

‘End of Ideology’

1960–1980

it, so party differences are minimal Political parties all now ‘pragmatic’ rather than ideological, particularly Leftist parties. Rise of ‘Catch-All’ non-ideological parties Neo-Corporatism

1970s

Governments have to negotiate policy with unions and business which have a veto so parties cannot act freely to effect their policies

‘Neo-Liberal Consensus’

1985–2015

Free markets rule so Leftist party governments provoke a crisis if

‘End of History’

1990–2000

[Neo-Liberal] democracy triumphs everywhere

Democratic Deficit

1990–2010

Parties hinder rather than aid expression of popular preferences

Leftist Populism

2010–2015

Destabilizes established parties and threatens stability of party

Authoritarian (Rightist) Populism

2016 onwards

they intervene to impose their policies

systems Parallels with 1930s; ‘lonely crowd’ turns to strong leaders, fascism and racism

120  Research handbook on political representation

REFERENCES Albright, M. (2018), Fascism: A Warning, London: Collins. Budge, Ian (2019), Politics: A Unified Introduction to How Democracy Works, Abingdon, UK and New York, NY, USA: Routledge. Budge, Ian, David Robertson and Derek J. Hearl (eds) ([1987] 2008), Ideology, Strategy and Party Change: Spatial Analyses of Post-War Election Programmes in 19 Democracies, Cambridge: Cambridge University Press. Budge, Ian, Hans Keman, Michael D. McDonald and Paul Pennings (2012), Organizing Democratic Choice, Oxford: Oxford University Press. Crewe, Ivor and David Sanders (eds) (2020), Authoritarian Populism and Liberal Democracy, London: Palgrave Macmillan. Klingemann, Hans-Dieter and I. Budge (2013), ‘Using the Manifesto Estimates to Refine Family Placements’, in Andrea Volkens, Judith Bara, Ian Budge and Michael D. McDonald (eds), Mapping Policy Preferences from Texts, Oxford: Oxford University Press, pp. 49–65. Klingemann, Hans-Dieter, Andrea Volkens, Judith Bara, Ian Budge and Michael D. McDonald (2006), Mapping Policy Preferences II, Oxford: Oxford University Press. Powell, G. Bingham Jr (2000), Elections as Instruments of Democracy, New Haven, CT: Yale University Press. Rokkan, Stein (1975), ‘Dimensions of State-Formation and Nation Building’, in Charles Tilly (ed.), The Formation of National States in Western Europe, Princeton, NJ: Princeton University Press. Soroka, Stuart and Christopher Wlezien (2010), Degrees of Democracy, Cambridge: Cambridge University Press.

10. Corporatism and representation Alan Siaroff

Corporatism is the formal role of (economic) interest groups in socio-economic policy-making, or as Lehmbruch (1979, p. 150) defines it, an ‘institutionalized pattern of policy formation in which large interest organizations co-operate with each other and with public authorities … in the implementation of such policies’. It thus differs from pluralism in that corporatism involves not only the representation of interests but the formal involvement of organized interests in decision-making and policy implementation, including the representation of recognized interest associations in public regulatory agencies (Williamson 1989, p. 213). These economic interest groups are hierarchical in nature and are monopolistic in terms of speaking for and aiding in the implementation of policies relating to these economic interests (Williamson 1989, pp. 209, 214), in contrast to the competing groups found under pluralism. Given that there is representation by groups – indeed specific groups – and not by individuals, corporatism would seem to overlap with or relate to collective and not individual patterns of representation. As Armingeon (2002, p. 143) notes: Consociational democracy and corporatism denote styles of decision-making of various actors in a democracy. Both styles share the element that decisions are usually not reached by plurality of votes or other power resources. Rather the actors seek to integrate a large number of actors or societal groups into a decision by negotiations and mutual concessions, respecting the autonomy of the various actors and their groups. Hence corporatism and consociational democracy are the major empirical instances of ‘negotiation democracies’.

Certainly the works of Lijphart (most recent version 2012) on majoritarian (Westminster) versus consensus democracy group pluralist versus corporatist interest group systems in an ‘executives–parties dimension’ along with cabinet types, executive–legislative relationships, party systems and electoral systems. However, this grouping has been criticized by Taagepera (2003), who does not see an empirical connection involving corporatism here (and for whom all other aspects of this dimension involve or follow from the electoral system). Consequently, Bernauer et al. (2014) exclude it from their analysis of dimensions of democracy. In his examination of parliamentary systems (only), Siaroff (2003) found pluralism to be part of a ‘fused parliamentarianism with policy centralization’ factor. That might imply a greater likelihood for corporatism in non-parliamentary systems, which is certainly true for Switzerland, though not for presidential systems it seems. In this analysis, the focus will be on the assumed affinity between corporatism and group representation, crucially in the context of ongoing coalitional party government in which there is a relevant party of the left. The chapter will begin by focusing on the concept and measurement of corporatism in democracies. Second, it will then show where (in which countries) corporatism has existed, both durably and for short periods. Third, it will review the aforementioned debates about how corporatism fits (or does not) in types of democratic systems. Fourth and finally, it will argue that coalitional party government in a party system with a relevant left party and/or initially a cross-class Christian democratic one is an addi121

122  Research handbook on political representation tional factor that facilitates corporatism, and logically relates to it in terms of representation by groups and policy compromise. However, this is but one route to corporatism, the other being left-wing dominance, so the affinity of corporatism with coalitional party government and group representation is limited. Some specific discussion will occur of countries which illustrate these points.

CORPORATISM: WHAT IS IT AND WHICH COUNTRIES HAVE IT? Corporatism, as noted, has at its core the notion of institutionalized tripartite (state, business, labour) policy formation and implementation in economic matters (Lehmbruch 1979, p. 150). However, the use of the concept can be broader or narrower. Detlef Jahn (2016) argues for a narrower use of the concept. His use of the concept involves three aspects: the first is structure; that is, the degree of hierarchical centralization; the second is form; that is, the degree of concertation with the state; and the third is scope; that is, the degree to which agreements encompass broader segments of society (Jahn 2016, p. 53, table 1). Given data constraints, his measurements focus on trade unions (using data from Visser 2011). Structure is measured in terms of the organization of trade unions, the structure of work council representation, and the rights of work councils. A high score occurs where the trade unions are centralized (into peak associations) and hierarchical, including the national unions having a direct impact on works councils. Function is measured in terms of wage bargaining and the involvement of unions and employers in government decisions. A high score occurs when governments participate in tripartite wage bargaining and do not impose wage settlements, where wage bargaining takes place at the national or central level, and where unions and employers are routinely involved in government decisions on social and economic policy. This last aspect can be seen as part of the more general participation of non-governmental actors in the preparation of legislation and regulations. Lastly, scope is measured in terms of the coordination of wage bargaining and the mandatory extension of collective agreements. A high score occurs when bargaining is economy-wide and based on enforceable agreements between peak associations or, failing that, key unions, and where mandatory extension of collective agreements by law to non-organized firms exists, is regularly applied, and affects a relevant share of the workforce (Jahn 2016, pp. 53–7). Jahn’s analysis is for a broad range of 42 countries and where possible for the entire period from 1960 to 2010. Factor analysis gives him a single z-score standardized summary variable of his eight measures (Jahn 2016, p. 58). For our purposes we are interested in countries that sustain a standardized score of at least 0.5 on his measure, as that is the cut-off here for (at least moderate) corporatism. Essentially there are four types of countries here: (1) those that are continually corporatist, either for the entire period of 1960 to 2010 (for example, Austria) or in the case of South Africa for the years measured; (2) conversely those never corporatist (for example, Bulgaria or Canada); (3) those initially non-corporatist but then becoming corporatist in the time period noted and remaining corporatist (for example, Denmark and Finland); and (4) those corporatist for a specific time period but where the corporatism ends (for example, Australia). Israel is the only country in this last group that began as corporatist right from 1960. One can note that many countries in group (3) have been corporatist for decades, making them similar to group (1). One should also note that given Jahn’s focus on organized labour, countries sometimes called ‘corporatist without labour’ such as Japan (Pempel and Tsunekawa 1979) and Switzerland do not score as corporatist.

Corporatism and representation  123 A broader measure of corporatism would be Siaroff’s (1999) measure of integration, which includes structural features, functional roles, behavioural patterns, and favourable contexts. Jahn (2016, p. 51) is especially critical of the last set of factors, which admittedly are more causal than definitional but the goal there was to create an ideal type. Indeed, Siaroff’s (1999, p. 179) contextual factors of ‘a consensual or even consociational political tradition, rather than a “majoritarian” one’ and ‘a long-term political role or even dominance of a united social democratic party’ are central to this analysis. The countries scored at least 3.0 out of 5 by Siaroff – thus indicating (at least moderate) corporatism – would qualify as corporatist in a dichotomous sense. Siaroff has mostly the same classifications as Jahn, even for countries with shorter periods of corporatism such as Australia and post-cold war Italy, but there are some important differences. Siaroff classifies Switzerland and Japan as corporatist, which Jahn does not. Jahn has Ireland as corporatist from the late 1990s, which Siaroff does not (though Siaroff has a higher score for Ireland in the late 1990s than previously for that country, and his analysis ends then). Jahn has corporatism in Israel ending in 1980, and corporatism in Luxembourg not really starting until the late 1970s; Siaroff has neither of these temporal points. There is also a slight time difference in the start of corporatism in Finland. Lastly, Jahn has both Greece and Spain as corporatist in the 1980s (a bit earlier in Spain), which Siaroff does not. This difference is partly due to Siaroff including strike levels which, though declining, were still comparatively high. Some of these differences are minor and may reflect the breakpoints of 0.5 for Jahn and 3.0 for Siaroff; others are more significant though. One could possibly consider as corporatist a country scored as such by either Jahn or Siaroff, but certainly the differences on Switzerland and Japan are conceptual. Concerning Israel, Mundlak’s (2007, p. 4) case study of Israeli labour law suggests 1987 as ‘the pivotal year’ in the shift from corporatism to pluralism – so later than Jahn but still ending. Indeed, the opposite approach from Jahn would be the varieties of capitalism distinction between a coordinated market economy and a non-coordinated liberal market economy of Hall and Soskice (2001), as this is especially focused on firms. Their classifications are given for 17 countries, and they also consider Japan and Switzerland to be coordinated. Overall, then, the European countries where there is consensus (or at least no disagreement) on long-term corporatism are Austria and Germany, the Benelux countries, the Nordic countries except for Iceland, and Slovenia. Conversely there is consensus on the long-term absence of corporatism in France, Malta, Portugal, the United Kingdom, and post-communist Europe outside of Slovenia. Short-term corporatism has existed in Italy and, following Jahn, in Greece and Spain. Disagreement exists with regard to Switzerland, where collective agreements have been sectoral and not ‘ideal-type corporatist’ national (Kriesi and Trechsel 2008, p. 107) but it has had a clear role for interest groups in policy-making – indeed this is now constitutionalized (Kriesi and Trechsel 2008, p. 118). Outside of Europe, corporatism in democracies has been much rarer, and there is consensus on the long-term absence of corporatism in Canada, New Zealand and the United States. Corporatism did exist for a longer period in Israel (see below) and for a short period in Australia. Jahn (2016, p. 62) also classifies Singapore and South Africa as having ongoing corporatism, though, as he notes, Singapore is not a democracy. Disagreement exists with regard to Japan, but certainly Japan lacks any national role for unions in policy-making (unlike Switzerland), which would make Japan pluralist. As for Lijphart (2012), he mostly follows Siaroff, but he also assesses developing countries and places Mauritius and Uruguay as cor-

124  Research handbook on political representation poratist, and Barbados as corporatist after 1993 (through 2010). Lijphart includes Papua New Guinea and Venezuela, but they are no longer democratic so are excluded here.

CORPORATISM AND TYPES OF DEMOCRACIES Lijphart in his writings famously contrasts two types of democracies: Westminster versus consensual, with the former being competitive and majority (effectively plurality) take-all and the latter involving power-sharing and negotiation. Likewise, Armingeon (2002) has used the term ‘negotiated democracies’. Our focus is on Lijphart’s first, executive-parties, dimension, which has five variables: cabinets, executive–legislative relations, party systems, electoral systems, and interest group systems. For Lijphart (2012, p. 60) though, the first difference here between single-party majority governments and broad multi-party coalitions ‘can also be seen as the most important and typical difference between the two models of democracy’. That said, he first discusses differences in party systems as the classification of cabinets ‘depends a great deal’ (2012, p. 61) on the definition of party systems. Indeed, Ganghof (2010, p. 683) criticizes Lijphart’s focus on single-party majorities above all, as this does not add anything systemic beyond the effective number of parties (as a high effective number of parties correlates even more with single-party minority governments, not coalition governments). As Ganghof (2010) notes, following Taagepera (2002), all we can really say is that when the effective number of parties is above four then single-party majorities are mathematically impossible. Perhaps, then, a better way of looking at these factors of party systems and cabinets, and obviously key for our purposes, is in terms of the breadth of group representation. In a multi-party system, there can be a party for specific groups (such as Catholics) or specific issues (such as environmentalism). Of course, the fit between issues represented and number of parties is not perfect (see Ganghof et al. 2015, p. 68, as cited in Ganghof and Eppner 2019, p. 127). Still, groups as groups can thus be specifically represented politically. Moreover, this representation can be both in parliament and in cabinet (if the latter is broad). In contrast, certainly in a two-party system, but even in a moderately multi-party system, each party (or at least each major party) is targeting a range of groups in a ‘brokerage’ way, giving each such group something but not focusing exclusively on it unless the focus is regional (such as Canada or Spain through 2011). The United States is a classic case of broad would-be voter majority coalitions in each of its two parties. To that end, one can note that Benoit and Laver’s (2006, p. 116) factor analysis finds that only a single underlying dimension is needed to describe the policy positions of the United States party system. More generally, Ganghof and Eppner (2019) argue that Lijphart’s executive–parties dimension actually speaks to approaches to achieving accountability and representation rather than the extent of consensus. One could have a multi-party system without much consensus between the parties; conversely one of Lijphart’s executive–legislative relations measures is cabinet duration which would score a durable cabinet as majoritarian – but said cabinet could be multi-party if said parties are indeed consensual. Lijphart (2012, p. 119), though, does not classify Switzerland based on its partisan coalitions but apparently on its annual change of head of state (or at least an equivalent low score). We shall suggest an alternative approach to consensus below. Of course, the trade-off for broad representation in this sense is less accountability and little identifiability of alternative governments during elections, specifically in terms of single-party

Corporatism and representation  125 government or even an allied two-party coalition (Carey and Hix 2011, as noted in Ganghof 2016, p. 221). Majorities under such systems of ‘complex majoritarianism’ are formed post-election, in cabinet negotiations with options not clearly identified pre-election, or even, in a country such as Denmark, formed over differing individual issues (Ganghof et al. 2015). These differences do go back to alternative electoral systems. As Carey and Hix (2011, p. 395) summarize: Consistent with the traditional view of electoral systems in political science, we find that SMD systems tend to produce a small number of parties and simpler government coalitions, but also have relatively unrepresentative parliaments. On the other side, electoral systems with large multimember districts have highly representative parliaments, but also have highly fragmented party systems and unwieldy multiparty coalition governments. In contrast, electoral systems with small multimember districts – with median magnitude between four and eight seats, for example – tend to have highly representative parliaments and a moderate number of parties in parliament and in government (emphasis in original).

The last type of electoral system is their preference but it does have less broad representation (especially in cabinet, see below) than large multi-member proportional representation. Moreover, there is no durable corporatism in most of Carey and Hix’s (2011, p. 384) ‘sweet spot’ countries – their examples being Costa Rica, Hungary, Ireland, Portugal and Spain.

CORPORATISM AND COALITIONAL PARTY GOVERNMENT Consequently, for breadth of representation one should measure the number of parties represented in a parliament and likewise in a cabinet. Lijphart (2012, pp. 33–4) mentions the second aspect specifically in his illustrations of the consensus model of democracy in Switzerland and Belgium, but does not use this measure systematically, relying instead on measuring temporally the opposite Westminster reality of one-party and bare-majority cabinets. We shall calculate these values for European countries (those in the EU or EFTA), other advanced industrial countries, and other democracies analysed by Jahn and/or Lijphart. Table 10.1 thus shows these values (the number of such parties) for all these countries. However, for this aspect, the analysis of party systems (or at least pattern if not yet a system) and related governments starts with the first post-war election and goes through to January 2020. It should be noted, though, that a few, mostly Nordic, countries have had their party systems/patterns shift to a highly multi-party type (from either a moderately multi-party or a two-and-a-half-party type). Given that the latter is much more likely to lead to coalitions, these countries (Belgium, Denmark, Iceland, India, Luxembourg, Norway, Spain and Sweden) are divided into two time periods. The second time period begins with the first election producing a highly multi-party pattern. Also, given the complete transformation of the Italian party system, it is divided into two time periods, with the second period starting in 1994.

126  Research handbook on political representation Table 10.1

Coalitional systems? Median P2%S

Median Parties in Government

Most Coalitions Cross-Class?

Austria

4

2

yes

Belgium until 1968

4

2

yes

Belgium since 1968

8.5

4

yes

Bulgaria

5

1

Croatia

4

2

no

Cyprus

4

2

no

Czech Republic

5

3

no

Denmark until 1973

6

1

Denmark since 1973

9

2

Estonia

6

2

no

Finland

7

4

yes

France Fourth Republic

5

4

no

France Fifth Republic

5

2

no

Germany

4

2

no

Germany since 1966

4

2

yes

European Countries

Greece

no

4

1

4.5

2

no

Iceland until 2016

5

2

no

Iceland since 2016

7

3

no

4.5

1

Italy until 1994

7

3

no

Italy since 1994

8

4

yes

Latvia

6

4

no

Liechtenstein

2

2

no

Lithuania

6

3

no

Luxembourg until 2018

5

2

yes

Luxembourg since 2018

7

3

yes

Malta

2

1

Netherlands

8

3

Norway until 1997

5

1

Norway since 1997

6.5

3

no

Poland

5.5

2

no

Portugal

4.5

1

Hungary

Ireland

no

Romania

5.5

2

Slovakia

6

3

yes no

Slovenia

7

4

yes

Spain until 2019

5

1

Spain since 2019

7

1

Sweden until 1991

5

1

Sweden since 1991

7

2

no

Switzerland

7

4

yes

United Kingdom

3

1

Corporatism and representation  127 Median P2%S

Median Parties in Government

Most Coalitions Cross-Class?

Australia

3

2

no

Canada

4

1

Israel

9

6

Japan

5

1

New Zealand

2

1

United States

2

1

Bahamas

2

1

Barbados

2

1

Botswana

2.5

1

Brazil

11

6

no

Chile

7.5

4

no

Colombia

4

1

Costa Rica

4

1

India until 1996

6

1

India since 1996

9

5

Jamaica

2

1

Mauritius

3

3

Mexico

4

1

South Africa

4

1

South Korea

4

1

Trinidad and Tobago

2

1

Uruguay

4

1

Other Advanced Industrial Countries

no

Other Countries

no yes

For our purposes the absolute number of parties is used, rather than the effective number, but this is limited to those parties with at least 2 per cent of the seats (P2%S) following Siaroff (2019, pp. 31–2). Table 10.1 provides P2%S values taken from Siaroff (2019), with updates and values for non-European countries added. These range from various two-party systems through moderately multi-party ones to highly multi-party systems, with several countries changing from moderately to highly multi-party, as noted. Table 10.1 then gives the median number of parties in government for each country/time period. To calculate this, each government is weighted equally, and non-partisan and caretaker governments are scored as 0 (parties). Here, the key distinction is between countries whose typical government is single-party and those which are typically coalitional. That said, certain multi-party and multi-party coalitional countries are pluralist (for example, Brazil, India and Latvia). Yet, for corporatism, a numerically broad coalition is not the issue as much as the inclusion of both business and labour, though this is more likely with a numerically broad coalition. Thus, what we ultimately want to identify is the pattern of a cross-class coalition, and moreover one including the main party of workers. To assess this, parties are divided into left (communist/left-wing populist, left socialist/new left, green, socialist/social democratic, left/ radical/social liberal, and national populist social democratic) parties and right (Christian democratic, right liberal, conservative, nationalist right-wing, populist radical right, and extreme right-wing/neo-fascist) parties, with agrarian and centre parties being classified in the middle (see Siaroff 2019, pp. 11–15 for these categories). The question then is whether most

128  Research handbook on political representation coalitions (of all coalitions, not all governments) in a country/time period (a) have a party from each of the left and right (so a social democratic – agrarian government would not count), and (b) whether the left party (or one of these) is the main party of workers, be this party socialist/ social democratic, left socialist (Iceland until 1999), communist/left-wing populist (Cyprus, Finland through the 1960s, France through 1981, Greece with SYRIZA, Iceland, Italy through the end of the PCI), or national populist social democratic (Slovakia, Bulgaria and Romania eventually). Only a few countries/party systems have had such cross-class coalitions: Austria, Belgium, Finland, Italy since 1994, Luxembourg, Romania and Slovenia. One can also add here Germany since 1966 (when the Social Democrats entered government for the first time in the post-war period), though this was not a new party system. Of these cross-class coalitions, Austria, Belgium through 1965, Germany since 1966, and Luxembourg involved two-party coalitions, with the rest being multi-party coalitions. Most importantly, these countries have all been corporatist except for Romania, and so it seems that a national populist social democratic party of workers does not facilitate corporatism. Nor do left-wing populist parties such as SYRIZA in Greece. Again, it is apparently not the number of parties in a coalition per se (Table 10.1) that matters for corporatism but that the coalition is cross-class.

TWO ROUTES TO CORPORATISM In fact, there are two different routes to corporatism. The first route, which is broader and more durable, involves cross-class coalitional party government. This route normally/eventually needs the main workers party to be relevant, that is, being: (1) one of the top two parties by seats and (2) being in the government – even if not often (or even ever, as in Luxembourg) leading the government. In this first route, though, a Christian Democratic party with a clear cross-class commitment – as has existed in Germanic Europe – can facilitate the start of corporatism even before the main workers party enters government (for example, Germany), with the latter aspect of cross-class coalition governments reinforcing corporatism. So in Germany this was done by the CDU which had a labour wing, and which was competing with the SPD for power. Consequently, the CDU-led coalition under Chancellor Adenauer brought in the Codetermination Act of 1951 and then the Works Council Act of 1952. As Korpi (2006, pp. 195–7) notes, the comparison here is with post-war Italy, which was not corporatist and where the PCI was never relevant in our terms: though the second-largest party in parliament, the PCI not only never led a national government, it was never even in a national government as it was always excluded from power (whereas the PCI’s post-cold war successor parties would lead governments, and Italy’s brief period of corporatism would occur then). As Armingeon (2002, p. 156) summarizes (drawing from Keman and Pennings 1995): ‘the [continental European] heartland of consensus-democracy comprises politically-fragmented countries with weak or organizationally fragmented trade unions and relatively weak social democratic parties – for example The Netherlands, Belgium or Switzerland.’ Relatively weak social democratic parties indeed (as noted below in Table 10.2, scoring ‘low’ on leading governments), but apparently not extremely weak or non-existent ones. In these countries there is definitely an overlap between corporatism and broad negotiated representation, stemming from a longstanding ‘politics of accommodation’ (Lijphart 1968) in heterogeneous societies. In the second route, corporatism ‘is not due to a political culture of accommodation but to the strength and power of the Left in parliament and in working life’ (Armingeon 2002,

Corporatism and representation  129 p. 157); that is, corporatism is a product of left-wing dominance, and representation reflects this dominance (often in single-party cabinets) rather than broad accommodation. For his part, Therborn (1992, pp. 36–7) makes a similar distinction of the countries with ‘relatively peaceful industrial orders’ (low strike volumes) between those with ‘an institutionalization of partnership and consensus’ of ‘social partners’ (the first route) and those with ‘an institutionalization … of conflict’ of autonomous labour market parties, with (potential) conflict being ‘orderly regulated positive-sum’ and government overseeing such regulation (the second route). For our purposes, left-wing dominance is defined as controlling the government usually in a single-party way over an extended period of time. This control could be coalitional too, but with social democratic dominance. Moreover, in recent decades in Sweden and Norway bipolarity exists in the form of pre-electoral coalitions so the left overall (rather than just the main left party) wins or loses. Table 10.2 speaks to the main party of workers in terms of both (potential) routes to corporatism. First, the table lists the specific parties, and notes where these are not social democratic/socialist. In a couple of cases no such parties on the left exist. Second, the table indicates the frequency of being one of the top two parties in terms of seats (or for presidential systems being one of the top two candidates in the first round of presidential elections) – classified into the following six categories: always; almost always (not always, but in at least 85 per cent of elections); usually (in 51 to 84 per cent of elections); sometimes (in 20 to 50 per cent of elections); rarely (has happened, but in fewer than 20 per cent of elections); and never. In a significant number of cases the main left party has always been one of the top two parties in parliament (or one of the top two presidential candidates, as in Brazil), and only in a few cases are such parties (assuming they exist) never or rarely in the top two. Again, this aspect is needed for the first route, along with the norm of cross-class coalitions (from Table 10.1). Table 10.2

Leftist strength Specific Main Party of Workers

Total Years Total Elections Under Analysis

One of Top Two

Time as Head of

Parties by Seats?

Government

medium

(socialist / social democrats unless otherwise noted) European Countries Austria

SPŐ

75

23

always

Belgium until 1968

BSP

22

7

always

low

Belgium since 1968

BSP / PS + SP

52

16

usually

very low

Bulgaria

BSPe

30

10

almost always

low

Croatia

SDP

28

8

usually

very low

Cyprus

AKELb

44

9

almost always

very low

Czech Republic

ČSSD

30

9

usually

medium

Denmark until 1973

SD

28

11

always

high

Denmark since 1973

SD

47

17

always

medium never

Estonia

M /SDE

28

8

never

Finland

SKDLb then SSDP

75

21

usually

low

Finland until 1966

SKDLb

21

6

sometimes

very low

Finland since 1966

SSDP

54

15

almost always

medium

France Fourth Republic

PCFb

13

5

always

never

France Fifth Republic

PCFb then PS from 1981

62

15

sometimes

low

Germany

SPD

71

19

always

low

Greece

PASOK then SYRIZAb

46

18

almost always

medium

Hungary

MSzP

30

8

usually

medium

130  Research handbook on political representation Table 10.2 (continued)

Leftist strength

Specific Main Party of Workers

Total Years Total Elections Under Analysis

One of Top Two

Time as Head of

Parties by Seats?

Government

(socialist / social democrats unless otherwise noted) Iceland until 2016

SPb / PAc / SDA

Iceland since 2016

SDA

Ireland

70

21

sometimes

very low

4

2

never

never

Labour

72

20

rarely

never

Italy until 1994

PCIb / PDS

48

12

almost always

never

Italy since 1994

PDS / DS / Daisy / PD

26

7

almost always

low

Latvia

TSP / PCTVL

27

9

rarely

never

Liechtenstein

(no such party)

75

22

(no such party)

never

Lithuania

LDDP /SDK /LSDP

28

7

usually

medium

Luxembourg until 2018

LSAP

73

15

almost always

never

Luxembourg since 2018 LSAP

2

1

never

never medium

Malta

MLP

54

12

always

Netherlands

PvdA

74

22

almost always

low

Norway until 1997

DNA

52

13

always

high medium

Norway since 1997

DNA

23

6

always

Poland

SLD

29

8

sometimes

low

Portugal

PSP

45

16

always

medium medium

Romania

FDSNb /PDSRe

30

7

always

Slovakia

HZDSd then Smer-SDd

28

9

always

high

Slovenia

SDP /ZL /ZLSD /SD

28

9

sometimes

very low

Spain until 2019

PSOE

43

13

always

medium

Spain since 2019

PSOE

1

2

always

very high

Sweden until 1991

SAP

43

14

always

very high

Sweden since 1991

SAP

29

8

always

medium

Switzerland

SPS/PSS

73

19

almost always

low

United Kingdom

Labour

75

21

always

medium

Australia

ALP

74

29

always

low

Canada

CCF /NDP

75

25

rarely

never

Israel

Mapai /Labour

72

22

usually

medium

Japan

JSP

73

26

usually

very low

Other Advanced Industrial Countries

New Zealand

Labour

74

25

always

low

United Statesa

(no such party)

72

18

(no such party)

never

Other Countries Bahamas

(no such party)

43

9

(no such party)

never

Barbados

(both main parties)

49

11

always

(very high)

Botswana

BNF since 1994

51

10

always

never

Brazila

PT

30

8

always

medium

Chilea

PSCh

30

7

sometimes

low

Colombiaa

PDA

46

12

rarely

never medium

Costa Ricaa

PLN

66

17

almost always

India

CPI (M)

68

17

never

never

Jamaica

PNP

58

12

always

medium

Mauritius

PTR

52

11

sometimes

medium

Corporatism and representation  131 Specific Main Party of Workers

Total Years Total Elections Under Analysis

One of Top Two

Time as Head of

Parties by Seats?

Government

(socialist / social democrats unless otherwise noted) Mexicoa

PRD

20

4

usually

very low

South Africa

ANC

26

6

always

very high

South Koreaa

DLP /UPP /JP

32

8

never

never

Trinidad & Tobago

UNC /PP

58

13

always

very low

Uruguaya

FA

35

8

sometimes

medium

Notes: presidential elections; communist/left-wing populist; left socialist; national populist social democratic; e communist initially then national populist social democratic. a

b

c

d

Table 10.2 then turns to whether the main party of workers is not just in government but indeed leading the government, with or without a coalition. Specifically, Table 10.2 speaks to this in terms of years in power (having the prime minister, or president in presidential systems), grouping countries/party systems into six categories: very high (above 80.0 per cent of the time); high (above 60.0 to 80.0 per cent of the time); medium (above 40.0 to 60.0 per cent of the time); low (above 20.0 to 40.0 per cent of the time); very low (above 0.0 to 20.0 per cent of the time); and never. This is a much harder criterion, as only Sweden in its first time period and South Africa score very high (actually all the time in the case of South Africa since 1994), and only Denmark and Norway in their first time periods and Slovakia score high. Several cases score as medium, such as Greece and Spain. Again, it is worth noting that some of these main left parties are not a traditional socialist/socialist democratic party. Specifically, be they left populist as in Greece or national populist social democratic as in Slovakia, these parties appear to be much less facilitating of corporatism in contrast to traditional socialist/social democratic parties in Scandinavia, Australia, Spain and so on. However, outside of the corporatist ‘heartland’ of Scandinavia (to use the term of Armingeon 2002, p. 156), said corporatism does not seem to last once such leftist dominance ends. We see this in the cases of Israel (which lasted the longest), Australia and Spain (and also Greece and Italy). As Mundlak (2007, p. 47) stresses on Israel, ‘[t]he fall of the Labor Party in 1977 and its very partial recovery since are symptomatic of the move away from the background conditions that are typical of or necessary to corporatism’. Consequently, both the finance and labor ministers from 1977 onward (again, with some exceptions) rapidly became detached from the corporatist system of social and political governance. The ministers were influenced by a different political climate, and they no longer viewed the corporatist system as an end in itself or as a particularly important instrument of governance. (Mundlak 2007, p. 47)

In Australia, corporatism arose due to the cooperation between the Australian Labor Party (ALP), which came to power in 1983 for what would prove to be its longest spell by far (lasting until 1996), and the Australian Council of Trade Unions. Prime Minister Bob Hawke was a former ACTU leader, which obviously facilitated matters. The key structure was the 1983 incomes policy Accord between these two, with business then being brought in. However, ongoing corporatism was conditional on these two leftist actors. Writing during the Accord period (which included its successor agreements), Archer (1992, p. 412) was prescient: ‘In the end, the only guarantee of Australian corporatism is the continued commitment of the two original Accord partners. If either union support waned or the government fell, corporatism would be seriously weakened. The second problem is more likely to strike first.’ Given that Australia corporatism arose via the second, leftist route it did indeed end along with the 1983–96 ALP government.

132  Research handbook on political representation As for Spain its corporatism was incomplete and often bipartite, not tripartite, but it did follow the second route, generally reflecting the PSOE government of 1982 to 1996. However, the Popular Party government continued corporatist policy-making through its first term (1996 to 2000), reflecting Prime Minister Aznar’s moderate conservatism. When the PSOE returned to power in 2004, Spanish corporatism returned, but it would not survive the economic crisis which began in 2008 (González Begega and Luque Balbona 2014). Even in Scandinavia, corporatism has declined in terms of policy preparation (Christiansen 2010). Indeed, in Sweden, Jahn (2016) notes a weakening in overall corporatism from around 1990 – paralleling the no-longer-hegemonic nature of the social democrats. Barbados is interesting, but there both main parties are social democratic. The endurance of corporatism after the end of social democratic dominance may largely be a feature of Scandinavia, but this situation may be not just specifically about Scandinavia but that corporatism there existed for decades and in good economic times (the post-war boom), and for these reasons was more durable. Of course, outside of Scandinavia, corporatism continues to exist where leftist parties are currently dominant – South Africa and Uruguay – but the implication is that corporatism would not survive a shift to the political right in these countries. Finland is perhaps the most interesting and illustrative case in terms of the two routes to corporatism, as corporatism there needed the rise of its social democrats – not to dominance but to strong cross-class multi-party governments. Although (barely) most coalitions through to 1968 were cross-class (10 out of 18), after then cross-class coalitions became quite standard (24 out of 30 through to January 2020) and more stable. As Helander (1982, p. 179) emphasizes: the increasingly close co-operation between the internally integrating organizations of employees and employers was not in itself enough to lead to the establishment of the incomes policy system in 1968. The most important political factors … were to be found at the parliamentary level. In the 1966 general election the Social Democrats, who had been out of office in the early 1960s, gained a remarkable victory. After the election, a strong coalition cabinet was formed with representatives from all the major parties, excluding the Conservatives. This ‘popular front coalition’ affected the trend towards the creation of an incomes policy system in two ways. First, it helped to integrate the labour movement in both the trade union and the political sectors. Second – and this was perhaps of more importance – the government was now strong enough to fulfil the duties it had to undertake on behalf of the state.

Table 10.2 thus subdivides Finland into these two time periods. One should also note that a Scandinavian-style pattern of pre-electoral coalitions has not developed in Finland (Ganghof et al. 2015). For its part, Iceland is thus a clear contrast to Finland. Post-war (more precisely, post-1947) corporatism in Iceland has been only sporadic, with one-off tripartite agreements in 1964 and 1990. The key contextual reasons are political: (1) the weakness of the left and conversely the post-war dominance of the conservative Independence Party; and (2) typical coalitions being not cross-class (as was Iceland’s green–red government of 1934 to 1938), but centre-right between the Independence Party and the agrarian Progressive Party (Jónsson 2014). Iceland thus had neither route to lasting corporatism.

CONCLUSIONS: PATTERNS, OUTLIERS AND CHANGING CONTEXTS So, yes, corporatism parallels and indeed arises from broad class and party representation in the countries of ‘Germanic’ Europe (Austria, Belgium, Germany, Luxembourg, Netherlands, Switzerland), but also Finland. However, corporatism has also existed in various other countries without such broad representation but with leftist dominance, in some countries durably

Corporatism and representation  133 (Scandinavia, South Africa for now) but in many such countries not durably. Pluralism overlaps well though with Westminster (competitive majoritarian) democracies with single-member electoral systems, as in Australia (pluralist most of the time), Botswana, Canada, Caribbean countries (except for corporatist Barbados), France, Malta (though it uses the single transferable vote), New Zealand (where the electoral system was changed in 1996 to mixed member proportional with resulting changes on other aspects of Lijphart’s executive-parties dimension except for the interest group system), and the United States. Returning to the issue of consensus versus majority representation, given that only the first route to corporatism involves coalitions and political consensus, the fit between corporatism and consensus democracy (Lijphart) or negotiated democracy (Armingeon 2002) is imperfect, and thus it is indeed not a full aspect of such a political system. Taagepera (2003) is thus correct on this point. Japan remains problematic and idiosyncratic, being integrated but not corporatist – or at least ‘corporatist without labour’ beyond the firm level. Certainly though, its cabinet representation is not cross-class coalitional. Indeed, Japan stands out in Korpi’s (2006) analysis of the 1945 to 1990 period, being the only coordinated market economy without either (1) confessional parties in governing coalitions (but not confessional dominance, as in post-war Italy); or (2) left dominance or left–right balance in its cabinets. It is also the only coordinated market economy without proportional representation. However, Japan does share with the (1) group of countries a state corporatist tradition (Korpi 2006, p. 197). In terms of the phenomenon of corporatism, there is a broad scholarly consensus that it has produced superior economic performance (Williamson 1989, pp. 152–3). More interestingly, given the corporatist focus on economics, is the finding of Scruggs (2003) that corporatist countries also have superior environmental performance – due in Scruggs’ (2003, pp. 140–41) analysis to the structural and also informational features of corporatism helping to overcome collective action problems. That said, there are many features that are found in corporatist countries (Siaroff 1999) and some of these have trade-offs in terms of representation. Specifically, the centralized and hierarchical nature of organized labour, and its willingness to compromise with capital, are criticized especially from a neo-Marxist perspective. Looking ahead, corporatist interest representation may also be under threat as (most) industrial countries shift from a left versus right politics based on social class more to an open versus closed politics based on education (Dalton 2018).

REFERENCES Archer, Robin (1992), ‘The Unexpected Emergence of Australian Corporatism’, in Jukka Pekkarinen, Matti Pohjola and Bob Rowthorn (eds), Social Corporatism: A Superior Economic System?, Oxford: Clarendon Press, pp. 377–417. Armingeon, Klaus (2002), ‘Interest Intermediation: The Cases of Consociational Democracy and Corporatism’, in Hans Keman (ed.), Comparative Democratic Politics: A Guide to Contemporary Theory and Research, London: SAGE, pp. 143–65. Benoit, Kenneth and Michael Laver (2006), Party Policy in Modern Democracies, London, UK and New York, NY, USA: Routledge. Bernauer, Julian, Nathalie Giger and Adrian Vatter (2014), ‘New Patterns of Democracy in the Countries of the Comparative Study of Electoral Systems 2’, in Jacques Thomassen (ed.), Elections and Democracy: Representation and Accountability, Oxford: Oxford University Press, pp. 38–59. Carey, John M. and Simon Hix (2011), ‘The Electoral Sweet Spot: Low-Magnitude Proportional Electoral Systems’, American Journal of Political Science, 55 (2), 383–97. Christiansen, Peter Munk, Asbjørn Sonne Nørgaard, Hilmar Rommetvedt, Torsten Svensson, Gunnar Thesen and PerOla Öberg (2010), ‘Varieties of Democracy: Interest Groups and Corporatist Committees in Scandinavian Policy Making’, Voluntas, 21 (1), 22–40. Dalton, Russell J. (2018), Political Realignment: Economics, Culture, and Electoral Change, Oxford: Oxford University Press.

134  Research handbook on political representation Ganghof, Steffen (2010), ‘Review Article: Democratic Inclusiveness: A Reinterpretation of Lijphart’s Patterns of Democracy’, British Journal of Political Science, 40 (3), 679–92. Ganghof, Steffen (2016), ‘Reconciling Representation and Accountability: Three Visions of Democracy Compared’, Government and Opposition, 51 (2), 209–33. Ganghof, Steffen and Sebastian Eppner (2019), ‘Patterns of Accountability and Representation: Why the Executive-Parties Dimension Cannot Explain Democratic Performance’, Politics, 39 (1), 113–30. Ganghof, Steffen, Sebastian Eppner and Katya Heeß (2015), ‘Normative Balance and Electoral Reform: A Finnish Puzzle and a Comparative Analysis’, West European Politics, 38 (1), 53–72. González Begega, Sergio and David Luque Balbona (2014), ‘Goodbye to Competitive Corporatism in Spain? Social Pacting and Conflict in the Economic Crisis’, Revista Española de Investigaciones Sociológicas, 148, Oct.–Dec., 79–102. Hall, Peter A. and David Soskice (eds) (2001), Varieties of Capitalism: The Institutional Foundations of Comparative Advantage, Oxford: Oxford University Press. Helander, Voitto (1982), ‘A Liberal-Corporatist Sub-system in Action: The Incomes Policy System in Finland’, in Gerhard Lehmbruch and Philippe C. Schmitter (eds), Patterns of Corporatist Policy-Making, London, UK and Beverly Hills, CA, USA: SAGE, pp. 163–87. Jahn, Detlef (2016), ‘Changing of the Guard: Trends in Corporatist Arrangements in 42 Highly Industrialized Societies from 1960 to 2010’, Socio-Economic Review, 14 (1), 47–71. Jónsson, Guðmundur (2014), ‘Iceland and the Nordic Model of Consensus Democracy’, Scandinavian Journal of History, 39 (4), 510–28. Keman, Hans and Paul Pennings (1995), ‘Managing Political and Societal Conflict in Democracies: Do Consensus and Corporatism Matter?’, British Journal of Political Science, 25 (2), 271–81. Korpi, Walter (2006), ‘Power Resources and Employer-Centered Approaches in Explanations of Welfare States and Varieties of Capitalism: Protagonists, Consenters, and Antagonists’, World Politics, 58 (2), 167–206. Kriesi, Hanspeter and Alexander H. Trechsel (2008), The Politics of Switzerland: Continuity and Change in a Consensus Democracy, New York, NY: Cambridge University Press. Lehmbruch, Gerhard (1979), ‘Liberal Corporatism and Party Government’, in Philippe C. Schmitter and Gerhard Lehmbruch (eds), Trends Toward Corporatist Intermediation, London: SAGE, pp. 147–83. Lijphart, Arend (1968), Politics of Accommodation: Pluralism and Democracy in the Netherlands, Berkeley, CA: University of California Press. Lijphart, Arend (2012), Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries, 2nd edn, New Haven, CT, USA and London, UK: Yale University Press. Mundlak, Guy (2007), Fading Corporatism: Israel’s Labor Law and Industrial Relations in Transition, Ithaca, NY and London, UK: Cornell University Press. Pempel, T.J. and Keiichi Tsunekawa (1979), ‘Corporatism Without Labor? The Japanese Anomaly’, in Philippe C. Schmitter and Gerhard Lehmbruch (eds), Trends Toward Corporatist Intermediation, London: SAGE, pp. 231–70. Scruggs, Lyle (2003), Sustaining Abundance: Environmental Performance in Industrial Democracies, New York, NY: Cambridge University Press. Siaroff, Alan (1999), ‘Corporatism in 24 Industrial Democracies: Meaning and Measurement’, European Journal of Political Research, 36 (2), 175–205. Siaroff, Alan (2003), ‘Varieties of Parliamentarianism in the Advanced Industrial Democracies’, International Political Science Review, 24 (4), 445–64. Siaroff, Alan (2019), Comparative European Party Systems: An Analysis of Parliamentary Elections Since 1945, 2nd edn, Abingdon: Routledge. Taagepera, Rein (2002), ‘Implications of the Effective Number of Parties for Cabinet Formation’, Party Politics, 8 (2), 227–36. Taagepera, Rein (2003), ‘Arend Lijphart’s Dimensions of Democracy: Logical Connections and Institutional Design’, Political Studies, 51 (1), 1–19. Therborn, Göran (1992), ‘Lessons From “Corporatist” Theorizations’, in Jukka Pekkarinen, Matti Pohjola and Bob Rowthorn (eds), Social Corporatism: A Superior Economic System?, Oxford: Clarendon Press, pp. 24–43. Visser, Jelle (2011), ICTWSS: Database on Institutional Characteristics of Trade Unions, Wage Setting, State Intervention and Social Pacts in 34 Countries between 1960 and 2007, Amsterdam: University of Amsterdam. Williamson, Peter J. (1989), Corporatism in Perspective: An Introductory Guide to Corporatist Theory, London: SAGE.

11. Technocratic representation Daniele Caramani

Technocracy has continuously been an integral feature of the nation-state. Next to the creation of a democratic nation, the nineteenth century witnessed the creation of a centralized, territorially organized, Weberian state bureaucracy. Technocratic elements of the state had to assure efficiency. Strongly influenced by the belief at the roots of the Industrial Revolution that society could be steered like a machine, and by the belief in progress and science, state formation implied a role for elites selected through merit rather than aristocratic privilege. A crucial aspect of the state formation project was thus to counterbalance the inclusion of the newly mobilized national masses with representative ‘government’, as opposed to radical versions of Rousseauian ‘democracy’ and institutions such as direct democracy, the imperative mandate and participatory democratic institutions. It involved restricted suffrage (initially), division of power, checks and balances, and sectors isolated from democratic control and political competition (such as the judiciary and central banks). The electoral selection of representatives instead of the lot aimed at balancing popular will and a rational, competent and independent government.1 Representatives as ‘trustees’, not as delegates of people’s will, are at the core of a representative system that values responsibility on a par with responsiveness.2 Modern parties played a crucial role in bridging the dual requirement of the nation-state: democratic self-rule and efficient government (Caramani 2017; 2020). Technocratic representation indicates a specific form of representation that is inherent in the nation-state to the present day. The goal of this chapter is to identify its components. Technocracy is an important element of the representative system, albeit not the only one. Its core principle – elites acting on the basis of expertise in the interest of the whole, independently of a popular mandate – is explored in the next section, which is followed by an examination of the dimensions that comprise it: depoliticization, neutrality, holism, anti-pluralism, responsibility and unaccountability, and autonomy. The third section addresses the manifestations of technocratic representation on a continuum between fully technocratic and fully democratic. The conclusion discusses the normative implications of technocratic representation.

THE PRINCIPLE OF TECHNOCRATIC REPRESENTATION Starting from Pitkin’s (1967) distinction between descriptive, symbolic and active (or substantive) representation, technocratic representation is a type of the latter. Substantive representation involves acting (‘doing’) on behalf of citizens in the interest of society. This action can have different sources of legitimacy, in particular those coming from a popular mandate (through either a numerical voting majority or the mystical consensual will of the people embodied by a leader). The source of legitimacy on which technocratic representation rests is expertise. It is through knowledge, skills, scientific procedures and rational speculation that society’s interest is established (Centeno 1993; Collins and Evans 2007; Fischer 1990; Habermas 2015). This fundamental principle has a number of consequences. 135

136  Research handbook on political representation First, establishing society’s interest through scientific procedures implies that they are objective – that is, not derived from subjective interests. Pitkin (1967) referred to this aspect of representation as ‘unattached’ interests. Society’s interest and the bases of policy action are decoupled from society’s preferences of and external to them. The course of action is discoverable. Discovery occurs through scientific procedures independently from society’s indications. Second, and consequently, society’s interest is defined in terms of being right or wrong, rather than in terms of responding or not responding to a mandate. Technocratic representation does not require a mandate from, or being responsive to, the preferences of society. On the one hand, citizens may not be knowledgeable enough for complex policy decisions, or may not have enough time to engage with them. Policy action thus becomes the specialized activity of an elite. On the other hand, citizens may act subjectively, based on ‘attached’, partisan interests (specific to given groups) and thereby distort society’s general interests. For the same two reasons – people not having the capabilities to judge the action of elites, and people acting on particularistic interests – the idea that people can sanction elites does not apply. Shielding decisions from subjective interests and not constraining elite’s actions makes it a trustee, unaccountable form of representation. Third, society’s interest is holistic and takes precedence over the interests of parts. It is only through an external, objective, ‘unattached’ procedure that the general interest can be identified. Action should be guided by the interest of the whole society, i.e. action that is ‘good for all’ as opposed to for a part only (Rehfeld 2009). The anti-pluralist nature of technocratic representation and the wholeness of society are intended also temporally. The general interest of society that representative technocratic action must cover includes that of future generations. Deriving the general interest objectively avoids the risk of having partisan interests (of the present generation) predominating over those of the future. Short-term interests not predominating over long-term ones is sometimes equated with the definition of responsibility itself (Goetz 2014; Magaloni 2006). To be sure, society is depicted in its complex interactions between components as being like a body or machine that needs to work efficiently. Parts, however, are not supposed to have interests, but rather roles and functions instrumental to the successful functioning of the whole in a corporatist way. As in nationalism and populism, the whole takes precedence over the particular. Unlike them, however, technocracy does not require the same degree of emotional mobilization through top-down symbolic representation securing loyalty to the system, although it may help in legitimizing restrictions on pluralist claims.3 Fourth, discovering society’s objective interest through rational procedures has the consequence that channels and actors for the expression of plural interests such as elections and parties are seen first of all as inefficient. Technocratic representation strives for efficiency. In addition, electoral competition is more likely to lead to distortions, mistakes and injustice as they may favour the part that has the power to impose its preferences. By avoiding competition and the aggregation of plural interests as grounds on which to base policy, technocratic representation is non-political, even anti-political, as it sees politics as an obstruction to efficiency and justice. Neutral technique, not power, guides action. Fifth, and finally, because entire societies cannot possess all the necessary skills and the independence to make informed and objective decisions, technocratic representation relies on an elite that specializes in this task. It is selected through merit – that is, for having superior knowledge and skills. Technocracy therefore does not involve descriptive representation (‘being’) in so far as merit, knowledge and competence make representatives necessarily dif-

Technocratic representation  137 ferent – intellectually and sociologically – from the composition of society. Their legitimacy comes precisely from being other than the population at large. To sum up, technocratic representation consists of an unaccountable elite acting efficiently and objectively based on expertise to discover and further the general interest.

THE ORIGINS OF TECHNOCRATIC REPRESENTATION As mentioned, technocratic representation is integral to the balance that representative government attempts to strike between democratic inclusion (input legitimacy) and efficiency (output legitimacy). Especially in contexts of high international competition between states, state structures require efficient and effective action. Historically, calls for radical forms of democracy (populism, direct democracy, citizens’ assemblies) have alternated with calls for technocratic solutions, expert action and a focus on output rather than input. Both types of calls challenge representative democracy. They are particularly strong during crises and failures of representative democracy to deliver output efficiently and effectively in a complex, multilevel and competitive globalized environment (Centeno 1994).4 This includes the need to ‘adjust’ to supranational integration, and compliance with financial regulations by the World Bank, the International Monetary Fund or the European Central Bank (Kahler 1992). It applies to various regions and periods, ranging from experiments of technocratic and corporatist states in Latin America during the twentieth century to technocratic cabinets in Europe more recently and the intervention of technocratic institutions in the wake of the financial crisis in the European Union in the 2010s.5 Such visions for society’s governance can resort to classical theories going back as far as Plato’s philosopher-kings, Francis Bacon’s New Atlantis and other utopias of running society like a machine and based on detailed knowledge (Wadia 1987). They become prominent in concomitance with Enlightenment, technological and scientific progress, and the positivist belief in rationality. The confidence that humans acquired from controlling energy during the Industrial Revolution and Taylorism spills over to managing social institutions. It is in the nineteenth century that social sciences develop, that is, the aim not only of analysing but also of steering societies scientifically. Thinkers from Saint-Simon to Comte and Karl Mannheim have supported such views (Fischer 2000; 2009). Technocratic representation has been applied by Marxism and by proponents of technocratic soviets or committees on technocracy in the United States (Thorstein Veblen and Walter Rautenstrauch, among many others).6 Simultaneously, elitist theories highlighted the shortcomings – the ‘failures’ – of parliamentary and representative democracy through authors such as Michels, Mosca and Schumpeter. Strong states in particular build administrative capacity based on planning and dirigisme, as well as schools for the meritocratic preparation of elites (Badie and Birnbaum 1983). Today, one finds elements of technocratic representation in concepts such as ‘epistemic community’, ‘epistocracy’ and the ‘regulatory state’ (Holst 2012; Majone 1994), and in various forms of non-majoritarian institutions that go under the label of type II governance (Hooghe and Marks 2003) – that is, independent from majoritarian competition and popular legitimation.

138  Research handbook on political representation

THE DIMENSIONS OF TECHNOCRATIC REPRESENTATION Starting with the definition of representation as ‘acting for’ is a useful way to address the features of technocratic representation and ask questions relating to: who acts, on behalf of whom, what is the link between those who act and those on whose behalf actions are taken, and in what way is the action determined? This allows us to single out the distinctive dimensions of technocratic representation (they appear in italics in the following subsections). Who Acts? Technocratic representation is elitist at its core; it is an elite acting. This elite is based on talent and competence, and is selected on the basis of merit. Technocratic representation is therefore based on a limited descriptive representation.7 Descriptive representation is minimal, with trust based on being different – in terms of education, intellect and competence – rather than similarity with the social composition of society. This principle is one of exclusion of the masses and scepticism towards the inclusion of less qualified groups. The focus is on ‘doing’, that is, acting (politics of ideas), rather than on ‘being’, that is, standing for (politics of presence). This is the distinction in German between vertreten (substantive action) and darstellen (descriptive presence) (Mansbridge 1999). Technocratic representation is therefore not intended as a ‘mirror’ of the diversity of society. This dimension of representation affects the mechanisms for the selection of the elite. It excludes the lot and citizens’ assemblies (or then only when accompanied by the advice of experts). Elections and party competition are also seen as unsuitable for selecting the elite, as this mechanism, too, is subject to the influence of the unqualified majority short-term considerations, turning them into popularity contests rather than contests about competence. On Behalf of Whom? The goal of technocratic representation is to act in the general interest. To the question of ‘who’ is represented, the answer is ‘society as a whole’. Technocratic representation therefore strives for a holistic (anti-pluralist) representation. Technocratic representation is about the common interest. This implies that the ‘whole’ takes precedence over the ‘parts’, and that what is good or bad applies to society in its entirety. To be sure, the whole has different components, similar to parts of a body, machine or system. These parts are not in competition with one another, but have their own roles and functions. They do not have rights of representation on their own, only in terms of fitting in the system.8 It is therefore anti-pluralist and corporatist in the sense that it does not allow the interest of the part to prevail over the interest of the whole. It is a form of representation detached from rights in so far as minorities (i.e. parts) have no protection against the whole.9 Technocratic representation is a type of aggregation, but not one that is in competition with alternative programmes of aggregation (as, for example, offered by parties in representative government). Competition is seen negatively as it pitches the interests of a part against those of the whole. Its pluralism is limited to that between those who are right and those who are wrong. To ask about the representation of ‘whom’ is limited; one needs to ask also on behalf of ‘what’. In this sense it is a form of representation not linked to rights. For example, tech-

Technocratic representation  139 nocratic representation allows action for forms of life without legal personality (nature) or individuals not born yet (future). What is the Link between Actors and Those on Whose Behalf they Act? At its core, representation concerns the link between the representatives (those who act) and the represented (those on whose behalf the action takes place). In this regard, technocratic representation displays a number of distinctive dimensions that set it apart (either qualitatively or as a matter of degree) from other forms of representation, in particular liberal, plural representation as well as populist representation. The guiding principle of technocratic action is responsibility, meaning that the link to the represented is not necessarily a responsive one. Technocratic representation does not rely on a mandate and representatives are not delegates, but rather ‘trustees’. Representation as ‘acting for’ (not ‘through’) the people thus requires a full Burkean/fiduciary model in which responsiveness takes a second-order position vis-à-vis responsibility and governing functions. It is not necessarily unresponsive, but technocratic action can disregard responsiveness to any sort of mandate from the electorate and can be incongruent with citizens’ preferences.10 Proximity with the distribution of the electorate’s preference does not guide policy. Output legitimacy trumps input legitimacy.11 The general interest trumps the general will. Technocratic representation is therefore a non-responsive representation. It is a responsible representation in the sense that it does not act to please, especially in the short term, and takes a long-term approach. Technocratic representation is also an unaccountable (unconstrained) representation. Since active representation is not based on a programme or mandate, it is impossible to ascertain whether or not the promise has been maintained. There is no ‘promissory’ mandate, as elites are not elected. Accountability as the possibility to sanction (not the sanction itself) is therefore not given. This applies also to an ‘anticipatory’ form of accountability (Mansbridge 2003) in which parties, for example, formulate programmes taking into account future sanctions. In principle, even if a programme was set out by elites through superior knowledge and expertise, it makes no sense to ask less competent voters to judge policies established by rational procedures. People do not possess the capabilities to judge the action of elites and thus to sanction them (vertical accountability). In addition, as technocratic representation acts in the interest of the whole, it should not be constrained by checks-and-balances, procedures or other forms of horizontal accountability. Yet one should not dismiss the possibility for sanctioning whether or not problems have been identified and solved, and if ‘mandates’ (from political principals) have been fulfilled by technocratic agencies (in terms of targets for example). Relying strongly on output and problem solving, technocratic representation must deliver. Furthermore, horizontal accountability may be conceived of through the plurality of experts’ views, quality control, accuracy of scientific protocols, data sources and fact checking. This relativizes the view of technocratic representation as totally unaccountable but does not fundamentally alter the view that it is not accountable to electoral preferences. Its goal is to find the ‘truth’ (the true general interest), not what people want (the subjective and plural people’s will). The relationship between technocratic representative action and those on whose behalf elites act must therefore be one of autonomy. Technocratic representation is an autonomous representation. This basically means that it isolates areas of decision-making from democratic control with the goal of ‘shielding’ decision-making from partisan influences,

140  Research handbook on political representation short-term electoral goals, or ‘pandering to the people’ for popular consent. This is the idea behind having independent central banks, for example. Autonomy means that technocrats are not simply ‘agents’ implementing decisions by a ‘principal’ (Max Weber’s ideal of setting the means but not the goals), but ‘actors’ (Dargent 2015). This makes technocracy a form of power, not simply an agency for implementation. As actors, technocrats do not act based on a mandate, whereas as agents they may implement a ‘mandate’ received from a political principal, which makes them technicians rather than technocrats. Autonomy – that is, the lack of political control and independence from a democratic principal – makes technocratic representation a non-democratic representation. To be sure, in all polities – including democracies – there is a recognition of autonomous (independent) spaces of decision-making beyond popular control through parliamentary majorities. The broader these areas, the larger the ‘democratic deficit’. How large this area of autonomy is in a given polity, is an empirical matter and a matter of degree. Under representative government, such autonomy is limited to certain areas. In a fully technocratic state, such autonomy is much more extensive. The asymmetrical expertise between politicians/citizens on the one hand and experts on the other gives the latter the possibility to advance their own interests and increase power, making them even more autonomous within a democratic state. On the other hand, acting as agents does not make technocracy necessarily democratic, such as when experts act on behalf of lobbies or international financial institutions (Centeno and Silva 1998), or when parties ask them to make painful decisions and do the ‘dirty job’ they prefer to avoid (for example, instigate reforms that would lead to electoral defeat). How are Actions Determined? Technocratic representation removes politics from policy, that is, actions. Being capable of discovering the correct way of action through expertise, rational speculation and scientific procedure does away with conflict, which is the essence of politics. If there is an undisputable truth on which to base action, and if facts, rather than opinions, are undisputable, then there is no need for dispute. The vision is one of government above the clashes of interests and a conflictless society. The procedures through which actions are determined are therefore ‘depoliticized’. Action is not determined politically through the competition between partisan (subjective) interests in elections and through majorities. Truth being incontrovertible and not a matter of opinion, technocratic representation is an anti-political representation. Actions are thus determined in an objective, impartial and neutral way. Policy is about information, not preferences, which are not objective and not neutral. This means that technocratic representation is an a-ideological representation that is not blinded by the fervour of interpretations of the social world, value systems, emotions, illusions, and so on. Its pragmatism means it focuses on what is possible and feasible (Putnam 1977; Adolph 2013). Neutrality is a problematic concept in this context as it implies alternatives about which technocratic action is neutral. In fact, depoliticized objectivity does not contemplate alternatives in the first place, so there is nothing to be neutral about. Also, the welfare state and the market are depoliticized logics (Bickerton and Invernizzi Accetti 2017; Bourdieu 2002; Laclau 2005; Pettit 2004; Rosanvallon 2011). Against such visions, some argue that in fact all production and transmission of knowledge reflects value systems and structures of power (Feindt and Oels 2005; Foucault 1980; Jasanoff 2012)

Technocratic representation  141 Depoliticization, the non-ideological nature of technocratic representation and pragmatism all point to the superfluity of ‘choice’. Policy does not involve choice in the technocratic vision as the course of action needs simply to be identified – not chosen. Policies are therefore often presented in terms of necessity, pointing precisely to an impossibility to choose (‘we have to do it’), for example because of the global context (Kahler 1992) or domestic economic conditions (Ruiz-Rufino and Alonso 2017; Crouch 2011) – choices nonetheless, but presented in the neutral language of reform, modernization and so on (more on discourse below). The fact that neutrality is an illusion is proved by the adoption of technocratic discourse by different ideological systems: from ‘developmentalist’ experts in Latin America in the 1970s to ‘neo-liberal’ economists in the 1990s (Babb 2001; Centeno and Maxfield 1992). The fact that its core can be combined with different content makes technocracy a sort of ‘thin ideology’. Finally, actions are determined in an efficient way. Part of the anti-politics nature of technocratic representation is its emphasis on the non-efficient way of reaching decisions based on discussions, confrontations, elections, party politics, and so on. Precisely those procedures set in place to choose are criticized, as the need to choose itself is questioned. They are perceived and presented as a loss of time when, on the contrary, the focus should be on solving problems that are evident and whose solution is also evident once analysed rationally. Technocratic representation is therefore also an efficient representation that sometimes causes admiration in democracies for polities in which decision-making is straightforward, without the bickering between politicians and the reversal of policies.

THE FORMS OF TECHNOCRATIC REPRESENTATION No polity – however democratic – completely excludes some form of technocratic representation. Every democracy has technocratic pockets in the form of state bureaucracy, experts in advisory positions, independent ministers and prime ministers, or entire executives. Indeed, unlike classical undemocratic arguments for technocracy, contemporary ones see technocratic bodies embedded in, and respectful of, democratic values and institutions, be it at the national or the supranational levels. Furthermore, technocratic representation may also be present in leaders’ attitudes, or in citizens’ support for technocratic decision-making, as well as in the public discourse by actors and the media. It is, therefore, not an ‘either/or’ question; polities, rather, incorporate technocratic representation to a certain degree. It is only from a certain degree that the level of technocracy in a polity makes it authoritarian. Technocratic representation as a continuum over time and across polities therefore has consequences for the authoritarian nature of the polity. The first and most obvious form of technocratic representation is the state itself. As mentioned, states are technocratic, and technocracy needs a state as it operates through its apparatus to act efficiently and effectively. The state’s ‘cybernetic function’ is to ‘think’, to acquire information, process complexity, anticipate consequences and put in place rational procedures to devise long-term goals and policies. It is also the state that acts internationally and implements decisions. The state is therefore not isolated from society, but insulated from politics (Centeno 1994). It needs a network and ‘feelers’ to act efficiently on societies and the environment. Approaching full technocracy is extremely rare, as all regimes include some form of popular mobilization, identification and support. Examples of high state capacity may include modern-

142  Research handbook on political representation izing countries such as Mexico in the early 1990s, China and Singapore today, or the Soviet Union (Hoffman and Laird 1985). Military regimes are sometimes considered technocracies, as are theocracies, but often such regimes are highly politicized and implement policies on the basis of specific ideologies such as the neoliberal ones in Chile in the 1970s (Silva 2008). Latin America in particular has been a region of experimentation with technocracy, and some still consider the region a case of ‘bureaucratic authoritarianism’ (O’Donnell 1973), in which the effects of autonomous technocrats on the quality of democracy are felt after democratic transitions.12 Under democratic regimes, ministerial bureaucracies are a case of technocratic apparatus, as well as central banks and various regulatory agencies. Expertise is present also in the legislative branch, with offices advising and supporting parliamentary works and preparing single legislators’ dossiers. Within the broad principle of separation of power, the independence of the judiciary and its increasing policy impact – that is, the ‘judicialization’ of politics (Shapiro 1964; Sadurski 2013) – is another such area of autonomous state activity. Parties also provide expertise and rely on their own research centres (for example, think tanks), or the expertise of trade unions or business associations, to which they are ideologically and organizationally close. One of the most visible manifestations of technocracy is technocratic cabinets – that is, the executive branch being technocratized beyond its ministerial bureaucracy. Technocratic and technocrat-led (only the prime minister is non-political) cabinets have increased in the last two decades as a consequence of ‘party failure’ in countries in both Eastern and Western Europe. These are cases opposed to the ideal-type of party government and in which the ‘non-partyness’ of cabinets is high (Andeweg 2000; Katz 2014). Empirical cases range between more or less technocratic, and cluster in types based on their composition, their output (remit) and their duration. Short-term caretaker cabinets have a mandate to maintain the status quo, others to change it by making policy.13 A lot of the discussion around the ‘technocraticness’ of polities and institutions depends on who qualifies as a technocrat. Technocrats are usually ‘civilian’ unelected personnel who are politicians. Technocrats are experts and rely on a non-political expertise derived from training, education or professional experience (for example, in business or academia). They are either insiders (i.e. from a state bureaucracy, such as a central bank) or outsiders called to occupy executive positions (Bersch 2016). Hence, expertise is not enough to characterize a technocrat. Typically, technocrats have power and are active in politics; they are experts who become political and cease to be merely experts. A técnico without power is thus not a tecnócrata (Meynaud 1969).14 Crucially, technocrats are not elected and, therefore, are not subject (accountable) to the constraints of electoral competition and partisanship (Tucker 2018; Vibert 2007). At the supranational level, the technocratic nature of policy-making appears more clearly – for example, in organizations such as the World Bank and the International Monetary Fund, as well as United Nations agencies. The European Union is also often considered a case of technocratic governance (Radaelli 1999). The non-elective nature of the personnel, their expertise and their policy-making functions makes them ideal-cases of technocracy, although their neutrality is a matter of empirical investigation. The policy areas in which they are active are extremely broad, from monetary policy to development, the environment, and so on. At the national level, the policy areas in which technocrats are active are more restricted. Economics is the crucial field, but also social policy (especially in Latin America). This means that the

Technocratic representation  143 technocratic nature of different policy fields varies, with experts being more or less involved in policy-making depending on the policy area. More pervasively, technocratic discourse is not exclusive to technocratic bodies. Elected politicians can also hold a technocratic discourse or make technocratic statements. This is a more diffuse form of technocratic representation, wherein leaders invoke their expertise and competence, cite statistics, facts and reports, or refer to procedures. Pandering to the public is framed as irresponsible. In such discourses (or even single statements), reference is made to expertise and competence. The general interest is invoked and policies are claimed to be for the common good in the long term. Such a discourse stresses necessity, lack of choice, efficiency, cost optimization and pragmatism. Problem solving and objectivity are emphasized in order to sidestep accusations of being partisan. Even their style can be more or less technocratic, including attire and jargon, with no dramatization of messages. At a very general level, such discourse can reveal how society sees itself, and how hegemonic technocratic ideas of representation are in a given society.15 The final form of technocratic representation consists of citizens’ attitudes towards it. Citizens may have more or less favourable attitudes towards technocratic governance and experts, in general or within specific institutions such as technocratic cabinets. Such attitudes reveal preferences for delegating decisions to experts. They also involve distrust towards people deciding on complex matters. Experts are perceived as having the necessary competence and as not being guided by short-term electoral interests. Technocratic attitudes point also to a fundamental distrust towards politicians who need to respond to an uninformed and partisan electorate. Elitism and a preference for expertise are thus combined with anti-politics sentiments.16 As new analyses reveal, such attitudes are linked to socio-economic traits, such as education, and other attitudinal traits, such as trust (Bertsou and Caramani 2020a).

CONCLUSION Representative government is an attempt to harness independent expertise, effectiveness and long-term responsible policy-making in the non-partisan interest of the whole within a democratic setting that secures responsiveness to people’s preferences and puts accountability mechanisms in the hands of citizens. Political parties in particular, within the responsible party model (Birch 1964; Rosenblum 2008; Schattschneider 1942), can play a crucial role in bridging responsibility and responsiveness (McCall Rosenbluth and Shapiro 2018). Polities oscillate in their attempt to maintain a balance between input and output legitimacy. The critique of party democracy as it developed over the past 200 years may either call for more radical forms of democracy, popular participation and self-rule (populism), or for more expertise and effective, non-partisan, depoliticized decision-making (technocracy). The beginning of this chapter linked this duality to the nature of the nation-state itself, a dynamic process of continuous balancing and correction, with such critiques surging recurrently in waves, one of which we inhabit in the present. No functioning polity has ever obliterated entirely some form of technocratic institutions that are autonomous from majoritarian politics, a possibility that is even less likely in a world that is increasingly complex and interconnected. The question from a normative point of view is at what point does the necessity of technocratic action stop being a corrective against populism and become an additional threat for liberal democracy as we have known it since

144  Research handbook on political representation state formation and nation-building in the nineteenth century? As a corrective to electoral obsession, the over-responsiveness of parties, the pandering to the masses’ instincts and so on, calls for expertise-driven, independent, rational judgements may appear as a healthy antidote. Yet the threat from calls for more technocracy is to incur the same dangers of de-legitimizing representative institutions based on procedures founded on the rule of law, deliberation and competition between proposals, and minority protection. Portraying fundamental features of the liberal polity as inefficient, partisan, motivated by short-term electoral success and disregarding the general interest can be profoundly de-legitimizing. Given past records, there are good reasons to believe in the flexibility and adaptability of representative institutions and parties as the main actors to cope with demands for more expertise while remaining responsive, accountable and attentive to society’s pluralism. Yet it would be a mistake to take the assets of representative institutions complacently and passively for granted.

NOTES 1. 2. 3. 4. 5.

6. 7. 8.

9. 10.

11. 12.

13.

Manin (1997), speaking of the ‘triumph’ of election over lot, expresses the duality between popular self-rule and efficiency through the oxymoron ‘democratic aristocracy’. The dichotomy between responsive and responsible government is found in Mair (2013), echoing the distinction between input and output legitimacy (Scharpf 1999). For an encompassing comparison between technocratic, populist and party democratic forms of representation, see Caramani (2017). On the technocratic challenge to democracy, see Bertsou and Caramani (2020b). Examples include the Chicago Boys during the Chilean military regime (Silva 1991), the Salinastrojka in Mexico (Centeno 1994) and various technocratic appointments in Argentina, Brazil, Mexico and Peru (Kaplan 2017; Teichman 1997; Weyland 2002). On technocratic cabinets in Europe, see note 13. On technocratic movements, see Akin (1977), Elsner (1967), and Laski (1931). The inclusion of groups on a descriptive base is only envisaged when it provides information. See Edmund Burke’s discussion on Irish Catholics and US colonists in his 1774 ‘Speech to the Electors of Bristol (at the Conclusion of the Poll)’. This whole may be represented symbolically. As with all types of representations, technocratic representation relies on loyalty of the parts towards the whole, imposed top down through socialization or indoctrination. This is the case for plural representation as well as for populism. Therefore, technocratic representation is also symbolic representation. However, this is not a distinctive feature of technocratic representation. That parts can be ‘sacrificed’ in pursuit of the common good can have far-reaching and tragic social consequences, such as eugenics programmes, for example. Technocracy may be responsive in so far as action coincides with citizens’ preferences, but also because there are other actors to be responsive to (minorities, supra-national organizations) and because it responds to risk and markets (Ezrow and Hellwig 2014; Gilad 2015; Rauh 2016). Congruence, as a static concept, does not correspond exactly to responsiveness, which is a dynamic concept. There is a large body of literature on the contrast between responsibility and responsiveness. See Caramani (2017) and Goetz (2014) for an overview. The influence of technocrats remains essential to elected politicians also after the end of authoritarian rule, protracting the power of more or less independent experts (Dargent 2015). See also Centeno and Silva (1998), Conaghan and Malloy (1994), Teichman (1997), Torre (2013), Williams (2006). On the criteria to assess the technocratic nature of executives see Brunclík and Parízek (2019), Costa Pinto et al. (2017), McDonnell and Valbruzzi (2014), Pastorella (2016) and Wratil and Pastorella (2018).

Technocratic representation  145 14. The influential Latin American literature speaks of técnicos (or tecnócratas, as opposed to políticos) and, similarly, Italian uses tecnici and tecnocrati. As these are usually experts in areas other than politics (say, economy, the environment, etc.), they are different from ‘technopols’ (Alexiadou and Gunaydin 2019; Joignant 2011). 15. Such discourse can appear also in symbolic representation stressing progress and social welfare. An example is the Brazilian flag, which is based on the 1889 modernizing project imbued with positivist references and leadership by an enlightened elite. 16. Empirical investigations measure such attitudes, for example through the concept of ‘stealth democracy’ (Hibbing and Theiss-Morse 2002; Ruiz-Rufino and Alonso 2017). Recently Bertsou and Pastorella (2017) conducted an empirical analysis. Bertsou and Caramani (2017) developed a new battery of questions administered in nine European countries.

REFERENCES Adolph, Christopher (2013), Bankers, Bureaucrats, and Central Bank Politics: The Myth of Neutrality, Cambridge: Cambridge University Press. Akin, William E. (1977), Technocracy and the American Dream: The Technocrat Movement, 1900–1941, Berkeley, CA: University of California Press. Alexiadou, Despina and Hakan Gunaydin (2019), ‘Commitment or Expertise? Technocratic Appointments as Political Responses to Economic Crises’, European Journal of Political Research, 58 (3), 845–65. Andeweg, Rudy B. (2000), ‘Political Recruitment and Party Government’, in Jean Blondel and Maurizio Cotta (eds), The Nature of Party Government: A Comparative European Perspective, London: Palgrave, pp. 119–40. Babb, Sarah (2001), Managing Mexico: Economists from Nationalism to Neo-liberalism, Princeton, NJ: Princeton University Press. Badie, Bertrand and Pierre Birnbaum (1983), The Sociology of the State, Chicago, IL: Chicago University Press. Bersch, Katherine (2016), ‘The Merits of Problem-Solving over Powering’, Comparative Politics, 48 (2), 205–25. Bertsou, Eri and Daniele Caramani (2017), Citizens’ Technocratic Attitudes: A Survey of Nine European Countries, Zurich: University of Zurich. Bertsou, Eri and Daniele Caramani (2020a), ‘People haven’t had Enough of Experts: Technocratic Attitudes among Citizens in Nine European Democracies’, American Journal of Political Science. Early View. Bertsou, Eri and Daniele Caramani (eds) (2020b), The Technocratic Challenge to Democracy, London: Routledge. Bertsou, Eri and Giulia Pastorella (2017), ‘Technocratic Attitudes: A Citizens’ Perspective of Expert Decision-Making’, West European Politics, 40 (2), 430–58. Bickerton, Christopher and Carlo Invernizzi Accetti (2017), ‘Populism and Technocracy: Opposites or Complements?’, Critical Review of International Social and Political Philosophy, 20 (2), 186–206. Birch, Anthony H. (1964), Representative and Responsible Government, London: Allen and Unwin. Bourdieu, Pierre (2002), ‘Against the Policy of Depoliticization’, Studies in Political Economy, 69 (1), 31–41. Brunclík, Miloš and Michal Parízek (2019), ‘When are Technocratic Cabinets Formed?’, Comparative European Politics, 17 (5), 759–77. Caramani, Daniele. (2017), ‘Will vs. Reason: The Populist and Technocratic Forms of Political Representation and Their Critique to Party Government’, American Political Science Review, 111 (1), 54–67. Caramani, Daniele (2020). ‘Introduction’, in Bertsou, Eri and Daniele Caramani (eds), The Technocratic Challenge to Democracy, London: Routledge, pp. 1–26. Centeno, Miguel Ángel (1993), ‘The New Leviathan: The Dynamics and Limits of Technocracy’, Theory and Society, 22 (3), 307–35.

146  Research handbook on political representation Centeno, Miguel Ángel (1994), Democracy within Reason: Technocratic Revolution in Mexico, University Park, PA: Pennsylvania State University Press. Centeno, Miguel Ángel and Sylvia Maxfield (1992), ‘The Marriage of Finance and Order: Changes in Mexican Political Elite’, Journal of Latin American Studies, 24 (1), 57–8. Centeno, Miguel Ángel and Patricio Silva (eds) (1998), The Politics of Expertise in Latin America, New York, NY: Macmillan. Collins, Harry M. and Robert Evans (2007), Rethinking Expertise, Chicago, IL: Chicago University Press. Conaghan, Catherine M. and James M. Malloy (1994), Unsettling Statecraft: Democracy and Neoliberalism in the Central Andes, Pittsburgh, PA: University of Pittsburgh Press. Costa Pinto, António, Maurizio Cotta and Pedro Tavares de Almeida (eds) (2017), Technocratic Ministers and Political Leadership in European Democracies, Cham: Palgrave. Crouch, Colin (2011), The Strange Non-Death of Neoliberalism, Cambridge: Polity. Dargent, Eduardo (2015), Technocracy and Democracy in Latin America: The Experts Running Government, Cambridge: Cambridge University Press. Elsner, Henry (1967), The Technocrats: Prophets of Automation, Syracuse, NY: Syracuse University Press. Ezrow, Lawrence and Timothy Hellwig (2014), ‘Responding to Voters or Responding to Markets?’ International Studies Quarterly, 58 (4), 816–27. Feindt, Peter H. and Angela Oels (2005), ‘Does Discourse Matter? Discourse Analysis in Environmental Policy Making’, Journal of Environmental Policy & Planning, 7 (3), 161–73. Fischer, Frank (1990), Technocracy and the Politics of Expertise, Newbury Park, CA: Sage. Fischer, Frank (2000), Citizens, Experts and the Environment: The Politics of Local Knowledge, London: Duke University Press. Fischer, Frank (2009), Democracy and Expertise: Reorienting Policy Inquiry, Oxford: Oxford University Press. Foucault, Michel (1980), Power/Knowledge, New York, NY: Pantheon Books. Gilad, Sharon (2015), ‘Political Pressures, Organizational Identity, and Attention to Tasks: Illustrations from Pre-Crisis Financial Regulation’, Public Administration, 93 (3), 593–608. Goetz, Klaus H. (2014), ‘A Question of Time: Responsive and Responsible Democratic Politics’, West European Politics, 37 (2), 379–99. Habermas, Jürgen (2015), The Lure of Technocracy, Oxford: Polity Press. Hibbing, John R. and Elizabeth Theiss-Morse (2002), Stealth Democracy: Americans’ Beliefs about How Government Should Work, Cambridge: Cambridge University Press. Hoffman, Erik O. and Robbin F. Laird (1985), Technocratic Socialism: The Soviet Union in the Advanced Industrial Era, Durham, NC: Duke University Press. Holst, Catherine (2012), ‘What is Epistocracy?’, in Simen Anderson Øyen, Tone Lund-Olsen and Nora Sørensen Vaage (eds), Sacred Science?, Wageningen: Wageningen Academic Publishers, pp. 41–54. Hooghe, Liesbet and Gary Marks (2003), ‘Unraveling the Central State, but How? Types of Multi-Level Governance’, American Political Science Review, 97 (2), 233–43. Jasanoff, Sheila (2012), Science and Public Reason, London: Routledge. Joignant, Alfredo (2011), ‘The Politics of Technopols: Resources, Political Competence and Collective Leadership in Chile, 1990–2010’, Journal of Latin American Studies, 43 (3), 517–46. Kahler, Miles (1992), ‘External Influence, Conditionality, and the Politics of Adjustment’, in Stephan Haggard and Robert Kaufman (eds), The Politics of Economic Adjustment, Princeton, NJ: Princeton University Press, pp. 89–136. Kaplan, Stephen B. (2017), ‘Partisan Technocratic Cycles in Latin America’, Electoral Studies, 45, 219–29. Katz, Richard S. (2014), ‘No Man Can Serve Two Masters: Party Politicians, Party Members, Citizens and Principal–Agent Models of Democracy’, Party Politics, 20 (2), 183−93. Laclau, Ernesto (2005), On Populist Reason, London: Verso. Laski, Harold (1931), The Limitations of the Expert, London: Fabian Society. Magaloni, Beatriz (2006), Voting for Autocracy: Hegemonic Party Survival and its Demise in Mexico, Cambridge: Cambridge University Press. Mair, Peter (2013), Ruling the Void, London: Verso. Majone, Giandomenico (1994), ‘The Rise of the Regulatory State in Europe’, West European Politics, 17 (3), 77–101.

Technocratic representation  147 Manin, Bernard (1997), The Principles of Representative Government, Cambridge: Cambridge University Press. Mansbridge, Jane (1999), ‘Should Blacks Represent Blacks and Women Represent Women? A Contingent “Yes”’, Journal of Politics, 61 (3), 628‒57. Mansbridge, Jane (2003), ‘Rethinking Representation’, American Political Science Review, 97 (4), 515‒28. McCall Rosenbluth, Frances and Ian Shapiro (2018), Responsible Parties: Saving Democracy from Itself, New Haven, CT: Yale University Press. McDonnell, Duncan and Marco Valbruzzi (2014), ‘Defining and Classifying Technocrat-Led and Technocratic Governments’, European Journal of Political Research, 53 (4), 654–71. Meynaud, Jean (1969), Technocracy, New York, NY: The Free Press. O’Donnell, Guillermo (1973), Modernization and Bureaucratic Authoritarianism, Berkeley, CA: University of California Press. Pastorella, Giulia (2016), ‘Technocratic Governments in Europe: Getting the Critique Right’, Political Studies, 64 (4), 948–65. Pettit, Philip (2004), ‘Depoliticizing Democracy’, Ratio Juris, 17 (1), 52–65. Pitkin, Hanna F. (1967), The Concept of Representation, Berkeley, CA: University of California Press. Putnam, Robert D. (1977), ‘Elite Transformation in Advanced Industrial Societies: An Empirical Assessment of the Theory of Technocracy’, Comparative Political Studies, 10 (3), 383–412. Radaelli, Claudio M. (1999), Technocracy in the European Union, New York, NY: Longman. Rauh, Christian (2016), A Responsive Technocracy?, London: Rowman and Littlefield. Rehfeld, Andrew (2009), ‘Representation Rethought: On Trustees, Delegates, and Gyroscopes in the Study of Political Representation and Democracy’, American Political Science Review, 103 (2), 214–30. Rosanvallon, Pierre (2011), Democratic Legitimacy: Impartiality, Reflexivity, Proximity, Princeton, NJ: Princeton University Press. Rosenblum, Nancy L. (2008), On the Side of the Angels: An Appreciation of Parties and Partisanship, Princeton, NJ: Princeton University Press. Ruiz-Rufino, Rubén and Sonia Alonso (2017), ‘Democracy without Choice: Citizens’ Perceptions of Government Autonomy during the Eurozone Crisis’, European Journal of Political Research, 56 (2), 320–45. Sadurski, Wojciech (2013), Rights before Courts, Dordrecht: Kluwer Academic. Scharpf, Fritz W. (1999), Governing in Europe: Effective and Democratic?, Oxford: Oxford University Press. Schattschneider, Elmer E. (1942), Party Government, New York, NY: Farrar and Rinehart. Shapiro, Martin (1964), Law and Politics in the Supreme Court, Glencoe: Free Press. Silva, Patricio (1991), ‘Technocrats and Politics in Chile: From the Chicago Boys to the Cieplan Monks’, Journal of Latin American Studies, 23 (2), 385–410. Silva, Patricio (2008), In the Name of Reason: Technocrats and Politics in Chile, University Park, PA: The Pennsylvania State University Press. Teichman, Judith (1997), ‘Mexico and Argentina: Economic Reform and Technocratic Decision Making’, Studies in Comparative International Development, 32 (1), 31–55. Torre, Carlos de la (2013), ‘Technocratic Populism in Ecuador’, Journal of Democracy, 24 (3), 33‒46. Tucker, Paul (2018), Unelected Power: The Quest for Legitimacy in Central Banking and the Regulatory State, Princeton, NJ: Princeton University Press. Vibert, Frank (2007), The Rise of the Unelected: Democracy and the New Separation of Powers, Cambridge: Cambridge University Press. Wadia, Pheroze (1987), ‘The Notion of Techne in Plato’, Philosophical Studies, 31 (1), 148–58. Weyland, Kurt (2002), The Politics of Market Reform in Fragile Democracies: Argentina, Brazil, Peru and Venezuela, Princeton, NJ: Princeton University Press. Williams, Mark E. (2006), ‘Escaping the Zero-Sum Scenario: Democracy versus Technocracy in Latin America’, Political Science Quarterly, 121 (1), 119–39. Wratil, Christopher and Giulia Pastorella (2018), ‘Dodging the Bullet: How Crises Trigger Technocrat-Led Governments’, European Journal of Political Research, 57 (2), 450–72.

12. Representation in authoritarian regimes Maurizio Cotta

REPRESENTATIVE INSTITUTIONS IN NON-DEMOCRATIC COUNTRIES Institutionalized representation, that is to say an established system of rules, processes (elections) and institutions (parliamentary assemblies) devised to ensure a two-way linkage between citizens and rulers, is typically associated with liberal and democratic regimes. It was in fact these regimes who invented and progressively implemented such political mechanisms. And these regimes are rightly defined as ‘representative democracies’. At the same time, a simple count shows that, as documented by the Interparliamentary Union (IPU) in its last report (https://​www​.ipu​.org/​), out of 193 world countries 179 have an institution recognized as a parliament. In most of these cases elections also take place (as documented in Hyde and Marinov 2011). However, according to the scientifically prevailing definitions a large number of these countries ‘with a parliament and elections’ do not qualify as democracies, but as authoritarian regimes of some type (Diamond 2002; Levitsky and Way 2010). What is then the meaning of such typical representative institutions in non-democratic regimes and what is the real role they play? To what extent do these elections and parliamentary bodies differ from democratic ones? Does their adoption mean that also authoritarian regimes need and implement some form of representation? But is it really representation – maybe of a peculiar nature – or is it a pure fake? These questions have received an increasing amount of attention in recent years. The study of representation and representative institutions in political regimes which cannot be defined as democracies is a relatively young but promising field in contemporary political science. Its developments closely follow the ongoing transformations in the broader field of regime studies and, in particular, the study of hybrid regimes (Gershewski 2013). The first stage in the study of regimes had concentrated on the polar division between democracy and totalitarianism (Friedrich and Brzezinski 1982). Analysts of totalitarianism, while aware of the existence of elections in such regimes as the Soviet Union, Nazi Germany or Italian fascism, tended to discard their relevance as pure window dressing and concentrated their attention on the role of the single party, of the leader and of manipulated mass mobilization. Elaborating on the distinction between totalitarianism and authoritarianism, Linz (1975) opened the way to a more articulate understanding of the large number of non-democratic regimes which do not fit the extreme model of totalitarianism, and of the fact that such regimes often count upon the support of a variety of important societal actors which in some sense are ‘represented’ in the regime dominant coalition. Despite this, more systematic comparative research on the bases and mechanisms of consent of such regimes did not develop significantly. This is probably due to the fact that the great waves of democratizations in the 1970s and in the last decade of the twentieth century diverted the attention from non-democratic regimes (seen as a declining phenomenon) and funnelled most of the research on transitions and democratization processes. 148

Representation in authoritarian regimes  149 More balanced assessments of the results of transitions and democratization processes, of their failures, or in any case of the non-deterministic nature of these itineraries (Carothers 2002; Geddes 2003; Geddes et al. 2014) have gradually reoriented the scientific landscape. An increasing amount of attention has been attracted by surviving non-democratic regimes, failed transitions ending in new forms of authoritarianism, and fuzzy regime forms that are difficult to classify. This has generated a growing number of studies exploring the configuration and internal functioning of new non-democratic regimes as well as of the intermediate area between democracy and non-democracy. The search for concepts able to catch this new reality, for a more articulate comprehension of the institutional elements shared across the divide between the two poles of democracy and authoritarianism, but also for a better understanding of the resilience of non-democratic regimes has burgeoned. Concepts such as hybrid regimes, illiberal democracies or competitive authoritarianisms have become of common use and they all suggest the presence in some countries of elements of democracy/representation alongside more typically authoritarian instruments (Collier and Levitsky 1997; Diamond 2002). The existence of a grey area between democracy and non-democracy, comprising countries where it is sometimes difficult to ascertain whether democratic or authoritarian elements prevail, suggests the need to establish with greater precision at the empirical level the features and functioning of those ‘representative elements’ – elections and parliamentary institutions – which are present in regimes of this grey area. But it also suggests the need to better understand the politics of authoritarianism.

NON-DEMOCRATIC OR SEMI-DEMOCRATIC ELECTIONS: DIMENSIONS OF VARIATIONS The study of non-democratic elections is not entirely new, but in the past it was mainly concentrated on the relatively few extreme cases of totalitarian regimes such as Stalinist and post-Stalinist Soviet Union (elections between 1936 and 1984), Nazi Germany (1933 and 1938 elections) or Italian fascism (1929 and 1934 elections). Elections conducted in these contexts displayed an extreme denial of the paradigmatic features of their democratic counterparts such as freedom, fairness, competitiveness and openness. With only one list, or one candidate, the voter could just say yes or no (and without guarantees of secrecy of the vote, the latter option was also severely limited). The regime left no space whatsoever for dissent. With regard to the Soviet Union (after the Stalinist period) some studies attempted to measure and evaluate the phenomenon of abstentions, the famous ‘missing one percent’ (Gilison 1968). Other authors suggested that in the Soviet Union the refusal to register (and thus to vote) could be used as a limited local form of protest about dysfunctional aspects of social life (Zaslavsky and Brym 1978; Karklins 1986). Both single-country and comparative studies of elections in non-democratic settings are by now offering much richer evidence on this apparently paradoxical phenomenon. The main research lines have concerned: (a) producing a more precise documentation of the form and nature of these elections; (b) investigating the reasons for holding them by regimes which, to say the least, do not favour opposition and accountability; (c) understanding the consequences they may have. The results of these studies show first of all the broadness of the phenomenon and the variety of forms it takes. The degree of pluralism and competitiveness, the amount of coercion

150  Research handbook on political representation applied on voters and candidates, and the degree of manipulation of results are the main elements of variation. At one extreme of this range of variations are the elections of some typical totalitarian regimes of the past. A typical example were the elections under Italian fascism in 1929 and 1934, or in the Soviet Union, or in Nazi Germany (Jessen and Richter 2011; Schedler 2002). In these cases, the absolute lack of pluralism in electoral lists and candidates and the strong amount of coercion (and eventually manipulation of results) applied to mobilize voters and to prevent any dissent or the appearance of it gave them a predominantly ritualistic meaning. Not very different from this model is what could be defined as the ‘predetermined pluralism’, applied in a number of Eastern European countries under Soviet control or today in Chinese elections. The peculiar pluralism of these elections is implemented through the presence of candidates of more than one party. These candidates, however, do not compete with those of the hegemonic party (Sartori 1976), but are preventively assigned a minority share of candidate positions and thus seats. East Germany (1950–86) provided a very clear example of this model with nine or more parties or movements participating in elections together with the Communist party (the Sozialistische Einheitspartei Deutschlands – SED) under the umbrella of a ‘national front’ which always won 100 per cent of seats. The apportionment of seats among the different parties was established beforehand and decided by the ruling SED. In China the multi-tier elections to the National People’s Congress (NPC) allow today a certain number of non-party members to reach the final stage and in the local tiers of elections the number of candidates may exceed by something like 10 per cent the number of seats to be allocated (allowing some filtering of unpopular candidates) (Diamond and Meyers 2001; Xia 2007). For instance, in the NPC elected in 2018, out of the total 2980 members, 2103 (or 70.6 per cent) are Communist Party members, 470 (15.7 per cent) independents and the others representatives of eight satellite parties of the CCP. More intriguing are elections where some space is left to competition and other actors beside the ruling party are allowed to participate with their candidates without a fully predetermined share of seats. A number of questions arise in these cases. What is really at stake in such elections? How can such electoral processes be acceptable to rulers that do not recognize the legitimacy of opposition? What are the advantages that ruling parties and leaders enjoy? Can citizens express their dissent from the ruling government and can oppositions really win some political space? Which are the obstacles that oppositions must face in these competitions? The increasing number of empirical analyses of such elections have shed more light on the variable set of tools adopted by the ruling actors in order to prevent elections from disrupting the existing allocation of power (Gandhi and Lust-Okar 2009). These tools can operate before, during and after the election itself. They vary from the selective exclusion of parties, or of individual candidates from participating in elections, to registration hurdles, to highly skewed allocations of campaign and propaganda resources (as access to media and funds), to control over voters through bribery or intimidation, to outright violence against opponents, to manipulation of results. Elections in Mexico during PRI dominance (Magaloni 2006), in Jordan and Morocco (Lust-Okar 2005), in Egypt under Mubarak (Blaydes 2010), in Iran (Ansari 2017), in Putin’s Russia (White 2011; Bader and van Ham 2015), in Vietnam (Malesky and Schuler 2010) and in many African countries (Lindberg 2006; Howard and Roessler 2006; Collier and Vicente 2012) display the whole gamut of means used for preventing the victory of opponents or simply to amplify the victory of incumbents. The use of such instruments drastically reduces for incumbents the uncertainty of results while at the same time allowing for a modicum of competition.

Representation in authoritarian regimes  151 The different limitations, as compared with the model of democratic elections, are often also used to define regime typologies (Diamond 2002). This shortcut, which is often used in large N comparative studies, underestimates the fact that the means used (and their effects) may vary sometimes significantly in a conjunctural way as more in-depth empirical analyses have shown (Hyde and Marinov 2011). We must therefore distinguish between large-scale static typological efforts and process-oriented dynamic analyses of regimes. Recent research has also tried to investigate the reasons for the choice of different tools. They depend on the one hand on the nature itself of the regime, as the resources available to different types of non-democratic regimes are variable, but on the other hand also on the specific internal and external circumstances that the regime is facing. Costs and benefits of different instruments vary and we may assume that non-democratic regimes will prefer less costly instruments as long as they seem sufficient to their purposes. The recourse to violence is probably more ‘expensive’ than other means such as co-optation through elections (Gandhi and Przeworski 2007). On the internal side the strengths and resources which opponents can control are also a relevant factor. Some evidence suggests that the pre-emptive use of violence will be more probable when authoritarian incumbents’ fears of being unsaddled are reinforced by negative public opinion forecasts (Hafner-Burton et al. 2014), and after elections when the opposition has enjoyed an unexpected success. To what extent international influences can also play a role is an open question. Again, this may depend on circumstances, and regularity of elections may depend on the greater or lesser need for the regime to obtain external help from democratic countries and international institutions who may want to link it to fair and free voting procedures. Authoritarian regimes counting upon rents from natural resources will be more able to resist international pressures than less resources-rich countries. Although non-democratic regimes make a great deal of effort to reduce the uncertainty of results to their advantage, their control of events may in the course of the electoral process prove insufficient, and unexpected results may materialize (Bunce and Wolchik 2010; Hyde and Marinov 2012). The case of the plebiscite called by Pinochet in 1988 in Chile, which resulted in a clear (and for the dictator unexpected) victory of the opposition and in the downfall of the regime is an outstanding case (Constable and Valenzuela 1989). In similar events the regime faces a choice between accepting the results or resorting to annulling them and possibly using violence to pursue this objective. The military junta of Chile after some internal wavering decided to accept the voters’ choice (and thus opened the way for a regime change); the Egyptian military adopted the opposite course in 2013.

PARLIAMENTARY ASSEMBLIES AND ELECTED PRESIDENTS IN NON-DEMOCRATIC REGIMES Elections in non-democratic regimes are sometimes linked to presidential offices, but even more importantly to the formation of parliamentary assemblies. Improving the understanding of these institutions and their functionality has therefore become an important point on the agenda of studies of authoritarian regimes. The purpose of parliaments in liberal democratic regimes is well established: they provide the institutional space for a pluralistic representation of the multiple interests and orientations which emerge in a free society; and the political relevance of these plural voices is assured by attributing to members of parliament a decisive say on legislation and the possibility of scrutinizing and questioning the executive. To what

152  Research handbook on political representation extent can this happen in the context of authoritarian regimes whose nature is fundamentally anti-pluralistic? To evaluate non-democratic legislatures, two aspects are crucial in this respect: on the one hand the degree of diversity (or limited plurality) admitted in the membership of the parliamentary body, and on the other hand the powers and the freedom of action allowed to the institution and its members. As for the first aspect, a parliamentary assembly, being typically larger than the executive, structurally allows a broader presence of actors in a national institution. We can assume that non-democratic regimes will exploit this opportunity as one among the instruments at their disposal for enhancing their survival chances (Boix and Svolik 2013), and they will do so by expanding in some direction their bases of support. Which variety of representatives this determines, is obviously linked to the practices adopted for elections. What we have seen about elections in non-democratic regimes shows that these elections either eliminate completely or drastically reduce the elements of choice. This means that the composition of such bodies is due more to co-optation (Löwenstein 1973) from above than to delegation from below. Their composition thus reflects the variable purposeful designs conceived by the authoritarian rulers. Some form of descriptive and symbolic representation can be produced by making the parliament into a ‘sample’ of the population and actively recruiting from sections of the population (as manual workers, peasants, women, or ethnic minorities) which are typically underrepresented in truly competitive parliaments, but also in the upper power levels of the authoritarian regime. This practice was typical of communist legislatures, but also of some more recent regimes (Millard 2004; Stockemer 2009). The parliament can be used to show that the regime faithfully ‘represents’ the whole population. Parliamentary membership may also be a way to promote the internal cohesion of the ruling party by recruiting to this position a broader selection of members from the national echelons of the party or by allowing space for local party elites (Malesky and Schuler 2010) and thus to reinforce the loyalty of potential internal dissidents (Magaloni 2008, p. 724). The co-optation strategy may also go beyond the ruling party and provide some limited space for representatives of social and economic elites or for members of the political opposition(s) (Lust-Okar 2005; Blaydes 2010). In some way these parliaments generate some elements of representation: in particular, symbolic representation, but also selective representation of interests. The second question to be addressed concerns the powers and resources that members of these institutions can enjoy. Formal descriptions of such powers are not adequate for this purpose as they do not provide a faithful description of real practices which, on the contrary, can be obtained only by direct observation. Unfortunately, sufficiently detailed empirical research on the working of such institutions is still relatively rare. Comparative studies have hypothesized that legislatures of authoritarian regimes are constructed to enable some participation in policy making as an instrument for gaining the loyalty of opposition forces (Gandhi and Przeworski 2007). Other studies are much more sceptical about the policy-making powers (Blaydes 2010 and Lust-Okar 2005). Systematic comparative evidence on this is still lacking. Some indirect evidence about the policy-making role can be gained simply by assessing the (generally short) duration of their working time, or the (limited) frequency and weight of dissenting votes. Such data, when available, suggest that the power of such bodies to influence decisions (and in particular important ones) should not be overestimated. The experience of the Chinese National People’s Congress indicates, however, that on the less politically charged policy issues parliamentary members may have their proposal accepted (Xia 2007; Truex 2016). Rather than policy making, other aspects may be more important in the life of

Representation in authoritarian regimes  153 legislative bodies. A few in-depth works show for instance that members of parliament, who are less close to the national party level, may exercise (as in Vietnam) some degree of scrutiny and pleading for local interests (Malesky and Schuler 2010). If the powers of such legislatures are generally extremely limited although not completely irrelevant, there is another aspect which should be considered in order to understand the role of these institutions. Being a member of parliament attributes a number of selective benefits (both legitimate and less legitimate), such as a relative immunity from judicial harassment, and privileged access to public goods (Blaydes 2010). An exchange between loyalty to the regime (or soft opposition) and these privileges becomes a relevant function of these institutions. Parliamentary assemblies do not exhaust the study of institutional mechanisms built with the purpose of creating authoritarian mechanisms of (selective) representation. Other collective bodies, such as consultative councils, commissions or more informal inner cliques may play an equally important role in broadening and cementing the bases of the authoritarian regimes.

WHAT ARE THE REASONS FOR ESTABLISHING THEM? AND WITH WHAT CONSEQUENCES? Why do non-democratic regimes incorporate the typical institutions of representative democracy? What is their purpose? Are these institutions instruments of representation or what else? And what are the consequences for the regime of having them instead of doing without them? In this perspective it may be interesting to go back to eighteenth-century France, which offers a historical and enlightening precedent of this contemporary phenomenon. In 1789, King Louis XVI of France, by all (current and past) standards an autocrat, decided to convene the États Généraux of France, that is to say the main representative (by the standards of the time) body of the kingdom, which had been dormant for more than 150 years (since 1614). A regime which was obviously not democratic nor even liberal or constitutional (in the modern sense), felt the need to make recourse to the representative institution then available. The reason was candidly declared by the King himself in the act of convocation: nous avons besoin du concours de nos fidèles sujets pour nous aider à surmonter toutes les difficultés où nous nous trouvons relativement à l’état de nos finances, et pour établir, suivant nos vœux, un ordre constant et invariable dans toutes les parties du gouvernement qui intéressent le bonheur de nos sujets et la prospérité de notre royaume.

This case is interesting not only because of its specific historical relevance, and because it produced an almost instant (albeit not in the end viable) transformation from pre-modern to modern representation, from estates to citizens’ representation, but also because it sheds some light on the possible reasons why a non-accountable ruler may want to activate his subjects politically. Representation was (re-)introduced by the king to generate popular support for his policies in hard times. In recent years a growing body of research has explored the contemporary phenomenon of non-democratic representation with the purpose of establishing the reasons behind it. Given that such institutions have some actual and potential costs for the incumbent regime, they must be compensated by benefits of some kind, otherwise the regime would do without them. We have already seen that authoritarian regimes do their best to limit costs of the adoption of representative institutions by trying to make elections predictable and by limiting the

154  Research handbook on political representation powers of parliamentary bodies. The benefits of this must then be clarified. On this a lively scientific debate has developed. If it may easily be assumed that the final and dominant goal of the regime is its preservation, the intermediate goals and their relevance are more debatable. The targets of the regime are on the one hand the broader public and on the other hand more specific groups and actors of society. Having regularly held elections and an established parliament are instrumental to conveying to the internal (but also the international) public opinion the image of a legitimate regime, based on constitutional norms and supported by large or even overwhelming majorities. Representation approximates here a ‘theatre-like’ model (Pitkin 1967, pp. 25–7). The regime stages a representation of itself as a ‘democratic’ and non-arbitrary regime. Experience shows that such a ‘representation’ (aided by strong control of the media) is often convincing for large parts of the population. And possibly also abroad. More delicate (and less theatrical) is the action directed to important social and economic actors and to potential opponents. Here the representative institutions are tools of selective divide et impera strategies and are oriented to strengthen the cohesion of the ruling coalition. Co-optation may be a crucial instrument in preventing a dangerous opposition from emerging by subtracting from it its leading elements and involving them in the institutional networks of the regime (Gandhi and Przeworski 2007). Along these lines it has been suggested that while commitments of authoritarian rulers towards other actors always entail elements of uncertainty, their credibility is increased by recruitment of such actors (who could turn into potential opponents) in a parliamentary institution (Wright and Escribà-Folch 2012, p. 308; Boix and Svolik 2013) and thus contributes to strengthening ruling coalitions and preventing opponents from destabilizing the regime. Another aspect to be underlined is that such institutions, by creating a limited area of openness and thus balancing to some extent the secrecy which surrounds the internal working of authoritarian regimes, offer to the rulers themselves an instrument to better assess strengths and weaknesses of potential or real opponents and thus to selectively pick and co-opt those who are most useful for the regime’s survival strategy (Blaydes 2010). This discussion is connected with the broader topic of responsiveness (see also Chapters 4 and 25 in this Handbook); a growing number of studies are trying to explore the extent to which authoritarian regimes are in some way responsive to the needs and demands of the population (or segments of it) and the formal and informal channels through which this happens. With special regard to China, recent works have used the concept of responsive authoritarianism to designate the efforts of the regime to respond with policy and administrative changes to expressions of public discontent enabled or even stimulated through controlled channels of communication (see for instance Truex 2016 or Qiaoan and Teets 2020). Further analyses have explored the connections between the presence of representative institutions in authoritarian regimes and economic development. The first findings indicate a positive relationship between the presence of multi-party legislatures and economic growth; this might be due not so much to the ability of such institutions to strongly constrain autocrats but to their potential of representing a plurality of relevant economic and social actors and thus contributing to more market-oriented policies (Wright 2008; Jensen et al. 2014). More in general, these analyses have contributed to the broader field of studies concerned with the stability (and avoidance on internal coups) of authoritarian regimes and with the chances of their democratization (Howard and Roessler 2006; Magaloni 2010).

Representation in authoritarian regimes  155

CONCLUSIONS The existence within non-democratic regimes of different types of institutions that are typical of liberal democracies and of their representative character raises some interesting puzzles that have attracted the interest of an increasing number of theoretical and empirical studies and brought back institutionalist and game theoretical (Svolik 2012) approaches to the study of such regimes. The real impact of these institutions is difficult to establish with precision, however. Do they produce ‘representation effects’, some modicum of bottom-up flows of information, transmission and delegation of interests; are they instruments of (limited) responsiveness and accountability? While democratic representation with all its limitations is a two-way vehicle, through which top-down influence is balanced by robust bottom-up pressure, and both support for the rulers and dissatisfaction can be expressed, the channels of representation adopted by non-democratic regimes are largely top-down exercises which try with a mix of subtle and brutal pressures to boast support and minimize dissent. Effectively manufactured elections can be used to show, in a process that is made to look free (at least to the less sophisticated citizens), that support is large and dissent is minoritarian. As the chance of displacing incumbent rulers is completely precluded (as it is with few exceptions the case) a crucial instrument of democratic representation is thus missing and accountability is forfeited. Both elections and (multi-party) parliaments in authoritarian regimes, however, play some role in delivering information about the mood of society to the rulers, but are even more instruments for managing the dominant coalition and reinforcing its cohesion. A crucial question in this research domain is whether such institutions are just epiphenomena, instruments of political manipulation fully under the control of ruling elites, or whether they are to some extent also constraints for them (Pepinsky 2014). The answers to this question are variable. Quite obviously the constraining strength of such institutions is significantly reduced compared with similar institutions in fully democratic environments. However, under special conditions they can play some role which is not totally subordinated. However, a satisfactory answer to this question requires further and broader empirical studies.

REFERENCES Ansari, Ali M. (2017), Iran under Ahmadinejad: The Politics of Confrontation, London: Routledge. Bader, M. and C. van Ham (2015), ‘What Explains Regional Variation in Election Fraud? Evidence from Russia: A Research Note’, in Post-Soviet Affairs, 31 (6), 514–28. Blaydes, Lisa (2010), Elections and Distributive Politics in Mubarak’s Egypt, Cambridge: Cambridge University Press. Boix, C. and M.W. Svolik (2013), ‘The Foundations of Limited Authoritarian Government: Institutions, Commitment, and Power-sharing in Dictatorships’, The Journal of Politics, 75 (2), 300–16. Bunce, V.J. and S.L. Wolchik (2010), ‘Defeating Dictators: Electoral Change and Stability in Competitive Authoritarian Regimes’, World Politics, 62 (1), 43–86. Carothers, T. (2002), ‘The End of the Transition Paradigm’, Journal of Democracy, 13 (1), 5–21. Collier, D. and Levitsky, S. (1997), ‘Democracy with Adjectives: Conceptual Innovation in Comparative Research? World Politics, 49 (3), 430–51. Collier, P. and P.C. Vicente (2012), ‘Violence, Bribery, and Fraud: The Political Economy of Elections in Sub-Saharan Africa’, Public Choice, 153 (1–2), 117–47. Constable, P. and A. Valenzuela (1989), ‘Chile’s Return to Democracy’, Foreign Affairs, 68 (5), 169–86.

156  Research handbook on political representation Diamond, L. (2002), ‘Thinking about Hybrid Regimes’, Journal of Democracy, 13 (2), 21–35. Diamond, Larry and Ramon H. Myers (eds) (2001), Elections and Democracy in Greater China, Oxford: Oxford University Press. Friedrich, C.J. and Z. Brzezinski (1962), Totalitarian Dictatorship and Autocracy, New York, NY: F.A. Praeger. Gandhi, J. and E. Lust-Okar (2009), ‘Elections under Authoritarianism’, Annual Review of Political Science, 12, 403–22. Gandhi, J. and A. Przeworski (2007), ‘Authoritarian Institutions and the Survival of Autocrats’, Comparative Political Studies, 40 (11), 1279–301. Geddes, B. (2003), Paradigms and Sand Castles: Theory Building and Research Design in Comparative Politics, Ann Arbor, MI: University of Michigan Press. Geddes, B., J. Wright and E. Frantz (2014), ‘Autocratic Breakdown and Regime Transitions: A New Data Set’, Perspectives on Politics, 12 (2), 313–31. Gerschewski, J. (2013), ‘The Three Pillars of Stability: Legitimation, Repression, and Co-optation in Autocratic Regimes’, Democratization, 20 (1), 13–38. Gilison, J.M. (1968), ‘Soviet Elections as a Measure of Dissent: The Missing One Percent’, American Political Science Review, 62 (3), 814–26. Hafner-Burton, E.M., S.D. Hyde and R.S. Jablonski (2014), ‘When do Governments Resort to Election Violence?’, British Journal of Political Science, 44 (1), 149–79. Howard, M.M. and P.G. Roessler (2006), ‘Liberalizing Electoral Outcomes in Competitive Authoritarian Regimes’, American Journal of Political Science, 50 (2), 365–81. Hyde, S.D. and N. Marinov (2011), Codebook for National Elections Across Democracy and Autocracy: NELDA, available from https://​nelda​.co/​#, accessed 22 March 2020. Hyde, S.D. and N. Marinov (2012), ‘Which Elections can be Lost?’, Political Analysis, 20 (2), 191–210. Jensen, N.M., E. Malesky and S. Weymouth (2014), ‘Unbundling the Relationship Between Authoritarian Legislatures and Political Risk’, British Journal of Political Science, 44 (3), 655–84. Jessen, Ralph and Hedwig Richter (eds) (2011), Voting for Hitler and Stalin: Elections under 20th Century Dictatorships, Frankfurt, Germany and New York, NY, USA: Campus Verlag. Karklins, R. (1986), ‘Soviet Elections Revisited: Voter Abstention in Noncompetitive Voting’, American Political Science Review, 80 (2), 449–69. Levitsky, Steven and Lucan A. Way (2010), Competitive Authoritarianism: Hybrid Regimes after the Cold War, Cambridge: Cambridge University Press. Lindberg, Staffan I. (2006), Democracy and Elections in Africa, Baltimore, MD: Johns Hopkins University Press. Linz, Juan J. (1975), ‘Totalitarian and Authoritarian Regimes’, in Fred I. Greenstein and Nelson W. Polsby (eds), Handbook on Political Science, Vol. 3, Reading, MA: Addison Wesley Publishing Company, pp. 175–411. Löwenstein, Karl (1973), Kooptation und Zuwahl: Über die Autonome Bildung Privilegierter Gruppen, Frankfurt: A. Metzner Verlag. Lust-Okar, Ellen (2005), Structuring Conflict in the Arab World: Incumbents, Opponents, and Institutions, Cambridge: Cambridge University Press. Magaloni, Beatriz (2006), Voting for Autocracy: Hegemonic Party Survival and its Demise in Mexico, Cambridge: Cambridge University Press. Magaloni, B. (2008), ‘Credible Power-Sharing and the Longevity of Authoritarian Rule’, Comparative Political Studies, 41 (4–5), 715–41. Magaloni, B. (2010), ‘The Game of Electoral Fraud and the Ousting of Authoritarian Rule’, American Journal of Political Science, 54 (3), 751–65. Malesky, E. and P. Schuler (2010), ‘Nodding or Needling: Analyzing Delegate Responsiveness in an Authoritarian Parliament’, American Political Science Review, 104 (3), 482–502. Millard, Frances (2004), Elections, Parties and Representation in Post-Communist Europe, New York, NY: Palgrave. Pepinsky, T. (2014), ‘The Institutional Turn in Comparative Authoritarianism’, British Journal of Political Science, 44 (3), 631–53. Pitkin, Hanna F. (1967), The Concept of Representation, Berkeley, CA: University of California Press.

Representation in authoritarian regimes  157 Qiaoan, R. and J.C. Teets (2020), ‘Responsive Authoritarianism in China: A Review of Responsiveness in Xi and Hu Administrations’, Journal of Chinese Political Science, 25 (1), 139–53. Sartori, Giovanni (2005 [1976]), Parties and Party Systems: A Framework for Analysis, Colchester: ECPR Press. Schedler, A. (2002), ‘Elections Without Democracy: The Menu of Manipulation’, Journal of Democracy, 13 (2), 36–50. Stockemer, D. (2009), ‘Women’s Parliamentary Representation: Are Women more Highly Represented in (Consolidated) Democracies than in Non-democracies?’, Contemporary Politics, 15 (4), 429–43. Svolik, Milan W. (2012), The Politics of Authoritarian Rule, New York, NY: Cambridge University Press. Truex, Rory (2016), Making Autocracy Work: Representation and Responsiveness in Modern China, Cambridge: Cambridge University Press. White, S. (2011), ‘Elections Russian-Style’, Europe-Asia Studies, 63 (4), 531–56. Wright, J. (2008), ‘Do Authoritarian Institutions Constrain? How Legislatures Affect Economic Growth and Investment’, American Journal of Political Science, 52 (2), 322–43. Wright, J. and A. Escribà-Folch (2012), ‘Authoritarian Institutions and Regime Survival: Transitions to Democracy and Subsequent Autocracy’, British Journal of Political Science, 42 (2), 283–309. Xia, Ming (2007), The People’s Congresses and Governance in China: Toward a Network Mode of Governance, London: Routledge. Zaslavsky, V. and R.J. Brym (1978), ‘The Functions of Elections in the USSR’, Soviet Studies, 30 (3), 362–71.

SECTION III THE INSTITUTIONAL ASPECTS OF REPRESENTATION

INTRODUCTION Maurizio Cotta In liberal democratic regimes, the idea of representation has been translated into specific and robust institutional devices which have deeply characterized the mechanisms of governance. Although institutional choices are never only the result of ideals, but also reflect variable configurations of political actors, balances of power, conveniences and even chance, we should not underestimate the role of some crucial political principles, such as the representative one, in guiding and legitimizing these choices. This section is devoted to systematically analysing how the idea of representation and its variable interpretations have inspired different institutional solutions and the problems inevitably arising in the passage from ideals to reality. Elections and the thick set of rules regulating them are the most obvious terrain on which the idea of representation, with its variable interpretations, meets real life politics. It is no coincidence that here the elaboration and experimentation of different rules have been significant, generating a very rich field of scientific discussion on the reasons and consequences of these choices. In the chapter on ‘Electoral systems and representation’ Chiaramonte analyses the main electoral systems and their formulas in detail, highlighting how they have been shaped by competing interpretations of political representation, by compromises between different views and by interest calculations of the relevant actors. After an in-depth discussion of the differences between majoritarian and proportional principles and of the electoral systems deriving from these systems, but also of the many adjustments of the latter, the chapter directs its attention to the consequences of different electoral systems on some crucial aspects of the practice of representation, such as the competitive behaviour of parties and the determinants of voters’ choices. Notwithstanding the importance of functional and non-territorial interests typically expressed through democratic party competition, the territorial dimension or, rather, different territorial dimensions play a crucial role in the institutional organization of representation and are reflected in the electoral systems as well as in the structure of parliaments. While 158

Section III  159 representation is typically predicated at the national level, elections in most cases follow a subnational territorial articulation and are based on constituencies of variable size. These choices respond to different priorities in the representative equation, such as the problem of proximity with the voters, respect for territorial articulations of interests, proportionality, and so on. At another level, as discussed in the chapter ‘Territorial and multilevel representation’ by Thorlakson, variable definitions of the territorial nature of a country (unitary, federal, regional) are also reflected in the institutional arrangements of representation. The choice between monocameral and different formats of bicameral parliaments (and in the latter case also the choice between different modes of election of the two chambers) may also be seen as the result of different interpretations of representation. In federal systems with bicameral parliaments, two channels of representation are made to coexist. Other sub-national and supra-national authorities can further complicate the representational landscape, opening the way to different mixes of cooperation and competition. Partly connected with the previous theme is that of the relationship between size and representation, which is in fact a continuation of one of the oldest debates in the study of politics. Since classical times, the relationship between the size of a political community and its politics, and in particular between size and democracy, has been widely discussed. In the chapter on ‘Local jurisdiction size and political representation’, Denters focuses on the question of how the size of political jurisdictions to be found at different tiers of governance, at the national, regional and local level, affects the functioning of systems of political representation. Different aspects of the representation game – such as proximity between representatives and represented, quality of the representatives, variety of interests represented, proportionality and pluralism – appear at stake with the choice of smaller or larger jurisdictions. Democratic systems have increasingly recognized that minority groups (ethnic, religious, linguistic, etc.) and also women – often described as a minority because of severe underrepresentation in spite of being numerically a majority – may not find what they deem to be an adequate space through standard systems of representation. This deficit is often attributed to the fact that these are generally not suitable for representing identities. As discussed by Krook in the chapter ‘Electoral quotas and group representation’, the growing mobilization of these groups, and in particular of women, has generated an increasing demand for descriptive representation, seen as better able to give recognition to sections of the electorate that define themselves through identities. The prevailing answer to this demand has been the introduction of group-based quotas. The chapter discusses key variations in their design and the impact on dynamics and quality of political representation as well as the normative dilemmas raised by quotas, particularly with regard to questions of equality, democracy and group recognition. The theme of women representation, while sharing some of the general problems of minority group representation, has special features of its own. As discussed by Lovenduski in the chapter ‘The political representation of women: a feminist institutionalist perspective’, it is one that requires special attention as it involves the most basic line of differentiation between human beings (and one which is highly charged, both symbolically and culturally). The persisting underrepresentation and weaker role of women in political institutions (not to mention economic ones) is not simply, as the author argues, a problem that can be tackled with quotas, but should involve a deeper reshaping of representative institutions which have been powerfully shaped over the years by the culture of masculinity. In the final chapter on ‘Ethnic minorities and representation’ Hänni and Saalfeld explore some old and new problems which these types of minority face in democratic life. The

160  Research handbook on political representation prevalence of the majority principle puts their political influence at risk and raises important normative questions. It also poses the problem of their peaceful integration in the democratic system. As descriptive representation, based on the assumption that the interests of minorities are better represented by advocates of their own ethnic group, is the most common answer to this problem, specially tailored electoral clauses provide the typical institutional solution. While this solution is often applied to autochthonous ethnic minorities, the problems of integration and representation of immigrant ethnics, which in many countries today make up a significant part of the population, are generally more complex and less easily solved as they are also mixed up with the problem of their citizenship.

13. Electoral systems and representation Alessandro Chiaramonte

According to Douglas Rae’s (1967) definition, an electoral system is a set of rules governing the process through which voters’ preferences are converted into votes and votes are translated into seats. By no means, however, does this definition imply that an electoral system is merely a technical tool. On the contrary, it is a very important link in the chain of political representation, in that it establishes how officials who make decisions on behalf of others are elected. An electoral system, in fact, affects the strategies of parties and candidates participating in an election and the way voters cast their ballots, as much as how votes are counted and count for the election of representatives and for the balance of power among political parties (Duverger 1951; Cox 1997). Furthermore, an electoral system extends its effects to the party system, that is, to the number of parties and the pattern of their interaction (Sartori 1976), and also to the government and its stability (Eckstein 1963; Lijphart 1999). The goal of this chapter is to see exactly how electoral systems connect to the idea of representation and, more specifically, how different types of electoral system connect to different ideas of representation. For this purpose, in the first section the literature on electoral systems is reviewed as it relates to their classification and the sometimes contradictory meanings associated to widely used terms such as ‘majoritarian’ and ‘proportional’. In the second section the focus of the analysis is more specifically on the majoritarian and proportional principles of representation, which, as we will see, are built into the electoral systems. The third section addresses the case of mixed electoral systems, which combine majoritarian and proportional arrangements, and discusses whether distinct principles of representation may coexist. In the concluding section the implications for democracy and its functioning of the choice of electoral systems and principles of representation are discussed.

CLASSIFYING ELECTORAL SYSTEMS: E PLURIBUS DUO? The search for optimal criteria to define and classify electoral systems has been a recurrent goal of research in political science as well as in the field of constitutional law. The complexity and variety of electoral systems make it almost impossible to reduce them to a few clearly identified categories. The different classifications that are found in the literature reflect this difficulty and testify that to date there is no complete agreement among scholars on how to deal with this issue. The simplest criterion used for classifying electoral systems is the type of electoral formula adopted, that is, the mathematical method for converting votes into seats. On this basis, electoral systems are qualified as majoritarian if they adopt the plurality or majority formulas, while they are qualified as proportional if they use formulas such as d’Hondt, Saint-Laguë, Hare, and so on.1 (Lakeman and Lambert 1955; Ross 1955; Van den Berg 1956; Bogdanor 1981; Butler 1981; Harrop and Miller 1987; Urwin 1987a; 1987b). 161

162  Research handbook on political representation This kind of classification certainly has the advantage of parsimony. Yet, it has been objected that in doing so, electoral formulas, not electoral systems themselves, are classified (Blais and Massicotte 1996; 1997). Indeed, as emphasized by many scholars (among others: Rae 1967; Taagepera and Shugart 1989; Lijphart 1994; Norris 2004), the formula is only one out of the many dimensions of which electoral systems consist and, as a consequence, a more comprehensive typology is needed to account for the multi-dimensionality of electoral systems. Here the operational definition of the concept of electoral system comes into play. Rae (1967) identifies the districts magnitude (the number of seats they return) and the ballot structure (the type of vote available to the voter), beyond the electoral formula, as the main dimensions of any electoral system. Further refinements to this operational definition have led to the identification of other distinct dimensions such as the legal thresholds of representation and the existence of multiple levels of seat allocation (Blais 1988; Taagepera and Shugart 1989; Gallagher and Mitchell 2005), as well as the assembly size, the possibility of malapportionment, whether parliamentary and presidential elections are held at the same time, and provisions regarding the apparentement among lists (Lijphart 1994). The typologies of electoral systems built on the basis (of some) of the constitutive dimensions of the electoral systems (Lakeman 1970; Nohlen 1978; 1995; Taylor and Johnston 1979; Newland 1982; Rose 1983; Reeve and Ware 1992; Lijphart 1994; Norris 1997; 2004; Farrell 2001) undoubtedly offer a wide range of information, but at the same time they have proved not to be parsimonious. So much so that even these scholars had no choice but to resort to the usual majoritarian–proportional dichotomy in order to identify the main types of electoral systems (see, for example, Blais 1988; 1991). An even more radical alternative to a classification of electoral systems based on the electoral formula utilized is represented by an approach that relies on the outputs (i.e., the ‘what it does’ of the electoral system), rather than on the legal inputs (i.e., the ‘how it is’ of the electoral system). Shifting the focus of the analysis from the inputs to the outputs has given rise to classifications based on the effects produced by electoral systems, particularly on the process of translating votes into seats and on the party system as a whole. Duverger (1951) was the first to move in this direction, although his goal was not precisely to provide a classification of electoral systems, but rather to analyse their consequences. As is well known, Duverger distinguishes between plurality systems on one side and (double-ballot) majority and proportional systems on the other side: the former tend to produce a two-party system, while the latter tend to produce a multi-party system.2 Beyond the methodological criticisms to his propositions, Sartori (1968; 1994) objected to the discriminating capacity of Duverger’s classification because the same party system format was associated with very different electoral systems such as the majority and proportional ones, and proposed that electoral systems should be distinguished on the extent of their manipulative capacity on voters’ choices, rather than on the majoritarian/proportional alternative. Thus, according to Sartori, electoral systems that are conventionally defined as proportional can instead be identified to be ‘strong’ like majoritarian electoral systems, for example by virtue of high legal thresholds or small-size districts. Along the same lines, Rose (1984) argued that electoral systems should not be classified into mutually exclusive categories, but be understood as points on a continuum defined by their degrees of proportionality. By doing so, he found that the differences between majoritarian and proportional systems are actually smaller than they were supposed to be. Although the idea of a continuum of electoral systems may seem convincing and has become widely accepted, the use of one criterion only – be it the degree of manipulative efficacy on

Electoral systems and representation  163 the voters as it is for Sartori, or the degree of proportionality, as it is for Rose – for the classification of electoral systems remains questionable. In particular, the degree of proportionality appears inadequate to capture the psychological effects that electoral systems exert before the election on voters’ choices and parties’ competitive strategies. Moreover, and even more importantly, it would be theoretically misleading to reduce the meaning of the ‘majoritarian’ concept to that of ‘non- (or dis-) proportional’.

‘MAJORITARIAN’ AND ‘PROPORTIONAL’ AS PRINCIPLES OF REPRESENTATION As we have seen in the previous section from the literature review on the classification of electoral systems, if we analyse the way in which the terms ‘majoritarian’ and ‘proportional’ are used and, therefore, the meaning that is explicitly or implicitly attributed to them, we derive that they are used to indicate, alternatively, or even simultaneously: (1) a distinct family of electoral formulas, that is, a set of mathematical methods for distributing and awarding seats; (2) a distinct family of electoral systems, characterized by a number of dimensions (mainly, the district magnitude and the vote structure) beyond the electoral formula; (3) an extreme of a continuum which includes the entire range of (dis)proportional effects that an electoral system can produce in the translation of votes into seats. From this brief examination, it should be concluded that these terms suffer from conceptual stretching and should be disregarded in that they have lost their original discriminating capacity. Yet, it seems impossible to do without them. Like few other terms in the field of political science, in fact, the terms ‘majoritarian’ and ‘proportional’ are as evocative as they are essential. This is so because they refer not only to specific technical tools such as the electoral formula or the other dimensions of which an electoral system consists, or to a measure of how the distribution of seats approximates the distribution of votes among political parties, but above all – as Nohlen (1984) puts it – to two distinct principles of representation. The idea of principle of representation is historically determined and comes first compared with that of electoral system. The kind of relationship existing between the two is a relationship between ends and means. In other words, an electoral system incorporates a principle of representation that defines its objectives. The majoritarian principle was originally conceived to solve the problem of who should represent communities like counties and boroughs, where socio-economic interests were located, and later – when civil and political rights were extended – of who should represent the persons. The representatives ought to be the ones receiving the largest support in terms of votes. In any case, this principle of representation is connected to the idea of the constituency, which is both a territorial unit and a unit of population (Reeve and Ware 1992; Blackburn 1995). With the emergence of modern political parties and the rise of mass politics, the majoritarian principle of representation further extended its scope to include the selection of the government. According to Nohlen (1984), the goal of a majoritarian electoral system is exactly that of reaching a majority for a party or a party alliance. The crucial element is to ensure that a political party (alliance) that has not obtained the majority of votes is still able to form a government majority. Ultimately, the essence of a majoritarian system (and the criterion for evaluating its effects) is its ability to produce governments.

164  Research handbook on political representation On the other hand, the proportional principle is inextricably linked to the development of representative democracy, and particularly to the suffrage extension and the emergence of mass parties. A proportional electoral system aims to reflect, as accurately as possible, the social groups and political forces of the population. For each party, the share of votes and the share of seats should correspond to each other approximately. This is the fundamental function of the principle of proportional representation and the criterion for evaluating the effects of a proportional electoral system.3 The two principles of representation permeate electoral systems mainly through the electoral formula. More specifically, the majoritarian principle postulates the indivisibility of the stakes (typically, seats) where they are awarded (usually, but not necessarily, in single-member districts), and therefore a zero-sum game type of competition with clear winners and losers.4 The proportional principle provides for an allocation of the stakes among the contestants (typically, party lists) proportionally to their share of votes. If we accept that electoral systems incorporate a historically determined principle of representation and it is with respect to it that they mainly differ, it follows that, contrary to Sartori, Rose and others, the distinction between majoritarian and proportional systems is a matter of quality and not of degree. In other words, the terms ‘majoritarian’ and ‘proportional’ cannot be conceived as the extremes of a continuum: they do not share a common dimension, but are incommensurable. As a consequence, it is not legitimate to evaluate the electoral systems incorporating one specific principle of representation through normative criteria deriving from the other. So, for example, a majoritarian electoral system should not be evaluated on the basis of a criterion such as the (dis)proportionality in the distribution of the seats with respect to the votes, because a low disproportionality in the transformation of votes into seats is not its objective. At the same time, a proportional electoral system should not be evaluated on the basis of its ability to ‘manufacture’ a majority of seats for a party or an alliance of parties that has only a plurality of votes, because its objective is not that of producing a government. Of course, the objectives defined by the built-in principle of representation may not always be achieved by an electoral system in its concrete application. In the event that third political forces emerge successfully, a majoritarian system could not be able to ensure an overall outcome with a winner – be it an individual party or coalition of parties – with an absolute majority of seats and to produce a government as a result. A proportional system with small-size districts and a high legal threshold could lead to a large disproportionality. These outcomes should be considered as wanted or unwanted, expected or unexpected empirical deviations from or approximations to the objectives associated with the electoral systems’ principle of representation. However, while as it relates to proportional electoral systems, disproportionality may derive only from the alteration of institutional elements such as the provision of small-size districts and a high legal threshold, the circumstance of majoritarian electoral systems not producing a government comes from the presence of a growing number of parties and/or from a certain cross-districts distribution of votes among them, which, in turn, may be the consequence of both social dynamics and institutional constraints.5 From this point of view, it must be noted that the achievement of the objective built in majoritarian systems is less predictable than that of the objective built in proportional systems.

Electoral systems and representation  165

BEYOND THE MAJORITARIAN/PROPORTIONAL DICHOTOMY? THE CASE OF MIXED ELECTORAL SYSTEMS Since the 1990s many electoral reforms have been characterized by the introduction of mixed electoral systems, which have a composite nature as they combine elements of majoritarian systems and proportional systems. The growing popularity of these systems has brought about a lot of scholarly attention (Massicotte and Blais 1999; Shugart and Wattenberg 2001a; Moser and Scheiner 2004; Ferrara et al. 2005; Batto et al. 2016), but also a certain degree of confusion. For the sake of our argumentation, it is particularly relevant to understand whether and how mixed electoral systems integrate two distinct and incompatible principles of representation such as the majoritarian and the proportional ones we described in the previous section, and whether and to what extent they represent an alternative to the traditional majoritarian/proportional dichotomy. To start with, Massicotte and Blais consider an electoral system to be mixed ‘if its mechanics involves the combination of different electoral formulas (plurality or PR; majority or PR) for an election to a single body’ (1999, p. 345; italics in original), provided that the representatives elected thanks to one of the two formulas are at least 5 per cent of those elected thanks to the other.6 That said, mixed electoral systems are also as such because they try to integrate the majoritarian and proportional representation principles, that is, two distinct logics of competition between parties and selection of representatives. This is not to say, however, that they give birth to a third principle of representation, other than the majoritarian and the proportional ones. Actually, as it regards the principles of representation, tertium non datur.7 Mixed electoral systems rather aspire to combine the best of both worlds (Lijphart 1984; Shugart and Wattenberg 2001c), that is to say, the best of the majoritarian ‘world’ as well as of the proportional ‘world’. But here the complications arise. Since these two principles of representation that we have so far considered incommensurable have to live together in the same electoral system, it is necessary to understand how this may be possible and what to expect from it. In some cases, the coexistence of the majoritarian and proportional principles of representation in the same electoral system is made possible by the fact that one of the two is hierarchically over-ordered. Let’s take for example the mixed electoral system currently in place in Germany for the election of the Bundestag. While providing for the election of half of the representatives in single-member districts under the plurality formula, a number of seats corresponding to the total are proportionally allocated among party lists. Thus, the ‘decisive’ tier for the purpose of calculating the share of seats due to the parties is the proportional one: it is at this level, in fact, that all the seats at stake are attributed to the parties, while the seats awarded to the candidates winning the single-member districts are subtracted from the number of seats proportionally due to the parties these candidates belong to. It is therefore clear that the proportional principle of representation – not by chance embedded in the German constitution – prevails in any case. As for the majoritarian principle, it is incorporated in the system, but it operates in so far as it does not conflict with the proportional principle. This is also true for all the cases that belong to the ‘dependent combination’ category in Massicotte and Blais’ (1999) typology of mixed electoral systems, which also includes the proportional systems with a majority premium such as the one that was adopted in Italy for the elections of the Chamber of deputies in 2006, 2008 and 2013 (Chiaramonte 2015). This latter system established that the party or coalition of parties with the largest number of votes nationally must obtain at least

166  Research handbook on political representation 55 per cent of the total seats. If the winning party or coalition does not receive this share of seats thanks to the proportional distribution (which never happened), the majority premium is awarded: 55 per cent of the total seats are then attributed to the winner, while the remaining seats are allocated proportionally to the other parties and coalitions. In contrast to the German system, here the majoritarian principle of representation is predominant, and the proportional principle can operate in so far as it does not conflict with the former. In some other cases, however, when the mixed electoral systems are of the ‘independent combination’ type (Massicotte and Blais 1999), the relationship between the two principles of representation is much more complex. In these electoral systems the majoritarian and proportional tiers operate independently of one another in selecting the representatives and in determining the allocation of seats to parties, as it is for example in Japan, and in Italy under the Rosato law.8 From the empirical point of view, the consequences of their application can be detected in terms of the degree of disproportionality and party fragmentation, and presence of manufactured majorities, and can be measured in terms of approximation to the ideal goals of each of the two principles of representation. In this sense, a trade-off between these two ideal goals, or, better, between the likelihoods to achieve them, may still be conceived. But from the normative point of view, the idea of a compromise of any kind is simply inconceivable. As a consequence, the effects of these mixed electoral systems may greatly vary and may not be consistent with any principle of representation. The contrast between the empirical and normative perspectives is reflected in the diversity of considerations on whether mixed electoral systems truly represent the best of both worlds. On the one hand, Shugart and Wattenberg claim that mixed electoral systems ‘simultaneously encourage divergent incentives that lead party systems to exhibit many of the features of an “efficient” and a desirable balance’ (2001c, p. 591); and add that ‘they permit myriad variations that can suit a specific political context, while still holding out the promise of providing the best of both worlds – i.e. the best of both identifiable governing blocs and proportionality’ (2001c, p. 595). On the opposite side, Sartori (1994) considers the coupling of the majoritarian and proportional principles of representation a bad marriage, which can only give birth to a ‘bastard’.

CONCLUDING REMARKS: ELECTORAL SYSTEMS, REPRESENTATION AND DEMOCRACY Elections are the core element of representative democracies. Electoral systems, which govern the process through which representatives are elected and seats are allocated to political parties, are therefore a crucial institution in any democracy. However, it is well known that electoral systems are not neutral. They can be designed in myriad ways and have very diverging consequences. Yet their diversity may still be reduced to a large extent to the majoritarian/proportional alternative, that is, to the choice between two alternative principles of representation that are historically embedded. As seen in the previous sections, the majoritarian principle of representation revolves around the ideas of local accountability – that is, representatives should be selected by and respond to territorial constituencies – and of a manufactured single-party (or party-alliance) government – that is, the plurality of votes received by the largest party or coalition should be transformed into a majority of seats, allowing for the formation of a government as a result.

Electoral systems and representation  167 The proportional principle of representation, on the contrary, is connected to the idea of fair representation – that is, the social diversity and diversity of opinions should be accurately reflected by (parties in the) representative assemblies. These principles of representation, which are built into the electoral systems, are able to shape the working of democracy. In as far as they create direct voter–government links allowing for manufactured single-party (or party-alliance) governments, majoritarian electoral systems ‘provide government stability, which in turn enhances political stability, government cohesion and thus stronger leadership, and finally, decisive elections, which allow greater accountability to the electorate’ (Blais 1991, pp. 242–3); proportional systems in turn provide ‘a more accurate mirror of opinions, which makes for a fairer and broader representation, thus ensuring responsiveness, legitimacy, and order’ (Blais 1991, p. 246). More in general, Lijphart (1999) argues that majoritarian systems are winner-takes-all methods and hence are a perfect reflection of the majoritarian philosophy permeating the Westminster model of democracy, while, in contrast, proportional systems fit well into the consensual model of democracy, whose consensus-building mechanisms foster harmony and order. In line with Lijphart, Powell (2000) maintains that the ‘majoritarian vision of democracy’ is characterized by electoral systems producing a clear mandate for one party and, thus, has the advantages of accountability and coherence, while the ‘proportional vision of democracy’, centred on proportional electoral systems, promotes negotiations and compromises, but is also able to produce a greater policy congruence between citizens and governments. Whatever it is, institutional engineers are warned that electoral reforms, particularly those implying a change in the principle of representation, may have broad consequences for the functioning of democracy and therefore should be conceived in conjunction with the overall institutional architecture of the political system.

NOTES 1. However, according to some scholars it remains unclear whether some formulas may be considered as majoritarian or proportional. This is the case, for example, of the single non-transferable vote (SNTV) or, more in general, of the variants of the so-called limited vote. Lakeman (1970), Bogdanor (1981), Lijphart (1984; 1994) and Reynolds and Reilly (1997) place these formulas within a group that they call semi-pro­portional. It is not so for Cox (1997), who considers these formulas as plurality-like and, thus, belonging to the majoritarian type. On this point see also Grofman (1999). 2. Such propositions have been the subject of an academic debate that has continued for years and has regarded both their exact scientific status (Duverger 1986; Riker 1982; Sartori 1968; 1986) and the direction of causality of the hypothesized relationships (Grumm 1958; Rokkan 1970). At any rate, they stimulated a flourishing of works not only about the variety of electoral systems applied in democracies, but also about their consequences on the party systems in terms of what Duverger regarded as ‘mechanical effects’ (mainly disproportionality and party fragmentation) (Rae 1967; Taagepera and Shugart 1989; Gallagher 1991; 1992; Lijphart 1994), and as ‘psychological effects’ (Cox 1997). 3. The two principles of representation may also be said to be linked to two distinct purposes of the election: the selection of representatives and the overall composition of the assembly by party (Dummett 1997). The majoritarian principle emphasizes the former purpose, while the proportional principle emphasizes the latter. 4. To put it more formally, if M seats are at stake in a given district, they are awarded to the top-M contestants with the highest number of votes. In some cases, the awarding of the seat(s) at stake requires the majority of votes, which may be obtained in one or more rounds of voting.

168  Research handbook on political representation 5. Indeed, given that party fragmentation has increased significantly for the past two to three decades in most democratic countries, in those adopting a parliamentary system and majoritarian electoral rules the goal of producing a government becomes harder to attain in the current time. 6. A slightly different definition is provided by Shugart and Wattenberg (2001b), according to whom mixed electoral systems (better, mixed-member electoral systems, in their words) consist of a superposition of a nominal tier and a list tier of seat allocation, where seats are awarded to individual candidates and (coalition of) party-lists respectively. 7. At the same time, the application of mixed electoral systems should not necessarily be expected to exert intermediate effects compared with those of majoritarian and proportional systems. For example, in terms of disproportionality in the translation of votes into seats, some mixed electoral systems may happen to result in being less disproportional than some proportional electoral systems and more disproportional than some majoritarian electoral systems. 8. In Japan, roughly 60 per cent of the total seats at stake for the election of the House of Representatives are attributed to candidates running in single-member districts according to plurality rule, while the remaining seats are allocated to party-lists through a proportional formula. The Rosato law, introduced in 2017 in Italy for the election of both the Chamber of Deputies and the Senate, is similar to the current Japanese electoral law, the main difference being the shares of plurality and proportional seats, which are about 35 per cent and 65 per cent, respectively.

REFERENCES Batto, N.F, C. Huang, A.C. Tan and G.W. Cox (eds) (2016), Mixed-Member Electoral Systems in Constitutional Context: Taiwan, Japan, and Beyond, Ann Arbor, MI: University of Michigan Press. Blackburn, R. (1995), The Electoral System in Britain, New York, NY: St Martin’s Press. Blais, A. (1988), ‘The Classification of Electoral Systems’, European Journal of Political Research, 16 (1), 99–110. Blais, A. (1991), ‘The Debate over Electoral Systems’, International Political Science Review, 12 (3), 239–60. Blais, A. and L. Massicotte (1996), ‘Electoral Systems’, in L. LeDuc, R.G. Niemi and P. Norris (eds), Comparing Democracies: Elections and Voting in Global Perspective, Thousand Oaks, CA: Sage Publications, pp. 49–81. Blais, A. and L. Massicotte (1997), ‘Electoral Formulas: A Macroscopic Perspective’, European Journal of Political Research, 32 (1), 107–29. Bogdanor, V. (1981), The People and the Party System: The Referendum and Electoral Reform in British Politics, Cambridge: Cambridge University Press. Butler, D. (1981), ‘Electoral Systems’, in D. Butler, H.R. Penniman and A. Ranney (eds), Democracy at the Polls: A Comparative Study of Competitive National Elections, Washington, DC, USA and London, UK: American Enterprise Institute for Public Policy Research. Chiaramonte, A. (2015), ‘The Unfinished Story of Electoral Reforms in Italy’, Contemporary Italian Politics, 7 (1), 10–26. Cox, G.W. (1997), Making Votes Count: Strategic Coordination in the World’s Electoral Systems, Cambridge: Cambridge University Press. Dummett, M. (1997), Principles of Electoral Reform, Oxford: Oxford University Press. Duverger, M. (1951), Les Partis Politiques, Paris: Colin. Duverger, M. (1986), ‘Duverger’s Law: Forty Years Later’, in B. Grofman and A. Lijphart (eds), Electoral Laws and their Political Consequences, New York, NY: Agathon Press, pp. 69–84. Eckstein, H. (1963), ‘The Impact of Electoral Systems on Representative Government’, in H. Eckstein and D.E. Apter (eds), Comparative Politics: A Reader, New York, NY: The Free Press of Glencoe, pp. 247–54. Farrell, D.M. (2001), Electoral Systems: A Comparative Introduction, Basingstoke: Palgrave. Ferrara, F., E.S. Herron and M. Nishikawa (2005), Mixed Electoral Systems: Contamination and its Consequences, New York, NY: Palgrave Macmillan.

Electoral systems and representation  169 Gallagher, M. (1991), ‘Proportionality, Disproportionality and Electoral Systems’, Electoral Studies, 10 (1), 33–51. Gallagher, M. (1992), ‘Comparing Proportional Representation Electoral Systems: Quotas, Thresholds, Paradoxes and Majorities’, British Journal of Political Science, 22 (4), 469–96. Gallagher, M. and P. Mitchell (2005), ‘Introduction to Electoral Systems’, in M. Gallagher and P. Mitchell (eds), The Politics of Electoral Systems, Oxford: Oxford University Press, pp. 3–23. Grofman, B. (1999), ‘SNTV, STV and Single-Member-District Systems: Theoretical Comparisons and Contrasts’, in B. Grofman, S.C. Lee, E.A. Winckler and B. Woodall (eds), Elections in Japan, Korea and Taiwan under the Single Non-Transferable Vote, Ann Arbor, MI: The University of Michigan Press, pp. 317–33. Grumm, J. (1958), ‘Theories of Electoral Systems’, Midwest Journal of Political Science, 2 (4), 357–76. Harrop, M. and W.L. Miller (1987), Elections and Voters: A Comparative Introduction, London: Macmillan. Lakeman, E. (1970), How Democracies Vote: A Study of Electoral Systems, London: Faber & Faber. Lakeman, E. and J.D. Lambert (1955), Voting in Democracies: A Study of Majority and Proportional Electoral Systems, London: Faber & Faber. Lijphart, A. (1984), ‘Trying to Have the Best of Both Worlds: Semi-Proportional and Mixed Systems’, in A. Lijphart and B. Grofman (eds), Choosing an Electoral System: Issues and Alternatives, New York, NY: Praeger Publishers, pp. 207–13. Lijphart, A. (1994), Electoral Systems and Party Systems: A Study of Twenty-Seven Democracies, 1945–1990, Oxford: Oxford University Press. Lijphart, A. (1999), Patterns of Democracy: Government Forms and Performances in Thirty-Six Countries, New Haven, CT: Yale University Press. Massicotte, L. and A. Blais (1999), ‘Mixed Electoral Systems: A Conceptual and Empirical Survey’, Electoral Studies, 18 (3), 341–66. Moser, R.G. and E. Scheiner (2004), ‘Mixed Electoral Systems and Electoral System Effects: Controlled Comparison and Cross-National Analysis’, Electoral Studies, 23 (4), 575–99. Newland, R.A. (1982), Comparative Electoral Systems, London: The Arthur McDougall Fund. Nohlen, D. (1978), Wahlsysteme der Welt: Daten und Analysen. Ein Handbuch, Munich: Piper Verlag. Nohlen, D. (1984), ‘Two Incompatible Principles of Representation’, in A. Lijphart and B. Grofman (eds), Choosing an Electoral System: Issues and Alternatives, New York, NY: Praeger Publishers, pp. 83–9. Nohlen, D. (1995), Elecciones y Sistemas Electorales, Fundacion Friedrich Ebert, Caracas: Nueva Sociedad. Norris, P. (1997), ‘Choosing Electoral Systems: Proportional, Majoritarian, and Mixed Systems’, International Political Science Review, 18 (3), 297–312. Norris, P. (2004), Electoral Engineering: Voting Rules and Political Behavior, Cambridge: Cambridge University Press. Powell, G.B. (2000), Elections as Instruments of Democracy: Majoritarian and Proportional Visions, New Haven, CT: Yale University Press. Rae, D.W. (1967), The Political Consequences of Electoral Laws, New Haven, CT: Yale University Press. Reeve, A. and A.J. Ware (1992), Electoral Systems: A Comparative and Theoretical Introduction, London: Routledge. Reynolds, A. and B. Reilly (1997), The International IDEA Handbook of Electoral System Design, Stockholm: International Institute for Democracy and Electoral Assistance. Riker, W. (1982), ‘The Two-Party System and Duverger’s Law: An Essay on the History of Political Science’, American Political Science Review, 76 (4), 753–66. Rokkan, S. (1970), Citizens, Elections, Parties, Oslo: Universitetsforlaget. Rose, R. (1983), ‘Elections and Electoral Systems: Choices and Alternatives’, in V. Bogdanor and D. Butler (eds), Democracy and Elections: Electoral Systems and their Political Consequences, Cambridge: Cambridge University Press, pp. 20–45. Rose, R. (1984), ‘Electoral Systems: A Question of Degree or of Principle?’, in A. Lijphart and B. Grofman (eds), Choosing an Electoral System: Issues and Alternatives, New York, NY: Praeger Publishers, pp. 73–81.

170  Research handbook on political representation Ross, J.F.S. (1955), Election and Electors: Studies in Democratic Representation, London: Spottiswoode & Eyre. Sartori, G. (1968), ‘Political Development and Political Engineering’, in J.D. Montgomery and A.O. Hirschmann (eds), Public Policy, vol. 17, Cambridge, MA: Harvard University Press, pp. 261–98. Sartori, G. (1976), Parties and Party Systems: A Framework of Analysis, New York, NY: Cambridge University Press. Sartori, G. (1986), ‘The Influence of Electoral Systems: Faulty Laws or Faulty Method?’, in B. Grofman and A. Lijphart (eds), Electoral Laws and their Political Consequences, New York, NY: Agathon Press, pp. 43–68. Sartori, G. (1994), Comparative Constitutional Engineering: An Inquiry into Structures, Incentives, Outcomes, London: Macmillan. Shugart, M.S. and M.P. Wattenberg (eds) (2001a), Mixed-Member Electoral Systems: The Best of Both Worlds?, Oxford: Oxford University Press. Shugart, M.S. and M.P. Wattenberg (2001b), ‘Mixed-Member Electoral Systems: A Definition and Typology’, in M.S. Shugart and M.P. Wattenberg (eds), Mixed-Member Electoral Systems: The Best of Both Worlds?, Oxford: Oxford University Press, pp. 9–24. Shugart, M.S. and M.P. Wattenberg (2001c), ‘Conclusion: Are Mixed-Member Systems the Best of Both Worlds?’, in M.S. Shugart and M.P. Wattenberg (eds), Mixed-Member Electoral Systems: The Best of Both Worlds?, Oxford: Oxford University Press, pp. 571–96. Taagepera, R. and M.S. Shugart (1989), Seats & Votes: The Effects and Determinants of Electoral System, New Haven, MA, USA and London, UK: Yale University Press. Taylor, P.J. and R.J. Johnston (1979), Geography of Elections, London: Penguin. Urwin, D. (1987a), ‘Choosing Representatives: Majority Electoral Systems’, Social Studies Review, 2 (4), 23–30. Urwin, D. (1987b), ‘Electing Representatives: Proportional Electoral Systems’, Social Studies Review, 2 (5), 1–12. Van den Berg, G. (1956), Unity in Diversity: A Systematic, Critical Analysis of All Electoral Systems, London: Batsford.

14. Territorial and multilevel representation Lori Thorlakson

In the modern state, we usually understand representation in territorial terms. Representative government typically has a territorial component, based in the spatial boundaries of the electoral districts. Already in pre-modern polities, this territorial element was present and was implemented by the representation afforded to localities such as counties and cities through local charters and petitions (Knights 2010). This conception of representation has expressed itself in the use of imperative mandates, which allowed territorially-defined constituencies to bind their representatives. These devices draw on delegate notions of representation, in which elected representatives are expected to act on behalf of the interests and preferences of their constituents in a Madisonian fashion, rather than as trustees left to act according to their best judgment, in a Burkean fashion (Pitkin 1967). The development of the modern democratic state led to the rise of an understanding of representation that attenuated this emphasis on territorial representation, including not only narrow territorial interests, but also the representation of the whole nation or people. Nevertheless, a strong territorial element of representation persists in modern democracies. Most fundamentally, citizenship is traditionally understood as a form of membership in the territorialized state. Spatial boundaries structure state authority and circumscribe the reach of institutions. Electoral districts are territorially defined, and, especially in single member constituencies, expectations of constituency service by elected representatives strengthen territorial representation (see Chapter 22 in this Handbook). In some political systems, such as federations, where territorially-delimited groups make claims for enhanced representation, territorial representation takes an even more central role. Territorial aspects of representation are so widespread in the modern state, that to identify their reach and parameters it is helpful to start by considering various forms of non-territorial representation. Non-territorial representation can refer to political representation in the deterritorialized state or the post-territorial state, linked to notions of citizens that transcend the state – such as transnational or extraterritorial citizenship (see Collyer 2014). Rights are attached to the citizen, and are not directly linked to the territorial state. Another example of this form of transnational citizenship can be found with European integration, which provides EU citizens with political rights beyond the borders of their member state of citizenship. Sometimes, political representation seeks to transcend the territorial state as a way of contesting the boundaries of the territorial state. This happens when parties use appeals based on kin-state nationalism to expand their representational role beyond territorial borders. In Hungary, for example, FIDESZ uses diaspora policies that seek to secure legal rights for Hungarian ethnic minorities in neighbouring countries. In Catalonia, the Republican Left party claims to represent Catalonians in Spain as well as across the border in France. Territorial representation can also be contrasted with different bases of representation, including functional representation. Functional representation occurs when political organization is based upon identities or cleavages that can cross-cut territory, such as class or religion. Another example is when social classes or professions are represented in bicameral legisla171

172  Research handbook on political representation tures (such as in the Irish upper house or Seanad Élreann, in which members of different vocational groups are represented). While political geographers note the importance of territorial and spatial politics in shaping relationships between centre and periphery and shaping political, economic and cultural life (Rokkan and Urwin 1983, p. 4), the development of modern party systems and the modern state has largely been marked by the gradual displacement of territorial interests by functional ones. Historic class conflicts have typically generated the dominant political cleavage structuring party systems in industrialized democracies. This has given rise to party organizations that represent economic interests and ideologies, and party systems that structure conflict in a way that counterbalances the particular interests shaped by territory. Daniele Caramani emphasizes the ‘dualism’ of ‘territoriality’ and ‘functionality’ in his account of how functional cleavages gradually displaced territorial ones in long-term processes of political developments that produced national electorates in European states (Caramani 2004, p. 15). The exception to this is in systems where territory has been politicized, creating regional and ethnic parties.

FEDERALISM: DECENTRALIZING DECISION-MAKING AND REPRESENTATION Federalism is a formal institutional means of enhancing territorial representation. It does this through providing a measure of sub-national territorial autonomy coupled with some form of territorial representation of the territorial units at the centre. The justification for providing enhanced territorial representation through federal arrangements is sometimes made on the grounds of efficiency. Federalism and decentralization allow policies to match local preferences. Decentralization brings representation closer to the people, allowing for effective and responsive ‘local’ representation because the spatial aggregation of interests occurs on a smaller scale, thus more faithfully capturing the policy preferences of voters. The representation of territory becomes especially important and meaningful when the territorial divisions reflect underlying federal societies. This creates an even stronger normative justification for institutional measures that enhance representation of territorial units. Territory acquires special significance in so-called ‘holding together’ federations (Stepan 1999), when territorially concentrated populations with ethnic or linguistic identities create a demand for formal structures of representation for these distinct societies. Federal systems have been adopted to provide regional autonomy to areas which have strong claims to self-government, based on national or ethno-linguistic identities. Federalism has also been used as a method to provide a strong form of territorial representation and preserve sub-national autonomy of units that had previously been self-governing, but which have come together through a process of confederation in order to secure economic or military advantage (Riker 1964). Here, the justification of strengthened territorial representation of units that were previously self-governing is based on a notion that the territorial unit, through the experience of self-rule, shapes and generates its own interests. An expression and justification of the sovereignty of the constituent territorial units of a federation appears in the Federalist Papers, where the American constitutional framers considered that not only citizens but also the constituent States possessed sovereignty as well as specific territorial interests. In the Federalist paper 32, Hamilton writes that the sovereignty of the original 13 states is to be limited, not ceded, and in Federalist paper 62, Madison argues that states have specific

Territorial and multilevel representation  173 interests. This justifies the representation afforded to the states through a bicameral legislature: ‘the equal vote allowed to each State is at once a constitutional recognition of the portion of sovereignty remaining in the individual States, and an instrument for preserving that residuary sovereignty’ (Madison et al. 2000 [1787]). Federalism, as a method for protecting and representing territorially-defined interests, can be contrasted with consociationalism, which structures power in a way that protects and gives primacy to the representation of non-territorial community interests. Federalism constitutionally entrenches sovereignty and a degree of autonomy for territorially-defined sub-national units. Federalism balances the desire for these territorial units to exercise autonomous power with shared decision-making as a larger federation. In practice, there are many ways in which federal constitutions achieve this balance. Not only do federations vary a great deal in the extent to which fiscal power and policy competence is decentralized, but federations also vary in the extent to which the sub-national territorial units exercise their power autonomously from the federal government (so-called ‘self-rule’) or in a highly interdependent manner (‘shared rule’). This affects territorial representation, both in terms of how agents are authorized and in terms of how accountability operates. The federal principle, as defined by K.C. Wheare, is ‘the method of dividing powers so that the general and regional governments are each, within a sphere, co-ordinate and independent’ (1963, p. 10). This federal principle emphasizes the duality and autonomy of both levels of government: ‘general and regional governments both operate directly upon the people; each citizen is subject to two governments’, with the two spheres of government that are ‘co-ordinate with each other’ (Wheare 1963, p. 2). The federal principle allows for considerable variation in how territorial representation is provided in practice, as I discuss below. One common feature of federalism, however, is that it safeguards territorial representation contractually. Federal constitutions typically contain a safeguard requiring concurrent majorities of people and constituent units supporting amendment. This prevents the federal arrangements from being unilaterally altered. In federalism and other forms of multi-level governance, political representation and accountability are complex because of what Tuschhoff (1999) refers to as compounded representation. In systems with compounded representation we find multiple principals and agents. Authority is transferred to more than one location of sovereignty, and delegated to more than one agent in the parliaments and governments of the federation. This creates citizens with multiple affiliations: a citizen in a federation can identify with multiple territories at the local, regional, federal and even supranational levels, and authorize multiple agents through voting in elections for government at the local, regional and federal levels. European integration and the evolving role of regional and national parliaments and sub-national organizations complicate these principal–agent relationships even further. Compounded territorial representation can be structured in different ways. Beyond the general federal principle of assigning authority to more than one territorial level, federations vary a great deal in the way in which the constitutional design allocates policy competence and provides for representation of the federal units at the centre. Federations vary in their degree of decentralization of policy competence and fiscal power, as well as in the extent to which they allow for a high degree of sub-national self-rule or provide for shared rule between the sub-national and federal governments over a range of policy areas. These variations in federal design influence the site of territorial representation in the state, whether it is the people, the

174  Research handbook on political representation governments or the parliaments of the territorial units that are represented, and the bargaining rules that shape the outcomes of territorial representation. Highly autonomous federal designs can create a high degree of self-rule for the sub-national units through the decentralization of policy competences. In these designs, common in ‘dual’ federations such as Canada and the United States, the sub-national units become the primary site for enhanced territorial representation, placing decision-making on a broad range of policy areas in the hands of the sub-national legislatures and governments. The upper houses of bicameral federal legislatures also provide a channel for representing territorial interests in the federal arena, as they are organized. This does not always guarantee that upper houses provide a strong channel of territorial representation. While the composition of upper houses in federal systems uses a territorial logic (discussed in more detail below), often providing an equal number of seats for representatives directly elected from each territorial unit, partisanship can provide a countervailing pressure attenuating the territorial representation of interests by senators. In other federal models, such as in Germany, the two levels of government coordinate more closely, exercising shared rule over a range of policy areas. In these models of joint or ‘cooperative’ federalism, sub-national governments tend to have a more limited range of autonomous policy competences, but have strong channels for representing territorial interests in federal decision-making. This is due to two aspects of federal design: the representation of sub-national governments or parliaments in federal upper houses, and decision rules that give these units veto power. In the German model of federalism, territorial interests of the sub-national units have the potential to be powerfully represented in federal policy making because Länder governments, rather than citizens, sit in the Bundesrat, the federal upper house, a point I discuss further below. When the decision rules provide veto power for sub-national units in the upper house or require supermajorities, they shape territorial representation by creating incentives for policy moderation and lowest common denominator decision-making, what Fritz Scharpf terms the ‘joint decision trap’ (Scharpf 1988). The European Union is a multi-level political system which provides territorial representation along the model of German federalism. Members of the European Parliament are directly elected by citizens within each member state, but territorial interests of the member states are represented in the European Council and in the Council of the European Union, where bargaining outcomes often reflect the compromise necessary to secure a qualified majority vote. Compound representation in the European Union is further complicated by the development of other institutions and practices that affect how the principal–agent relationship is structured and whose interests are represented. For example, the large member states in the EU tend to have strong coordinating mechanisms between government and MEPs. The permanent representative issues voting lists and sometimes sends amendment requests to MEPs. This introduces a principal–agent relationship between the national government and its MEPs, competing with the direct relationship between voters and MEPs (Mühlböck 2016, p. 78).

BICAMERALISM: TERRITORIAL REPRESENTATION IN THE CENTRE In unitary states, representation in upper houses has been based on a range of rationales, including offering representation to specific social classes. In federal states, upper houses

Territorial and multilevel representation  175 of legislatures are a key institutional element, designed to provide sub-national territorial representation. Territorial representation in upper houses has historically been considered to be more legitimate than representation based on competing rationales, such as social class (Tsebelis and Money 1997, p. 33). All federations use an upper house to represent sub-national units, although the method of selection and strength of the upper house varies. Some non-federal states, including Bolivia, the Netherlands and Spain, also use upper houses to provide representation for sub-national units. Beyond federal states, we see a principle of territorial representation in decentralized unitary states, such as in Italy. The use of bicameralism to provide territorial representation creates mixed representation, with two different rationales and selection methods in the lower and upper houses. This generates different preferences in these two chambers of the legislature, making different legislative outcomes possible (Buchanan and Tullock 1965). The degree to which preferences in the upper and lower houses are congruent or incongruent is shaped by the selection method, such as whether members of the upper house are elected from districts of different sizes, or if they are appointed to represent the governments or parliaments of the constituent units. Even with a similar composition, the preferences of the upper and lower houses might be different, as the strategies in each chamber are shaped by institutional rules (Tsebelis and Money 1997, p. 54). The implication of this is that bicameralism matters because it can have significant impacts on legislative outcomes (Buchanan and Tullock 1965; Tsebelis and Money 1997). For example, legislation that can command a winning coalition of votes in the lower house may not secure support in the upper house. In this way, bicameralism can safeguard territorial interests of the federal units. The way in which bicameralism provides territorial representation is affected by three aspects of institutional design of bicameralism. The first aspect of institutional design is the basis of representation of the constituent territorial units. This refers to whether it is the people, governments or legislatures of sub-national units that are represented. In federations such as the United States, Australia and Switzerland, representatives of the upper chamber are directly elected by the people and so the people of the sub-national territorial units are directly represented. By contrast, some upper chambers, such as those in Germany and Russia, specifically represent sub-national governments. In Austria, the sub-national legislatures are represented in the upper house. Each Austrian Land parliament can send delegates to the Federal Council, with the number of delegates weighted according to the size of the Land. The delegations reflect the composition of Land parliaments, including delegates from both governing and opposition parties. The norms regarding how the delegation in the upper house votes, whether representatives vote independently, as in the United States or Austria, or as a single weighted vote, as in the case of Germany, is the second aspect affecting the nature of territorial representation (Russell 2001). In Germany, the combination of a Land delegation to the Bundesrat composed entirely of Land governments, together with a requirement that the Land delegation votes as a bloc, represents territorial interests as defined by Land governments, and represents these in a strong fashion. This model of bicameralism in a federal state represents the institutional interests of the sub-national units and hereby involves the Land governments in the federal legislative process. In the German model, the influence of partisanship can counterbalance the extent to which territorial, rather than partisan interests are represented in the upper house. Scholars have noted that while party loyalty has normally guided voting in the Bundesrat, there have been periods of time when the pendulum has swung the other way, allowing Bundesrat votes

176  Research handbook on political representation to represent the interests of the territorial units, even if this conflicted with positions held by co-partisans in other Land governments (Lehmbruch 1978; Jeffery 1999). By contrast, the Austrian Federal Council is organized around the representation of Land parliamentary interests, but these interests are also cross-cut by differences of partisan affiliation, which can lead to voting differences within a Land parliamentary delegation. As a result, partisan representation tempers territorial representation. In the US model, the states are represented in the senate by directly elected members. The US senate, in which each state is equally represented by two members, does not represent state governmental or legislative interests, but instead serves as an additional channel for territorial representation. The equality of representation of the constituent units and the power of the upper house is the third aspect affecting territorial representation. Equality of representation refers to whether large and small units are provided with equal representation or weighted according to population. An upper house with powers equal to the lower house is what Lijphart termed congruent bicameralism (Lijphart 1984 and 1999). In some models, like the United States and Switzerland, sub-national territorial units have equal representation (two senators per state in the US and two members of the Council in Switzerland) regardless of size, resulting in sparsely populated states and cantons being overrepresented in terms of population. This aspect of the design of second chambers has historically been quite contentious, especially, as in the case of the US and Switzerland, where equal representation is combined with an upper house with co-equal powers. In the United States and Switzerland, populous states or cantons have resisted equal representation, and small states and cantons have pushed for this to prevent their interests being overruled by more populous units (Russell 2001, p. 106; Tsebelis and Money 1997). The US and Switzerland are examples of both equal state representation and co-equal powers. In other federations, such as Germany, the representation of the Länder is weighted (Tsebelis and Money 1997, p. 32). In the German Bundesrat, the 69 members (and votes) are allocated among the Länder, partially weighted according to population, with the number of members (and votes) per Land ranging from three for small Länder to six for larger Länder. Land governments are represented in the German Bundesrat, and the Land delegations do not include opposition parties. The German Bundesrat possesses co-equal powers to the lower house, the Bundestag, only in certain areas, where Land interests are most at stake, allowing the Bundesrat to represent territorial interests most powerfully in those policy areas. The Council of Ministers in the EU, the ‘upper chamber’ in the EU legislative process, similarly uses weighted representation, and in certain areas in which territorial representation is deemed to be essential, it possesses veto power. While Canada has a highly decentralized federation, with powerful provinces, its Senate does not provide for the powerful representation of provincial interests. Unusually for a federation, territorial representation in Canada’s upper house was originally designed on a regional, and not a provincial basis. The senate provides for equal representation of four regions: the Maritimes, Quebec, Ontario and the Western region are each allocated 24 seats. Newfoundland is allocated six seats and the territories each have one seat. This creates equal regional representation in the upper house but unequal provincial representation. While the Canadian senate possesses formal veto power, this is a power it has rarely exercised, in part because, as an appointed body, it is seen to lack the democratic legitimacy to overturn the will of the elected parliament.

Territorial and multilevel representation  177

ACCOUNTABILITY AND TERRITORIAL REPRESENTATION To this point we have examined the ways in which political institutions structure and channel territorial interests in representative democracies. This relates to the authorization aspect of representation. Another facet of the concept of representation emphasizes responsiveness and accountability (Pitkin 1967). A growing body of literature focuses on this supply side and relates to how territorial interests are represented and ultimately reflected in legislation, focusing on responsiveness to the policy preferences of voters, as well as the way in which voters hold their elected governments to account. While, in theory, federal arrangements optimize territorial representation because they allow for a separate sphere of decision-making to occur closer to the people, especially in spheres of policy that tend to relate to personal services, such as health, education and social services, there is little evidence that elections in the sub-national arena provide an efficient channel for communicating voters’ assessments of sub-national government performance. Instead, at the individual level, federalism is known to impede the electoral accountability mechanism. Similar to the way in which coalition government is known to obscure clarity of responsibility, federalism makes it difficult for voters to see which level of government is responsible for specific policy outcomes (Powell and Whitten 1993). Cross-national evidence has found that federalism (and decentralization in general) is associated with lower levels of economic voting in federal or national elections – an outcome that indicates voters are not holding governing parties to account for poor economic performance (Anderson 2006). At the sub-national level, ‘barometer voting’ is a common finding, in which sub-national elections are used to reward or punish the co-partisans of the federally incumbent party. This means that partisanship (and functional representation) limits the potential for effective territorial representation through the institutional channels of federalism. The literature regards policy voting – a situation where voters have positions on issues and these positions guide vote choice – to be essential to representation (Van der Eijk et al.1999, p. 161). Compounded representation can frustrate this. Research on the European Union on the correspondence between the policy positions of MEPs and citizens has found that while generally there is a strong congruence between voter and MEP positions, this is found mostly with respect to issues that structure national-level political competition. There is a much greater incongruence on issues related to European integration (Marsh and Wessels 1997; Thomassen and Schmitt 1999). The implications of this for territorial representation are that formal structures for territorial representation in multi-level systems contribute to territorial representation mostly in a general sense. In the EU, the EP provides an additional level of representatives from member states who can serve as the agents representing voter interests in a supranational legislature. However, federal and multi-level institutions are designed, at least in theory, to provide territorial representation in a more nuanced way, allowing voters to delegate decision-making to agents on specific issues that correspond to the constitutional assignment of competences to each territorial level. There is little evidence that territorial representation through the European Parliament captures these nuances. Research that examines legislative bargaining outcomes yields a more optimistic prognosis for the potential of policy responsiveness to the interests of member states. Examining the Council of the European Union, where governments represent territorial interests of the member states, Wratil (2019) has found that territorial representation in the EU operates through a mechanism in which policy formation is responsive to territorial variations in policy

178  Research handbook on political representation position and salience. Empirically, there is correspondence between policy output and public opinion in jurisdictions adjusted for salience and clarity, rather than a direct correspondence to average policy position. Bargaining in the Council produces outcomes that are responsive to high salience issues, where the issue is important to some constituencies. It shows that while policy change is not responsive to average public opinion positions on policy, the closest match between public opinion and policy occurs when salience is high and voter preferences are clear (Wratil 2019). This form of policy responsiveness is shaped by the institutional provisions for territorial representation: bicameralism in the EU, where the Council represents member state interests, allows for vote-trading behaviour which generates this differential responsiveness.

TERRITORIALIZATION OF POLITICS THROUGH PARTIES The discussion of territorial aspects of representation has so far focused on the formal representation of territorial interests in the state through institutional means such as federalism and bicameralism. Another important dimension of territorial representation is how mediating institutions, such as parties and party systems, can shape the framing and mobilization of territorial interests. If political parties are organized on the basis of territorial interests, they can serve as powerful agents to mobilize and politicize territorial identities and cleavages. Regional parties are widespread across the world. While they often arise from regional cleavages, Dawn Brancati shows that strong regional cleavages are not necessary, and instead, the development of regional parties is encouraged by the decentralization of the state, an institutional accommodation that enhances territorial representation (Brancati 2008). The presence of regional parties can also in turn impact territorial representation. Not only do they seek to advance policy goals framed primarily in terms of regional interests, rather than functional ones, but they often seek to promote devolution or decentralization of the state, which would deliver stronger formal channels of territorial representation. Sometimes party systems can be territorially concentrated as a result of geographic imbalances in the basis of party support, rather than deliberate regional strategy. When a party’s electoral support is territorially concentrated, it can affect the ability of the party to represent wider national interests, and can exacerbate territorial cleavages. Parties that are not regional parties may go through periods of territorial concentration. For example, in the United States the Democratic Party experienced solid control of the south from reconstruction until the civil rights movement. In Canada, the electoral support of both major federal parties has experienced periods of strong territorial concentration, with the federal Liberal Party shut out of the western provinces in the early 1980s and shut out of the western provinces of Alberta and Saskatchewan after the 2019 federal election. This territorial concentration of the party system had consequences for the ability of the federal system to balance territorial and functional representation (Keating and Thorlakson 2017). It can inflame regional tensions as territorial boundaries increasingly coincide with, rather than cross-cut partisan divisions. Territorial conflict is channelled into the national legislative arena. Second, a territorially concentrated party system can force governments to innovate in order to achieve territorial representation in cabinet. When territorial concentrated party support has left the governing party without regional representation in cabinet, Canadian prime ministers have appointed senators to cabinet to provide regional representation.

Territorial and multilevel representation  179 The internal organization of parties in multi-level systems can also shape the ability of parties to represent the interests of sub-national territorial units. Parties vary in the extent to which their sub-national and federal organizations are vertically integrated, through formal structures, resource sharing, cooperation and through shared goals, as well as in the extent to which the party organization affords sub-national party organizations the autonomy to pursue their own policies (Thorlakson 2009 and 2013). In Canada, for instance, a formal organizational truncation between the provincial and federal Conservative parties, and, in many provinces, between the provincial and federal Liberal arties, allows provincial parties to pursue policies that sometimes put them at odds with their federal counterparts. Typically, such policy differences emerge from regional economic interests, such as the management of natural resource royalties or pipeline approvals.

MULTI-LEVEL GOVERNANCE AND NEW MODES OF TERRITORIAL REPRESENTATION Multi-level governance approaches, which focus on the multi-level and overlapping allocation of competence in the state, conceptually and empirically broaden our perspective on territorial representation. They highlight the role of regional organizations and opportunities for influence of multiple territorial levels in the governance process at various stages, ranging from agenda setting, legislative process to implementation. The shifting loci of authority in the state due to European integration have led to the development of additional forms of territorial representation. The development of formal and informal institutions to represent regional interests in supranational policy making has created a channel for sub-national territorial interests to influence policy advocacy and policy implementation. The EU has provided an opportunity for the representation of sub-state regions. Organizing and representing their interests at the EU level allows them to bypass the national level or increase their impact vis-à-vis the national government. The Treaty of Maastricht created a formal channel for the representation of sub-national territorial interests in the Committee of the Regions (CoR). In the CoR, sub-national governments seek to influence both legislation that impacts sub-national governments, as well as the EU funding for regions (Callanan and Tatham 2013). With only a consultative role, this body has little power. In addition to this, sub-state territorial interests are represented by a range of sub-national associations that act like interest groups, including liaison offices and transnational associations (Donas and Beyers 2013). These work outside of formal legislative representation channels to lobby for the territorial interests of Europe’s regions. By the mid-1990s, more than 50 regional and local government organizations had appeared in Brussels, providing an informal channel for the representation of sub-national territorial interests to the European Parliament and the European Commission (Marks et al. 1996). These organizations represent territorial interests by raising issues with the Parliament and Commission and lobbying, and providing both the sub-national governments and the Commission with valuable information (Marks et al. 1996). Territorial representation through these offices is uneven, with larger and richer regions leading the push for the establishment of offices in Brussels (Blatter et al. 2008; 2009). The EU funds and encourages the organization of regional interests. These regional offices to represent sub-national territorial interests in Brussels tended to appear when national constitutions assigned overlapping competences to the national and regional levels in policy areas in

180  Research handbook on political representation which the EU was involved, as well as when regional governments had an opportunity to use the supranational arena to advance political conflicts with the central state (Marks et al. 1996). While we see many regional organizations and interest organizations appearing in the 1990s to represent territorial interests, we see a turning point with the Laeken Declaration, Constitutional Convention and Lisbon Treaty. The 2001 Laeken Declaration called for a new commitment to subsidiarity. The subsequent Constitutional Convention process allowed regional authorities to provide input into the deliberations, thus fragmenting the channels for sub-national territorial interest representation, which previously were limited to the Committee of the Regions in formal terms. The Lisbon Treaty gave a boost to sub-national territorial representation through strengthened commitments to cooperation and consultation in the development of legislation, which includes consultation with regional and local authorities (Mandrino 2008, p. 521). Notably, the Committee of the Regions has legal recourse if regions are not consulted. The development of a multiplicity of representative channels in the EU has increased the complexity of territorial representation in the EU. We also see examples of an increasingly complex environment for territorial representation in the United States. Here, in addition to formal institutions of bicameralism, territorial representation may occur through intergovernmental organizations and interest group activity. For example, in the US, the National Governors’ Association and the National Conference of State Legislatures provide a forum for promoting sub-national territorial interests. While sub-national territorial representation beyond the borders of the traditional state is still limited, we still find processes through which regions can exercise influence. In the US, regions can mobilize interests inside agreements in trade deals (Gress 1996). This demonstrates the increasing complexity of territorial representation, generated by formal federal institutional channels, intergovernmental channels, and interest groups. In unitary states, similar institutional structures produce channels for sub-national territorial representation. In Italy, for example, the Conferenza Stato-regioni and the Conferenza Stato-città provide channels for regional and municipal participation in policy discussions.

CONCLUSION Territorial representation is a central and enduring aspect of representation in the modern state. From an institutional perspective, there is a substantial body of literature that addresses how federal and legislative institutions channel the representation of territorial interests. The range of federal institutional designs generate a great deal of variation in the way in which sub-national interests are represented – whether sub-national territorial interests representation occurs most powerfully and effectively in highly autonomous sub-national arenas, or whether these interests are channelled into policy making in the central state. The design of bicameralism also shapes whether governments are given a key role in the articulation and representation of sub-national territorial interests in central policy making. Territorial representation is not reserved for federal states. Multi-level unitary states also have institutional means for the representation of sub-state territorial interests. Whatever institutional design structures territorial representation, it creates complexity, by introducing compounded representation in the state, with multiple principal–agent relationships. Research on democratic accountability has taken a different perspective on territorial representation, being interested instead in assessing the quality of territorial representation

Territorial and multilevel representation  181 in terms of how it delivers policy responsiveness across a multi-jurisdictional space. From a sociological perspective, research on territorial representation has turned to the question of the drivers of demand for territorial representation, and how regionally or territorially defined interests, identities and conflicts are mobilized. Here, parties and party systems have a role to play as intermediary actors (or institutions) that can shape conflict. Finally, while territorial representation often takes the territorial state and its spatial boundaries for granted, scholars are addressing the challenges of understanding representation in a world in which opinion and policy formulation take place beyond the borders of the territorial state.

REFERENCES Anderson, Cameron (2006). ‘Economic Voting and Multilevel Governance: A Comparative Individual-Level Analysis’, American Journal of Political Science, 50 (2), 449–63. Blatter, J., M. Kreutzer, M. Rentl and J. Thiele (2008), ‘The Foreign Relations of European Regions: Competences and Strategies’, West European Politics, 31 (3), 464–90. Blatter, J., M. Kreutzer, M. Rentl and J. Thiele (2009), ‘Preconditions for Foreign Activities of European Regions: Tracing Causal Configurations of Economic, Cultural, and Political Strategies’, Publius: The Journal of Federalism, 40 (1), 171–99. Brancati, D. (2008), ‘The Origins and Strengths of Regional Parties’, British Journal of Political Science, 38 (1), 135–59. Buchanan, J.M. and G. Tullock (1965), The Calculus of Consent, Ann Arbor, MI: The University of Michigan Press. Callanan, M. and M. Tatham (2013), ‘Territorial Interest Representation in the European Union: Actors, Objectives and Strategies’, Journal of European Public Policy, 21 (2), 188–210. Caramani, D. (2004), The Nationalization of Politics: The Formation of National Electorates and Party Systems in Western Europe, Cambridge: Cambridge University Press. Collyer, M. (2014), ‘A Geography of Extra-territorial Citizenship: Explanations of External Voting’, Migration Studies, 2 (1), 55–72. Donas, T. and J. Beyers (2013), ‘How Regions Assemble in Brussels: The Organizational Form of Territorial Representation in the European Union’, Publius, 43 (4), 527–50. Gress, F. (1996), ‘Interstate Cooperation and Territorial Representation in Intermestic Politics’, Publius: The Journal of Federalism, 26 (1), 53–71. Jeffery, C. (1999), ‘Party Politics and Territorial Representation in the Federal Republic of Germany’, West European Politics, 22 (2), 130–66. Keating, M. and L. Thorlakson (2017), ‘Party Systems and Party Competition’, in Michael Keating and Guy Laforest (eds), Constitutional Politics and the Territorial Question in Canada and the United Kingdom, London: Palgrave, pp. 135–58. Knights, Mark (2010), ‘Participation and Representation Before Democracy: Petitions and Addresses in Premodern Britain’, in Ian Shapiro, Susan Stokes, Elisabeth Jean Wood and Alexander Kirshner (eds), Political Representation, Cambridge: Cambridge University Press, pp. 35–57. Lehmbruch, G. (1978), ‘Party and Federation in Germany: A Developmental Dilemma’, Government and Opposition, 13 (2), 151–77. Lijphart, Arendt (1984), Democracies: Patterns of Majoritarian and Consensus Government in Twenty-One Countries, New Haven, CT: Yale University Press. Lijphart, A. (1999 [2012]), Patterns of Democracy: Government Forms & Performance in Thirty-six Countries, New Haven, CT: Yale University Press. Madison, James, Alexander Hamilton and John Jay (2000 [1787–8]), The Federalist, William Brock (ed.), London: Phoenix Press. Mandrino, C. (2008), ‘The Lisbon Treaty and the New Powers of Regions’, European Journal of Law Reform, 10 (4), 515–33. Marks, G., F. Nielsen, L. Ray and J. Salk (1996), ‘Competencies, Cracks and Conflicts: Regional Mobilization in the European Union’, Comparative Political Studies, 29 (2), 164–92.

182  Research handbook on political representation Marsh, M. and W. Wessels (1997), ‘Territorial Representation’, European Journal of Political Research, 32 (2), 227–41. Mühlböck, Monika (2016), Voting Unity of National Parties in Bicameral EU Decision-Making, London: Palgrave. Pitkin, Hanna (1967), The Concept of Representation, Berkeley, CA: University of California Press. Powell, G.B. and G.D. Whitten (1993), ‘A Cross-national Analysis of Economic Voting: Taking Account of Political Context’, American Journal of Political Science, 37 (2), 391–414. Riker, William (1964), Federalism: Operation, Origin, Significance, Boston, MA: Little, Brown and Company. Rokkan, Stein and Derek Urwin (1983), Economy, Territory, Identity: Politics of West European Peripheries, London: Sage. Russell, M. (2001), ‘The Territorial Role of Second Chambers’, Journal of Legislative Studies, 7 (1), 105–18. Scharpf, F. (1988), ‘The Joint-Decision Trap: Lessons from German Federalism and European Integration’, Public Administration, 66 (3), 239–78. Stepan, A. (1999), ‘Federalism and Democracy: Beyond the US Model’, Journal of Democracy, 10 (4), 19–33. Thomassen, Jacques and Schmitt, Hermann (1999), ‘Issue Congruence’, in Hermann Schmitt and Jacques Thomassen (eds), Political Representation and Legitimacy in the European Union, Oxford: Oxford University Press, pp. 186–208. Thorlakson, L. (2009), ‘Patterns of Party Integration, Influence and Autonomy in Seven Federations’, Party Politics, 15 (2), 157–78. Thorlakson, L. (2013), ‘Measuring Vertical Integration in Parties with Multi-level Systems Data’, Party Politics, 19 (5), 713–34. Tsebelis, George and Jeannette Money (1997), Bicameralism, Cambridge: Cambridge University Press. Tuschhoff, C. (1999), ‘The Compounding Effect: The Impact of Federalism on the Concept of Representation’, West European Politics, 22 (2), 16–33. Van der Eijk, Cees, Mark Franklin and Wouter van der Brug (1999), ‘Policy Preferences and Party Choice, Political Representation and Legitimacy in the European Union’, in Hermann Schmitt and Jacques Thomassen (eds), Political Representation and Legitimacy in the European Union, Oxford: Oxford University Press, pp. 161–85. Wheare, Kenneth (1963), Federal Government, 4th edn, London: Oxford University Press. Wratil, C. (2019), ‘Territorial Representation and the Opinion-Policy Linkage: Evidence from the European Union’, American Journal of Political Science, 63 (1), 197–211.

15. Local jurisdiction size and political representation Bas Denters

INTRODUCTION In the growing body of literature on the effects of size and democracy most attention so far has been given to size effects on citizens’ psychological orientations and their active involvement, either via direct participation or indirectly via voting in elections (see for instance Gaardsted Frandsen 2002; Rose 2002; Saglie and Vabo 2009; Kelleher and Lowery 2009; Denters et al. 2014; Van Houwelingen 2017; McDonnell 2019). Studies of the effects of jurisdiction size on political representation are less common (for a few exceptions see Denters et al. 1990, pp. 95–119; Van der Kaap 2006; Charron et al. 2013). This is remarkable, because contemporary democracies at both the national and at the sub-national level are typically based on principles of representative rather than direct democracy. Although many Western democracies in recent years have introduced reforms allowing for more direct citizen participation (Cain et al. 2003; Smith 2009; Michels 2011; Newton and Geissel 2012), these democratic innovations have not corroded the primacy of representative democracy. Even in Switzerland, a pure form of direct democracy in citizens’ assemblies is only found in two cantons, and in most small municipalities. At the federal level, in most cantons and in larger municipalities, however, Swiss government is based on a system of semi-direct democracy, combining representative democracy with strong instruments of direct democracy (popular initiatives and referendums) (Ladner 2002). In this chapter I focus on the question of how the size of political jurisdictions affects the functioning of systems of political representation. In this question, political jurisdictions are the units of analysis. Political jurisdictions can be found at different tiers of governance: at the national, regional and local level. These political jurisdictions vary in size. Here – as in most of the literature – I will focus on the population size of jurisdictions. Variations in size occur between units operating at different levels of government and between units operating at the same level. In this chapter the focus is on the latter type of size differences.1 Because of my personal expertise and research interests, the focus here will be further narrowed down to the effects of the size of local political jurisdictions.2 This issue is particularly relevant for two reasons. First, in many European countries there is a heated debate about the pros and cons of amalgamation reforms aimed at strengthening the performance and democratic quality of local governments by increasing the size of local jurisdiction (see for instance Steiner et al. 2016). Second, a well-functioning democracy at the local level, according to several authors – from the present (Dahl 1997 and Vetter 2002) and the past (e.g. John Stuart Mill and Alexis de Tocqueville) – is an essential resource for the democratic legitimacy of larger political systems at the national and the supranational levels. 183

184  Research handbook on political representation In this chapter, I will provide an overview of the most important theoretical arguments that are relevant for answering our research question and the political debates on issues of size and democracy. In this way I propose an agenda for future empirical research. In this chapter, political representation refers to various mechanisms that secure democratic responsiveness, that is, the correspondence between citizens’ needs and preferences and the results of public decision-making. In the next section, I will explain four mechanisms that should contribute to democratic responsiveness, and in a subsequent section I will present two rival perspectives on possible effects of jurisdiction size on the operation of these mechanisms.

REPRESENTATION BY ELECTED REPRESENTATIVES In democratic theory, direct participation of citizens in governing and the representation of citizens by their elected representatives are conceived as two methods to secure concurrence between the preferences and needs of the demos and major political decisions in political jurisdictions. As Dahl (1989, pp. 13–23) has argued, initially – in the democracies established in several Greek city-states (poleis) during the first half of the fifth century BC – direct participation in citizen assemblies was the primary channel for securing concurrence.3 The use of this channel, however, was ‘inherently limited to small-scale systems’ (Dahl 1989, p. 23). Therefore, when at the turn of the eighteenth to the nineteenth century democratic reformers (like James Mill or James Madison) were facing the challenge of moulding democratic ideals into political institutions fit for the large-scale political jurisdictions of their time, they had to look for an alternative channel to secure concurrence (Dahl 1989, pp. 28–30). In this context they considered the medieval institution of representation as an attractive alternative to democratize the modern nation-state. From then on, debates on representation centred not on whether democracy should be representative, but on key mechanisms through which representative assemblies and their members could best contribute to the realization of the democratic ideal of concurrence. In the literature on political representation we find three main views on representation. In each of these views a particular mechanism is identified that should secure democratic responsiveness (Pitkin 1967; Thomassen 1991; Birch 1993, pp. 69–79; Judge 1999). The first interpretation stipulates that the representative assembly should be a microcosm of society at large. The representatives should as much as possible reflect the electorate in terms of socio-demographic features, such as class, gender, age and education (descriptive representation). The challenge to assure this type of representativeness is essentially the same as the problem a statistician faces when asked to draw a representative sample from a particular universe (statistical population) of objects. The underlying idea of microcosmic representation is that ‘if direct participation by “the people” is impractical, then as a second preference, representation should at least attempt as close a reflection of the constituent elements of “the people” as possible’ (Judge 1999, p. 22). Responsiveness – in this view – is secured by the “identity” of the represented (cosmos) and the representatives (microcosm). Inspired by Edmund Burke (1999 [1774]) and similar ideas of James Madison in a second interpretation, representation is no substitute for direct participation but an alternative superior to direct participation. In this view, the elected representative should act as a trustee of voters.

Local jurisdiction size and political representation  185 As a trustee the representative should, on the one hand, be responsive to the demands and preferences of the electors. But on the other hand, the representatives should also ‘refine and enlarge the public views by passing them through the medium of a chosen body of citizens whose wisdom may best discern the true interest of their country’ (Hamilton et al. 1987 [1788], Federalist N.10). In this view, independent representatives – with discretion and room for personal judgement – should act ‘in the interest of the represented, in a manner responsive to them’ (Pitkin 1967, p. 209). The third view conceives of representatives as agents of political parties. The representative should primarily act as a party-delegate. Citizens in this view do not maintain a personal relationship with a representative. Instead they have an affinity or identify with a particular political party – often based on the party’s ideology or platform. On the basis of this affinity citizens vote for a party in periodic elections. Party-representatives – united in a party-faction – clearly communicate their party’s views to the voters, and after the elections they collectively work to realize the party’s political priorities. Here responsiveness to citizens is secured by prospective (selecting preferred party based on future promises) and retrospective voting (holding parties to account for their political achievements in the pre-election period) and responsible political parties (Thomassen 1991). In addition to these traditional mechanisms we also distinguish a fourth mechanism. The previously discussed models all imply that representatives articulate the needs and preferences of the represented. But in response to new participation demands of increasingly emancipated and critical citizens who can and do often speak for themselves (Klingemann and Fuchs 1995; Norris 1999) governments across the Western world have initiated democratic reforms complementing the representative system with channels for direct citizen participation (e.g. Cain et al. 2003; Smith 2009; Michels 2011; Newton and Geissel 2012). In light of these changes, the role of the representative might change towards becoming a democratic facilitator and auditor. As a facilitator the representative might create new or improve existing channels for direct citizen participation. And as an auditor he or she can assure that new participatory arenas will offer a level playing field where all citizens can articulate their needs and preferences. This latter function is especially important because there is increasing evidence that direct participation might exacerbate existing political inequalities in favour of already politically active citizens (Fiorina 1999; Michels and de Graaf 2010; Bovens and Wille 2017). In this new interpretation, responsiveness is realized by representatives who create noise-free channels through which all citizens can make themselves heard, and monitor the proper functioning of these channels for direct participation.

SIZE AND POLITICAL REPRESENTATION: BROBDINGNAG AND LILLIPUT REVISITED Our central question is how the size of a political jurisdiction affects the functioning of its system of political representation. Theoretically, the answer to this question might and does vary. First, hypothesized size effects may vary depending on what view one takes at representation. Size effects on microcosmic representation might very well be different from size effects on party-based representation. But, secondly, even if we focus on one view on representation (e.g. the trustee model) expected size effects may differ, depending on theoretical perspectives on size effects.

186  Research handbook on political representation In a previous publication (Denters et al. 2014), we identified two main theoretical traditions in the literature on jurisdiction size and citizens’ political orientations and their participation in local democracy. In the first tradition it is hypothesized that a larger size of municipalities will offer favourable conditions for a well-functioning political system that is capable of satisfying citizens’ political demands and their need for high-quality facilities and services. Because of this, local democracy is likely to become more meaningful for people. Therefore, increased size will not only lead to more satisfied citizens but also to more citizen engagement. With a wink to Jonathan Swift’s Gulliver’s Travels, I will refer to this as the Brobdingnag argument.4 Claims like this are often used in making the case for municipal amalgamation reforms. Instead, in a second tradition it is argued that small jurisdictions offer many advantages. Small size not only goes hand in hand with a strong sense of social community, but is also likely to have positive consequences for people’s political interest, knowledge and trust. At the same time, small size will also make it easier (in terms of necessary resources) to become politically active. Henceforth, I will refer to this perspective as the Lilliput argument. On the basis of the literature we can construct two rather similar theoretical narratives about the effects of jurisdiction size on political representation. In the following subsections I will discuss Brobdingnagian and Lilliputian arguments pertaining to size effects on the functioning of the four representation mechanisms introduced in the previous section. Microcosmic Representation and Jurisdiction Size Microcosmic representation aims at securing maximum possible descriptive representativeness. If we want to determine the effect of population size on the accuracy of representation in the microcosm (i.e. the representative assembly), just as in statistical sampling (for example Henry 1990, pp. 117–22), we need to consider three additional factors: 1. the sampling procedure; in political representation this is equivalent to the procedures for recruitment and the electoral system used for (s)electing representatives; 2. the sample size; in political representation this relates to the size of the representative assembly; 3. the degree of heterogeneity in the population. The procedures (1) for the (s)election of representatives are, to an important extent, determined by formal institutional rules included in the laws applicable to the jurisdiction. Sometimes, for example in unitary states like the Netherlands, these rules are uniform across the country. But as Van der Kolk (2007) demonstrates in his excellent overview of local electoral systems across Europe, the electoral system can also vary within a country. Sometimes these differences are between small and large municipalities, as is the case in countries such as France, Greece, Italy and Luxembourg. But there are also regional differences, as is the case within federal states, such as Germany, Austria and Switzerland. A major institutional factor is the electoral system. On the one hand, a local election may be contested in one multimember district on the basis of proportional representation (PR). Such a system allows – all other things being equal – for optimal representation of party-political preferences. On the other hand, an election in a jurisdiction with several single-member districts based on a first-past-the-post (FPTP) system will guarantee optimal representation of territorial interests. But such a system might lead to a seriously flawed representation of

Local jurisdiction size and political representation  187 ideological differences. Likewise, these differences in electoral systems may also have repercussions for other forms of representativeness. For example, research has demonstrated that PR systems (as compared with FPTP systems) have positive effects on the gender balance in elected assemblies (for instance Norris 2000). Just like the electoral system, the size of local representative assemblies (2) is determined by national or sub-national legislation. In most unitary countries the size of municipal councils is determined in the Local Government Act. This implies that there are no regional variations in these countries. This does not, however, mean that there is no intermunicipal variation. This is because, as is the case in the Netherlands, it is ruled that the size of municipal councils increases with the population size of the municipality. As Table 15.1 indicates, the increase of council size in the Netherlands – as in several other countries – is by no means proportional to increasing population size. This leads to rather big differences in the so-called representation ratios. In the Dutch case, for example, a councillor in the smallest Dutch municipality of Schiermonnikoog (with 947 residents) represents about 105 residents. In Amsterdam, every councillor represents well over 19 000 local residents. Table 15.1

Population size and council size as regulated in article 8 of Dutch Municipal Law and average representation ratio per class of municipalities Population Size Category

Council Size

Average Representation Ratio*

0–3000

9

167

3001–6000

11

409

6001–10000

13

615

10001–15000

15

833

15001–20000

17

1.029

20001–25000

19

1.184

25001–30000

21

1.310

30001–35000

23

1.413

35001–40000

25

1.500

40001–45000

27

1.574

45001–50000

29

1.638

50001–60000

31

1.774

60001–70000

33

1.970

70001–80000

35

2.143

80001–100000

37

2.432

100001–200000

39

3.846

> 200000

45

11.111

Note: * This ratio was computed by dividing the mid-class size for a municipality in the relevant class of municipalities by the number of councillors for that class.

But intermunicipal differences may not only be the result of national legislation. In Denmark for example, the central government has only specified a minimum number of councillors (9 for municipalities with fewer than 20 000 inhabitants and 19 for the larger municipalities) and a maximum number of councillors (31 for all municipalities; with the exception of Copenhagen, which has 55 councillors). Within these broad ranges municipalities are free to decide on the size of their councils.

188  Research handbook on political representation In federal countries – like Germany, Austria and Switzerland – the council size is normally regulated by the federal state (or other regional governments, like the Swiss cantons). In these countries, such regional variations may come on top of size-related intermunicipal differences. From a Brobdingnagian perspective we hypothesize that: B1: In larger jurisdictions the accuracy of representation of interests and preferences may be better than in smaller jurisdictions, but this positive effect is contingent upon the electoral system and the assembly size. In formulating this hypothesis we consider the council as a sample of the population. From statistical sampling theory we know that, in general, the precision of samples becomes larger if the absolute size of the sample increases. These increases in sampling precision are especially strong when the absolute sample size is rather small (Henry 1990, pp. 117–18) – as is the case with most local assemblies; in the Netherlands for example the minimum council size is 9 and the maximum is 45. On this basis we expect, as stated in the first part of the hypothesis, that accurate representation is easier to achieve in the larger municipalities where generally the councils (i.e. the sample sizes) consist of more councillors. But this positive effect may be weakened or strengthened by institutional provisions that determine the electoral formula and rules regarding the assembly size. For example, in larger Dutch municipalities – as a consequence of their relatively larger council and the prevalent PR system – there are better opportunities for the representation of ideological and other minorities (based for instance on ethnicity, lifestyle or sexual preferences). And, likewise, Danish municipalities – that operate under a PR system – apparently have used the previously discussed choice options to increase the assembly size in order to offer territorial communities in the larger amalgamated municipalities better representation.5 But this conclusion is only preliminary because thus far we have not yet considered the third factor affecting representativeness, population heterogeneity (3). It is obvious that heterogeneity affects the accuracy of representation. If, for example, in a given population everyone is Catholic, then a sample of one (one representative) and any sampling procedure will do, for achieving an adequate representation of this population in terms of its religion. If the proportion of Catholics in the population were 20 per cent, however, it would require a larger sample (larger representative assembly) and appropriate sampling procedures to secure adequate representation. When considering the effects of jurisdiction size on the accuracy of microscopic representation, population heterogeneity is particularly relevant because there is typically a strong relationship between the population size of a jurisdiction and the socio-demographic heterogeneity of its population. As Dahl and Tufte in their 1973 classic book Size and Democracy have argued, ‘the larger a community the greater the categorical diversity with respect both to the historic cleavages of language, religion and culture and to […] socioeconomic diversity’ (1973, p. 35). Likewise the urban sociologist Louis Wirth (1938, p. 8) has emphasized that in cities size and heterogeneity go hand in hand. In jurisdictions heterogeneity is not only found at the individual level, but also at the level of community organizations. As a consequence of individual diversity, a process of institutional differentiation will take place, leading to an increase in the number of social and political organizations active in the jurisdiction (Baglioni et al. 2007; Dahl and Tufte 1973, pp. 30–40; Verba and Nie 1987 [1972], pp. 229–47; Denters et al. 2014, pp. 77–8).

Local jurisdiction size and political representation  189 This increasing heterogeneity has important consequences for political representation in larger jurisdictions. Brobdingnagians – as we just saw – point to the opportunities that increasing size – under the right institutional conditions – might offer for better representation of citizens’ interests and preferences. But, from a Lilliputian perspective, it is important to realize that, as a consequence of heterogeneity, larger jurisdictions also face a greater need of adequately representing diversity. This implies a Lilliputian counter-hypothesis: L1: In larger jurisdictions – as a consequence of greater cultural and socio-economic heterogeneity of their populations – it will be more difficult to adequately represent the interests and preferences of residents than in smaller jurisdictions. Of course – as Table 15.1 shows for the Dutch case – in many countries the council size is likely to increase with increasing population size. But it remains to be seen whether this increase will be proportional to the greater needs to represent the cultural and socio-economic heterogeneity in these larger jurisdictions. Lilliputians will argue that as a consequence of increasing representation ratios (Table 15.1, column 3) the net effect of increased jurisdiction size on the accuracy of microcosmic representation is likely to be negative. Trustee Representation and Jurisdiction Size The trustee interpretation of representation is based on an aversion to pure (that is to say) direct democracy. On this basis, James Madison in the Federalist Papers – especially in Federalist N. 10 – makes a case for popular government in which public views are not pronounced directly by the people themselves but by independent, qualified representatives who, on the basis of their wisdom, can ‘guard against the confusion of a multitude’ (Hamilton et al. 1987). James Madison claims that a large jurisdiction is likely to provide optimal conditions for a well-functioning republic (the term that Madison uses to refer to a representative democracy based on the trustee model). His argument is based on two relatively straightforward propositions. First, Madison argues that in a large jurisdiction representation is likely to be of a superior quality because, in a larger jurisdiction, conditions for finding enough ‘proper guardians of the public weal’ are more favourable than in a small jurisdiction (Hamilton et al. 1987 [1788], Federalist N. 10). Similar arguments about the need to recruit enough ‘people of the right calibre’ to serve as elected representatives – in the second half of the twentieth century – were used again, now to make a case for larger jurisdictions in British local government (Dearlove 1979). Second, according to Madison, in large units it is not only easier to find these ‘proper guardians’. It is also more likely that because of ‘the greater variety of parties and interests’ in larger units representatives will be able to act independently and not be carried away by a dominant majority – which is much more likely to emerge in more homogeneous jurisdictions (Hamilton et al. 1987 [1788], Federalist N. 10). It is interesting to see that – in the context of the Madisonian trustee interpretation of representation – heterogeneity is no longer seen as a problem for representation but as an asset. In combination, Madison’s arguments amount to a Brobdingnagian hypothesis:

190  Research handbook on political representation B2: In larger jurisdictions trustee representation is likely to function better (in terms of the calibre and independence of representatives) than in smaller jurisdictions. But from a Lilliputian perspective some objections can be raised. Centuries before James Madison expressed his concerns about ‘pure democracy’, in much a similar vein Aristotle expressed his aversion to Athenian democracy. And just like Madison, the Athenian advocated a mixed constitution, combining elements of aristocracy and democracy. But when considering the question of the optimal scale for constitutional government, Aristotle comes to a rather different conclusion. He argues that a state should not be too large, because ‘if citizens of the state are to judge and to distribute offices according to merit, they must know each other’s characters; where they do not possess this knowledge both the election of offices and the decision of lawsuits will go wrong’ (Aristotle 1984, vol. 2, p. 2105). Whereas Madison emphasizes the potentials of large jurisdictions in securing adequate representation (in the trustee interpretation of this concept), Aristotle is sceptical regarding the societal conditions under which this potential is to be realized. After all, the relation of trust between trustor and trustee is probably less likely to emerge if there is only a distant relation between both. Second, it can also be argued that if the jurisdiction is actually too large, it may be impossible for representatives to inform themselves adequately about the needs and preferences of their constituents. Actually, this issue is – reluctantly – raised by Madison, when he acknowledges that if you make the polity too big ‘you render the representative too little acquainted with all their local circumstances and lesser interests’ (Hamilton et al. 1987 [1788], Federalist N. 10). In the same spirit Dahl and Tufte (1973, p. 87) have argued that the smaller the scale of a jurisdiction, the more likely it is that elected representatives: ●● ‘gain their information about citizens’ wants by direct observation and communication’ and ●● ‘have occupations or roles outside politics that provide them with opportunities to engage in politics but less than full-time’. Together with the effects of jurisdiction size on the representation ratio (see Table 15.1, column 3) these concerns and objections give rise to a Lilliputian rival hypothesis: L2: In larger jurisdictions trustee representation is likely to function less well (in terms of the mutual trust between citizens and their elected representatives and in terms of the responsiveness of trustees to the needs of their constituents) than in smaller jurisdictions. Party-democracy and Jurisdiction Size The party-delegate model is assumed to secure concurrence between citizens’ needs and preferences and policy decisions and outputs when a number of conditions are satisfied. In part these conditions pertain to party behaviour. Parties (a) should develop distinct ideological and programmatic profiles that are clearly communicated to the voters; and (b) party-delegates should act in unison to adopt and implement the party platform. Voters subsequently (c) should vote on the basis of their own preferences and evaluations of the platform and past performance of the parties competing in the election. The party-delegate model requires a well-developed, differentiated and competitive party-system. As was already discussed when we considered the effects of jurisdiction size on

Local jurisdiction size and political representation  191 heterogeneity, larger jurisdictions as a consequence of institutional differentiation are likely to have a larger, more differentiated system of civic and political organizations (Baglioni et al. 2007; Dahl and Tufte 1973, pp. 30–40; Verba and Nie 1987 [1972], pp. 229–47; Denters et al. 2014, pp. 77–8). As a corollary, Dahl and Tufte have hypothesized that ‘[a]mong political units within a democratic country, diversity and competition among political parties increase with the size of the unit’ (1973, p. 100). This is especially relevant in offering voters distinct and visible political choice options (condition (a)). At the same time, in larger jurisdictions, there is also a greater likelihood that parties will attract sufficient members and volunteers to develop a strong well-functioning local party-organization (condition (b)). Moreover, in larger jurisdictions citizens may also have more incentives and resources to make an informed vote based on their local political preferences and an evaluation of party platforms and performance. On the one hand it has been argued that as a consequence of, amongst others, their greater administrative capacity and the more competitive party landscape, local politics in larger jurisdictions might be more salient and interesting for citizens (Denters et al. 2014, pp. 122–4). Moreover, in larger jurisdictions citizens may also be more likely to be informed about local politics and government through a richer network of civic organizations than in smaller units (Denters et al. 2014, p. 237). On this basis it might be expected that voters in larger jurisdictions, better than electors in smaller units, will make an informed choice based on relevant local information. On the basis of such considerations, it is possible, in a Brobdingnagian spirit, to formulate the following hypothesis: B3: Larger jurisdictions offer better opportunities (in terms of competitive and strong parties, and voter choice) for well-functioning party-delegate representation than in smaller jurisdictions. But, based on a different set of arguments, this hypothesis is by no means undisputed. The main debate here centres on voters’ choices in local elections. Amongst those that disagree with hypothesis B3 we can distinguish two positions. First, Lilliputians may argue that because of the closer proximity and closer interpersonal networks in smaller jurisdictions, citizens’ interest in and knowledge of local political affairs may be greater than in larger jurisdictions, rather than the other way round (Denters et al. 2014, pp. 124–5 and 135–7). This Lilliputian argument is in accordance with similar arguments made by classic political theorists such as de Tocqueville (1990 [1838]) and J.S. Mill (1974 [1861]). Smallness in their view would be conducive to acquiring political interest and knowledge, and would make local politics into a ‘school for democracy’. On this basis one might formulate a Lilliputian counter-hypothesis: L3: In larger jurisdictions party delegate representation is likely to function less well (because of lower voter interest and knowledge) than in smaller jurisdictions. Second, it has been argued that in municipal elections it is unlikely that the condition is fulfilled that voters should base their vote on local political preferences and information on the programmes and performance of local parties. Local elections are characterized as second-order elections (Reif and Schmitt 1980). In such second-order elections voters are

192  Research handbook on political representation typically unaware of local political debates and the local performance of parties and cast their vote primarily. To the extent that this view is correct, both hypotheses (B3 and L3), in as far as they pertain to voter choice, are wrong and voter choice is unaffected by jurisdiction size.6 Facilitating and Monitoring Democratic Governance and Jurisdiction Size In the second half of the twentieth century, party democracy and the corresponding party delegate interpretation of representation were dominant in local governance in most Western countries. But gradually, in response to new participation demands of citizens in many countries in parallel to this dominant model, elements of what Vetter (2009) has termed ‘citizen-democracy’ have been added to the system, providing citizens with more, direct channels to make their voice heard. As Table 15.2 indicates, there are important variations in the degree to which European local governments have adopted citizen-democratic innovations. To the extent that such innovations have been implemented, the role of councillors may also have to change: away from a party delegate towards a role of democratic facilitator and guardian. In the latter role the councillor is shifting away from someone speaking for citizens, to someone who creates new channels for meaningful direct participation of citizens, and secures equal political opportunities for citizens to make themselves heard. At the same time, this new role is procedural – oriented towards arranging new participatory channels and monitoring the proper use of these channels – rather than substantive. As before, for the three traditional interpretations of political representation, we will now ask how jurisdiction size will affect the functioning of councillors in this new role. Theoretically, here we are very much left to our own devices, because there is – as far as we know – no previous work that we can build upon. Nevertheless, we think that again there may be two rival expectations about the effects of jurisdiction size. From a Brobdingnagian perspective, it might first be argued that large-scale units may offer a fertile political climate for the adoption of such a more procedural role orientation. As Dahl and Tufte have argued, there are important differences in the nature of the political process in small and large jurisdictions: whereas in small systems interactions are more based on personal interactions and informal procedures, politics in larger jurisdictions is based on ‘more elaborate, more formal, more highly organized institutions’ (1973, p. 91). Likewise, Denters et al. (1990) have argued that, theoretically, the procedural quality of decision-making in large municipalities might be higher. First, because the distance between administration and the public is larger, allowing for a more impartial treatment (without any discrimination, nepotism or clientelism). Second, because larger jurisdictions have the administrative capacity (legal expertise) to develop and implement proper formal procedures. The new democratic facilitator/monitor role – with its focus on proper and fair procedures is probably more in tune with the politico-administrative culture of large jurisdictions than with the more informal ethos that prevails in smaller units. Moreover, as we argued before, in large municipalities the opportunities for effective citizen participation based on direct personal contacts with local politicians and administrators are more limited than in smaller units (e.g. Dahl and Tufte 1973, pp. 66–88). Hence, in larger municipalities there is a greater need to adopt formal procedures to satisfy citizens’ demands for opportunities to participate directly (instead of the direct access to politicians and administrators possible in smaller jurisdictions). The role of democratic facilitator (and monitor) is thus probably more needed in larger than in smaller jurisdictions. On the basis of these arguments we hypothesize that:

Local jurisdiction size and political representation  193 B4: In larger jurisdictions the need for councillors (in the absence of informal channels for direct citizen access) to adopt the role of democratic facilitator and monitor is greater than in smaller jurisdictions. At the same time, Lilliputians might argue that in smaller jurisdictions the prospects for actually adopting a democratic facilitator/monitor role may be better than in larger units. Here two – previously discussed – considerations are relevant. First, in larger units, the party system is likely to be more fully developed. Hence it is to be expected that representatives in large units are more likely than their colleagues in smaller jurisdictions to subscribe to principles of party democracy and are less sympathetic to citizen democracy. Second, in smaller jurisdictions representatives are more likely to be part-time, amateur politicians who are likely to be closer to their fellow citizens and less likely to identify with the professional politicians and administrators in the executive branch of local government. In larger units, representatives are more likely to be (close to) full-time, professional politicians (as was, for instance, argued by Dahl and Tufte 1973, p. 87). On the basis of these considerations it might be argued that: L4: In larger jurisdictions councillors (because of the strong party system and their professionalization) are less likely to adopt the role of democratic facilitator/monitor than in smaller jurisdictions. Table 15.2

Country scores on citizen democratic institutions (minimum 0.00–maximum 1.00)

Country

Citizen Democratic Institution score (0–1) (high score = high degree of citizen democracy)

Sweden

0.03

Norway

0.08

Austria

0.08

The Netherlands

0.12

England

0.27

Belgium

0.27

Croatia

0.32

Czech Republic

0.38

Spain

0.39

Greece

0.45

France

0.45

Germany

0.51

Poland (municipalities ≤ 20000 inh.)

0.58

Israel

0.64

Italy

0.66

Poland (municipalities > 20000 inh.)

0.66

Switzerland

0.67

Notes: Scale based on the degree to which four citizen-democratic forms were implemented in a country’s local government system: directly elected mayors; citizen-initiated binding referendums; degree to which councillors are elected on the basis of personal vote; degree to which councillors represent local party lists or ran as independent candidates. For Israel the scale score was computed on the basis of only three indicators. For Poland there are two entries: the scores differ because institutional provisions differ between large and small municipalities. Source: Denters and Klok (2013, p. 669).

194  Research handbook on political representation

CONCLUSIONS In this chapter we have discussed the question of how jurisdiction size affects the functioning of local systems of political representation. Studying these effects is not only interesting against the backdrop of the debates on the merits of local government amalgamations that were recently implemented or are currently discussed in many European countries (see for instance Steiner et al. 2016). This question is also interesting, because both current and past scholars (such as Dahl 1989; Vetter 2002; John Stuart Mill; Alexis de Tocqueville) have considered a well-functioning local democracy as an important resource for the democratic legitimacy of larger political systems such as the nation state and supranational public authorities. In this chapter I have provided an overview of the most important theoretical arguments that are relevant for answering this question. In doing so, I have considered four different interpretations of political representation. In Table 15.3 I have summarized the two main lines of argument, which can be identified in the prevalent literature on size and democracy (e.g. Denters et al. 2014). Table 15.3

Overview of hypotheses implied in two contending theoretical perspectives on the effects of jurisdiction size on representative democracy Brobdingnag

Lilliput

Microcosmic

In larger jurisdictions the accuracy of representation In larger jurisdictions – as a consequence of greater

representation

of interests and preferences may be better than

cultural and socio-economic heterogeneity of their

in smaller jurisdictions, but this positive effect

populations – it will be more difficult to adequately

is contingent upon the electoral system and the

represent the interests and preferences of residents

assembly size

than in smaller jurisdictions

In larger jurisdictions trustee representation is

In larger jurisdictions trustee representation is likely

likely to function better (in terms of the calibre and

to function less well (in terms of the mutual trust

independence of representatives) than in smaller

between citizens and their elected representatives

jurisdictions

and in terms of the responsiveness of trustees to

Trustee representation

the needs of their constituents) than in smaller jurisdictions Party delegate

Larger jurisdictions offer better opportunities (in

In larger jurisdictions party-delegate representation

representation

terms of competitive and strong parties, and voter

is likely to function less well (because of lower voter

choice options) for well-functioning party delegate

interest and knowledge) than in smaller jurisdictions

representation than in smaller jurisdictions Democratic facilitator

In larger jurisdictions the need for councillors (in

In larger jurisdictions councillors (because of the

and monitor

the absence of informal channels for direct citizen

strong party system and their professionalization) are

access) to adopt the role of democratic facilitator and less likely to adopt the role of democratic facilitator/ monitor is greater than in smaller jurisdictions

monitor than in larger jurisdictions

The Brobdingnag argument, which also underlies pleas for amalgamation reforms, emphasizes the better structure of political opportunities that larger units have to offer. Instead, Lilliputians claim that rather than political structures the social dynamics in civic communities are more conducive to the actual well-functioning of the democratic system. As we have argued before, empirical research on the actual effects of jurisdiction size on various aspects of representation is rather limited. Where there is a growing body of empirical evidence on the effects of size on citizens’ political orientations and participation in (local)

Local jurisdiction size and political representation  195 politics, little is as yet known about the effects of jurisdiction size on various aspects of political representation. With this chapter we have developed a theoretically grounded agenda for future research on this under-researched topic. Future research will have to tell whether size has essentially positive effects (as the Brobdingnagians argue) or whether the Lilliputians are right in their more negative expectations.

NOTES 1. The reason is that when comparing units across different levels the size effects are likely to be convoluted by effects of other factors (such as functional responsibilities of jurisdictions) that are also likely to vary across different tiers of government. 2. Dahl and Tufte (1973) in their pathbreaking Size and Democracy have focused on the broader issues of size and democracy, as does the more recent study by Gerring and Veenendaal (2020). 3. But in addition to direct participation by (assembled) citizens, Greek city states like Athens also used elections and sortition (lotteries) to select representatives that were to govern on behalf of the citizens (see Manin 1997). 4. In Swift’s novel, Brobdingnag is the land of the giants. In our previous book (Denters et al. 2014) we have used Brobdingnag as an imaginative label for the ‘Big is Beautiful’ theoretical narrative. On the other hand, Lilliput is the land of the tiny people. In contrast to the Brobdingnag argument I will use Lilliput as the label for the alternative narrative (‘Small is Beautiful’). 5. As Denters et al. (1990, pp. 77–80) and Berghuis et al. (1995, pp. 78–9) have shown, after Dutch amalgamations independent parties representing territorial communities find it increasingly difficult to get representation on the newly elected councils under a system of PR. As we argued before, this type of representation is easier to achieve in multiple single member districts in a FPTP system. 6. Of course jurisdiction size might positively or negatively affect the degree to which second-order effects on voter choice occur.

REFERENCES Aristotle (1984), in J. Barnes (ed.), The Complete Works of Aristotle (2 vols), Princeton, NJ: Princeton University Press. Baglioni, Simone, Bas Denters and Laura Morales (2007), ‘City Size and the Nature of Associational Ecologies’, in William Maloney and Sigrid Rossteutscher (eds), Social Capital and Associations in European Democracies: A Comparative Analysis, London, UK and New York, NY, USA: Routledge, pp. 224–43. Berghuis, J.M.J., M. Herweijer and W.J.M. Pol (1995), Effecten van Herindeling, Groningen: Kluwer. Birch, Anthony H. (1993), The Concepts and Theories of Modern Democracies, London: Routledge. Bovens, Mark and Ankhrit Wille (2017), Diploma Democracy: The Rise of Political Meritocracy, Oxford: Oxford University Press. Burke, Edmund (1999 [1774]), ‘Speech at Mr. Burke’s Arrival in Bristol’, in I. Kramnick (ed.), The Portable Edmund Burke, Harmondsworth: Penguin Books. Cain, Bruce E., Russell J. Dalton and Susan E. Scarrow (eds) (2003), Democracy Transformed? Expanding Political Opportunities in Advanced Industrial Democracies, Oxford: Oxford University Press. Charron, Nicholas, José Fernández-Albertos and Victor Lapuente (2013), ‘Small is Different Size, Political Representation and Governance’, in Santiago Lago-Peñas and Jorge Martinez-Vazquez (eds), The Challenge of Local Government Size: Theoretical Perspectives, International Experience and Policy Reform, Cheltenham, UK and Northampton, MA, USA: Edward Elgar Publishing, pp. 55–82. Dahl, Robert A. (1989), Democracy and its Critics, New Haven, CT: Yale University Press.

196  Research handbook on political representation Dahl, Robert A. (1997), ‘A Democratic Dilemma: System Effectiveness Versus Citizen Participation’, in Robert A. Dahl (ed.), Toward Democracy: A Journey; Reflections 1940–1997, volume 1, Berkeley, CA: Institute of Government Studies Press, University of California, pp. 429–41. Dahl, Robert A. and Edward R. Tufte (1973), Size and Democracy, Stanford, CT: Stanford University Press. Dearlove, John (1979), The Politics of Policy in Local Government: The Making and Maintenance of Public Policy in the Royal Borough of Kensington and Chelsea, Cambridge: Cambridge University Press. Denters, B. and P.-J. Klok (2013), ‘Citizen Democracy and the Responsiveness of Councillors: The Effects of Democratic Institutionalisation on the Role Orientations and Role Behaviour of Councillors’, Local Government Studies, 39 (5), 661–80. Denters, Sebastianus A.H., H.M. De Jong and Jacques J.A. Thomassen (1990), Kwaliteit van Gemeenten: Een Onderzoek naar de Relatie Tussen de Omvang van Gemeenten en de Kwaliteit van het Lokaal Bestuur, ’s-Gravenhage: VUGA. Denters, Bas, Michael Goldsmith, Andreas Ladner, Poul E. Mouritzen and Lawrence E. Rose (2014), Size and Local Democracy, Cheltenham, UK and Northampton, MA, USA: Edward Elgar Publishing. Fiorina, M.P. (1999), ‘Extreme Voices: A Dark Side of Civic Engagement’, in Theda Skocpol and Morris P. Fiorina (eds), Civic Engagement in American Democracy, Washington, DC: Brookings Institution Press & Russell Sage Foundation, pp. 395–425. Gaardsted Frandsen, A. (2002), ‘Size and Electoral Participation in Local Elections’, Environment and Planning C: Government and Policy, 20 (6), 853–69. Gerring, John and Wouter Veenendaal (2020), Population and Politics: The Impact of Scale, Cambridge: Cambridge University Press. Hamilton, Alexander, James Madison and John Jay (1987 [1788]), The Federalist Papers, Harmondsworth: Penguin Books. Henry, Gary T. (1990), Practical Sampling, Newbury Park, CA: Sage. Judge, David (1999), Representation: Theory and Practice in Britain, London: Routledge. Kaap, Harm van der (2006), Politieke Representatie en Lokale Democratie, dissertation, Assen: Universiteit Twente. Kelleher, C.A. and D. Lowery (2009), ‘Central City Size, Metropolitan Institutions and Political Participation’, British Journal of Political Science, 39 (1), 59–92. Klingemann, Hans-Dieter and Dieter Fuchs (eds) (1995), Citizens and the State, Oxford: Oxford University Press. Ladner, A. (2002), ‘Size and Direct Democracy at the Local Level: The Case of Switzerland’, Environment and Planning C: Government and Policy, 20 (6), 813–28. Manin, Bernard (1997), The Principles of Representative Government, Cambridge: Cambridge University Press. McDonnell, J. (2019), ‘Municipality Size, Political Efficacy and Political Participation: A Systematic Review’, Local Government Studies, 45, 1–20. Michels, A. (2011), ‘Innovations in Democratic Governance: How does Citizen Participation Contribute to a Better Democracy?’, International Review of Administrative Sciences, 77 (2), 275–93. Michels, A. and L. De Graaf (2010), ‘Examining Citizen Participation: Local Participatory Policy Making and Democracy’, Local Government Studies, 36 (4), 477–91. Mill, John S. (1977), Considerations on Representative Government, Indianapolis, IN: Bobbs-Merrill. Newton, Kenneth and Brigitte Geissel (2012), Evaluating Democratic Innovations: Curing the Democratic Malaise? Abingdon, UK and New York, NY, USA: Routledge. Norris, Pippa (ed.) (1999), Critical Citizens: Global Support for Democratic Governance, Oxford: Oxford University Press. Norris, Pippa (2000), ‘Women’s Representation and Electoral Systems’, in Richard Rose (ed.), The International Encyclopedia of Elections, Washington, DC: CQ Press, pp. 348–51. Pitkin, Hanna F. (1967), The Concept of Representation, Berkeley, CA: University of California Press. Reif, Karl H. and Hermann Schmitt (1980), ‘Nine Second-order National Elections: A Conceptual Framework for the Analysis of European Election Results’, European Journal of Political Research, 8 (1), 3–44.

Local jurisdiction size and political representation  197 Rose, Lawrence E. (2002), ‘Municipal Size and Local Nonelectoral Participation: Findings from Denmark, the Netherlands and Norway’, Environment and Planning C: Government and Policy, 20, 829–52. Saglie, J. and S.I. Vabo (2009), ‘Size and e-Democracy: Online Participation in Norwegian Local Politics’, Scandinavian Political Studies, 32 (4), 382–401. Smith, Graham (2009), Democratic Innovations: Designing Institutions for Citizen Participation, Cambridge: Cambridge University Press. Steiner, Reto, Claire Kaiser and Grétar T. Eythórsso (2016), ‘A Comparative Analysis of Amalgamation Reforms in Selected European Countries’, in Sabine Kuhlmann and Geert Bouckaert (eds), Local Public Sector Reforms in Times of Crisis, London: Palgrave, pp. 23–42. Thomassen, Jacques J.A. (1991), ‘Kiezers en Verkiezingen’, in Jacques J.A. Thomassen (ed.), Hedendaagse Democratie, Alphen aan den Rijn: Samsom H.D. Tjeenk Willink, pp. 187–212. Tocqueville, Alexis de (1990 [1838]), Democracy in America, New York, NY: Vintage. Van der Kolk, H. (2007), ‘Local Electoral Systems in Western Europe’, Local Government Studies, 33 (2), 159–80. Van Houwelingen, P. (2017), ‘Political Participation and Municipal Population Size: A Meta-study’, Local Government Studies, 43 (3), 408–28. Verba, Sidney and Norbert H. Nie (1987 [1972]), Participation in America: Political Democracy and Social Equality, Chicago, IL: The University of Chicago Press. Vetter, Angelika (2002), ‘Lokale Politik als Ressource de Demokratie in Europa?’, in Angelika Vetter, Lokale Autonomie, Lokale Strukturen und die Einstellungen der Bürger zur lokalen Politik, Opladen: Leske + Budrich, pp. 191–202. Vetter, A. (2009), ‘Citizens versus Parties: Explaining Institutional Change in German Local Government, 1989–2008’, Local Government Studies, 35 (1), 125–42. Wirth, L. (1938), ‘Urbanism as a Way of Life’, American Journal of Sociology, 44 (1), 1–24.

16. Electoral quotas and group representation Mona Lena Krook

Academic and policy debates around the world increasingly recognize the importance of descriptive representation, observing that a variety of groups are under-represented in the electoral process in relation to their share of the general population. In a growing number of countries, these discussions have resulted in the introduction of quotas to encourage or guarantee the participation of selected groups. The vast majority of these measures address women’s representation, with political parties and national parliaments in more than 130 states mandating the greater inclusion of women as political candidates and/or elected officials. Policies for other groups – based on language, religion, ethnicity, nationality, race, caste, age, expatriation, profession and disability – exist in approximately 40 countries (Krook and O’Brien 2010). This chapter provides an introduction to group-based quotas and their impact on dynamics of political representation. The first section maps these measures globally, outlining key variations in their design and the contexts in which they have been adopted. The second section identifies three normative dilemmas raised by quotas related to questions of equality, democracy and group recognition. The next four sections review the state of existing research on quotas and the four dimensions of representation theorized by Pitkin (1967): formalistic, descriptive, substantive and symbolic. Based on this review, the final section charts new frontiers in quota research. As most of the current literature focuses on gender quotas, the discussion highlights the need for more studies on the adoption and impact of quotas for other groups. It also points to the importance of further research on the intersectional effects of quotas, in order to better understand how quotas may mitigate or exacerbate other group-based inequalities.

QUOTAS AROUND THE WORLD Quotas have been implemented for a wide range of groups worldwide. Despite theoretical works developing criteria to identify groups in need of descriptive representation (Mansbridge 1999; Williams 1998), as well as empirical studies linking quota types to innate features of group identities (Htun 2004), global patterns reveal a great deal of diversity in the groups recognized as politically salient – as well as their modes of recognition. Closer analysis of the circumstances surrounding quota adoption, moreover, points to a variety of circumstances in which quotas are introduced. Together, these trends suggest that proposals for group quotas emerge through the active construction of ‘relevant’ political identities, processes which may vary widely across national and regional contexts (Krook and O’Brien 2010). Quotas for Women Electoral gender quotas take three main forms: reserved seats, party quotas and legislative quotas (Krook 2009). Reserved seats set aside places in political assemblies for women. Early policies of this type – appearing in countries such as India, Pakistan and Taiwan – mandated 198

Electoral quotas and group representation  199 a low proportion, less than 10 per cent. More recent measures guarantee as many as one-third to half of all seats for women, as in Rwanda, Tanzania, and some states in India. Party and legislative quotas, in contrast, apply to the share of women among political candidates rather than the proportion of women elected. Party quotas entail pledges by individual parties to include a minimum percentage of women, most commonly between 25 and 50 per cent, on their electoral slates. In Norway, for example, the Norwegian Labour Party has a 50 per cent quota, while the Socialist Left Party, the Centre Party, and the Christian People’s Party all have 40 per cent quotas. Legislative quotas are measures passed by national parliaments requiring that all parties nominate a certain proportion of women. In the 1990s and 2000s, most quota laws mandated 30 per cent female candidates. Over the last ten years, however, many countries, especially in Latin America, have increased their policies from 30 to 50 per cent. At the same time, a growing number of first-time adopters, such as Senegal and Tunisia, are opting for parity from the start.1 As a result of their widespread global diffusion, gender quotas appear in countries with varied institutional, social, economic and cultural characteristics. Nonetheless, some patterns are evident in relation to quota adoption. First, specific quota types prevail in different regions. Reserved seats are concentrated in Africa, Asia and the Middle East. Party quotas are most common in Europe, as well as among leftist parties around the world. Legislative quotas dominate in Latin America, but are also found in Africa and Europe, especially in post-conflict contexts. Second, adoption dates indicate clustering over time. Ten states established policies for women between 1930 and 1980, followed by 12 countries in the 1980s. In the 1990s, however, measures for women appeared in more than 50 countries, which have been joined by over 40 more since 2000 (Krook 2006). Third, intersecting with these trends, specific quota types have predominated at distinct moments in time. Reserved seats were the main quota type between 1930 and 1970. In comparison, party quotas first appeared in the early 1970s, but grew more widespread in the 1980s and 1990s. Legislative quotas emerged in the 1990s, but soon gained momentum and constitute the majority of quotas adopted today. These regional and temporal patterns point to the possibility of learning across national borders, as well as the consolidation of a new global norm in favour of gender-balanced decision-making (Hughes et al. 2015). Quotas for Racial and Ethnic Minorities Quotas for racial and ethnic minorities exist in fewer states, but with a few exceptions2 they take the form of reserved seats (Htun 2004). Due to national and regional variations in definitions of ‘race’ and ‘ethnicity’, however, the specific groups identified for special measures range widely from one context to another. In countries such as Colombia and New Zealand, seats are allocated to members of indigenous groups (Geddis 2006; Van Cott 2005). Across Eastern Europe, ethnic policies ensure representation for small national communities. In Croatia, these include seats for ethnic Hungarians, Italians, Czechs and Slovaks, and Serbs, among others (Krook and O’Brien 2010). In still other cases, groups receiving guarantees are defined by religion, language and class. In Pakistan, a small percentage of seats are reserved for Hindus and Christians, for instance, while in Lebanon, all seats in parliament are divided equally between Christians and Muslims (Salloukh 2006). Belgium divides seats in its upper house among three linguistic groups, Flemish, French and German (O’Neill 1998). India, finally, allocates seats for members of Scheduled Castes and Tribes (Galanter 1984).

200  Research handbook on political representation As reserved seats, these measures address the number of individuals elected, making them stronger guarantees of presence than most quotas for women. These measures tend to have one of two goals: protection or power-sharing. Protection entails allocating seats to groups which constitute a relatively small contingent within the population, including indigenous peoples, members of minority religions and nationalities, and class- or caste-based groups. These provisions are generally minimal, involving as few as 1 or 2 per cent of seats (Reynolds 2005). Aiming to compensate for past oppression, including colonialism, these measures typically over-represent the minority in question (Htun 2004). Power-sharing arrangements, in contrast, involve dividing up most or all seats between two or more factions, defined by ethnicity, religion or language. These policies entail a higher proportion of seats, often as much as 25 to 70 per cent, with the goal being to ensure democratic stability in a divided society – often in the wake of conflict as a means of national reconciliation – by granting group members a guaranteed voice in the political system (Krook and O’Brien 2010). While distinct, the two categories of measures share similar patterns of adoption over time, being popular in the years following World War II and then again in the 1990s (Reynolds 2005). Quotas for Young People Youth quotas have been established in a small but growing number of countries in Africa, Asia, Europe, Latin America and the Middle East. These measures take the same three forms as quotas for women: reserved seats, party quotas and legislative quotas. In Rwanda and Morocco, nearly 8 per cent of seats are reserved for young people, defined as under the ages of 35 and 40, respectively. In other cases, such as Kenya and Uganda, the share of seats is far smaller, between 1 and 3 per cent of all seats. Party quotas are the most common form of youth quota, appearing in party statutes in more than 15 countries. These policies are highly diverse, ranging between 10 and 30 per cent of candidates, and encompass a wide range of age limits, from under 30 to under 40. Youth quotas in Sweden – where multiple parties have instituted 25 per cent youth quotas and targets – appear to be the most effective, based on the share of parliamentarians under the age of 30 (Inter-Parliamentary Union 2018, p. 26). Legislative quotas have been introduced in a more limited array of states. These measures vary from 15 to 25 per cent and apply to candidates under the ages of 35 to 40. In Tunisia, for example, parties must place at least one candidate under the age of 35 among the first four candidates on their lists. In Egypt, parties must nominate a minimum of 16 candidates under the age of 35 across the four electoral districts. Countries have followed several different paths to adoption. In some contexts, youth quotas are perceived by elites as a way to promote greater inclusivity and stability in the wake of dramatic political upheavals (Belschner 2018; Muriaas and Wang 2012). Youth quotas in Uganda, Rwanda and Kenya surfaced after these countries emerged from armed conflict. In Tunisia, Morocco and Egypt, quotas for young people were included in constitutional reforms enacted after the Arab Spring. Elsewhere, young activists have been strong advocates for youth quotas, as in Peru, where they framed these measures as a crucial alternative to a ‘gerontocratic’ political system (Reyes 2015). A consistent pattern across these diverse cases, however, is that quotas for youth tend to be adopted together with, or subsequent to, quotas for women (Belschner 2020). Notably, all of the existing reserved seat provisions have a gender component, mandating that a woman occupy at least one of the seats reserved for youth. In two instances, a party quota in Nicaragua and the legislative quota in the Philippines, the policies

Electoral quotas and group representation  201 establish a single quota for women and youth together. In other cases, youth and gender quotas apply in a parallel but separate fashion (Inter-Parliamentary Union 2018). Quotas for People with Disabilities Quotas for people with disabilities are less common than quotas based on gender, race and ethnicity, and youth. Nonetheless, they coexist with these measures in several countries. One of the first states to establish such a quota was Uganda, where the 1995 Constitution mandated that the state ensure the fair representation of marginalized groups in all government bodies. Currently, five of the 458 seats in parliament are reserved for people with disabilities (approximately 1 per cent of all seats), and every village, parish, sub-county and district council must include at least one man and one woman with a disability. A national electoral college of people with disabilities, with four representatives from each electoral district, selects those who will occupy the five reserved seats in parliament.3 Rwanda and Egypt similarly reserve about 1 per cent of seats in parliament for people with disabilities. While the method for choosing these representatives in Rwanda is similar to that employed in Uganda, the system in Egypt is more complex. Of the 596 members of parliament, 448 seats are filled through majoritarian elections, 120 are elected through party lists, and 28 are appointed by the president. For the purposes of the list component, the country is divided into four constituencies: two 15-member constituencies and two 45-member constituencies. In the former, at least one person with disabilities must be included on each list; in the latter, at least three must be nominated. Parties receiving more than 50 per cent of the vote are awarded all the constituency’s seats, ensuring the election of at least eight candidates with disabilities to parliament (Völkel 2017). Countries with quotas for people with disabilities tend to establish quotas for a wide range of social groups, suggesting that the goal of these measures is to ensure broad-ranging social diversity in political institutions. In Uganda, for example, seats in parliament are also reserved for women, youth, organized labour, and the army. In Egypt, the list of groups is even longer, including women, Copts (an ethnoreligious minority), youth, workers and farmers, and expatriates. Quotas for Other Groups A number of additional groups also enjoy reserved seats in parliaments around the world. These include nomadic tribes, for whom reserved seats are necessary because they lack fixed addresses, and thus would not be adequately represented through a traditional geography-based constituency system. Instances include the Kuchi nomads in Afghanistan (4 per cent of seats) and Bedouin tribes in Jordan (8 per cent). For slightly different geographic reasons, other countries – spread across Africa, Asia, Europe and Latin America – reserve seats in parliament for expatriates. The greatest share of such seats (8 per cent) is found in Cape Verde, a small island nation off the coast of West Africa with a large global diaspora. A final, far more limited, category of representation is occupation. In addition to reserved seats for the army in Uganda and workers and farmers in Egypt, Rwanda reserves seats in parliament for university professors (Krook and O’Brien 2010).

202  Research handbook on political representation

NORMATIVE DILEMMAS Although group-based quotas have been widely adopted around the world, doubts are often expressed about their legitimacy and legality. Quotas are controversial in at least three ways: (1) they encourage affirmative action in candidate selection procedures, provoking a conflict between competing principles of equality; (2) they promote identities over ideas, leading some to argue that they undermine democracy; and (3) they recognize some, but not all, identities as categories of political representation (Krook et al. 2009). Awareness of the ‘false universalism’ behind these neutral-sounding concepts, however, offers a means forward by highlighting how dynamics of privilege and oppression are inflected in these debates, with traditional understandings reinforcing – rather than undermining – prevailing patterns of exclusion (Young 1990). Equality ‘Equality’ is a highly contested concept. Competing views, however, typically contrast equality of opportunities with equality of results (Dahlerup 2007). Stressing equal opportunities tends to favour individual-level explanations of unequal outcomes. From this perspective, citizens enjoy legal equality – and thus, their inability to achieve equality in practice stems from their own personal failures. Prospects for change, in turn, are viewed solely in terms of individual initiative. Emphasizing equal results, in comparison, entails recognizing that some inequalities may in fact be structural. In the face of such systemic barriers, legal equality may not suffice to guarantee substantive equality. Instead, group-based solutions like quotas may be necessary to tackle the broader structures – including discrimination and prejudice – preventing truly equal opportunities for all. Democracy Notions of democracy are typically rooted in concepts of substantive representation, or what Phillips (1995) terms the ‘politics of ideas’, whereby representatives act on behalf of and according to the ideas of those who are represented. Quotas, however, seek to acknowledge the importance of the ‘politics of presence’, suggesting that representatives – as a whole – should mirror the composition of the population in terms of relevant characteristics. Until recently, the lack of descriptive representation was not viewed as a problem for the legitimacy of political institutions – or, indeed, for definitions of ‘democracy’ itself (Paxton 2000). Despite the fact that many scholars view ‘inclusion’ as central to the normative concept of democracy, this insight has not generally been incorporated into analyses assessing the democratic character of different regimes (but see Wang et al. 2017). The widespread diffusion of quotas has pushed these debates in new directions over the last two decades. On the one hand, quotas have been criticized as ‘artificial’ solutions that violate the rights of voters to select their preferred candidates (Dahlerup 2007), overlooking the fact that parties play a strong role in structuring the options available to voters through candidate selection procedures. On the other hand, quotas have become increasingly central to democracy promotion efforts worldwide (Bush 2011). Yet quotas may serve a variety of strategic purposes for political elites in both democratic and authoritarian contexts, both deepening inclusion and political engagement as well as reinforcing the authoritarian status quo (Bush

Electoral quotas and group representation  203 and Jamal 2015; Muriaas and Wang 2012). Democracy may thus be needed for quotas to be meaningful, given that high representation via quotas may not translate into real political power in authoritarian regimes. Group Recognition Theoretical work on group representation has sought to identify criteria to determine which groups might have a ‘case’ for increased descriptive representation. Williams (1998), for instance, outlines four such group characteristics: where patterns of inequality are structured along the lines of group membership; membership is not usually experienced as voluntary; membership is not usually experienced as mutable; and negative meanings are assigned to group identity by the broader society. In the case of the United States, she argues that African Americans and women offer ‘paradigmatic examples’ of groups with ‘strong moral claims for recognition’ (1998, p. 17). Yet there are significant variations across countries in terms of which groups actually receive representational guarantees, raising questions about innate characteristics and group recognition. What patterns of quota adoption suggest, instead, is that groups gain quotas when they are actively constructed as ‘relevant’ political identities – a tendency that becomes evident when comparing quota adoption patterns across cases, and especially across groups (Krook and O’Brien 2010).

QUOTAS AND FORMALISTIC REPRESENTATION In her seminal book, Pitkin (1967) theorizes four types of political representation: formalistic, descriptive, substantive and symbolic. Formalistic representation refers to methods of authorization and accountability, conferring the authority to act on behalf of others, together with the means to hold representatives to account for their actions. In democratic systems, elections provide these mechanisms, shaping who is elected and re-elected (or not) to political office. Research on gender quotas quickly identified electoral systems as a key factor influencing the effectiveness of these policies (Jones 1998). This work noted that proportional representation (PR) systems tended to facilitate the introduction and application of quotas, given the existence of party lists and the possibility to nominate multiple candidates in each district. In contrast, majoritarian, first-past-the-post (FPTP) systems were organized around single candidates in each district, making it less clear how to institute quotas – as well as bringing into sharper focus their zero-sum nature, with male aspirants in some instances needing to give way in favour of female candidates. Differences in electoral systems, together with variations in policy design, have produced wide variations in quota effectiveness; some countries have witnessed strong increases, while others have seen more modest changes or even setbacks in the proportion of women elected (Krook 2009). Quotas tend to be more successful in increasing the share of women elected when implemented in PR systems with closed lists and high district magnitudes (Caul 1999; Jones 2009). Some scholars argue, however, that variations across closed- and open-list PR are due to institutional variables other than electoral systems per se, most notably the use of placement mandates for quotas in closed-list systems that are not possible when quotas are applied in open-list systems (Schmidt 2009). This perspective is consistent with studies showing that quotas are most effective when they require at least 30 per cent female candidates, include

204  Research handbook on political representation placement provisions ensuring that women are placed in winnable list positions, and impose strong sanctions for non-compliance (Schwindt-Bayer 2009; Rosen 2017). Given a plethora of case-study evidence indicating strong resistance among elites to implementing quota provisions (Krook 2016), such mechanisms are crucial for translating quotas into practice, even in presumably favourable institutional contexts. At the same time, some quotas – notably reserved seats – can also contribute to an increased share of women in parliament in countries using FPTP (Rosen 2017; Yoon 2008), suggesting that quota design can overcome some of the barriers associated with electoral system dynamics.

QUOTAS AND DESCRIPTIVE REPRESENTATION Descriptive representation concerns the composition of political assemblies, particularly in relation to its social characteristics. Initial work on gender quotas focused primarily on the numbers of women elected as a result of quota policies (Krook 2009). Emerging research on youth quotas has also taken up this question as it relates to young members of parliament (MPs). These measures employ a wide range of age thresholds. Data from the Inter-Parliamentary Union (2018) finds that youth quotas, as a whole, appear to have a largely negligible impact on the share of MPs under 30, with only a third of countries with youth quotas electing more than the global average: El Salvador, Kyrgyzstan, Lithuania, Mexico, Montenegro, Romania, Tunisia, Sweden and Ukraine. However, states with youth quotas elect a far greater share of parliamentarians under the age of 40 – a pattern suggesting that the individuals benefiting from these quotas are most likely to be near the upper limits of their age groups. Research on other group quotas is less developed, in contrast, but this probably stems from the fact that most of these measures involve reserved seats, which guarantee the descriptive presence of these groups. More recent developments in the gender quota literature, however, point to promising new research directions on the topic of descriptive representation for all groups. This work explores who is elected as a result of gender quotas, taking into account other aspects of group identity. In France, for example, Murray (2010) finds that women entering parliament for the first time as a result of gender quotas were on average younger than new male entrants. Analysing trends in more than 80 countries, Hughes (2011) observes that gender quotas typically lead to the election of majority-ethnic women, while quotas for ethnic minorities usually contribute to the election of minority-ethnic men. When ‘tandem quotas’ are used, however, ethnic minority women tend to benefit. Going one step further, some studies consider further attributes like education and occupation. In Morocco, nearly all female MPs elected via quotas had received higher education and nearly half had earned a doctoral degree (Sater 2007). In Argentina, many women entered parliament with professional backgrounds as teachers, shifting the balance of occupations among legislators overall (Franceschet and Piscopo 2012).

QUOTAS AND SUBSTANTIVE REPRESENTATION Substantive representation refers to attention to group interests in policy-making. Franceschet and Piscopo (2008) theorize two ways in which quotas might alter existing dynamics of policy advocacy. One possibility is a ‘mandate effect’, whereby female legislators elected through

Electoral quotas and group representation  205 quotas sense a special obligation to act on behalf of women. These perceived mandates emerge as a result of arguments made during quota campaigns that female representatives are needed in order to introduce new perspectives in policy-making, including passing laws beneficial to women as a group. The other possibility, however, is a ‘label effect’, whereby women promoted via quotas experience a feeling of stigma attached to the mode of selection, leading them to avoid pursuing women’s issues in an effort to establish themselves as ‘serious politicians’. These dynamics are rooted in negative stereotypes about ‘quota women’ – as less experienced, less capable, and blindly loyal to male party leaders – affecting how they are viewed by colleagues and the public at large. Franceschet and Piscopo (2008) propose that individual women may respond to these two effects in different ways, shaping prospects for substantive representation of women’s policy concerns in the wake of quota adoption. Research finds support for both effects. In Rwanda, a comparison of female parliamentarians in reserved and non-reserved seats finds that those elected through quotas were far more likely to report feeling obligated to act for women as a group (Schwartz 2004). Employing a different research design, Xydias (2008) observes that party quotas in Germany led female legislators to speak more frequently and extensively on women’s issues in parliamentary debates, above and beyond effects attributable to gender differences alone. Evidence from Uganda tempers these conclusions, however, finding more differences in legislative behaviour – measured in terms of participation in parliamentary debates on women’s issues – between male and female MPs than among quota and non-quota women (Clayton et al. 2017). In terms of label effects, interviews in the United Kingdom reveal that – at least initially – women elected via quotas in the late 1990s felt a need to dissociate themselves from these measures and women’s issues more broadly, leading to charges that they had ‘failed women’ (Childs 2004). Interviews conducted in later years with these same women, however, indicate a change in perceptions over the course of their parliamentary careers. Many women moved in the direction of mandate effects as they became more experienced and the perceived stigma attached to being a quota woman lessened over time (Childs and Krook 2012). At the same time, studies disaggregating the policy process provide a more nuanced picture of women’s policy impact; in countries as diverse as Rwanda and Argentina, quotas are associated with a sharp rise in the introduction of women-friendly policy proposals, even if they have rarely altered policy outcomes (Devlin and Elgie 2008; Franceschet and Piscopo 2008).

QUOTAS AND SYMBOLIC REPRESENTATION Symbolic representation captures the cultural meanings and ramifications of the representative process. Existing work on gender quotas operationalizes symbolic representation in two ways: public attitudes towards women’s political leadership and trends in women’s political engagement. Research in the first vein explores whether quotas alter traditional gendered views of women as political actors. The most extensive evidence on this question comes from India, where the lottery system used to designate districts reserved for women creates a ‘natural experiment’, allowing scholars to isolate the effects of quotas versus other factors. Comparing districts where seats had never been reserved for women, those where seats had been reserved once, and those where seats had been reserved twice, Beaman et al. (2009) find that longer exposure to female leaders as a result of quotas can weaken gender stereotypes and eliminate negative bias in how the performance of female leaders is perceived among male constituents.

206  Research handbook on political representation However, work on other cases, such as Finland and Belgium, reveals that public acceptance of quotas can mask continued resistance, especially among male elites, regarding their legitimacy as a measure for overcoming gender inequality (Holli et al. 2006; Meier 2008). According to data from Latin America, citizen support for quotas – especially among men – is higher in countries with higher good governance indicators, signalling the importance of institutional performance in legitimizing quotas as a mechanism of representation (Barnes and Córdova 2016). Studies on women’s political engagement theorize that the election of more women via quotas may signal greater inclusiveness of the political system, inspiring ordinary women to get more politically involved. Although early work on Latin America found that quotas had little or no effect on women’s political activities (Zetterberg 2009), more recent research suggests that the adoption of quotas has contributed to diminishing gender gaps in political participation across sub-Saharan Africa (Barnes and Burchard 2013). Further, evidence from India indicates that quota adoption encourages women to enter politics, acquire political skills, and develop sustained political ambitions. When districts reserved for women were subsequently de-reserved in the next elections, women were far more likely to run for office in these constituencies than in districts that had never been reserved for women – an effect driven largely by women’s decisions to run for re-election, this time against men (Bhavnani 2009). Some research also observes effects on women’s empowerment outside the political sphere. In Rwanda, the existence of quotas for women in parliament and local government has encouraged women more broadly to participate more actively in community life, work outside the home, speak in public meetings, and demand greater equality in their intimate relationships (Burnet 2011).

NEW FRONTIERS IN QUOTA RESEARCH Quotas are increasingly employed around the world as a measure to overcome inequalities in political representation. While the literature on gender quotas is extensive (and still growing), little work currently examines the effects of quotas for other groups in a systematic fashion (for an exception, see Crisp et al. 2018, on Maori seats in New Zealand). An emerging body of studies, however, has begun to explore questions of intersectionality in relation to quotas – asking, for example, how quotas for women and ethnic minorities affect the representation of ethnic minority women (Hughes 2011) and how those for women and youth shape the electoral prospects of young women (Belschner 2020). This work offers reasons for both optimism and pessimism with regard to the impact of quotas on wider patterns of inequality. In cases where gender and youth quotas are introduced simultaneously, for example, progress for young women can be interpreted in two ways. On the one hand, the double counting of young women under both sets of quota requirements can serve to promote young female candidates, counteracting the double discrimination they face. On the other hand, such policies are susceptible to abuse by elites (mainly older men), who can thus limit the number of seats they might otherwise have to give up to newcomers. After the 2014 elections in Tunisia, for example, women under 45 occupied more than 80 per cent of the seats held by that age group. The share of older men stayed roughly the same while those of younger men and older women went down (Belschner 2020). In addition to mapping the representational effects of quotas for a variety of groups, therefore, future research should also attend to their intersectional impact – highlighting how quotas contribute (or not) to greater diversity in political representation overall.

Electoral quotas and group representation  207

NOTES 1. For details on policies mentioned in this chapter (as well as information on quotas in place in other countries), see the Gender Quotas Database at https,//www​.idea​.int/​data​-tools/​data/​gender​-quotas. 2. In Singapore, at least one minority ethnic candidate must be fielded in multi-member constituencies based on the party list plurality bloc vote system – an approach similar to a legislative quota (Tan 2014). 3. For more details, see https,//zeroproject​.org/​policy/​uganda​-2/​.

REFERENCES Barnes, T.D. and S.M. Burchard (2013), ‘“Engendering” Politics: The Impact of Descriptive Representation on Women’s Political Engagement in Sub-Saharan Africa’, Comparative Political Studies, 46 (7), 767–90. Barnes, T.D. and A. Córdova (2016), ‘Making Space for Women: Explaining Citizen Support for Legislative Gender Quotas in Latin America’, Journal of Politics, 78 (3), 670–86. Beaman, L., R. Chattopadhyay, E. Duflo, R. Pande and P. Topalova (2009), ‘Powerful Women: Does Exposure Reduce Bias?’, Quarterly Journal of Economics, 124 (4), 1497–540. Belschner, J. (2018), ‘The Adoption of Youth Quotas after the Arab Uprisings’, Politics, Groups, and Identities, DOI, 10.1080/21565503.2018.1528163. Belschner, J. (2020), ‘Empowering Young Women? Gender and Youth Quotas in Tunisia’, in Drude Dahlerup and Hanane Darhour (eds), Double-Edged Politics on Women’s Rights in the MENA Region, New York, NY: Palgrave, pp. 257–78. Bhavnani, R.R. (2009), ‘Do Electoral Quotas Work after they are Withdrawn? Evidence from a Natural Experiment in India’, American Political Science Review, 103 (1), 23–35. Burnet, J.E. (2011), ‘Women Have Found Respect, Gender Quotas, Symbolic Representation, and Female Empowerment in Rwanda’, Politics & Gender, 7 (3), 303–34. Bush, S.S. (2011), ‘International Politics and the Spread of Quotas for Women in Legislatures’, International Organization, 65 (1), 103–37. Bush, S.S. and E. Gao (2017), ‘Small Tribes, Big Gains: The Strategic Uses of Gender Quotas in the Middle East’, Comparative Politics, 49 (2), 149–67. Bush, S.S. and A.A. Jamal (2015), ‘Anti-Americanism, Authoritarian Politics, and Attitudes about Women’s Representation: Evidence from a Survey Experiment in Jordan’, International Studies Quarterly, 59 (1), 34–45. Caul, M. (1999), ‘Women’s Representation in Parliament: The Role of Political Parties’, Party Politics, 5 (1), 79–98. Childs, S. (2004), New Labour’s Women MPs: Women Representing Women, London: Routledge. Childs, Sarah and Mona Lena Krook (2012), ‘Labels and Mandates in the United Kingdom’, in Susan Franceschet, Mona Lena Krook and Jennifer M. Piscopo (eds), The Impact of Gender Quotas, New York, NY: Oxford University Press, pp. 89–102. Clayton, A., C. Josefsson and V. Wang (2017), ‘Quotas and Women’s Substantive Representation: Evidence from a Content Analysis of Ugandan Plenary Debates’, Politics & Gender, 13 (2), 276–304. Crisp, B.F., B. Demirkaya, L.A. Schwindt-Bayer and C. Millian (2018), ‘The Role of Rules in Representation, Group Membership and Electoral Incentives’, British Journal of Political Science, 48 (1), 47–67. Dahlerup, D. (2007), ‘Electoral Gender Quotas: Between Equality of Opportunity and Equality of Result’, Representation, 43 (2), 73–92. Devlin, C. and R. Elgie (2008), ‘The Effect of Increased Women’s Representation in Parliament: The Case of Rwanda’, Parliamentary Affairs, 61 (2), 237–54. Franceschet, S. and J.M. Piscopo (2008), ‘Gender Quotas and Women’s Substantive Representation: Lessons from Argentina’, Politics & Gender, 4 (3), 393–425.

208  Research handbook on political representation Franceschet, Susan and Jennifer M. Piscopo (2012), ‘Gender and Political Backgrounds in Argentina’, in Susan Franceschet, Mona Lena Krook and Jennifer M. Piscopo (eds), The Impact of Gender Quotas, New York, NY: Oxford University Press, pp. 43–56. Galanter, Marc (1984), Competing Equalities, Berkeley, CA: University of California Press. Geddis, A. (2006), ‘A Dual Track Democracy? The Symbolic Role of the Maori Seats in New Zealand’s Electoral System’, Election Law Journal, 5 (4), 347–71. Holli, A.M., E. Luhtakallio and E. Raevaara (2006), ‘Quota Trouble: Talking about Gender Quotas in Finnish Local Politics’, International Feminist Journal of Politics, 8 (2), 169–93. Htun, M. (2004), ‘Is Gender Like Ethnicity? The Political Representation of Identity Groups’, Perspectives on Politics, 2 (3), 439–58. Hughes, M.M. (2011), ‘Intersectionality, Quotas, and Minority Women’s Political Representation Worldwide’, American Political Science Review, 105 (3), 604–20. Hughes, M.M., M.L. Krook and P. Paxton (2015), ‘Transnational Women’s Activism and the Global Diffusion of Gender Quotas’, International Studies Quarterly, 59 (2), 357–72. Inter-Parliamentary Union (2018), Youth Participation in National Parliaments, 2018, Geneva: IPU. Jones, M.P. (1998), ‘Gender Quotas, Electoral Laws, and the Election of Women’, Comparative Political Studies, 31 (1), 3–21. Jones, Mark P. (2009), ‘Gender Quotas, Electoral Laws, and the Election of Women’, Comparative Political Studies, 42 (1), 56–81. Krook, M.L. (2006), ‘Reforming Representation: The Diffusion of Candidate Gender Quotas Worldwide’, Politics and Gender, 2 (3), 303–27. Krook, Mona Lena (2009), Quotas for Women in Politics, New York, NY: Oxford University Press. Krook, M.L. (2016), ‘Contesting Gender Quotas’, Politics, Groups, and Identities, 4 (2), 268–83. Krook, M.L. and D.Z. O’Brien (2010), ‘The Politics of Group Representation: Quotas for Women and Minorities Worldwide’, Comparative Politics, 42 (3), 253–72. Krook, M.L., J. Lovenduski and J. Squires (2009), ‘Gender Quotas and Models of Political Citizenship’, British Journal of Political Science, 39 (4), 781–803. Mansbridge, J. (1999), ‘Should Blacks Represent Blacks and Women Represent Women? A Contingent “Yes”’, Journal of Politics, 61 (3), 628–57. Meier, P. (2008), ‘A Gender Gap Not Closed by Quotas: The Renegotiation of the Public Sphere’, International Feminist Journal of Politics, 10 (3), 329–47. Muriaas, R.L. and V. Wang (2012), ‘Executive Dominance and the Politics of Quota Representation in Uganda’, Journal of Modern African Studies, 50 (2), 309–38. Murray, R. (2010), ‘Second Among Unequals? A Study of Whether France’s “Quota Women” are up to the Job’, Politics & Gender, 6 (1), 93–118. O’Neill, M. (1998), ‘Re-imagining Belgium, New Federalism and the Political Management of Cultural Diversity’, Parliamentary Affairs, 51 (2), 241–58. Paxton, P. (2000), ‘Women’s Suffrage in the Measurement of Democracy’, Studies in Comparative International Development, 3 (3), 92–111. Phillips, Anne (1995), The Politics of Presence, New York, NY: Oxford University Press. Pitkin, Hanna Fenichel (1967), The Concept of Representation, Berkeley, CA: University of California Press. Reyes, C.P. (2015), ‘A Democratic Revolution for Youth: the “Youth Tithe” as a Doctrine’, Intergenerational Justice Review, 1 (1), 63–4. Reynolds, A. (2005), ‘Reserved Seats in National Legislatures’, Legislative Studies Quarterly, 30 (2), 301–10. Rosen, J. (2017), ‘Gender Quotas for Women in National Politics’, Social Science Research, 66, 82–101. Salloukh, B.F. (2006), ‘The Limits of Electoral Engineering in Divided Societies: Elections in Postwar Lebanon’, Canadian Journal of Political Science, 39 (3), 635–55. Sater, J.N. (2007), ‘Changing Politics from Below? Women Parliamentarians in Morocco’, Democratization, 14 (4), 723–42. Schmidt, G.D. (2009), ‘The Election of Women in List PR Systems: Testing the Conventional Wisdom’, Electoral Studies, 28 (2), 190–203. Schwartz, Helle (2004), Women’s Representation in the Rwandan Parliament, M.A. thesis, University of Gothenburg, Sweden.

Electoral quotas and group representation  209 Schwindt-Bayer, L.A. (2009), ‘Making Quotas Work: The Effect of Gender Quota Laws on the Election of Women’, Legislative Studies Quarterly, 34 (1), 5–28. Tan, N. (2014), ‘Ethnic Quotas and Unintended Effects on Women’s Political Representation in Singapore’, International Political Science Review, 35 (1), 27–40. Van Cott, D.L. (2005), ‘Building Inclusive Democracies, Indigenous Peoples and Ethnic Minorities in Latin America’, Democratization, 12 (5), 820–37. Völkel, J.C. (2017), ‘Sidelined by Design: Egypt’s Parliament in Transition’, Journal of North African Studies, 22 (4), 595–619. Wang, Y.-T., P. Lindenfors, A. Sundström, F. Jansson, P. Paxton and S.I. Lindberg (2017), ‘Women’s Rights in Democratic Transitions: A Global Sequence Analysis, 1900–2012’, European Journal of Political Research, 56 (4), 735–56. Williams, Melissa S. (1998), Voice, Trust, and Memory, Princeton, NJ: Princeton University Press. Xydias, C.V. (2008), ‘Inviting More Women to the Party: Gender Quotas and Women’s Substantive Representation’, International Journal of Sociology, 37 (4), 52–66. Yoon, M.Y. (2008), ‘Special Seats for Women in the National Legislature: The Case of Tanzania’, Africa Today, 55 (1), 61–86. Young, Iris Marion (1990), Justice and the Politics of Difference, Princeton, NJ: Princeton University Press. Zetterberg, P. (2009), ‘Do Gender Quotas Foster Women’s Political Engagement? Lessons from Latin America’, Political Research Quarterly, 62 (4), 715–30.

17. The political representation of women: a feminist institutionalist perspective Joni Lovenduski1

Central to understanding political representation are the almost universal overrepresentation of men and the institutionalized dominance of traditional masculinity in politics. Their political underrepresentation is evidence of women’s relative exclusion from democracies. Yet only feminist political scientists have given sustained academic attention to this problem. Men are typically overrepresented in positions of decision making regardless of the type of political system. Standard contextual variables such as levels of democratization, level of socio-development, women’s participation in the labour market, and type of political system do not explain such overrepresentation. This does not mean that contextual factors are unimportant, rather that they do not explain inequality (Celis and Lovenduski 2018, pp. 152–3). Yet we need to understand it. Unequal representation, like all inequalities, is corrosive, it produces bad outcomes and allows unfair and inefficient patterns of dominance to continue unquestioned. Most important, it undermines the quality of democracy, which depends on adequate systems of political representation. To understand such phenomena, feminist political scientists examined representative institutions to consider how and why gender equality principles are so often overlooked, and how and why representatives and processes of representation fail women. Since the 1970s, feminist research on male dominance in political representation moved from a focus on numbers of women representatives to assessments of their impact on policy to a recent consideration that brought numbers, substance and action together in consideration of how gendered institutions constrain representative processes and outcomes. Recently there has been a further change of emphasis from a concentration on the absence of women to consideration of the overrepresentation of men, a frame shift that highlights male dominance (Bjarnegård and Murray 2018). While the research considers most political institutions, I concentrate here on studies of legislatures, drawing on the feminist institutionalist approach. Representation takes place in an array of institutions at various levels of the system, including political parties, interest groups, bureaucracies, governments and government agencies. Legislators vary in their importance to women’s advocacy. I focus on them here both for reasons of space and for their symbolic importance. In terms of constitutional arrangements, it is clear that key policy decisions are eventually, at least formally, decided in the legislature. In some policy systems, women’s advocates are especially dependent on interventions by legislators (Lovenduski and Guadagnini 2010). However, legislators are constrained by party discipline and party manifestos, institutions that historically are dominated by men and male concerns. In such systems women’s movements and officials appointed to represent women may actually have more scope to represent women than elected legislators bound by party discipline (Guadagnini 2007; Cowley and Childs 2003). 210

The political representation of women  211 Feminist institutionalism enables scholars to understand gendered power relationships in representative politics and to make proposals about strategies for equality. While well theorized, the approach demands a concentration on evidence. It facilitates attention to institutional cultures, to how institutions work, to how they change as they accommodate more women and to how they respond to claims for equal representation for women. Accordingly, in this chapter I first describe the problem of men’s overrepresentation in democracies. I then consider how feminist political scientists have explained and understood how political institutions shape the representation of and by women. Political representation is a process whereby some individuals act on behalf of (a normally larger group of) others. The many arguments about how it does and should work are discussed elsewhere in this volume. Here I am interested in two kinds of political representation within democracies: descriptive and substantive as defined by Hannah Pitkin in 1967. For Pitkin, descriptive representation takes place when the representative stands for a group by virtue of sharing similar characteristics such as race, sex, ethnicity or residence, while substantive representation occurs where the representative seeks to advance a group’s policy preferences and interests (Pitkin 1967). Whilst Pitkin was dismissive of descriptive representation, feminists, broadly defined as advocates of women’s equality, regard it to be important. Although representative democracy can be simply defined as a governing system in which politics are organized around an elected representative assembly, in fact most are complex sets of institutions that have developed through struggles about inclusion and exclusion, inequality and accountability. These ideas take institutional form in legislatures, elections, political parties, constitutions and laws, judiciaries, executives, the separation of powers and freedom of the press and of expression. Acknowledging this complexity, the feminist political science of representative democracies has become an extensive field of research in which male dominance is exposed and the benefits of democracy to women are contested. For example, Drude Dahlerup argues that the level of democracy in a system is not a reliable predictor of women’s representation and raises questions of whether democracy is failing women (Dahlerup 2017) not only in terms of numbers but also in terms of substantive policy outcomes. Understanding institutional factors helps to explain how gender politics works and is essential to the development of proposals for solutions to women’s political inequality. Paralleling and drawing on the institutional turn in political science feminist institutionalism is a variant of new institutional theory which seeks to analyse women as actors in political processes, to ‘gender’ institutionalism and to move the research agenda towards questions about the interplay between gender and the operation and effect of political institutions (Mackay et al. 2010, p. 574). It draws on a distinction between gender as a social category and gender as a process. Thinking of gendering as a process highlights the way in which the formal and informal rules, norms and practices that constitute institutions have the effect of sustaining and reproducing gender regimes that support certain kinds of dominant masculinity2 (Lovenduski 2005; 2015). Feminist institutionalism both assumes and demonstrates that all political institutions are gendered. More specifically, feminist institutionalism exposes the gendered logics of political institutions adapting the conceptual repertoires of institutional theory including concepts of institutional norms, formal and informal rules, path dependence and logics of appropriateness. In addition, feminist institutionalists argue that institutional dynamism provides a basis on which activists may re-gender institutions in order to remove obstacles to women. Hence this growing area of research also considers how institutions change under pressure from women’s advocates. Concepts such as drift and displacement as well as change strategies including lay-

212  Research handbook on political representation ering and conversion are employed in analyses of changing gender power relations in politics (Chappell 2006; Palmieri 2018; Waylen 2018). The approach requires that we consider the roles and impacts of political representatives in their operating contexts (which are not necessarily democracies) and identify the gendered processes that affect them. It is well established amongst feminist political scientists that women’s presence in legislatures is necessary to their fair democratic representation (Phillips 1995). The notion of fairness incorporates a commitment to diversity. So, for example, to be descriptively representative, legislatures and assemblies should be broadly representative of the population of women in terms of class, race, age, sexuality as well as the standard marker of political party (Celis 2008). Inevitably there will be competition among the different groups claiming representation, but the basic point holds. Such contentions are equally applicable to men. In short, feminists are as aware of differences among women and among men as they are of differences between women and men. The scholarship acknowledges that such differences make it impossible to identify a single women’s interest although political disadvantage is common to all women and most men. Feminists have analysed institutions in consideration of who represents women, how they do that, how effective they are and what else has changed as the numbers of women representatives have changed. The feminist institutionalist approach recognizes that institutions, defined as rules and processes, shape debate and decision-making not least by determining who is and is not a representative, and how decisions are made. Recognizing that institutions distribute and maintain power, their work has specifically explored the place of women and the achievements of women’s advocates in various sites of representative politics. The work considers the configurations of systemic, practical and normative institutions to establish whether they ‘facilitate or hinder’ women’s descriptive representation (Childs and Lovenduski 2013; Krook 2009; see also Kenny and Mackay 2009) and to consider if they affect women’s substantive representation (McBride and Mazur 2010; Lovenduski and Guadagnini 2010; Mackay and Waylen 2009; Franceschet and Piscopo 2008; Lovenduski 2005; Dodson 2006; Wangnerud 2000). More recent work asks how, if at all, institutions have changed to accommodate feminist demands (Celis and Lovenduski 2018; Palmieri 2018; Erikson and Josefsson 2019).

NUMBERS The majority of the political representatives of women are men. At the beginning of September 2019, men were 76 per cent of legislators worldwide. Regionally, the proportions ranged from 56 per cent in the Nordic states to 83.3 per cent in the Pacific regions, with variations in each region.3 The data on women’s presence in national legislatures indicate decline in the proportion of men and growth in the number of legislatures. Between 1945 and 1985 the number of legislatures worldwide rose from 26 to 136, while the percentage of men deputies in them fell from 97 per cent to 88 per cent, rising slightly by 1995 when they constituted 88.4 per cent of 176 legislatures. By 2005 men made up 84.1 per cent of deputies in 184 parliaments and by 2019 they constituted 74.5 per cent of 188 national legislatures.4 The legislatures vary considerably. Not all are situated in democracies. While we might reasonably expect women to fare better in more democratic systems, this turns out not to be the case in either descriptive or substantive terms. If we examine levels of women’s representa-

The political representation of women  213 tion in relation to the quality of democracy we find no correlation between how countries are ranked in standard indices of democratic practice such as the Economist Intelligence Unit Democracy Index, and how they rank in terms of the level of women’s representation (Economist Intelligence Unit 2019; and see Dahlerup 2017). In terms of the presence of women in the legislature, the highest-ranking countries in 2019 were Rwanda (61 per cent), Cuba (53 per cent), Bolivia (53 per cent), Andorra (50 per cent) and Mexico (48 per cent). According to the Economist Index of Democracy, in 167 states in 2018, Rwanda ranked 128th, Bolivia 83rd, Cuba 142nd and Mexico 71st (Andorra was not ranked). By the same measure the most democratic political systems were Norway, Iceland, Sweden, New Zealand and Denmark. At 16th, 25th, 7th, 16th and 21st respectively, none of them was in the top ranked countries in the IPU’s table of women’s presence in legislatures in 2019. The numbers reflect only binary differences. A more detailed account would bring differences among women and among men into consideration. Feminists scholars of political representation place diversity at the heart of their analyses in consideration of which women become elected representatives, noting that elected women rarely share social backgrounds with women in the electorate, a pattern that also holds for men, but which is not often problematized. European women elected representatives are more likely to be highly educated, middle class, elite women (Mateo Diaz 2005; Celis 2008; Celis and Childs 2011; Celis and Lovenduski 2018). Conceptually, difference is often expressed in terms of intersectionality (Weldon 2006), according to which various bases of oppression interact in multiple forms of discrimination against women. Although women’s heterogeneity, and how it might affect conceptions of representation, is quite well theorized (see Phillips 1995; Young 2002; Dovi 2007), empirical work has proved difficult to do, not least because so little data on the representation of the various social categories of women and men is available. What data are collected reflect cultural and political characteristics of particular systems. We have good information on the presence of women from different political parties in the Anglo-American and European democracies, on race and ethnicity in the USA; on race and class in the UK, on caste in India, but relatively poor information on other aspects of difference. Difference itself can be politically sensitive, sometimes too sensitive to be officially recorded. In post-genocide Rwanda, for example, the election of a majority of women to the legislature is held to be an indication that tribal divisions are being healed, but it is illegal to ascribe tribal identity to anyone (Childs and Lovenduski 2013). The politics of difference frequently impacts on women’s advocacy. In the USA and UK, divisions of both race and class come into play as the observation is made that the presence only of white, middle class, elite women increases inequalities of race and class. Historically, European socialist men opposed women’s suffrage and women’s movements for much the same reason (Lovenduski 1986). In India, the movement for women’s political representation was criticized for the failure of its middle-class leaders to form alliances with lower caste women. Htun (2004) takes up this point when she argues that candidate quotas are the suitable solution for women’s descriptive underrepresentation, whilst reserved places are best for minority ethnic groups. When reserved places are used to promote women, they risk division in minority communities. Indeed, during the Indian movement for women’s political representation, a prominent politician said the relevant bill was for ‘balkati auraten’, or short-haired women, a reference to upper class urban feminists (Htun 2004, p. 448). In

214  Research handbook on political representation short, feminist claims for presence are often mistakenly labelled as essentialist by potential supporters.

SUBSTANCE Turning to policy issues, at least two major factors impede the translation of women’s descriptive representation into identifiable effects on their substantive representation. These are the lack of general agreement that there is such a thing as women’s interests and the institutional cultures in which representation takes place. The concept of women’s interests is contested and difficult, not least because of intersectionality, which creates cross-cutting differences among women. Hence, what advocates should press for is not completely agreed. However, a feminist women’s policy agenda has been developed by advocates of women’s rights that includes equal pay and equal treatment at work, reproductive rights (including abortion), child care, domestic labour, sexual freedom, freedom from violence, prostitution, trafficking, abolition of female genital mutilation, and political representation.5 Advocacy of some or all of these issues or elements of them is found wherever women have mobilized on behalf of women (RNGS 2012: Htun and Weldon 2012; Weldon 2006; Franceshet and Piscopo 2008). One reason for such common policy concerns across countries may be the transmission of collective ‘women’s issues’ via international women’s movements and feminist activists and international actors such as the UN. The Convention on the Elimination of All Forms of Discrimination against Women (CEDAW 1979), adopted in 1979 by the UN General Assembly,6 covers, inter alia, prostitution, political representation, education, employment, women’s health, women’s economic and social benefits, and marriage and family life; the Beijing Platform for Action (1995) addressed women’s poverty, education, health, violence against women, conflict, economic inequality, power sharing and politics, women’s human rights and the rights of the girl child.7 We do not have sufficient data to analyse the effects of women’s representation on the whole of this policy agenda. There is evidence that women legislators are effective at getting some of these issues on the agenda in some systems (Lovenduski and Guadagnini 2010; Childs and Lovenduski 2013; Palmieri 2018). More broadly, it is possible to assess, albeit fairly roughly, how countries rank in terms of women’s substantive representation. The Annual UN Human Development Report routinely includes a Gender Equality Index (UN Human Development Report Index). This can be used as a proxy measure of women’s substantive representation. Its top five ranked countries in 2015 were Norway, Switzerland, Australia, Ireland and Germany, none of which is in the top five of the Inter Parliamentary Union (IPU) ranked list of proportions of women in legislatures. They were respectively 17th, 38th, 47th, 89th and 46th (IPU 2019).

DO WOMEN REPRESENTATIVES PURSUE FEMINIST POLICY AGENDAS? Numerous mediating factors, identified in a range of studies, are found to affect representatives’ actions (see Dodson 2006), including the external political environment, institutional norms, the impact of party affiliation, ideology and cohesion, differences amongst women

The political representation of women  215 representatives, representatives’ newness, institutional position, for example front- and back-bench and government or opposition membership, committee appointment and leadership, women’s caucus presence, the existence of a women’s policy machinery and the vagaries of policy making.8 The attitudes and behaviour of women and men may converge (Swers 2002a, p. 10), as gender roles alter, or because women’s presence within legislatures causes men to act for women (Reingold 2000, p. 50; Mateo Diaz 2005). Feminists are often disappointed by what they perceive as failure by women representatives to advocate for women in sufficiently robust terms. Yet institutional practices and norms include conventions about how to argue policy positions and how to develop agendas. These shape the claims that representatives can make. Advocates often find that women’s issues and interests can be represented with greater success when they fit the dominant views and discourses about the type of issues and interests that representatives should be concerned with. However, these have a gendered and elitist bias. The strategic framing decisions that are made to fit with these dominant conceptions by definition distort the selection of the types of issues and interests that will be represented (Lombardo et al. 2009). In processes of representation, actors inevitably anticipate favourable or hostile reception to their claims. This leads to the use of strategic framing in order to increase the acceptability of their demands (Franceschet 2011). Hence, institutional norms and practices are such that policy proposals will be mediated as claims that run against dominant ideas and discourse are apparently discarded in favour of claims that are more likely to be supported. Such rhetorical entrapment may lead to the assimilation of women’s advocates who may, through this process, adopt the masculine norms of the institution (Squires 2008). It is worth noting however that institutional analysis suggests that what looks like not very much substantive representation may, in fact, be quite considerable given the effort needed to achieve feminized change in a particular setting (Dodson 2006, p. 29; Childs 2008, p. 170).

INSTITUTIONS, SOCIAL LOGICS AND INSTITUTIONAL CULTURES There are at least two other obstacles to the equality of women’s substantive representation: the continuing effects of the traditional separation of public and private life that excludes women from public presence and consideration, and the barriers of institutional masculinity that are embedded in political institutions. Failures to bring about equality between women and men are not accidental. The political inequality is not only protected by centuries of tradition and powerful institutions, it is rooted in private life, in arrangements designed for reproduction and the preservation of property, and seemingly in biological certainties whereby sex is binary, motherhood is a natural phenomenon but fatherhood more of a rational, legal construct, until recently difficult to prove, but always important to the social order. These institutions are based on a relationship between public and private life that trapped women in the private sphere, as others, as different, as less than men. In these systems women were first treated as dependants, as minors, and more recently as a minority, despite constituting more than half the population (Lovenduski 2019). Efforts by women to get political equality fail because the operating institutions of representative democracy are inappropriate to accommodate ascribed and real differences between women and men and do not overcome the continuing sexual divisions of labour in society. An

216  Research handbook on political representation apparent naturalness of masculine dominance is hardwired into our thinking and part of the DNA of our organizations (Krook and Mackay 2011). Institutional cultures are important to the way that representation takes place, how effective it is and who it benefits. Political institutions are saturated with gender biases, often expressed as institutional masculinity, a powerful concept that draws on feminist research on gender regimes and on institutions. The argument here is compelling. Legislatures were originally designed by and for the men who occupied them. Historically, these were powerful men drawn from exclusively male political elites. The formal and informal rules, processes and logics of the institutions suited well-educated, confident, domineering, authoritative men whose domestic or private lives had no place in their public worlds. Representatives argued, negotiated, compromised and decided policy in sets of established rules that assumed a citizenry of men. This was possible not only because women were not enfranchised, eligible to be representatives or indeed citizens, but also because the private world of domestic life was treated as invisible. For decades the dominance of certain kinds of masculinity persisted as organizational rules and practices were developed, as political agendas were set and as decisions were made on policy. Each part of the process favoured male characteristics and interests. Accepted notions of authority, of appropriate dress, of styles of debate and manners of speech led to a situation in which masculine habits become standards of behaviour, hence the ‘natural’ politician was a man. Legislative chambers favoured voices that carried well, something reinforced by the architecture. Debate was typically peppered with military metaphors, references to human rights that were men’s rights. Physical concepts of heroism and ceremonies recalled war and aggression. The art on the walls and in the chambers of assemblies mainly depicted men. Men came to legislatures expecting to meet with other men with whom they forged alliances and made agreements Their business was organized on the unspoken assumption that they were supported by domestic services mainly provided by women. Working hours, meeting regimes, schedules that took no account of the needs of children or of any domestic caring responsibilities were standard. Rituals evoked histories in which women were not present. Frames of decision making envisaged a population of households led by male breadwinners. Women were not part of such visions except in so far as they related to men as fathers, husbands, brothers, sons. In this way legislatures incorporated social norms about gender appropriate roles according to which women are not thought to be ‘qualified’ for leadership or even for the parliamentary positions from which leaders are most often drawn. That men historically have controlled/dominated women has effects that are embedded in institutions and, despite demands for equality, continue to be so. Any legislature, assembly, government or government agency has considerable regulatory capacity, its own logic, little space for innovation and is underpinned by the ability to assimilate new members including dissidents (Mateo-Diaz 2005, p. 225), Moreover, even in new parliaments established with goals of gender neutrality, gender-biased notions of how to perform as a representative soon emerge (Mackay 2014; Palmieri 2018). Indeed, Kenny and Evans (2020) recently found that such institutional legacies of masculinity are also to be found in the UK’s newly founded (2012) Women’s Equality Party. Recent research on legislatures as gendered workplaces identifies specifically gendered practices, rules and norms. This research is a logical continuation of feminist institutionalist work on political representation and a rich source of information and examples of continuing masculinity in political representation (Erikson and Josefsson 2019; Childs 2016). Institutional legacies of masculine norms are path dependent. Legislatures and political parties were mainly

The political representation of women  217 formed on the assumption that their members would be men. Long excluded, when women arrived not only were they not welcome, they did not fit in even when they adapted their behaviour to institutional norms. Those norms persisted long after the arrival of women, and often continue today in the form of a heteronormative nostalgia. Where women did adapt, they were required to perform masculinity to be effective but of course were routinely criticized for behaving in an unfeminine way. Research shows considerable and effective resistance to attempts by women to change normative regimes to less gender-biased cultures. At the core of this resistance is sexism that occurs in the legislatures themselves and also is supported by wider political communities (Celis and Lovenduski 2018). Despite rhetorical commitments by governments to gender mainstreaming, policies about economic management, housing, healthcare, transportation and pensions tend to harm women more than men for various structural reasons (Women’s Budget Group 2019). In the institutions themselves, such norm change as has occurred tends not to be well established. Palmieri shows how new gender equality norms must compete, often from a position of considerable disadvantage with other policy logics (Palmieri 2018). The sexism that underpins male dominance is a given in contemporary politics. Julia Gillard found international resonance and support as her ‘misogyny’ speech detailing and condemning the almost daily sexist abuse she endured while she was Prime Minister of Australia went viral (Gillard 2012). Gillard’s Australia was no exception. Evidence of institutional resistance to women’s inclusion in male-dominated institutions is widespread. Examples include political parties that commit to sex equality without the necessary revisions of recruitment processes, equality agencies established without the funding necessary for effectiveness, rhetorical commitments to women-friendly policies that are not actioned, marginalized women’s ministries, the delegitimation of women representatives through tokenism, insistence on assimilation to existing (masculine) norms, appointment preferences for anti-feminist women, and toleration of violence against women politicians (Celis and Lovenduski 2018, pp. 157–9). Women representatives frequently report that they are advised to drop the feminism/sex equality work if they want to be promoted. They experience considerable intimidation. Recent research brings evidence of violence to light. Violence against women is a global problem in legislatures. For example, in 2016 an IPU survey of women deputies reported that more than 80 per cent of those surveyed had experienced some form of psychological violence including death, rape, beating and abduction threats as well as threats to kidnap their children and other forms of intimidation (IPU 2016; Palmieri 2018; Krook 2018).

CONCLUSIONS Feminist demands to equalize women’s political representation are demands to share power. Women throughout the world have operated a dual strategy to press their demands. Not only have they worked through the processes to achieve presence in existing institutions, they have also established new institutions within and outside them to secure political space from which to make claims and have had some success at increasing the presence of women and also altering the policy agenda to incorporate more of the issues that are important to women’s movements. However, the glass remains half empty. While women have repeatedly challenged established arrangements to seek equality in political institutions, they have nowhere

218  Research handbook on political representation succeeded. There are two insuperable obstacles. Women are required to play by the rules of the established order, determined by dominant groups of men. As they learn and use the rules, they risk becoming assimilated, weakening their ability to pressure for change. Secondly, their goals are not achievable without comprehensive system change. Indeed, change is not actually enough. What is needed is a paradigm shift whereby roles and institutions are re-imagined and re-engineered. Thus, women’s advocates have so far had to settle for relatively limited regulatory innovation in institutions that are well able to reproduce themselves and preserve their systems of privilege. It is doubtful that sufficient institutional change to permit women’s political equality has been or can be achieved in the face of such deep-rooted political masculinity. It is not clear what, if anything, women should do about it (Lovenduski 2019). If there is a single conclusion to be drawn from this account of the political representation of women, it is that their exclusion from politics is ubiquitous, operated through layer upon layer of established, male-dominated institutions that are insulated by layer upon layer of formal and informal rules of exclusion. The tasks involved in achieving equality in political institutions are enormous. There is no single reason for women’s under-representation, hence no single or simple solution. Moreover, it is not enough to secure simple parity of presence in legislatures; the institutions themselves must not only be re-gendered, they need also to become sites of diversity. Whilst gender roles are changing, change has been resisted and is insufficient and too slow. While overlap in the behaviour of women and men representatives is now more extensive, the path dependence of legislative institutions is such that certain kinds of masculinity and certain groups of men continue to be dominant, not least because the way things were shapes the way things are.

NOTES 1. 2.

3. 4. 5. 6. 7. 8.

I draw on state of the field essays written with Sarah Childs and Karen Celis (Childs and Lovenduski 2013; Celis and Lovenduski 2018). These are noted throughout the text. In this research the distinction between sex and gender is fundamental. Briefly, sex is defined as a more or less fixed biological category, often simplified as a binary difference, while gender is a social construction and process that is usually expressed in terms of a continuum of masculinity and femininity. In 2019, outliers were Bolivia, Cuba, Rwanda and Mexico where men held 39, 47 and 47 per cent respectively of seats in the elected lower house. IPU. Archive accessed 31 October 2019. Celis et al. (2010), citing Dodson and Carroll (1995); Swers (1998); Reingold (2000); Taylor-Robinson and Heath (2003); Wolbrecht (2002); Bratton (2005); Dodson (2006). Currently, 186 countries have ratified or acceded to CEDAW (http://​ treaties​ .un​ .org/​ Pages/​ ViewDetails​.aspx​?src​=​TREATY​&​mtdsg​_no​=​IV​-8​&​chapter​=​4​&​lang​=​en). http://​www​.un​.org/​womenwatch/​daw/​beijing/​platform/​plat1​.htm. Hawkesworth (2003); Kathlene (1995); Carroll (2001); Trimble and Arscott (2003); Gotell and Brodie (1991); Dodson (2001); Chaney (2006); Thomas (1994); Weldon (2002); Reingold (2000); Swers (2002b); Childs and Withey (2006); Childs (2008); Bratton and Ray (2002); Reingold (2008); Dodson (2006); Lovenduski (2005), Mateo-Diaz (2005); Bratton and Ray (2002).

The political representation of women  219

REFERENCES Bjarnegård, E. and R. Murray (2018), ‘Revisiting Forms of Representation by Critically Investigating Men’, Politics & Gender, 14 (2), 265–70. Bratton, K. (2005), ‘Critical Mass Theory Revisited: The Behavior and Success of Token Women in State Legislatures’, Politics & Gender, 1 (1), 97–125. Bratton, K.A. and L.P. Ray (2002), ‘Descriptive Representation, Policy Outcomes, and Municipal Day-care Coverage in Norway’, American Journal of Political Science, 46 (2), 428–37. Carroll, Susan J. (2001), The Impact of Women in Public Office, Bloomington, IN: Indiana University Press. CEDAW (1979), Chapter IV: 8. Human Rights. Convention on the Elimination of All Forms of Discrimination against Women, United Nations Treaty Collection, accessed 31 October 2019 at http://​ treaties​.un​.org/​Pages/​ViewDetails​.aspx​?src​=​TREATY​&​mtdsg​_no​=​IV​-8​&​chapter​=​4​&​lang​=​en. Celis, Karen (2008), ‘Gendering Representation’, in Gary Goertz and Amy Mazur (eds), Politics, Gender, and Concepts, Cambridge: Cambridge University Press, pp. 71–93. Celis, K. and S. Childs (2011), ‘The Substantive Representation of Women: What to do with Conservative Claims?’, Political Studies, 60 (1), 213–25. Celis, K. and J. Lovenduski (2018), ‘Power Struggles: Gender Equality in Political Representation’, European Journal of Politics and Gender, 1 (1–2), 149–66. Celis, Karen, Sarah Childs, Johanna Kantola and Mona Lena Krook (2010), ‘Women’s Interests’, unpublished paper. Chaney, P. (2006), ‘Critical Mass, Deliberation, and the Substantive Representation of Women: Evidence from the UK’s Devolution Programme’, Political Studies, 54 (4), 691–714. Chappell, L. (2006), ‘Comparing Political Institutions: Revealing the Gendered “Logic of Appropriateness”’, Politics & Gender, 2 (2), 223–35. Childs, Sarah (2008), Women and British Party Politics, London: Routledge. Childs, Sarah (2016), The Good Parliament, London: House of Commons Library. Childs, Sarah and Joni Lovenduski (2013), ‘Representation’, in G. Waylen, K. Celis, J. Kantola and L. Weldon (eds), Oxford Handbook of Gender and Politics, Oxford: Oxford University Press. Childs, Sarah and Julie Withey (2006), ‘The Substantive Representation of Women: The Case of the Reduction of VAT on Sanitary Products’, Parliamentary Affairs, 59 (1), 10–23. Cowley, P. and S. Childs (2003), ‘Too Spineless to Rebel? New Labour’s Women MPs’, British Journal of Political Science, 33 (3), 345–65. Dahlerup, Drude (2017), Has Democracy Failed Women? Cambridge: Polity Press. Dodson, Debra L. (2001), ‘Acting for Women: Is What Legislators Say, What they Do?’, in Susan Carroll (ed.), The Impact of Women in Public Office, Bloomington, IN: Indiana University Press. Dodson, Debra. L. (2006), The Impact of Women in Congress, Oxford: Oxford University Press. Dodson, D.L., S.J. Carroll, R.B. Mandel, K.E. Kleeman, R. Schreiber and D. Liebowitz (1995), Voices, Views, Votes: The Impact of Women in the 103rd Congress, Brunswick, NJ: Center for the American Woman and Politics (CAWP), Eagleton Institute of Politics, Rutgers. Dovi, Suzanne (2007), The Good Representative, Malden, MA: Blackwell. Economist Intelligence Unit (2019), Democracy Index, accessed 9 December 2019 at https://​www​.eiu​ .com/​topic/​democracy​-index. Erikson, J. and C. Josefsson (2018), ‘The Legislature as a Gendered Workplace: Exploring Members of Parliament’s Experiences in the Swedish Parliament’, International Political Science Review, 40 (2), 197–214. Franceschet, Susan (2011), ‘Gendered Institutions and Women’s Substantive Representation: Female Legislators in Chile and Argentina’, in Mona Lena Krook and Fiona Mackay (eds), Gender, Politics, and Institutions, London: Palgrave, pp. 58–68. Franceshet, S. and J. Piscopo (2008), ‘Gender Quotas and Women’s Substantive Representation: Lessons from Argentina’, Politics & Gender, 4 (3), 393–425. Gillard, Julia (2012), ‘Julia Gillard’s misogyny speech’, accessed on 31 October 2019 at https://​www​ .youtube​.com/​watch​?v​=​SOPsxpMzYw4. Gotell, Lise and Janine Brodie (1991), ‘Women and Parties: More than an Issue of Numbers’, in Hugh G. Thorburn (ed.), Party Politics in Canada, Scarborough: Prentice-Hall Canada, pp. 53–67.

220  Research handbook on political representation Guadagnini, Marila (2007), ‘The Reform of the State in Italy’, in Melissa Haussmann and Birgit Sauer (eds), Gendering the State in an Age of Globalisation: Women’s Movements and State Feminism in Postindustrial Democracies, Plymouth: Rowman and Littlefield. Hawkesworth, M. (2003), ‘Congressional Enactments of Race-gender: Toward a Theory of Raced-gendered Institutions’, American Political Science Review, 97 (4), 52–50. Htun, M. (2004), ‘Is Gender Like Ethnicity? The Political Representation of Identity Groups’, Perspectives on Politics, 2 (3), 439–58. Htun, M. and S.L. Weldon (2012), ‘The Civic Origins of Progressive Policy Change: Combating Violence against Women in Global Perspective, 1975–2005’, American Political Science Review, 106 (3), 548–69. Inter-Parliamentary Union (2019), accessed on 1 November 2019 at http://​archive​.ipu​.org/​english/​home​.htm. Kathlene, Lyn (1995), ‘Position Power Versus Gender Power: Who Holds the Floor?’, in Georgia Duerst-Lahti and Rita Mae Kelly (eds), Gender Power, Leadership, and Governance, Ann Arbor, MI: University of Michigan Press, pp. 167–94. Kenny, M. and E. Evans (2020), ‘Doing Politics Differently? Applying a Feminist Institutionalist Lens to the U.K. Women’s Equality Party’, Politics & Gender, 16 (1), 1–22. Kenny, M. and F. Mackay (2009), ‘Already doin’ it for Ourselves? Skeptical Notes on Feminism and Institutionalism’, Politics & Gender, 5 (2), 271–80. Krook, Mona Lena (2009), Quotas for Women in Politics: Gender and Candidate Selection Reform Worldwide, New York, NY: Oxford University Press. Krook, M.L. (2018), ‘Westminster too: On Sexual Harassment in British Politics’, The Political Quarterly, 89 (1), 65–72. Krook, Mona Lena and Fiona Mackay (2011), Gender, Politics and Institutions: Towards a Feminist Institutionalism, Basingstoke: Palgrave Macmillan. Lombardo, E., P. Meier and M. Verloo (2009), The Discursive Politics of Gender Equality: Stretching, Bending and Policymaking, New York, NY: Routledge. Lovenduski, Joni (1986), Women and European Politics, Brighton: Harvester. Lovenduski, Joni (2005), Feminizing Politics, Cambridge: Polity. Lovenduski, Joni (2015), Gendering Politics: Feminising Political Science, Colchester: ECPR Press. Lovenduski, Joni (2019), ‘Feminist Reflections on Representative Democracy’, in Tony Wright and Andrew Gamble (eds), Rethinking Democracy, Oxford: Wiley. Lovenduski, Joni and Marila Guadagnini (2010), ‘Political Representation’, in Dorothy McBride and Amy Mazur (eds), The Politics of State Feminism, Philadelphia, PA: Temple University Press. Mackay, F. (2014), ‘Nested Newness, Institutional Innovation and the Gendered Limits of Change’, Politics & Gender, 10 (4), 549–71. Mackay, F. and G. Waylen (2009), ‘Feminist Institutionalism’, Politics & Gender, 5 (2), 237–80. Mackay, F., M. Kenny and L. Chappell (2010), ‘New Institutionalism through a Gender Lens: Towards a Feminist Institutionalism?’, International Political Science Review, 31 (5), 573–88. Mateo Diaz, Mercedes (2005), Representing Women: Female Legislators in West European Parliaments, Colchester: ECPR Press. McBride, Dorothy and Amy Mazur (2010), The Politics of State Feminism: Innovation in Comparative Research, Philadelphia, PA: Temple University Press. Palmieri, S. (2018), ‘Gender Sensitive Parliaments’, Oxford Research Encyclopaedia of Politics, Oxford: Oxford University Press. Phillips, Anne (1995), The Politics of Presence, Oxford: Clarendon Press. Pitkin, Hanna F. (1967), The Concept of Representation, Berkeley, CA: University of California Press. Reingold, Beth (2000), Representing Women, Chapel Hill, NC: University of North Carolina Press. Reingold, Beth (2008), Legislative Women, Getting Elected, Getting Ahead, Boulder, CO: Lynne Rienner. Research Network on Gender Politics and the State (RNGS) (2012), accessed on 16 November 2019 at https://​pppa​.wsu​.edu/​research​-network​-on​-gender​-politics​-and​-the​-state/​. Squires, J. (2008), ‘The Constitutive Representation of Gender: Extra-parliamentary Representations of Gender Relations’, Representation, 44 (2), 187–204. Swers, M. (2002a), The Difference Women Make: The Policy Impact of Women in Congress, London: The University of Chicago Press.

The political representation of women  221 Swers, M. (2002b), ‘Transforming the Agenda: Analyzing Gender Differences in Women’s Issue Bill Sponsorship’, in C.S. Simon Rosenthal (ed.), Women Transforming Congress, Norman, OK: University of Oklahoma Press, pp. 260–83. Swers, M.L. (1998), ‘Are Women more likely to Vote for Women’s Issue Bills than their Male Colleagues?’, Legislative Studies Quarterly, 23 (3), 435–48. Taylor-Robinson, M.M. and R.M. Heath (2003), ‘Do Women Legislators Have Different Policy Priorities than their Male Colleagues? A Critical Case Test’, Women and Politics, 24 (4), 77­­–100. Thomas, Sue (1994), How Women Legislate, New York, NY: Oxford University Press. Trimble, Linda and Jane Arscott (2003), Still Counting: Women in Politics across Canada, Ontario: Broadview. United Nations Human Development Report (1990–2018), available at http://​hdr​.undp​.org/​en accessed 9 December 2019. Wangnerud, L. (2000), ‘Testing the Politics of Presence: Women’s Representation in the Swedish Riksdag’, Scandinavian Political Studies, 23 (1), 67­–91. Waylen, Georgina (2018), ‘A Feminist Institutionalist Approach to IPE and Gender’, in J. Elias and A. Roberts (eds), Handbook on the International Political Economy of Gender, Cheltenham, UK and Northampton, MA, USA: Edward Elgar Publishing Ltd, pp. 113–24. Weldon, L.S. (2002), ‘Beyond Bodies: Institutional Sources of Representation for Women in Democratic Policymaking’, The Journal of Politics, 64 (4), 1153­–74. Weldon, L.S. (2006), ‘The Structure of Intersectionality: A Comparative Politics of Gender’, Politics & Gender, 2 (2), 235–48. Wolbrecht, Christina (2002), ‘Female Legislators and the Women’s Rights Agenda’, in Cindy Simon Rosenthal (ed.), Women Transforming Congress, Norman, OK: University of Oklahoma Press. Women’s Budget Group (2019), accessed on 31 October 2019 at https://​wbg​.org​.uk/​. Young, Iris Marion (2002), Inclusion and Democracy, Oxford: Oxford University Press.

18. Ethnic minorities and representation Miriam Hänni and Thomas Saalfeld

INTRODUCTION: IMMIGRANTS AND ETHNIC MINORITIES – DIFFERENT TRAJECTORIES, SIMILAR MECHANISMS In this chapter, we focus on the political representation of ethnic minority groups in democratic societies. Ethnic groups can be defined as groups of people who share a common origin, language, culture or religion and are conscious of their group membership based on these shared attributes (Cederman et al. 2010b; Fearon 2003). An ethnic minority is an ethnic group living in a country or region where most other people stem from a different ethnic origin. In the current literature on the representation of ethnic minorities we can discern two main strands: scholarship on ethnic politics and on the incorporation of immigrants. These two strands of research have remained largely separate because the two types of ethnic minorities differ in the sources of minority status (typically the history of state formation for autochthonous ethnic minorities versus patterns of immigration and settlement for immigrants). Nevertheless, there are also similarities, especially for autochthonous ethnic-minority groups and large groups of immigrants who settled, became citizens, have remained relatively segregated and retained their cultural identity (e.g., Latinos in the United States, Turks in Germany, Afro-Caribbeans and South Asians in the United Kingdom or Northern Africans in France). For both types of ethnic minorities, ethnic group consciousness may be a crucial source of political participation and mobilization (Junn and Masuoka 2008; McKenzie 2004; Stokes 2003). Both types of minorities may share an experience of discrimination and shared threats based on their ethnicity. Not least, institutional mechanisms to mitigate the under-representation of ethnic minorities have been relatively similar, for example electoral systems (Epstein and O’Halloran 1999; Ferwerda et al. 2018). Therefore, we bring together these two strands of research – on autochthonous ethnic minorities and immigrant-origin ethnic minorities – and discuss similarities and differences.

REPRESENTATION AND ETHNIC MINORITIES: SOME CONCEPTUAL AND THEORETICAL THOUGHTS Modern liberal democracies are ‘representative’ if control over public policy is constitutionally vested in elected officials who are chosen and removed in regular, general, free and fair elections in a chain of delegation and accountability (Strøm 2000). A crucial question in the study of the representation of ethnic minorities of both types has been whether ethnicity matters in this process. This question was discussed classically by Pitkin (1967, pp. 8–9), who defines representation as the activity of ‘making present in some sense of something which is nevertheless not present literally or in fact’ (emphasis in the original). In this process, ‘substantive’ representation refers to any actions (speeches or votes in a representative assembly) taken in the interest of the represented. The personal or ethnic identity of the representative 222

Ethnic minorities and representation  223 is not crucial in this context. In ‘descriptive’ representation, by contrast, ‘representatives are in their own persons and lives in some sense typical of the larger class of persons whom they represent’ (Mansbridge 1999, p. 629). Some models of citizenship have sought to suppress any notion of ethnicity being a legitimate source of representation. The French ‘Republican’ model of citizenship is an example of this (Brubaker 1992). Empirical research has shown, however, that the identity of those represented in politics is not only a matter for normative political theory but has concrete political implications (Bochsler and Hänni 2017). Ethnic groups excluded from effective representation are more likely to be involved in civil conflict with the government, especially during democratization periods (Cederman et al. 2010a) and in times of elections (Cederman et al. 2013). This not only applies to countries in transition, but also to electoral democracies. Ethnic minority groups who lack inclusion, and whose preferences are not represented in politics, are more likely to protest against the government – both violently and non-violently (Hänni 2017a). But even when dissatisfaction and exclusion do not translate into actions against the state, they have potentially harmful consequences for democracy, as excluded or underrepresented groups exhibit less support for the democratic regime (Hänni 2017b), participate less in politics (Burns et al. 2001; Phillips 1995), and are generally less politically involved (Bühlmann and Schädel 2012). This applies not only to ethnically distinct groups (Banducci et al. 2004; Pantoja and Segura 2003), but also to women (Bühlmann and Schädel 2012) or to socio-economically poorer groups (Schäfer 2013). Even where developed representative democracies maintain effective safeguards against ethnic discrimination and guarantee certain minority rights, there remains an uneasy tension between the legitimization of decisions through majority votes and the preferences of minorities. Voting is the main institutional mechanism to terminate disagreement when preferences are not aligned perfectly. Deciding votes by applying the majority principle may be highly problematic if there are distinctive, long-standing minorities at risk of being outvoted permanently (Lijphart 1999). In political theory, the risks of an unconditional primacy of the median voter have been highlighted in discussions about the possible tyranny of the majority. Unconditional rule by the majority risks neglecting strongly-held preferences of minorities. This critique particularly applies to societies with stable identity-based cleavages where the prerequisite of majority rule does not hold; namely the assumption that today’s majority may be tomorrow’s minority (Guinier 1994, Chapter 1). This problem is aggravated if minorities hold intensive preferences very strongly and simple majorities are insufficient to legitimize decisions (Sartori 1987). Issues involving ethnicity and identity are often such decisions as they tend to be harder to resolve than controversies that lend themselves to compromises along the lines of a ‘divide-the-dollar game’ (Jenne 2014). In this context, majority rule easily transforms into the rule of the majority over those citizens identifying as the minority (Bochsler and Hänni 2017). At the same time, well-organized ethnic ‘sub-constituencies’ with clearly articulated and strong preferences may also lead to a ‘tyranny of the minority’ when they veto important collective decisions or influence the outcomes of elections and votes to ensure the success of decisions that are not supported by a majority (Bishin 2009). There has been a long-standing tradition of research on institutions designed to protect the interests of minorities. Lijphart (1999), for example, advanced his model of ‘consociational democracy’, arguing that this model is more suitable for the management of conflict in ethnically divided societies than ‘majoritarian democracy’. Rather than relying on a majoritarian ‘winner-takes-all logic’, he proposes power sharing and veto powers in a constitution charac-

224  Research handbook on political representation terized by proportional representation, coalition government, the territorial decentralization of political power and a system of checks and balances. Proportional representation in elections has been an important institutional device ensuring the representation of ethnic minorities irrespective if they are autochthonous or of immigrant origin. By creating many veto points in the political system, consociational constitutions effectively require super-majorities to change the status quo, thus enhancing the bargaining leverage of minorities (McGann 2004; McGann and Latner 2013). Therefore, consociational institutions have been advocated frequently for the management of conflicts between ethnic majorities and minorities, but they have not always been successful (Seaver 2000).

REPRESENTATION OF (AUTOCHTHONOUS) ETHNIC MINORITIES Many democracies are ethnically heterogeneous and characterized by the presence of ethno-national minority groups. Despite the democratic ideal of equal opportunities for participation, even ethnic minority groups with full citizenship rights1 tend to be less represented (Bochsler and Hänni 2017; Hänni 2017a; Ruedin 2013) and less engaged in the political process (Pantoja and Segura 2003) than majority groups. Table 18.1 compares the participation in elections between members of ethnic minority and majority groups in the 12 multi-ethnic democracies included in the sixth wave of the European Social Survey. In more than half of these countries, individuals belonging to a different ethno-national group than the majority population are less likely to participate in elections than the titular nation. However, even if members of these minority groups participate in elections, they remain underrepresented in most democracies. Figure 18.1 presents the average level of descriptive representation across a country’s democratic period. It is evident that many minority groups are severely underrepresented or entirely excluded from the legislature (values