Download Book

Challenges to Democracy Political Studies Association Yearbook Series

Titles include: Keith Dowding, James Hughes and Helen Margetts (editors) CHALLENGES TO DEMOCRACY Ideas, Involvement and Institutions Chris Pierson and Simon Tormey (editors) POLITICS AT THE EDGE The PSA Yearbook 1999

Political Studies Association Yearbook Series Standing Order ISBN 978-0-333-91373-4 (outside North America only) You can receive future titles in this series as they are published by placing a standing order. Please contact your bookseller or, in case of difficulty, write to us at the address below with your name and address, the title of the series and the ISBN quoted above. Customer Services Department, Macmillan Distribution Ltd, Houndmills, Basingstoke, Hampshire RG21 6XS, England Challenges to Democracy Ideas, Involvement and Institutions

The PSA Yearbook 2000

Edited by

Keith Dowding Professor of Political Science London School of Economics James Hughes Senior Lecturer in Comparative Economics London School of Economics and Helen Margetts Professor of Political Science Director, School of Public Policy University College London

in association with Political Studies Association © Political Studies Association 2001 Softcover reprint of the hardcover 1st edition 2001 978-0-333-78982-7 All rights reserved. No reproduction, copy or transmission of this publication may be made without written permission. No paragraph of this publication may be reproduced, copied or transmitted save with written permission or in accordance with the provisions of the Copyright, Designs and Patents Act 1988, or under the terms of any licence permitting limited copying issued by the Copyright Licensing Agency, 90 Tottenham Court Road, London W1P 0LP. Any person who does any unauthorised act in relation to this publication may be liable to criminal prosecution and civil claims for damages. The authors have asserted their rights to be identified as the authors of this work in accordance with the Copyright, Designs and Patents Act 1988. First published 2001 by PALGRAVE Houndmills, Basingstoke, Hampshire RG21 6XS and 175 Fifth Avenue, New York, N. Y. 10010 Companies and representatives throughout the world PALGRAVE is the new global academic imprint of St. Martin’s Press LLC Scholarly and Reference Division and Palgrave Publishers Ltd (formerly Macmillan Press Ltd). ISBN 978-1-349-41862-6 ISBN 978-1-4039-4017-9 (eBook) DOI 10.1007/978-1-4039-4017-9 This book is printed on paper suitable for recycling and made from fully managed and sustained forest sources. A catalogue record for this book is available from the British Library. Library of Congress Cataloging-in-Publication Data Challenges to democracy : ideas, involvement, and institutions : the PSA yearbook 2000 / edited by Keith Dowding, James Hughes, and Helen Margetts. p. cm. — (Political Studies Association yearbook series ; 2000) Includes bibliographical references and index.

1. Democracy. I. Dowding, Keith M. II. Hughes, James, 1955– III. Margetts, Helen. IV. Series. JC423 .C5124 2001 306.2—dc21 2001021881

10987654321 10 09 08 07 06 05 04 03 02 01 Contents

Notes on the Contributors vii

Introduction xi

PART I CHALLENGES FOR DEMOCRACY: EQUALITY AND SATISFACTION

1 Political Equality in the Coming Century 3 Robert A. Dahl

2 Levelling Down 18 Jonathan Wolff

3 The Political Economy of Human Happiness 33 Benjamin Radcliff

4 Freedom and the Achievement of Happiness 46 Jan Ott

PART II CHALLENGES FROM SOCIETY: CHANGING PATTERNS OF POLITICAL INVOLVEMENT

5 Is There a Crisis of Democracy in Great Britain? Turnout at General Elections Reconsidered 61 Ron Johnston and Charles Pattie

6 Social Capital, Civic Literacy and Political Participation: Explaining Differences in Voter Turnout 81 Henry Milner

7 A Virtuous Circle? The Impact of Political Communications in Post-Industrial Democracies 100 Pippa Norris

8 Hacktivism: Direct Action on the Electronic Flows of Information Societies 118 Tim Jordan

9 Deliberative Democracy and Referendums 131 John Parkinson

v vi Contents

10 Creating Spaces of Deliberation in Barcelona: It’s Good to Talk? 153 Georgina Blakeley

PART III CHALLENGES FROM THE STATE: CHANGING INSTITUTIONS

11 External Inﬂuences on Party Development and Transnational Party Cooperation: The Case of Post-Communist Europe 169 Geoffrey Pridham

12 Promoting Parties and Party Systems in New Democracies: Is There Anything the ‘International Community’ Can Do? 188 Peter Burnell

13 Regulatory Accountability: Towards a Single Citizen-Consumer Model? 205 Martin Lodge

14 E-governance: Weber’s Revenge? 220 Perri 6

15 Decentralization and Development: The New Panacea 237 Elinor Ostrom

Index 257 Notes on the Contributors

Georgina Blakeley is currently Lecturer in Politics at the University of Huddersﬁeld. She has recently submitted her doctoral thesis on participation and democracy in Spain. She has published articles on democracy and civil society in Spain and Chile as well as on aspects of Spanish politics more generally.

Peter Burnell is a Professor of Politics at the University of Warwick. His interests include democratization, the political economy of international assistance and politics in Zambia. He is joint founding editor of the journal Democratization.

Robert A. Dahl is Sterling Professor of Political Science, Emeritus, at Yale University and the author of numerous books on democracy, including A Preface to Democratic Theory (1956), Who Governs? Democracy and Power in an American City (1961), Democracy and Its Critics (1989), and On Democracy (1998).

Keith Dowding is Professor of Political Science at the London School of Economics. He has published widely in political theory, public administration and urban politics. His books include Rational Choice and Political Power, The Civil Service, Power and he has co-edited Preferences, Institutions and Rational Choice. He is also co-editor of the Journal of Theoretical Politics.

James Hughes is Senior Lecturer in Comparative Politics. He is the author of two monographs and numerous journal articles on Russian history and politics. His current research projects include developments in post-Soviet Russian federalism and a comparative study of the role of local elites in transition in Central and Eastern Europe.

Ron Johnston is a Professor in the School of Geographical Sciences at the University of Bristol. He has published widely on various aspects of electoral studies; the most recent work (co-authored with Charles Pattie, Danny Dorling and David Rossiter) is From Votes to Seats: the UK’s Electoral System in Operation since 1950 (Manchester University Press, 2001).

Tim Jordan is a lecturer in Sociology at the Open University. He is the author of Cyberpower: the Culture and Politics of Cyberspace and the Internet (London:

vii viii Notes on the Contributors

Routledge) and co-editor of Storming the Millennium: the New Politics of Change (London: Lawrence & Wishart). He has published work about on-line social movements, hackers and social theory, as well as working on recent popular protests and social movements. E-mail: [email protected]

Martin Lodge is Lecturer in Public Policy in the School of Public Policy, Economics and Law at the University of Ulster at Jordanstown. His research focuses on comparative regulation, in particular railway regulation, comparative public administration and German and EU public policy.

Helen Margetts is Professor of Political Science and Director of the School of Public Policy, University College London. She is the author of Information Technology in Government: Britain and America (Routledge) and has published widely on electoral systems, the relationship between the state and the Internet and other information technologies, administrative reform and political participation.

Henry Milner is a political scientist afﬁliated with Laval University in Canada and the University of Umea in Sweden. He is co-editor of Inroads, a Canadian journal of opinion and policy, and has published a number of books on Scandinavian and Canadian politics. His most recent book is Civic Literacy: How Informed Citizens Make Democracy Work, University Press of New England, 2001.

Pippa Norris is Associate Director (Research) of the Shorenstein Center on the Press, Politics and Public Policy, and Lecturer at the John F. Kennedy School of Government, Harvard University. The author or editor of two dozen books, her latest are A Virtuous Circle: Political Communication in Postindustrial Societies (Cambridge University Press, 2000) and Digital Divide: Civic Engagement, Information Poverty and the Internet in Democratic Societies (Cambridge University Press, 2001). More details about research and publications can be found at pippanorris.com.

Elinor Ostrom is Co-Director of the Workshop in Political Theory and Policy Analysis and the Center for the Study of Institutions, Population, and Environmental Change at Indiana University. Her major area of study is institutional analysis. She is currently studying with a network of colleagues the ways in which government, private, and common property affect user incentives and forest conditions.

Jan Ott is carrying out PhD research into freedom and happiness at Erasmus University, with Ruut Veenhoven. He previously studied sociology and law, Notes on the Contributors ix specializing in development problems, social economic policies, labour relations and government administration, and acting as a student-assistant in social psychology. Since 1984 he has been a policy adviser in labour relations for the Dutch Ministry of Social Affairs and Employment.

John Parkinson is a doctoral student in the Research School of Social Sciences at the Australian National University, previously at Melbourne University. He specializes in democratic theory and public policy, especially popular participation in decision-making. His current research examines legitimacy problems of deliberative designs.

Charles Pattie is a Professor in the Department of Geography at the University of Shefﬁeld. In collaboration with Ron Johnston he has published widely on various aspects of electoral geography. He is currently co-editor, with Pat Seyd, of an ESRC-funded project into citizenship and political participation in Great Britain.

Geoffrey Pridham is Professor of European Politics and Director of the Centre for Mediterranean Studies (CMS) at the University of Bristol. He has published widely on problems of post-authoritarian democratization in Southern Europe and, more recently, in Central and Eastern Europe. Among his publications may be counted ‘Democratization in Eastern Europe’, with Tatu Vanhanen (Routledge, 1994); ‘Stabilising Fragile Democracies: Comparing New Party Systems in Southern and Eastern Europe’, with Paul Lewis (Routledge, 1996); ‘Experimenting with Democracy: Regime Change in the Balkans’, with Tom Gallagher (Routledge, 2000); and, ‘The Dynamics of Democratization: A Comparative Approach’ (Continuum, 2000). He is now working on a project looking at EU enlargement and domestic politics in ﬁve post-Communist states with particular attention to Slovakia.

Benjamin Radcliff is Associate Professor of Government at the University of Notre Dame, Indiana. His present research agenda focuses on the links between mass politics, governmental policies, and quality of life. Radcliff’s articles have appeared in the British Journal of Political Science, the American Political Science Review, the American Journal of Political Science, the Journal of Politics, and a variety of other journals. E-mail: [email protected]

Perri 6 is a Senior Research Fellow in the Department of Government at the University of Strathclyde. His most recent books include Governing in the Round: Strategies for Holistic Government (with D. Leat, K. Seltzer and G. Stoker, 1999), Morals for Robots and Cyborgs: Ethics, Society and Public Policy in the Age of Autonomous Intelligent Machines (1999), The Future of Privacy (2 vols, 1998), x Notes on the Contributors

Holistic Government (1997) and The Contract Culture in Public Services (ed. with J. Kendall, 1997). He is currently ﬁnishing a new book on ‘joined-up’ government and research on e-governance.

Jonathan Wolff is Professor of Philosophy at University College London (UCL). He is the author of Robert Nozick: Property, Justice and the Minimal State (1991) and An Introduction to Political Philosophy (1996). Introduction

The twentieth century was the century of democracy. As Robert Dahl points out in his essay in this collection, at the beginning of the twentieth century only eight nations had the basic institutions of representative democracy and only one of those had female suffrage. By the end of the century universal suffrage existed in 85 nations which together include nearly 60 per cent of the world’s population. The century saw the break-up of the great European empires; the rise and fall of fascism in Europe; the rise and fall of communism in Europe; newly emerging democracies in Latin America as their authoritarian and dictatorial regimes started to crumble; and democracy beginning to emerge in South-East Asia. There is a long way to go for fully developed democracies around the world, and there are many problems emerging for newly formed democracies in East and Central Europe, and continuing problems in Africa, Latin America and elsewhere. We must also not be complacent in Britain, whose democracy may be old but is showing signs of tiredness, with a devastated local government system, great swathes of unelected bodies with enormous public powers, an unelected upper chamber, and falling turnout at all levels of elections. The democratic deﬁcit in the European Union is not being addressed as a federal state is carved out without the kind of constitutional assembly which saw the start of the US march towards democracy. The US also has democratic problems of its own, with low levels of participation, high levels of cynicism and, like the UK, increasing levels of inequality. Each chapter in this book is concerned with the challenges facing us as we enter a new century. They examine ideas about democracy, new ways of involving citizens in government and the institutions which form democracy’s core. The book is divided into three parts. The ﬁrst – ‘Challenges for Democracy: Equality and Satisfaction’ – is concerned with the challenges posed for democracy by inequality and providing for the happiness of citizens. Inequality and democracy are inextricably linked. The very basis of democratic thought is one person one vote – equality in franchise. But liberty is also a fundamental component of a democratic system; liberty and equality often clash. State institutions which can help to overcome great inequalities are also those which cut across economic and personal freedoms. One of the key questions of this section is the level of satisfaction provided under different types of democracy – those which provide redistribution and strong welfare systems and those which concentrate upon economic freedom. The second part – ‘Challenges from Society: Changing Patterns of Political

xi xii Introduction

Involvement’ looks at the participation of citizens in a democracy from various standpoints – from falling turnout in mature democracies, through the use of referendums and deliberative devices, to new forms of participation and protest using the Internet and new technologies. Can citizens truly be involved in democratic decision-making? Is falling turnout a result of alienation from representative systems where voters feel powerless? Can more direct forms of democratic decision-making such as referendums provide a genuine channel which enhances democracy? Can the ideas of those arguing for more deliberative forms of democracy actually work in practice? All these questions are addressed in this part. Institutional change in both emergent and established liberal democracies is a further challenge to democracy – yet institutional development is also necessary to keep pace with a changing world. These concurrent challenges are covered in the final part of the book – ‘Challenges from the State: Changing Institutions’. The new democracies of Eastern and Central Europe are struggling to produce party systems recognizable to Western ideas of democracy. Can the West help in the production of competitive party systems? Can the state find new ways of regulating itself and harness new technologies to aid this process? Should the state intervene in local affairs to enforce new participatory modes or should it allow groups to find their own ways of working together and regulating their life?

Challenges for democracy: equality and satisfaction

How much equality is required for a fully developed democratic system is a fundamental issue for democracy. It is an issue which Robert Dahl picks up in the essay which opens this volume. Dahl notes the spread of democracy and equality during the twentieth century. Quoting Tocqueville’s account of the spread of equality and democracy throughout the past 800 years, Dahl shows how these values have increased during the last century. Dahl argues, contra Tocqueville, that there is an inexorable rise in rights and liberties under democracy, and notes how stable mature democracies have become. However, that is not to say that there are not challenges facing democracy in the new century. While the advanced welfare democracies of the late twentieth century started to bring about greater equality, these advances have been reversed in the last twenty years. But Dahl argues that a strong measure of equality is required for a well-functioning democracy. One challenge is that provided by an economic order of market capitalism and the problems of system effectiveness with high levels of citizen participation. He also sees the challenges to liberty, equality and democracy posed by globalization, and the challenges we face in making all organizations in society democratic. Introduction xiii

Jonathan Wolff takes up an issue that has long troubled egalitarians. Critics of egalitarianism often claim that the attempt to make people equal in the end makes all worse off. Modern egalitarians are agreed that they do not oppose Pareto improvements, that is they do not oppose making one person better off as long as no one else is thereby made worse off. In Wolff’s terms modern egalitarians all oppose ‘levelling down’. Wolff argues that there are specific circumstances where egalitarians should not oppose levelling down. These circumstances arise in cases where the principle of equality is symbolic of the equality of different people, even where someone else’s gain is not specifically another’s loss. Wolff’s argument suggests that there may be a trade-off between maximizing satisfaction in a democratic society, and promoting equality. Both Benjamin Radcliff and Jan Ott look at public satisfaction levels in both mature and emerging democracies in relationship to equality and liberty. They both suggest that the nature of democracy and democratic competition may have dramatic effects on satisfaction levels within nations. Both studies utilize empirical evidence on citizens’ stated preferences over their general levels of satisfaction or ‘happiness’. Radcliff uses both pooled time series data from the Eurobarometer of member states of the EU, and cross-sectional data from a wider group of industrialized nations from the World Values Study. Both data sets suggest that higher levels of welfare spending by the state increase overall satisfaction levels. When looking at party control he also finds that the longer a left-wing party is in control the happier the citizens claim to be. This evidence is entirely on stated preferences, and of course citizens do vote out left-wing governments and vote in right-wing ones. It does not show that citizens are ‘objectively’ better off in high welfare spending regimes and under left-wing governments, but people do report that they are more satisfied. This challenges a key component of those who argue that democracy must first and foremost protect individual liberties. Jan Ott’s chapter complements Radcliff’s essay. Ott’s data are from a broader set of nations and includes both developed and developing nations. Using data from a number of sources, he examines the relationship between satisfaction and different types of freedom. He shows that democracy and democratic values are correlated with satisfaction levels. He finds significant correlations between ‘happiness’ and sets of indicators from economic, political and press freedom. However, he finds a difference between rich and poor nations. The correlation between economic freedom and satisfaction holds across the full set, but political freedom and press freedom is correlated with satisfaction levels only in relatively rich nations. In poorer nations, satisfaction is higher where government consumption is low, there are lower levels of redistribution, fewer state-owned enterprises and a generally lower level of state intervention in the market. It may be that government intervention dampens xiv Introduction market behaviour and expectations, and in poor nations satisfaction is correlated with higher expectations about a better life in the future established by a more vibrant economy. Transfers will be relatively low, and poor nations cannot afford the welfare benefits to provide proper social insurance. However, in richer nations, people appreciate the social insurance of strong state welfare programmes even if they dampen the market slightly. Here the trade-off is worth it. Or, to put it in a nutshell, only the rich can afford the welfare state.

Challenges from society: changing patterns of political involvement

If democracy rests on the key principles of popular control and political equality, then a major challenge must be securing the continued willingness of individual citizens to be involved in the project. Liberal democracies in the twenty-first century must maintain such involvement in the face of changing patterns of political participation. The twentieth century was the age of the election – the century which started with mass enfranchisement and ended, in Britain at least, with concerns over citizens’ enthusiasm for using their vote and participating in political institutions or political decision-making. Ron Johnston and Charles Pattie explore one such concern, examining patterns of turnout in elections in the last fifty years of the twentieth century, asking whether the phenomenon of low voter turnout is really new, or whether it reflects patterns that have existed since the 1950s. They conclude the latter, finding a trendless fluctuation from 1959 around an average turnout of 75 per cent. Voters rationally choose to vote when they think there will be a close outcome and 1997 reflected the expectation of a clear victory for Labour. This rationality does not appear to extend to marginality at constituency level, where the dominant party in a constituency (Labour voters have traditionally been more difficult to mobilize) and the age of voters (younger voters have always been less willing to vote) seem to remain the significant factors. Thus at the national level, outcomes reflect in part the electoral system used – if the outcome is a foregone conclusion, more people abstain. Henry Milner examines voter turnout across OECD countries through the lens of social capital, defined by Robert Putnam as features of social life – networks, norms and trust – that enable participants to act together to pursue shared objectives. He observes that the burgeoning debate on the relevance of social capital has marginalized the most basic expression of democratic citizenship: voting. Contrary to the arguments of the social capitalists, Milner finds little basis for the presumption that people in high trust societies or those with high levels of organizational participation vote in national or local elections more than do those in low trust ones. He introduces the concept of ‘civic lit- Introduction xv eracy’ to explain the absence of a relationship between local voting turnout and the standard indicators of social capital, arguing that the link between social capital and political participation passes through political knowledge, measured through literacy levels and newspaper circulation, representing the ‘demand’ side of civic literacy, while the choice of electoral system (proportional or otherwise) again emerges as a key factor on the ‘supply’ side. Pippa Norris and Tim Jordan investigate the reaction of citizens to the major new forums for political activity to emerge in the second half of the twentieth century: the mass media and, more recently, the Internet. Pippa Norris outlines the claims and counterclaims that have been made for ‘the media malaise’, the argument that the news media is to be blamed for growing public disengagement and mistrust of government. She argues that contrary to what authors such as Putnam have claimed, use of the news media is positively associated with a wide range of indicators of political knowledge, trust and mobilization – people who watch more TV news, read more newspapers, surf the net are more knowledgeable, trusting of government and participate more. She identifies a ‘virtuous circle’ where the news media serve to ‘activate the active’, yet have less power to reinforce the disengagement of the disengaged. Tim Jordan looks at the Internet and more radical forms of participation, showing how individuals undertaking a kind of technological apoliti- cal participation increasingly show off their technological skills by breaking into the computer systems of governments and large corporations. These ‘hacktivists’ are developing connections with radical social movements (ecological, animal rights and anti-capitalist) that have existed since the 1960s to form a new electronic direct action movement. Thus technological game playing in the electronic nervous systems of liberal democracies becomes political action, with fundamental implications for the future of popular protest. Two chapters look at new ways of involving citizens in decision-making: direct and deliberative forms of democracy. John Parkinson looks at the potential of referendums to embody the principles of popular control and political equality and thereby to support ‘direct’ democracy in the twenty-first century by examining a series of referendums and petitions in Switzerland since the late 1960s. He questions the openness of this democratic device; participation rates are low (especially among younger, liberal, lower income and less educated members of society); initiating referendums is costly; information is distorted by the media; citizenship is hard to obtain and foreigners are excluded. He concludes from the Swiss experience that while referendums may appear deliberative in isolation, their social and institutional setting introduces fundamental distortions which make genuine deliberation highly unlikely: ‘issues come pre-framed, wielding cultural categories in ways which are not readily open to challenge.’ Georgina Blakeley examines a form of xvi Introduction deliberative democracy, using examples from a participatory experiment carried out in Barcelona in the 1980s and 1990s, where a series of sectoral and territorial advisory councils were used to facilitate dialogue and interaction between the associational fabric of the city and the local council. The experiment illustrates the dangers of ‘top-down’ democratic initiatives, where a key objective is the securing of cooperation and the support of associations and, arguably, too many functions are loaded onto one democratic device. Deliberative democracy does not prove itself as a model of democracy, but simply an element within a wider theoretical framework. Wherever disagreement appears to be an insurmountable barrier to deliberation, participants resort to other more effective mechanisms to influence public administration. And deliberative democracy does not have the means to address the key question of how people can be encouraged to deliberate in the first place, even in Barcelona where there was a political will to facilitate spaces of deliberation and a historical tradition of associational activity.

Challenges from the state: changing institutions

Other challenges to democracy come from changes in the shape of the state and political institutions, particularly given the constant battle of newly emergent liberal democracies to stabilize their political systems. Key to such a process must be the institution of the political party, also an invention of the twentieth century. Geoffrey Pridham investigates the influence of transnational party cooperation between member states of the European Union and new democracies on party development in post-communist Europe, challenging the assumption that party development is essentially a concern of domestic political arenas. He finds that the pattern in the 1990s has been for West European party-political habits to imprint themselves on parties from countries across Central and Eastern Europe. The coming years seem likely to see this pattern of influences developing into more of a two-way process, especially when some of these countries join the European Union, to the point where it might be called a process of Europeanization. Peter Burnell outlines the challenge of weakly institutionalized party systems to ‘third wave’ democracies and investigates the options for the international community in terms of political reform and help with party-building and electoral competition. Democratic assistance ranges from the ‘soft end’ of transfer of resources to incentives, inducements and rewards, which can be applied to individual parties or the party system as a whole – and can be effective or non-effective in democratic terms. Other democratic challenges coming from the state are the legacy of the fast- moving public management change of the last twenty years of the twentieth Introduction xvii century. Martin Lodge looks at the impact of privatization on the contemporary state, investigating the democratic implications of the privatization and subsequent regulation of utilities across Britain, Ireland, Sweden, Germany and New Zealand. Rather than a uniform model of the so-called regulatory state, he finds evidence of the persistence of diversity with varying use of choice, voice, information and representation as the key tools of national regulatory models. Perri 6 looks at another key trend – the widespread use of information and communication technologies in the policy-making process (e-governance) and the extent to which such technologies are used to reinforce Weberian rationality. His conclusion is that decision-makers shape technology to pre-existent ends, and that its use introduces new rituals and symbols into the policy toolkit, rather than new methods of bureaucratic control. In the concluding chapter, Elinor Ostrom identifies institutions as both the key and the ultimate challenge to successful decentralization in developing countries, in which ‘developed countries’ assist ‘developing’ countries in the process of economic and political development by providing resources for local initiatives. Applying common pool arguments to natural resource systems such as fishery, forestry and irrigation, Ostrom concludes that decentralization requires systems where citizens are able to organize not just one but multiple governing authorities at differing scales, where each unit exercises considerable independence to make and enforce rules within a circumscribed scope of authority for a specified geographical area. Development of effective institutions is a process that takes time and depends upon problem-specific factors. The process of institutional development cannot be jump-started through substantial development assistance – but requires a solid theoretical understanding of institutional development. Ostrom’s ‘polycentric’ political systems may look messy and chaotic – but they outperform neat hierarchies of unified governmental layers. Learning better how to analyse such complex systems stands out as the challenge for political science in the twenty-first century. Her chapter highlights the grave dangers of imposing systems on polities, and warns that decentralization, any more than centralization, should not be seen as the new panacea to solve collective action problems.

All of the essays in this book were presented at the 50th Anniversary Annual Conference of the Political Studies Association in London, 10–13 April 2000. The chapters by Robert Dahl and Elinor Ostrom were plenary addresses and we were particularly pleased to be able to include them in this collection. The other chapters were chosen from among the ﬁve hundred plus papers given at that conference. The choice of these papers was based both on their quality and the extent to which they ﬁt the theme of the conference and of this book: Challenges to Democracy in the 21st Century. xviii Introduction

We would like to thank everyone who helped to make the conference such a success, particularly the organizing team of Razmik Panoussian, Jurgen de- Wispelaere (Web Master) and Kennedy Stewart (Graduate Conference Convenor).

Keith Dowding James Hughes Helen Margetts Part I Challenges for Democracy: Equality and Satisfaction 1 Political Equality in the Coming Century Robert A. Dahl1

Is greater political equality among citizens a highly desirable goal for people in democratic countries to pursue in the twenty-first century? A convincing answer requires convincing responses to at least four questions: Is political equality a feasible goal? Is it a desirable goal? Are other feasible processes more desirable? And finally, if the answers to these questions reveal serious problems, can we find satisfactory solutions?

Is political equality feasible?

As to the ﬁrst question, I need hardly remind this audience of the enormous and persistent barriers to political equality, and indeed to human equality in general. For example, many Europeans – should I say, other Europeans? – would probably agree that the people of the highly advanced democratic country in which I speak today have been, over many centuries, concerned more passionately than in any other western European country with maintaining social inequalities in the form of differences in class and status. The recent report on university admissions reveals that these inequalities still have powerful consequences. I might point out that until only a few months ago this country, unlike any other democratic country in the world, maintained the astonishing anachronism of an upper house in its national parliament consisting overwhelmingly of hereditary peers. My own country, I hasten to add, provides telling examples of the huge gap between the rhetoric and reality of political equality. In the second paragraph of a document that is otherwise a rather tedious listing of the ‘repeated injuries and usurpations’ inﬂicted by the King of Great Britain, we encounter the famous assertion of a supposedly self-evident truth, that all men are created equal. The authors of the American Declaration of Independence and the 55 delegates to the Second Continental Congress who voted to adopt it in July 1776 were, of course, all men, none of whom had the slightest intention

K. Dowding et al. (eds.), Challenges to Democracy © Political Studies Association 2001 4 Robert A. Dahl of extending the suffrage or many other basic political and civil rights to women, who, in legal contemplation, were essentially the property of their fathers or husbands. Nor did the worthy supporters of the Declaration mean to include slaves, or, for that matter, free persons of African origin, who were a substantial fraction of the population in almost all the colonies that now claimed the right to become independent self-governing republics.2 The principal author of the Declaration, Thomas Jefferson, owned several hundred slaves, none of whom he freed during his life; and he freed only five on his death.3 It was not until more than four score and seven years later (to borrow a poetic phrase from Lincoln’s Gettysburgh Address) that slavery was legally abolished in the United States by force of arms and constitutional enactment. And it took yet another century before the rights of African-Americans to participate in political life began to be effectively enforced in the American South. Now, two generations later, Americans white and black still bear the deep wounds that slavery and its aftermath inflicted on human equality, freedom, dignity and respect. Nor did our noble Declaration mean to include the people who for some thousands of years had inhabited the lands that Europeans colonized and came to occupy. We are all, I think, familiar with the story of how the settlers denied homes, land, place, freedom, dignity and humanity to these earlier peoples of America, whose descendants even today continue to suffer from the effects of their treatment throughout several centuries when their most elementary claims to legal, economic and political – not to say social – standing as equal human beings were rejected, often by violence, a lengthy period followed more recently by neglect and indifference. All this in a country that visitors from Europe like Tocqueville portrayed, quite correctly, I think, as displaying a passion for equality stronger than they had ever observed elsewhere. Yet despite the obvious fact that equality has often been denied in practice everywhere, to a quite remarkable extent over the past several centuries many claims to equality, including political equality, have come to be strongly reinforced by institutions, practices and behaviour. Although this monumental historical movement toward equality is in some respects world wide, it has been most conspicuous, perhaps, in democratic countries like Britain, France, the United States, the Scandinavian countries, the Netherlands and many others. In the opening pages of the first volume of Democracy in America Tocqueville pointed to the inexorable increase in the equality of conditions among his French countrymen ‘at intervals of fifty years, beginning with the eleventh century’. Nor was this revolution taking place only in his own country: ‘Whithersoever we turn our eyes,’ he wrote, ‘we shall witness the same con- tinual revolution throughout the whole of Christendom.’ Political Equality in the Coming Century 5

The gradual development of the equality of conditions [he goes on to say] is … a providential fact, and it possesses all the characteristics of a Divine decree: it is universal, it is durable, it constantly eludes all human interference, and all events as well as men contribute to its progress. (Tocqueville 1961, p. lxxxi)

We may wish to grant Tocqueville a certain measure of hyperbole in this passage. We may also want to note that in his second volume several years later he was more troubled by what he viewed as some of the undesirable consequences of democracy and equality (Tocqueville, 1961b). Even so, he did not doubt that a continuing advance of democracy and equality was inevitable. If today we look back to the changes since his time, we, like Tocqueville in his own day, may well be amazed at the extent to which ideas and practices that respect and promote political equality have advanced across so much of the world – and, for that matter, aspects of a broader human equality as well. As to political equality, consider the incredible spread of democratic ideas, institutions and practices during the century just ended. In 1900, 48 countries were fully or moderately independent countries. Of these, only eight pos- sessed all the other basic institutions of representative democracy, and in only one of these, New Zealand, had women gained the right to vote. Furthermore, these eight countries contained no more than 10–12 per cent of the world’s population. At the opening of our present century, among some 190 countries the political institutions and practices of modern representative democracy, including universal suffrage, exist in around 85, at levels comparable to those in Britain, western Europe and the United States. These countries include almost six out of every ten inhabitants of the globe today.4 In Britain, as we all know, the working classes and women were enfranchised, and more. Men and women of middle, lower middle and working-class origins not only gained access to the House of Commons and its facilities but to the Cabinet and even the post of prime minister. And the hereditary peers in the House of Lords have, after all, at last been sent packing – well, most of them. In the United States, too, women were enfranchised; the Civil Rights Act of 1964 protecting the right of African-Americans to vote did in fact become law; the law was actually enforced; and African-Americans have become a signiﬁcant force in American political life. I wish I could say that the miserable condition of so many Native Americans had greatly changed for the better, but that sad legacy of human injustice remains with us. Failures and all, if we simply assume that beliefs about equality are always hopelessly anaemic contestants in the struggle against the powerful forces that generate inequalities, I do not see how we could account for the enormous gains for human equality over the past two centuries. Although no brief 6 Robert A. Dahl summary can do justice to the historical variations and complexities in the process by which changes toward equality came about – and here I have in mind mainly political equality – a summary of the main elements in an explanation would probably run something like this:

• Despite the fervent efforts of elites to promote views intended to give legitimacy to their superior power and status, and their own unquestioning belief in the rightness of their entitlements, many members of subordinate groups doubt that the inferior position they are assigned by their self- proclaimed superiors is really justiﬁed. James Scott (1990) has shown pretty convincingly, I think, that people who have been relegated to subordinate status by history, structure and elite belief systems are much less likely to be taken in by the dominant ideology than members of the upper strata are prone to assume.5 • Next, given the open or concealed rejection of the elite ideology by members of the subordinate groups, a change in conditions, whether in ideas, beliefs, structures, generations or whatever, offers the subordinate groups new opportunities to express their grievances. • Given these new opportunities and moved by anger, resentment, a sense of injustice, a prospect of greater individual or group opportunities, group loyalty or other motives, some members of the subordinate groups begin to press for change by whatever means are available. • Some members of the dominant group begin to support the claims of the subordinate strata. Insiders ally themselves with outsiders. Insiders may do so for a variety of reasons: moral convictions, compassion, opportunism, fear of the consequences of disorder, dangers to property and the legitimacy of the regime, and even the real or imagined possibility of revolution.6 • So a seismic shift occurs: democratization, extension of the franchise, legal protection of basic rights, political competition by leaders of hitherto subordinate groups, election to public ofﬁce, changes in law and policy, and so on.

For example, once the Civil Rights Acts of 1964 were passed and enforced, African-Americans seized their opportunities to vote – and among other things soon tossed out the police ofﬁcials who had violently enforced their subordination. In India the scheduled castes have begun to vote in substantial numbers for leaders and parties drawn from their own strata and committed to reducing discrimination against them.

Is political equality desirable?

So much for history, one might say. But is movement toward greater political equality necessarily a good thing? Is it really a desirable goal? Among others, Political Equality in the Coming Century 7

Tocqueville himself expressed some worries. Nonetheless, if we are prepared to make two assumptions, I believe that the case for political equality and democracy becomes extraordinarily powerful. And each of these assumptions is, in my view, difficult to reject in reasonable and open public discourse. The first is the moral judgement that all human beings are of equal intrinsic worth, that no person is intrinsically superior in worth to another, and that the good or interests of each person must be given equal consideration (Dahl, 1989, pp. 84ff; Benn, 1967, pp. 61–78). Let me call this the assumption of intrinsic equality. Yet even if we accept this moral judgement, the deeply troublesome question immediately arises, who or what group is the best qualified to decide what the good or interests of a person really are? Pretty clearly the answer will vary, depending on the situation, the kinds of decisions and the persons involved. But if we restrict our focus to the government of a state, then it seems to me that the safest and most prudent assumption would run something like this: among adults no persons are so definitely better qualified than others to govern that they should be entrusted with complete and final authority over the government of the state (see Dahl, 1989, pp. 105ff; Dahl, 1998, pp. 74ff). Although we might reasonably add refinements and qualifications to this prudential judgement, it is difficult for me to see how any substantially different proposition could be supported, particularly if we draw on crucial historical cases in which substantial numbers of persons were denied equal citizenship. Does anyone really believe today that when the working classes, women and racial and ethnic minorities were excluded from political participation, their interests were adequately considered and protected by those who were privileged to govern over them? I am not saying that the reasons I have given were in the minds of the persons who brought about greater political equality. I am simply saying that if we ask ourselves why we ought to believe in political equality as a desirable goal or ideal, the reasons I have offered provide a sufficient answer. But what about the well-known objection I mentioned earlier, that like many desirable goals, political equality may conflict with other important goals or values, may indeed do harm to them, and consequently our pursuit of that goal should be tempered by our justifiable desire to attain other goals? I have in mind mainly liberty and fundamental rights. Does political equality conflict with liberty? Like many others, Tocqueville appears to have believed so. But before I turn to his remarks, I cannot forbear adding that I am frequently amazed by an assertion about the supposed conflict between liberty and equality that makes no mention of what would seem to me to be an absolutely essential requirement of any reasonable discussion about the 8 Robert A. Dahl relation between the two. Whenever we talk about liberty, freedom or rights, are we not obliged to answer the question: liberty or rights for whom? Indeed, to discuss equality and liberty with much precision we would need a vocabulary that scarcely exists except by implication. Quite possibly a precise vocabulary would block discussion, but I cannot help thinking that any argument about liberty and equality is literally meaningless unless the terms are carefully specified. Let me illustrate what I have in mind by referring to a grammar of equality that Douglas Rae offered some years ago. Rae proposed that in speaking about equality we must at a minimum specify the subjects of equality, or the set of persons who are to be treated as equal; the domain of equality, or the classes of things that are to be allocated equally; whether we mean absolute versus relative equalities, and so on (Rae, 1981; see also Sen, 1992, pp. 12ff). In much the same way, when we speak of freedom, liberty or rights it seems to me absolutely essential that we go beyond answering the question ‘What liberty or right?’ The answer to that question only specifies the domain of liberty. But we are also obliged to answer the question, ‘Liberty for whom?’ In short, we need distinctions roughly like those that Rae proposes for equality. Amartya Sen seems to me entirely correct when he says:

It is, I believe, arguable that to have any kind of plausibility, ethical reasoning on social matters must involve elementary equal consideration for all at some level that is seen as critical. The absence of such equality would make a theory arbitrarily discriminating and hard to defend … Libertarians [he goes on to say] must think it important that people should have liberty. Given this, questions would immediately arise regarding: who, how much, how distributed, how equal? (Sen, 1992, pp. 17, 22)

Keeping in mind the importance of distinctions like these, let me return to Tocqueville. His view, if I understand him correctly, was roughly this: an equality of condition among a people helps to make democracy possible, perhaps even inevitable. But since the very essence of democratic government, according to Tocqueville, is the absolute sovereignty of the majority, which nothing in democratic states is capable of resisting, a majority necessarily has the power to oppress a minority. Just as a man with absolute power may misuse it, he points out, so may a majority. Given an equality of condition among citizens, we may expect that in democratic countries a wholly new species of oppression will arise. Among citizens all equal and alike, the supreme power, the democratic government, acting in response to the will of the majority, will create a society with a network of small complicated rules, minute and uniform, that none can escape. Ultimately, then, Political Equality in the Coming Century 9 the citizens of a democratic country will be reduced to nothing better than a flock of timid and industrious animals, of which the government is the shepherd.7 If I have fairly summarized Tocqueville, how should we interpret his forecast in the light of subsequent developments? After all, we have the advantage, as he did not, of two centuries of experience with modern democratic institutions. Some readers have interpreted Tocqueville in these passages as foreshadowing mass society, while to others Tocqueville expected that mass democracy would be the seed of twentieth-century authoritarian and totalitarian systems. Yet if we now read these passages of Tocqueville as a forecast of the way in which democratic countries would tend to evolve, I think we are bound to conclude that Tocqueville was just dead wrong. When we examine the course of democratic development over the past two centuries, and particularly over the century just ended, what we find is a pattern of democratic development that stands in total contradiction to such a prediction. We find instead that as democratic institutions become more deeply rooted in a country, so do fundamental political rights, liberties and opportunities. As a democratic government matures in a country, the likelihood that it will give way to an authoritarian regime approaches zero. Democracy can, as we all know, collapse into dictatorship. But breakdowns are extraordinarily rare in mature democracies; they occur instead in countries that encounter times of great crisis and stress when their democratic institutions are relatively new. Crisis appears to be inevitable in the life of every country. Even mature democratic countries have had to face wars, economic depression, large-scale unemployment, terrorism and other challenges. But they did not collapse into authoritarian regimes. In the twentieth century, on something like 70 occasions, democracies have given way to non-democratic regimes. Yet with very few exceptions, these breakdowns have occurred in countries where democratic institutions were very new – less than a generation old. Indeed, the only clear-cut case of a democratic breakdown in a country where democratic institutions had existed for 20 years or more seems to be Uruguay in 1973. In the same year, Chile provided a less clear-cut case because of restrictions on the suffrage that had only recently been lifted. The Weimar Republic had existed less than 14 years before the Nazi takeover. In all three countries the path to collapse bore no relation to the Tocquevillean scenario. Nor is that scenario confirmed by the 21 or so countries in which democratic institutions have now existed continuously for the past half century or more, the older or mature democracies, if you will. In these countries – Iceland, Britain, Norway, France, Switzerland, Australia, New Zealand, the United States, to name a few – have the fundamental rights and liberties of citizens grown steadily narrower or less secure over the past half century? I do 10 Robert A. Dahl not see how an affirmative answer to this question could be seriously maintained. Far from being a threat to fundamental rights and liberties, political equality, as it is expressed through democratic institutions, requires them.8 To see why this is so, let us begin by viewing democracy as, ideally at least, a political system designed for citizens of a state who are willing to treat one another, for political purposes, as political equals. Citizens might view one another as unequal in other respects. Indeed, they almost certainly would. But if they were to assume that all of them possess equal rights to participate fully in making the policies, rules, laws or other decisions that all citizens are expected (or required) to obey, then the government of their state would, ideally, have to satisfy several criteria. Let me list them here without amplification. To be fully democratic a state would have to provide rights, liberties and opportunities for effective participation, voting equality, acquiring adequate understanding of the relevant alternative policies and their likely consequences, and means by which the citizen body could maintain adequate control of the agenda of government policies and decisions. Finally, as we now understand the ideal, in order to be fully democratic a state would have to ensure that all or at any rate most adult permanent residents under its jurisdiction would possess the rights of citizenship. As we all know, the democratic ideal I have just described is too demanding to be fully achieved in the actual world of human society. To achieve it so far as may be possible under the imperfect conditions of the real world, then, certain political institutions for governing the state would be required. Moreover, since the eighteenth century these institutions would have to be suitable for governing a state encompassing a large territory, such as a country. There is no need to describe here the basic political institutions of a modern democratic country, but by now it should be obvious that just as in the ideal so too in actual practice, democratic government presupposes that its citizens possess a body of fundamental rights, liberties and opportunities. These include the rights to vote in the election of officials in free and fair elections; to run for elective office; to free expression; to form and participate in independent political organizations; to have access to independent sources of information; and rights to other freedoms and opportunities that may be necessary to the effective operation of the political institutions of large-scale democracy. It is obvious, then, that both as an ideal and as an actual set of political institutions, democracy is necessarily a system of rights, liberties and opportunities. These are required not merely by definition. They are required in order for a democratic system of government to exist in the real world. If we consider these political rights, liberties and opportunities as in some sense funda- Political Equality in the Coming Century 11 mental, then in theory and practice, democracy does not conflict with liberty. On the contrary, democratic institutions are necessary to the existence of some of our most fundamental rights and opportunities. If these political institutions, including the rights, liberties and opportunities they embody, do not exist in a country, then to that extent the country is not democratic. When they disappear, as they did in Weimar Germany, Uruguay and Chile, then democracy disappears; and when democracy disappears, as it did in these countries, then so do these fundamental rights, liberties and opportunities. Likewise, when democracy reappeared in these countries, so, necessarily, did these fundamental rights, liberties and opportunities. The connection, then, is not in any sense accidental. It is inherent. The links between political equality, democracy and fundamental rights, liberties and opportunities run even deeper. If a country is to maintain its democratic institutions through its inevitable crises, it will need a body of norms, beliefs and habits that provide support for the institutions in good times and bad – a democratic culture transmitted from one generation to the next. But a democratic culture is unlikely to be sharply bounded. A democratic culture will not only support the fundamental rights, liberties and opportunities that democratic institutions require. People who share a democratic culture will, I think inevitably, also endorse and support an even larger sphere of rights, liberties and opportunities. Surely the history of recent centuries demonstrates that it is precisely in democratic countries that liberties thrive.

Are other feasible processes more desirable?

Yet is it not possible that when our descendants look back at the close of this new century they may view the connection between democracy and fundamental rights and liberties as less robust than I have suggested? To look forward, let me first look backward, to Tocqueville’s belief that an increasing equality of condition was an inexorable force in France and elsewhere, and his heavy emphasis on the equality of condition he thought existed among Americans – among white male Americans, at any rate. Whatever we may think of the accuracy of his observations in his own time, it is altogether possible that during the coming century citizens in existing democratic countries may become more unequal in their capacities to influence the conduct of their government – or governments. This outcome may result, at least in part, from the increased adoption of alternative processes at the expense of democracy and political equality. Democracy is not the only process for making collective decisions that we accept as legitimate. To oversimplify a vastly complicated reality, depending on our needs, aims and circumstances we all accept three major alternatives as legitimate. These are hierarchy, or decision-making by leaders; bargaining, or 12 Robert A. Dahl decision-making among leaders; and markets, or decision-making by and among leaders and consumers (see Dahl and Lindblom, 1953). The relations between democracy and these legitimate alternatives are often difficult and changing, and each system threatens the place of the others. I have no doubt that under certain appropriate conditions the three main alternatives to making collective decisions by democratic means can be justified, though not necessarily in their commonly prevailing forms. One can readily imagine ways in which the expansion of one or more of these alternatives might erode support for political equality as a desirable and feasible goal during this new century. One such development is the triumph, all but world wide, of market economies in which goods and services are produced mainly in privately owned and managed firms, that is predominantly capitalist market economies. By the end of the twentieth century, the major structural alternatives to market capitalism proposed by its critics had largely been vanquished or marginalized. I do not mean highly centralized and grossly inefficient economic orders like those of the Soviet Union and Maoist China. What is more germane here, visions of alternatives that many advocates believed would steadily reinforce political equality and democracy had also been effectively rejected. One such vision was American; the other was European. The American vision of agrarian republicanism portrayed by Thomas Jefferson and others foresaw a society consisting predominantly of citizens – white male citizens, to be sure – who were free and independent property-owning farmers, substantially equal socially, economically and politically. Alas for the Jeffersonian vision, the American agrarian republic had no future. Before the end of the nineteenth century the society of free farmers, such as it was, had been largely superseded by a triumphant commercial, financial and industrial capitalism. Meanwhile, in Britain and Europe the inequalities and injustices of nineteenth-century market capitalism had produced a radically different vision: a democratic socialist society with an economy in which goods and services would be produced in enterprises that were to be collectively owned in some fashion. In some fashion, too, the enterprises would be democratically controlled. Finally, the enterprises would somehow share their surplus revenues in ways that would sustain a fair equality of condition among all citizens. Yet just as the Jeffersonian vision had become essentially irrelevant by the end of the nineteenth-century, so too by the end of the twentieth century the idea of a democratic socialist economy had largely lost its content and support. Although that vision may linger on as a vague ultimate end, the end now seems to stand completely isolated from any feasible means of achieving it. The relation between democracy and market capitalism is far too complex for me to describe and appraise here. Perhaps it is enough to say that in my view democracy and market capitalism help one another in some ways and in Political Equality in the Coming Century 13 some ways each inflicts harm on the other. One particular aspect of that complex relation is worth singling out because it bears heavily on prospects for political equality. Market capitalism inevitably generates inequalities among citizens in their access to basic resources of many kinds, from income and wealth to status, information, knowledge, control, access to leaders and others. These unequally distributed resources are all readily convertible into political resources, that is resources for influencing governments and political life. Thus inequalities in political resources are convertible into inequalities in power and influence. Although we have no satisfactory way of measuring the conversion rates from resources to power and influence, surely we must agree that significant inequalities in political resources produce important inequalities in power and influence; and important inequalities in power and influence impose serious barriers to political equality among the citizens of a democratic country. The limits to political equality set by the unequal distribution of resources are old and almost too familiar to us. Yet this old problem, which measures taken during the last century had somewhat mitigated, has recently been intensified in many advanced democratic countries by an increased inequality in the distribution of incomes. If we assume, as I do, that for the foreseeable future an economic order of market capitalism will prevail in all democratic countries, then this obstacle to political equality left over from the last century will remain with us in the new century. Indeed, the problem may be even more acute, and not only because of the recent widening of the gap between the best and worst off. We also face the possibility that in contrast to moves toward political equality undertaken directly by governments, such as the expansion of the suffrage, previous shifts toward economic equality may now be more easily reversible by the seemingly non-political actions of markets and corporate hierarchies.9 The coming century also presents a newer obstacle to achieving a high degree of political equality among citizens in advanced democratic countries. Or rather, an old problem is now greatly magnified by a new setting. From its very beginnings democracy has been troubled both in theory and practice by a perplexing dilemma: the dilemma of system effectiveness versus citizen participation.10 In general, a smaller democratic political unit might provide greater opportunities for its citizens to participate in political life in ways that would influence the conduct of their government, yet in the smaller unit important matters may lie beyond the capacity of the government to deal with effectively. Conversely, the government of a larger democratic unit may be able to deal more effectively with certain matters of importance to its citizens, but because of its size, the citizens of the larger unit will have fewer opportunities to participate effectively in political life. This dilemma has been present, if not always acknowledged, from the Greek city-states down to our present century. 14 Robert A. Dahl

Increasing interdependence requires larger units in order for governments to possess the capacity to deal effectively with many problems that are important to the lives, liberties and well-being of their citizens. During the twentieth century interdependence increased as our world steadily shrank. Surely the trend toward interdependence will continue indefinitely. It is obvious to all of us that as our world shrinks, we are more and more affected by actors outside our national boundaries: whether individuals, groups, corporations, markets, countries or international organizations. To avoid the costs and gain the benefits of our interdependence, we need transnational organizations.11 Although many of these organizations make decisions that touch directly or indirectly on our lives, none is to any significant degree democratic. Their decisions are arrived at instead by hierarchy and bargaining among political and bureaucratic elites that operate with a very high degree of autonomy within broad limits set by charters, treaties or other international agreements. Can international organizations be made democratic? Some observers believe so, or at least hope so. For my part, I am sceptical (Dahl, 1999). I believe that their decisions will continue to be made largely by systems of hierarchy and bargaining among highly autonomous political and bureaucratic elites.

Can we ﬁnd solutions?

If so, then, in this new century political equality faces formidable challenges as an idea, an ideal and a relevant and feasible goal. To those who believe that political equality is neither desirable nor feasible the absence of solutions might seem to pose no challenge. But anyone who rejects political equality as a goal must face an even more formidable task: to formulate a superior alternative, to demonstrate how that alternative may reasonably be justiﬁed, and to show how it may feasibly be achieved. It goes without saying that I remain extremely doubtful about the success of any programme along these lines. Yet those of us who believe that political equality is a desirable goal need to search for ways of meeting the challenges I have suggested, if only to prevent a regression toward greater inequality. We may well ﬁnd that it takes all the running you can do, as the Red Queen admonished Alice, just to keep in the same place. Fortunately, some scholars have already taken up the search for ways by which democracy and political equality might be strengthened. Let me close by mentioning a few of the questions they are attempting to answer.

• How can national governments can be made more responsive to their citizens?12 Political Equality in the Coming Century 15

• How can international organizations be rendered more accountable to the people of the countries they represent?13 • How might ordinary citizens engage in more effective methods of deliberation on major issues of public policy?14 • How might democratic values be extended in appropriate ways to important associations other than the state?15 Could economic enterprises, for example, be governed more democratically?16 • How might the Internet serve democratic values?17

Finally, I want to pose a question that, I believe, needs urgent attention in the older democratic countries:

• How can immigrants from highly diverse cultures be integrated into the citizen body in ways that respect basic human rights and do not lead to deep cultural cleavages that produce highly destructive cultural conﬂicts?18

These are just a small selection of the important questions that we need to explore. Among the many scholars and teachers throughout the world who pursue this ancient and noble calling, the study of politics, I hope that many more will undertake a search for solutions to the problem of achieving a satisfactory degree of political equality in the twenty-ﬁrst century.

Notes

1. This chapter was given as a plenary address to the Political Studies Association annual conference, London School of Economics, 12 April 2000. 2. In 1790, when the ﬁrst census was taken, in a total population of 3.9 million, Negroes numbered 757 000 of whom 698 000 were slaves. In the nine northern states from Maine to Pennsylvania, in a total population of just under 2 million, Negroes numbered 67 000, of whom 40 000 were slaves. Historical Statistics of the United States, Colonial Times to 1957 (Washington, DC, 1960), Series A123–180, pp. 12–13, and Series A59070, p. 9, fn. 2. 3. With the exception of several of his surviving children by his mistress, a slave, Sally Hemings and her half-brothers. Although the issue of paternity is disputed, Annette Gordon-Reed (1997) provides strong circumstantial evidence that Thomas Jefferson fathered Hemings’ children. For her ‘Summary of the Evidence’, see pp. 210ff and see also Appendix B, ‘The Memoirs of Madison Hemings’, pp. 245ff. DNA tests provide additional circumstantial, though not conclusive, evidence (see Smith and Wade, 1998). 4. I have drawn these estimates from Karatnycky (2000), Dahl (1989), Table 17.2, p. 240, and Vanhanen (1984), Table 22, p. 120. 5. As one example (Scott, 1990, p. 117) writes, that ‘among the untouchables of India there is persuasive evidence that the Hindu doctrines that would legitimize caste domination are negated, reinterpreted, or ignored. Scheduled castes are much less likely than Brahmins to believe that the doctrine of karma explains their present 16 Robert A. Dahl

condition; instead they attribute their status to their poverty and to an original, mythical act of injustice.’ 6. The late Joseph Hamburger showed that to secure the expansion of the suffrage (and ultimately the passage of the Reform Act of 1832), James Mill, though opposed to violence as a means, deliberately set out to create a fear of revolution among members of the oligarchy. ‘Since Mill wished to achieve fundamental reforms without violence, it became necessary to devise means by which an oligarchy would be led to grant concessions out of self-interest … [T]here were only two alternatives: “[The people] can only obtain any considerable amelioration in their government by resistance, by applying physical force to their rulers, or at least, by threats so likely to be followed by performance, as may frighten their rulers into compliance.” Since the use of physical force was to be avoided, Mill built his hopes on the second alternative … Mill was proposing that revolution be threatened. He assumed that the threat would be sufficient and that it would not be necessary to carry it out’ (Hamburger, 1963, pp. 23–4). 7. These sentences are a close paraphrase of his statements in Tocqueville (1961a), pp. 298 and 304, and Tocqueville (1961a), pp. 380–1. 8. The following section is adapted from Dahl (1999a). 9. I am indebted to Ian Shapiro for suggesting this possibility to me. 10. I draw here from Dahl (1994). 11. A recent count shows these to number at least 82, ranging alphabetically from the Association of Southeast Asian Nations to the World Bank, World Health Organization, World Trade Organization and others. The estimate is from the University of Michigan website on international organizations. 12. Budge (1996) explores the possibilities of referendums in national decisions. 13. See Schmitter (forthcoming). 14. James Fishkin (1997) argues for the use of ‘deliberative polls’ composed of panels of randomly selected citizens to discuss and deliberate on a major public issue. He has employed them in both Britain and the United States. 15. Shapiro (1999, p. 37) argues that ‘everyone affected by the operation of a particular domain of civil society should be presumed to have a say in governance’. Though hierarchies may be necessary and desirable, ‘there are good grounds for prima facie suspicion of them’. He goes on to explore the possibilities of ‘democratic justice’ in major parts of the human life cycle: family, education, work, and ‘life’s ending’. 16. Cheney (1999) analyses the operation of the renowned Mondragon cooperatives since their founding in 1941 and their recent adaptation to international politics. I explore the issues concerning the application of democracy to firms in Dahl (1985). 17. One highly promising possibility is a computer program that would apply parliamentary rules of order to ‘meetings’ taking place via the Internet. Such a program is under development at a major firm of communications technology (private communication). 18. I raised this question in a paper presented at the 13th Korean Association for International Studies conferences in 1994 (Dahl, 1995), revised as Dahl (1997). Brian Barry (2000) critically examines ‘multiculturalism’ as a solution.

References

Barry, Brian (2000) Culture and Equality: An Egalitarian Critique of Multiculturalism (Cambridge, MA: Harvard University Press). Political Equality in the Coming Century 17

Benn, Stanley I. (1967) ‘Egalitarianism and the Equal Consideration of Interests’, in J. R. Pennock and J. W. Chapman (eds), Equality (Nomos IX) (New York: Atherton Press). Budge, Ian (1996) The New Challenge of Direct Democracy (Cambridge: Polity Press). Cheney, George (1999) Values at Work: Employee Participation Meets Market Pressure (Ithaca, NY: Cornell University Press). Dahl, Robert A. (1985) A Preface to Economic Democracy (New Haven, CT: Yale University Press). Dahl, Robert A. (1989) Democracy and Its Critics (New Haven, CT: Yale University Press). Dahl, Robert A. (1994) ‘A Democratic Dilemma: System Effectiveness Versus Citizen Participation’, Political Science Quarterly, vol. 109, no. 1, pp. 23–4. Dahl, Robert A. (1995) ‘From Immigrants to Citizens: A New Challenge to Democracies’, in Sun Chul Yung (ed.), Democracy and Communism (Seoul: KAIS). Dahl, Robert A. (1997) ‘From Immigrants to Citizens: A New Challenge to Democracies’, in Robert A. Dahl, Toward Democracy: A Journey. Reﬂections 1940–1997, vol. 1 (Berkeley: University of California). Dahl, Robert A. (1998) On Democracy (New Haven, CT: Yale University Press). Dahl, Robert A. (1999a) The Past and Future of Democracy, Occasional Paper No. 5, Centre for the Study of Political Change, University of Siena. Dahl, Robert A. (1999b) ‘Can International Organizations Be Democratic? A Skeptic’s View’, in Ian Shapiro and Casiano Gordon-Hacker (eds), Democracy’s Edges (Cambridge: Cambridge University Press), pp. 19–35. Dahl, Robert A. and Charles E. Lindblom (1953) Politics, Economics, and Welfare (New York: Harper H. Hamilton, 1953). Fishkin, James (1997) The Voice of the People: Public Opinion and Democracy (New Haven, CT: Yale University Press). Gordon-Reed, Annette (1997) Thomas Jefferson and Sally Hemings, An American Controversy (Charlottesville and London: University of Virginia Press). Hamburger, Joseph (1963) James Mill and the Art of Revolution (New Haven, CT: Yale University Press). Karatnycky, Adrian (2000) ‘The 1999 Freedom House Survey: A Century of Progress’, Journal of Democracy, vol. 11, pp. 187–200. Rae, Douglas (1981) Equalities (Cambridge, MA: Harvard University Press). Schmitter, Philippe (forthcoming) How to Democratize the European Union and Why Bother? Scott, James (1990) Domination and the Arts of Resistance (New Haven, CT and London: Yale University Press). Sen, Amartya (1992) Inequality Reexamined (Cambridge, MA: Harvard University Press). Shapiro, Ian (1999) Democratic Justice (New Haven, CT: Yale University Press). Smith, Dinitia and Nicholas Wade (1998) ‘DNA Test Finds Evidence of Jefferson Child by Slave’, New York Times, 1 November. Tocqueville, Alexis de (1961a) Democracy in America, Vol. I, trans. Henry Reeve (New York: Schocken Books). Tocqueville, Alexis de (1961b) Democracy in America, Vol. II, trans. Henry Reeve (New York: Schocken Books). Vanhanen, Tatu (1984) The Emergence of Democracy, A Comparative Study of 119 States, 1850–1879 (Helsinki: Finnish Academy of Sciences and Letters). 2 Levelling Down Jonathan Wolff 1

Egalitarianism, we were once told, is the ‘politics of envy’. It is better, so egalitarians were alleged to believe, to make everyone equal than to allow inequalities, even if some or all would be better off. Thus egalitarians were said to favour ‘levelling down’, jealously refusing to allow, or even undoing, Pareto improvements over equality: improvements that made at least one person better off without making anyone else worse off. In the worst case, it is said, egalitarians will recommend dragging everyone down to the same level, even if everyone is worse off than they would have been in an unequal society. This accusation now seems rather dated. Post Rawls (1971), Frankfurt (1987) and Parfit (1998) most of those who call themselves egalitarians would say that they too would oppose levelling down, all things considered, preferring some form of sufficiency or prioritarian position. Some theorists put things this way: although justice requires equality, considerations of Pareto efficiency always trump considerations of justice. Others set out their view another way: the Pareto principle is itself a principle of justice, and so levelling down can never be required by justice. I wonder, though, whether it is time to re-examine this knee-jerk antipathy to levelling down. Can it never be right to level down? In the abstract levelling down sounds mean-spirited or wasteful, or both, but when we look at examples need this always be the case?

The ‘real-Pareto’ maxim

We should note, from the outset, that if a theorist favours some redistribution, then some forms of levelling down, in some circumstances, may be inevitable. To explain, once we have chosen our ‘currency’ of justice, legitimate redistribution in that currency may require levelling down in some other currency. Suppose, for example, we want to maximise the preference satisfaction of the worst off. Doing this may require us to move to a form of

K. Dowding et al. (eds.), Challenges to Democracy © Political Studies Association 2001 Levelling Down 19 society in which everyone, including the worst off, has fewer material resources. For example, the worst off may now get better use of, and thus more preference satisfaction from, their smaller bundle of resources because of reduced overcrowding effects. Preferring a lower total stock of material goods may in one way seem inefficient or wasteful but this is irrelevant. The point to note is that a change that is Pareto efficient in terms of one currency may commonly be Pareto inefficient in terms of another. Once this point is recognized we see that the accusation that a proposal will be Pareto inefficient may, in itself, be a very weak objection; it all depends on its efficiency in the correct currency. Few, I think, will have reason to object to anything I have said so far. But many will be strongly attracted to the following line of thought: although levelling down in ‘irrelevant’ currencies is sometimes acceptable, levelling down in the correct currency never is. Once we know what the real currency of justice is, there is never a sufficient reason for levelling down. Call this the real-Pareto maxim. Is the real-Pareto maxim correct? Its attraction is based, I think, on a type of moral individualism. The guiding thought is that the only things in the world that are good or bad are good or bad for particular or individual human beings (or other members of the moral community). Consequently to show that something is bad one must show that it is bad for a given individual. If we cannot point to an individual who is worse off under situation x than they would be under situation y then x cannot be worse than y.2 If you agree with this apparently appealing thought then you accept the real-Pareto maxim. Is Pareto efficiency required by justice or by efficiency? A view associated with Cohen (1989, p. 911) is that, properly speaking, justice requires strict equality (equal access to advantage) but where justice leads to Pareto inefficiency, it is wrong to insist on justice. So on a strong statement of this view, there are two values in play – justice and efficiency – and efficiency (at least real-Pareto efficiency) trumps justice. The socially best outcome may, in some respects, be unjust. The alternative view is that the Pareto efficiency is a principle of justice, and so there is no outweighing of different values, just a more complex view of justice. This alternative view is highly implausible, at least if taken in its full gener- ality.3 Suppose that by whatever is our preferred currency we have achieved equality. Suppose now that by some economic freak we have two choices: we can either stay as we are, or we can rearrange things so that one particular person, Harry, now has ten times as much as anyone else, although no one else’s share has changed, even in the correct currency. If we refuse to permit this change we might be accused of being mean-spirited, wasteful or narrow- minded, but it does not ring true to say that we have been unjust to Harry (assuming, of course, he has no other special claim for the resources), except 20 Jonathan Wolff perhaps in the most extended sense of justice, where to act unjustly is simply to do something wrong. Now if this example is accepted it does not show that Pareto improvements are never required by justice. All it shows is that some Pareto improvements are not required by justice. Thus this on its own does not establish that the real-Pareto maxim is purely a maxim of efficiency. However, I think we have this much: if we wish to insist that levelling down in the correct currency is always wrong, then we have conceded that in a certain range of cases, however small, requirements of efficiency trump requirements of justice where they conflict. But should we accept this? The view we are considering, then, is that where we can make at least one person better off in the relevant sense without making anyone worse off in that sense then we should always do so. As we have seen, in broad terms the explanation for this is that in a given range of cases efficiency always trumps equality. However, it is worth noting that the explanation can be elaborated – i.e. the relevant range of cases can be specified – in importantly different ways:

• The absolute view: efficiency always trumps equality in cases where efficiency gains for some do not lead to losses for anyone. (These are the real-Pareto cases.) However, we must never let the position of the worst off fall. • The relative view: efficiency always trumps equality in the real-Pareto cases. However, we can also allow losses, even to the worst off, if the losses are relatively slight when compared to the gains that can be made elsewhere. (The small-loss/great-gain cases.)

The absolute view, then, says that the only efficiency gains that are justifiable are Pareto improvements. The relative view – one version of which is Parfit’s prioritarianism – adds that some utility gains are also acceptable, even if they are not Pareto improvements. Although the relative view might be defended on intuitive grounds, it does seem that there are also theoretical reasons that make the absolute view the harder to defend. The absolute view appears to make at least two assumptions:

• Justice is so important that we should never make the worst off even worse off (however trivially) for the sake of others.4 • Efﬁciency is so important that whenever we have the chance we should make someone better off, provided it does not make anyone (or at least the worst off) worse off.

In other words, in one range of cases justice trumps efficiency (when the efficiency gains are partly at the expense of the worst off), while in another Levelling Down 21 range of cases efficiency trumps justice (when the efficiency gains are not at the expense of the worst off). Now there is a certain mathematical elegance about this, and I would not claim that there is any inconsistency in holding this combination of views. The difficulty is at a different level. What plausible argument could justify this combination? What good reason can be given for allowing the trumping to ‘switch’ on a hair-trigger in this way? Consider again the case of Harry. We are required to make him ten times better off than others, provided no one else is thereby made worse off. But if a single person loses a fraction of a per cent of what they already have then this change becomes impermissible. Now there is no difficulty is showing that the combination of theses set out above has this consequence. The difficulty is explaining what reason there can be for holding the combination of theses that gives such importance to such a boundary. Why should we pay such overwhelming attention to the actual current level of the worst off? Thus we can see the relative view as tacitly responding to this line of rhetor- ical questioning. Once the importance of efficiency is acknowledged, the claim that there is any range of cases where justice trumps efficiency is dropped. Now it is important not to misunderstand this: we need to be clear about the distinction between ‘trumping’ and ‘beating’. As I understand the distinction, value x trumps value y iff in any case of conflict value x always beats value y and, on a case-by-case basis, no further justification need be given to explain why. In cases of ordinary beating, one value beats another because in that particular case the combination of the weight and extent of one value is greater than the combination of the weight and extent of another. In such cases we have to go through the argument and comparison before we can reach a resolution. In trumping cases such a procedure is unnecessary. In consequence, to say that justice never trumps efficiency does not entail that efficiency always wins. In fact it is consistent with the outcome that in the cases under consideration efficiency never wins. The point is that we need to look at the details of each case to come to a resolution. Thus on the relative view, there is a range of cases where justice generally wins. But there is no range where justice trumps efficiency. The absolute view, as we have seen, assumes that there is a range of cases where efficiency trumps justice, and another range of cases where justice trumps efficiency. This leads to what I called hair-trigger switching of trumping, which seems hard to justify. So in apparent recognition of this difficulty the relative view abandons the claim that there is a range of cases where justice trumps efficiency. However, we now have something else in need of explanation: why was that the correct abandonment? Why not, instead, abandon the claim that efficiency trumps justice? Or, as I would prefer, why not give up both claims, and thus look at the issues on a case-by-case basis? 22 Jonathan Wolff

Diagnostically, we might suggest that those who assert the real-Pareto maxim, including holders of the relative view, have been so concerned about the levelling down objection that they have wanted to set out a view where levelling down is, in principle, never acceptable.5 This seems to me an over- reaction. Suppose we could come up with a view in which levelling down was acceptable only in certain very special cases, and in such cases we can be clear that, first, the circumstances do indeed justify levelling down and, second, they are sufficiently special that there is no reason to believe that this reasoning will spread to other types of cases. Wouldn’t that be a sufficient reply to the levelling-down objection? The most likely response to this alternative proposal is that we simply will not find any such cases. If so, first, even those of egalitarian sympathies can agree with their critics that levelling down is never justified, and, second, it may be that the best explanation of such a generalization is that (real-Pareto) efficiency trumps justice. The only way of settling this is to look at some apparent cases of levelling down. But first it is worth recalling a point made early on in this paper. Whether or not a change is a Pareto improvement is always relative to currency. Therefore what may be a levelling down in one currency may be a Pareto improvement in another. From this it follows that from a given example in which there appears to be a levelling down we cannot conclude that it is a genuine case. For when the right currency is specified it might turn out that no levelling down was involved after all. Indeed this makes the discussion methodologically complex. Suppose I present an example in which it looks as if levelling down is legitimate. We might take this as proving my case that efficiency does not trump equality. But we could just as easily take it to show that we have not yet put our finger on the correct currency of justice. Thus there are two ways of taking any apparent example of levelling down: first, as a counter-example to the real- Pareto maxim; second, as evidence that we still have not yet achieved clarity on the question of the currency of justice. Some will feel that there is something ad hoc, perhaps question-begging, about the second strategy. After all, it will render the real-Pareto maxim close to true by definition, or at least treat it as having axiomatic status. Cutting off debate in this way seems both unphilosophical and premature. But in response, it seems that no one has yet provided a definitive statement of the ultimate currency of justice (hereafter ‘real well-being’ which I shall simply use as a placeholder6) – and so we need any help we can get. If a discussion of apparent levelling down can yield insight into the nature of real well-being, that is a valuable enough result in itself. If the price of this is to treat the real- Pareto maxim as an axiom, that need not be too high a price to pay. The methodology, then, would be to continually refine our understanding of real Levelling Down 23 well-being until apparent counter-examples to the real-Pareto maxim disappear. And, indeed, this strategy is far from risk free. For there are other constraints on an acceptable account of real well-being, and the question then would be whether they can consistently be met together with the real-Pareto constraint. In particular the following two seem plausible:

• It must be ﬁnitely statable. • It must be such that it is possible to provide at least a partial order of well- being so that we can identify the worst off.

If, after all our efforts, we cannot come up with an account of real well- being which satisfies these constraints and can be made consistent with our intuitions about apparently acceptable levelling down, then the most plausible strategy must be to give up this approach, and to admit that the real- Pareto maxim should be abandoned. But we are, of course, a long way from that point yet, and we will not conclusively reach it in this paper, for I will not attempt to state an account of real well-being that satisfies the real-Pareto maxim. But I will attempt to identify some difficulties that stand in the way.

Richard Norman on social equality

Let us begin our study of cases by considering the following argument from Richard Norman, who, like me, suspects that the levelling down objection has been overplayed. Norman argues for what he calls Socially-Located Egalitarianism (SE) which he deﬁnes thus:

Equality is a socially-located value, a conception of social justice i.e. egalitarians should object to inequalities of well-being between people in the same community. (Norman, 1998, p. 38) It is worth quoting the argument at length:

We can imagine circumstances in which, from the standpoint of SE, equality at a lower level of well-being might be seen as preferable to inequality at a higher level of well-being for everyone. Imagine an egalitarian community at a fairly low level of economic development whose members, though not experiencing great hardship or absolute poverty, have a simple life style. Given the opportunity of economic development which would make them all better off but introduce substantial inequalities, they might prefer to remain less prosperous but equal. I am not thinking here of the typical attendant evils of industrialisation such as crime and social conﬂict and 24 Jonathan Wolff

environmental pollution which would enable us to explain their choice by saying that they would not really be better off. I am supposing that they would acknowledge that they would be better off with economic development, but they still prefer equality. It is not, as it might appear . . ., a crazed obsession with uniformity and symmetry and neatness. It is a preference for certain kinds of social relations. They may fear that, with greater inequality, they will become more distanced from one another, their society will become less cooperative, the more prosperous among them will become disdainful and supercilious and the less prosperous will become either more servile or more resentful, and they will no longer be united by shared experience and a shared condition. (Norman, 1998, p. 51)

This initially plausible account of apparently justified levelling down allows us to bring out several issues. First, it is worth noting that Norman is clearly aware of a point mentioned above: levelling down may be more apparent than real. If inequality brought crime and pollution, perhaps this would make the worst off worse off still. They could, then, object to the inequality on the grounds that it makes them worse off. This could not be construed as ‘real levelling down’, a violation of the real-Pareto maxim. To avoid this response, Norman stipulates that the example involves no such loss. Nevertheless, his opponents will claim that Norman has not done enough to register the importance of this kind of point. We must note that several ways of assessing levels of well-being seem implicit here. Economic development makes everyone better off, it is said. Certainly there will be improvements in standard of living: cars and washing machines replace bicycles and hard, unfulfilling toil. In material terms, then, inequality improves the lot of everyone: they have more resources and their standard of living is higher. Nevertheless, SE is defined in terms of well-being, not material resources or standard of living. That there may be occasions when we must level down in material resources in order to boost the well-being of the worst off is only to be expected. Arguably it might also be the case that we may need to level down in terms of standard of living in order to increase the well-being of the worst off: a car may be insufficient consolation for alienation from one’s fellow human beings. Clearly Norman thinks his example is a significant one, and introduces it as one where the equal society involves a lower level of well-being for everyone. Nevertheless the people ‘prefer to remain less prosperous but equal’. There are various complexities here. First the language has shifted somewhat. Is ‘prosperity’ a welfare term, a standard-of-living term or material resources term? If either of the latter two, then, once more the real-Pareto maxim is not challenged. So we must read ‘less prosperous’ as ‘having a lower level of well- Levelling Down 25 being’. But now we must ask, who prefers this? We are told that ‘they’ do, where this is most naturally read as whoever it is who is empowered to speak for the society as a whole. Should we then read this as ‘everyone’? Certainly as we read on, we can be persuaded that neither future-rich nor future-poor could much like the prospect of anticipating what they might become: who wants to become disdainful and supercilious or servile and resentful? But here the same strategy of defence bites again: the real-Pareto theorist will say that what we are agreeing to when we agree with Norman is that these gains in well-being from equality are more apparent than real. When we weigh an easier life but distant social relations against honest toil in equality we realize we are better off as we are. So, the real-Pareto theorist will argue, there is no levelling down after all. We stick with equality for fear of making people worse off in what really counts. In conclusion, then, this example does not show that real levelling down can be acceptable. Can a more convincing illus- tration be found?

Race and the swimming pool

You are the mayor of a small town in the Southern states of the US. Your town has a swimming pool which is open to all. Against your opposition, your state Senate passes a new law: swimming pools must be racially segregated. If there is only one pool, then it must be made available to whites only. As mayor, you do not have funds to build another swimming pool, but in any case object to racial segregation. However, if you try to disobey the new law, you will be removed from office and replaced with a state official. But rather than allowing a whites-only swimming pool you decide to shut it down completely. Your opponents then accuse you of levelling down. So two related questions arise. First, is this really a case of levelling down in any interesting sense? Second, if so, is it justified? But let me be dogmatic. I will assume in this case that it is justified. So if it is a case of levelling down it is a case of justified levelling down. There are some fairly obvious things that can be said, immediately, to make it appear that, although this is a levelling down in short-term access to swimming pools, it is not a real levelling down, that is a violation of the real-Pareto maxim. Presumably the most promising way of making out such a case would be to argue that there is a sense in which blacks are better off with the swimming pool closed than open only to whites. And no doubt there are ways of expanding on the details to make this clearly so. The opposing view is that at least on one expansion of the case things are better (the state of affairs is better justified) with the swimming pool closed even though the blacks are not better off. But is there such an expansion? Let us first remove some distractions. 26 Jonathan Wolff

Strategic posturing It could be suggested that this policy is not so much a case of levelling down as an attempt to put pressure on the authorities – the state – to repeal their new law. If the state wants to improve conditions for whites and sees that this is now tied together with improved conditions for blacks, then they may have no alternative but to repeal the law. However, it is perfectly possible that this is simply the latest move in an incremental policy of state racism, and no such tactics, when attempted in the past, have paid off. Thus as mayor you have no reason to believe that the authorities will be swayed to any degree by anything you do.

Reinforcement of inequality In some cases a policy of segregation or exclusion can have accumulating effects. Consider a debate that sometimes takes place about whether a golf club can properly exclude Jews, or blacks, or women. Sometimes it is said that a private club can set its own rules: if people want to mix only with a certain type of person than this is up to them. Many replies to this are often made, but one prominent response is that a golf club, typically, is not simply a place where people go to play golf, but also to make social and business contacts. Thus anyone who is a member of the club is further advantaged by access to a level of opportunity that is not readily available to non-members. So non-members may see their well-being decline both relative to the members and in absolute terms as the members consolidate their social and business advantages. In this way, then, non-members are not only excluded from golf but lose further in life’s competitive struggle. Consequently closing the golf club would make non-members better off in at least one signiﬁcant respect. In general this is an important argument. But we can, I think, stipulate that it does not apply to the case of the swimming pool. At the swimming pool people just swim. No deals are struck, there is no bar or cafe in which people enjoy a rich social life and consequent opportunities. So these longer-term accumulating effects simply do not apply. There is no reason to believe that this non-competitive inequality will lead to a competitive inequality – an inequality in a ﬁxed supply of goods – in which those who do badly are thereby made worse off than they would be under conditions of equality.

Moral virtue and solidarity Suppose that the white users of the pool were extremely sympathetic to the plight of the blacks, and felt that if the blacks were excluded they, the whites, would not want to use the pool anyway to show solidarity, or at least to avoid the moral taint of taking advantage of an unfair situation. Hence in one sense closing the swimming pool does not harm anyone. The whites would feel Levelling Down 27 better if it were closed, and are better off in terms of what really counts – virtue – and the blacks are better off for this expression of solidarity. On this reading closing the pool is very far from an example of levelling down: indeed it may be a strong Pareto improvement! In reply it can be conceded that we can imagine such preferences, but that the case for closing the swimming pool is not premised upon them. Consider the case where the whites greatly resent the closing of the pool and set up a protest against it. The blacks do not beneﬁt from increased solidarity because none is shown. If, even in this case, it is justiﬁed to close the swimming pool, then considerations of virtue and solidarity are not the prime justifying factor.

Fairness as well-being Nevertheless, it may be replied that even if the whites do not show solidarity, there is still at least one sense on which the blacks are better off for the closing of the pool. For that situation is fair and the previous situation was unfair. Being treated unfairly is a way of being made worse off. This is a claim, though, that can come in importantly different versions, and of different degrees of plausibility, and it is necessary to make some distinctions. Consider the following claims:

(a) The fact that a situation is unfair to you makes you worse off. (b) The fact that you correctly believe a situation to be unfair to you makes you worse off. (c) The fact that you believe a situation (whether correctly or incorrectly) to be unfair to you makes you worse off. (d) A combination of your belief that a situation is unfair to you, and your feeling resentment or envy or anger at this unfairness, makes you worse off. (e) Envy or resentment of others who have more, even though this is not believed to be unfair, makes you worse off.

Let us take these in order. As to the ﬁrst, although the claim has some plausibility there seem to be theoretical reasons to avoid it. For it is natural to think that the fairness of a situation is at least partly determined by the well-being of the people in that situation. If, then, the fairness of the situation partly determines well-being levels then, for many cases, indeterminacy threatens. So for reasons of conceptual clarity it is sensible to deny (a), which means devising a measure of well-being which does not allow that the mere fact that a situation is unfair to you makes you worse off. It might be thought, though, that when we add the correct belief that the situation is unfair the case for claiming that there is an impact on well-being becomes stronger. For here we have something undeniably internal to the 28 Jonathan Wolff agent. Yet in reply, it should be said that if the belief is correct then taking it into account is a type of double-counting: unfairness must already be taken into account, and so what does the belief in unfairness add? If the belief is incorrect it seems quite bizarre to take it into account at all (cf Dworkin, 1981, pp. 198–201). And aside from moral concerns, indeterminacy threatens again. This may not seem enough to head off this challenge. After all, a belief that a situation is unfair to you, whether a correct belief or not, can eat away at your life and generate misery of the most literal sort. However, if this is true we have moved to a different claim: the source of the well-being loss is the belief plus the further effect this belief has on one’s mental life. So we should now consider the fourth claim above, having dismissed the ﬁrst three. On a natural theory of well-being, suffering from envy and resentment are clear forms of lack of well-being. Thus a theory of well-being that excludes them must be theoretically, rather than analytically, motivated in that respect.7 But before continuing with this line of thought it is worth reminding ourselves of the place in the argument. The question we are addressing is that of whether closing the swimming pool makes the blacks better off. The immediate version of this question is whether closing it makes them better off in the following respect: it turns an unfair situation into a fair one. I have argued that this is not a way of making people better off, and thus this is not a way of avoiding describing the situation as one of levelling down. We now have a different suggestion: closing the swimming pool reduces the anger or resentment or envy felt by the blacks, and thus makes them better off in one respect. For this reason, once more, we need not describe the case as levelling down. For the purposes of my main argument, however, we can remain agnostic on the question of whether envy affects well-being. For it seems to me that the case for closing down the swimming pool is not premised on the blacks having a feeling of these or any other sort. Although there could be cases where the blacks are envious of the whites this need not be the case. If this is conceded and it is still conceded that the swimming pool should still be shut, even if the blacks do not care whether it is shut, then we still have a case of apparent levelling down. Further investigation is necessary. However, although not strictly necessary to the argument, the question of whether envy or similar emotions can affect well-being in our theoretically pure sense of well-being is undeniably an interesting question and worth pur- suing. Here the consensus view seems to be that it is not a relevant determinant: Rawls (1971, pp. 530–41) and Nozick (1974, p. 162) both suggest that arguments from envy have no place in the theory of justice. However, given that such feelings can cripple a life it seems harsh to judge that they should simply be ignored. Levelling Down 29

It seems to me that the idea that envy should not be taken into account may be based on the following argument:

(a) If we were to take envy into account as a determinant of well-being we would have to compensate those who were envious. (b) To compensate those who are envious means taxing the non-envious. (c) It is highly counter-intuitive and morally unattractive to redistribute from the non-envious to the envious just in virtue of that difference.

Therefore:

(d) Envy should not be considered a determinant of well-being.

Now it would be possible to contest (c), but I will not consider that. Rather, I think (a) requires examination. It is based on what could be called the ‘compensation’ paradigm: that if there is injustice then compensation of some sort should be made. This goes hand in hand with what we might think of as the ‘thermometer’ model of well-being: that one can model well-being as one does temperature, and if it falls then it should be restored to previous levels by the simplest and easiest method – normally a compensating cash payment. But well-being may be a highly complex notion, and compensation can be quite inappropriate in some cases. To take the case of envy, if A is envious of B, and one thinks this an undesirable situation, then there seem to be at least four ways of remedy:

(a) Compensate A. (b) Remove from B whatever it is that is the cause of the envy. (Levelling resources is one example of this, although if envy is a determinant of well- being this would not be real levelling down.) (c) Induce false beliefs in A so that A no longer has the beliefs that gave rise to the envy. (For a real example of this, see Wolff, 1991, p. 125.) (d) Induce character changes in A so that A is no longer an envious person.

Once this is spelled out it seems to me that we should acknowledge that envy can be detrimental to well-being, but it should not be assumed to be in the same category of well-being loss as, say, hunger or lack of shelter. It may not call for compensation, but it may call for remedy of the fourth type, and there may be reasons to tax everyone – envious and non-envious alike – to try to establish ‘envy clinics’ for the most serious cases, just as there are clinics for other character disorders. But if these fail it does not follow that there is a case for any compensation and still less for complete levelling down. 30 Jonathan Wolff

It follows from this that any mere fact that the blacks are envious of the swimming opportunities of the whites is in itself no reason for closing the swimming pool. But in any case, as I have suggested, the intuitive plausibility of the case for closing the pool does not rest on considerations about envy.

Symbolic value Some may have been uncomfortable with my use of the categories ‘blacks’ and ‘whites’ in discussing this case, and would have preferred that I used the niceties often followed in such discussions, describing the groups as the blues and the greens or the bigfeet and the smallfeet. But the examples do not work so well as fiction: it is vital, I think, to the discussion that there is an implied history and background. To explain, let me contrast this with a different case. Suppose the state edict was not that blacks should be excluded, but that people with red hair should be banned, on health grounds. Perhaps the situation is that it has been discovered that red hair contains a chemical which reacts with water to induce sickness in those who do not have red hair. (Ignore the question of why this has only been found out now.) Just as before the state insists that swimming should be segregated, this time on public health grounds, and if there is only one pool it should be available only to the majority group – the non-red- haired. Again, as before, you as mayor, do not have the resources to build a second pool. Now it seems to me that it could be a perfectly reasonable decision of the people as a whole to close down the pool to express solidarity. But it also seems that the case in justice for doing so is very weak indeed – almost to the point of vanishing altogether. In this clash between Pareto efficiency and fairness, efficiency seems the clear winner in this case. But superficially the two cases look very similar. How can we make the distinction between the cases? The answer will, of course, be obvious. The black/white case is a racist policy, based on a standing pattern of discrimination against a group which is already worse off. The red hair case does not have these features. One feels much more prepared to ensure that no one is relatively disadvantaged by deliberately unjust treatment than to ensure that no one is relatively disadvantaged by a public health policy which turns out to be uneven in its effects. Is it, then, that in the one case we feel entitled to stand up to the malice of the rulers, whereas in the other there is no malice present? But this is not quite right. Suppose, to adjust the example somewhat, the racist policy is a legacy of past legislation that no one believes in but no one has bothered to repeal. Thus although there may be negligence, there is no malice. But this hardly seems to make a difference to the acceptability of levelling down. Why, then, does one inequality, related to skin colour, matter so much when another, related to hair colour, matters much less, when the inequalities are, in some sense, the same? Levelling Down 31

To use an argument from Ann Phillips, it is not true that the inequalities that matter most are the ones that are most unfair in themselves. Rather, certain inequalities have a further symbolic meaning or function and can express an explicit or implicit ranking of citizens into groups of different worth (Phillips, 1999). To adapt one of Phillips’ examples, it would not normally matter very much if eye surgeons happened to be paid less than equally trained, skilled and dedicated ear surgeons. But if eye surgery happened to be a job performed mostly by women or by members of a minority and ear surgery by white men, then things look very different. Purely contingent unfairness is much easier to accept than a systematic pattern which is to the disadvantage of a previously disadvantaged group. What we see here is a type of intersection between political equality and economic equality. Sometimes it seems right to level down in economic or, at least, well-being terms, in order to achieve political equality or, at least, to remove clear barriers to political equality. Now, with these reasons on display we have to revisit the question of whether this is well represented as a case of levelling down after all. Three positions appear possible:

(a) Yes – this is a genuine case of levelling down. (b) No – this is a case where the worst off are made better off in an already well understood sense. (c) No – this is a case where the worst off are made better off in a subtle new category of well-being.

If one treats the real-Pareto maxim as axiomatic it is necessary to attempt to defend (b) or (c). Suppose, though, on the basis of further reﬂection (a topic for a further occasion), we conclude (a). Sometimes, then, we should level down. But I have only suggested that this is relevant when there are symbolic factors at play, which send messages of deep political inequality. This is not the politics of envy, or a cancer that will spread to allow all sorts of levelling down. Thus I would provisionally conclude that levelling down can be reasonable in a very special sort of case. Those sympathetic to equality should not be ashamed of this.

Notes

1. My thanks to Serena Olsaretti, Michael Otsuka, John O’Neill and Richard Norman for their written comments on an earlier draft. I would also like to thank the participants at PSA 2000 for their discussion of this paper, and the members of the equality workshop in London 1999 for discussion of the swimming pool example. 2. This principle loses its grip in respect of policy choices that affect the make-up of future populations. (See Parﬁt, 1984, part 4, for the classic discussion.) 3. So I have been persuaded by Michael Otsuka, to whom I owe the following example. 32 Jonathan Wolff

4. Some will say that justice allows the worst off to become worse off still when it is their own fault. I want to ignore that complication in what follows as it does not affect the substance of the argument. 5. Thus I would say that this is a case where the development of an egalitarian position has been hampered by taking too much notice of right-wing criticism. For another example see Wolff (1998). 6. Thus I do not intend to beg the question against those who believe that the correct currency is to be formulated in terms of primary goods or resources or advantage or basic capabilities or anything else. 7. I should acknowledge that those who argue that the current currency is resource- based, to the exclusion of well-being, have no reason to pursue this issue, at least in these terms. Thus the following paragraphs should concern only those who believe that well-being is at least part of the currency of justice.

Bibliography

Cohen, G. A. (1989) ‘On the Currency of Egalitarian Justice’, Ethics, 99, 906–44. Dworkin, Ronald (1981) ‘What is Equality? Part 1: Equality of Welfare’, Philosophy and Public Affairs, 10, 185–246. Frankfurt, Harry (1987) ‘Equality as a Moral Ideal’, Ethics, 98, 21–43. Norman, Richard (1998) ‘The Social Basis of Equality’, in A. Mason (ed.), Ideals of Equality. Oxford: Basil Blackwell. Nozick, Robert (1974) Anarchy, State, and Utopia. Oxford: Basil Blackwell. Parﬁt, Derek (1984) Reasons and Persons. Oxford: Oxford University Press. Parﬁt, Derek (1998) ‘Equality and Priority’, in A. Mason (ed.), Ideals of Equality. Oxford: Basil Blackwell. Phillips, Ann (1999) Which Equalities Matter? Oxford: Polity Press. Rawls, John (1971) A Theory of Justice. Oxford: Oxford University Press. Wolff, Jonathan (1991) Robert Nozick: Property, Justice and the Minimal State. Oxford: Polity Press. Wolff, Jonathan (1998) ‘Fairness, Respect and the Egalitarian Ethos’, Philosophy and Public Affairs, 27, 97–122. 3 The Political Economy of Human Happiness Benjamin Radcliff 1

Introduction

Proponents of democracy from Pericles to Robert Dahl have stressed its potential for building a society in which people enjoy better lives. As the promise of a better life is indeed a powerful mechanism for winning political support, electoral competition in modern democratic societies has come largely to embody ideological disagreements over the best method for improving quality of life. All shades of public philosophy, from the social democratic left to the libertarian right, implicitly predicate their policy appeals on the claim that their approach will increase the extent to which citizens find life agreeable. As Esping-Andersen (1990) comments, the democratic state is of necessity ‘preoccupied with the production and distribution of social well-being’. The dominant mode in the liberal democracies might thus profitably be conceived of as the politics of human happiness. It is hardly surprising, then, that one of the principal concerns of empirical democratic theory is the search for evidence that the actual practice of democratic politics does in fact matter for the lives of ordinary people. The goal has been to determine how policy outputs – the means to the end of a better life for more citizens – are affected by democratic politics. While much has been learned about how politics affects public policies, rather less attention has been devoted to understanding the determinants of the presumptive end of democratic politics: citizens who are more satisfied with the quality of their lives. As yet there has been little systematic effort within the discipline to identify the political factors which determine the final variable of interest, namely the general level of the subjective quality of life.2 This paper attempts such an examination through an analysis of mean levels of life satisfaction in the industrial democracies. Such an endeavour is especially called for given that the dominant theoretical approaches to the subject deny that cross-national differences in satisfaction can be attributed to

K. Dowding et al. (eds.), Challenges to Democracy © Political Studies Association 2001 34 Benjamin Radcliff politics (Easterlin, 1974; Inglehart, 1990). To anticipate what follows, I argue that certain political outcomes – those which insulate citizens against the worst forms of market dependence – dramatically affect a nation’s general level of satisfaction independent of economic or cultural conditions. This chapter is a condensed, non-technical summary of the arguments and empirical evidence reported in Radcliff et al. (2000). Readers are directed toward that paper for a more comprehensive literature review, the details of the econometric procedures, a discussion of various methodological issues and the tables containing the statistical estimates.

The study of life satisfaction

Conceptually, life satisfaction refers to the degree to which individuals evaluate positively the quality of their life in total. While subtle differences are sometimes made between ‘satisfaction’, ‘happiness’, ‘contentment’ and other variants of subjective well-being, these terms are often argued to be conceptually and empirically equivalent (e.g. Veenhoven, 1993). Most contemporary work on life satisfaction relies on a single, direct question that asks a respon- dent to comment on ‘how satisfied’ they feel with their lives ‘in general’ or ‘on the whole’. This simple item is generally agreed to perform as well as or better than more complex formulations (e.g. Veenhoven, 1993).3 Comparing mean levels of satisfaction across nations using this indicator has become commonplace (for a review, see Oswald, 1997).4 There are three broad theoretical styles applied to studying differences in satisfaction across nations. The first, comparison theory, is associated most strongly with Easterlin’s (1974, 1995) analyses of the relationship between economic development and national levels of satisfaction. He argues that richer countries display no greater mean happiness than poor ones. This is because individual assessments of one’s life are not absolute but rather relative to the ‘consumption norm’. This norm provides a common reference standard by which individuals appraise their well-being. Those above the standard are more happy, those below it less so. The overall level of satisfaction will thus tend to be neutral, in that the number of people happy relative to the reference standard should be about the same as those unhappy. Any observed national differences would reflect only measurement error or deviation in the location of a country’s reference standard relative to its own consumption norm. Political conditions within a country – such as whether the right or left tends to dominate electorally – would have no effect. Another perspective focuses on cultural norms. Different countries manifest different levels of happiness because of powerful and enduring ‘national creeds’ or ‘national character traits’ (Inkeles, 1997). Political interventions that affect objective living conditions are, as in comparison theory, de-emphasized The Political Economy of Human Happiness 35 in place of the ‘cultural glasses’ through which one looks at life. Thus Inglehart (1990: 30) suggests that international differences in life satisfaction ‘reflect cognitive cultural norms, rather than individual grief and joy’. Cultural theory thus differs from comparison theory by admitting the possibility of meaningful differences in satisfaction, but agrees with it in denying that objective conditions play a significant role in determining those differences. It is important to note that neither comparison nor trait theory leaves a role for politics. Positive political outcomes would just move the consumption norm, leaving people no happier than before. Political interventions similarly could not change deeply held cultural norms, at least over the relatively modest periods of time for which data are available. The dominance of these approaches may explain why political scientists (and to a lesser extent, political sociologists) have devoted such little effort to assessing how politics affects subjective well-being. To find a theoretical perspective that allows for a role for politics, we must look to work in the tradition of Maslow’s (1970) focus on such human ‘needs’ as physical and economic security. As Schyns (1998) nicely puts it, the con- tention implicit in this approach is that ‘the more needs are satisfied, the happier people will be’. Thus, contrary to the theories described above, in this view subjective appreciation of life depends upon the actual ‘objective’ quality of life. This approach has been most fully developed by Veenhoven in what he aptly calls ‘livability theory’ (see especially Veenhoven, 1993, 1995). As he notes, it is merely the application of the ‘common-sense’ view that people will lead better lives when they live in better, or more ‘livable’, societies. All of these theoretical approaches are plausible, but the literature derived from each is plagued by a variety of conceptual and methodological problems, at least vis-à-vis the effort to assess the impact of politics (for a discussion, see Radcliff et al., 2000). In the end, it is safest to conclude that little is known about the political determinants of subjective well-being across nations (but see Ott, this volume). In what follows, I argue that salient features of democratic politics do profoundly impact the subjective quality of life in the industrial democracies.

Markets versus politics

It is widely agreed that the two mechanisms available for the distribution and generation of well-being are markets and politics (e.g. Polanyi, 1944; Weber, 1958; Lindblom, 1977). In the liberal democracies, the conﬂict between these allocating mechanisms typically manifests itself in the struggle to replace the perceived inequalities of market distribution with the presumed equality of citizenship rights, i.e. to make ‘citizen entitlements . . . rather than the market 36 Benjamin Radcliff contract’ the basis of allocation (Esping-Andersen, 1985: 159). Those favouring political ‘entitlements’ in place of the market may do so for a variety of reasons, but all ultimately reduce to the fact that markets are ‘indifferent to the fate of individuals’ (Lane, 1978). The mechanisms by which markets contribute to unhappiness are most commonly argued to result from the economic insecurity and personal loss of autonomy that accompany market economies. Echoing Marx, Polanyi (1944) and Lindblom (1997), among others, Esping-Andersen (1990: 36) ‘observes that, whatever the many positive and commendable aspects of a capitalist economy, “the market becomes to the worker a prison within which it is imperative to behave as a commodity in order to survive”. This vision of the market as a prison, it should be noted, is hardly unique to class theorists or their neo-pluralist allies: Weber (1958), for example, refers to capitalism as an “iron cage” for the human spirit. It is certainly the case that the great mass of citizens in the industrial world depend for their livelihood upon the sale of their labour power as a commodity. The market for that commodity is characterized by uncertainty, even for the managerial and professional classes (e.g. Kuttner, 1997, ch. 3). Thus, as Lindblom (1977: 82) notes, ‘a pertinent objection to markets is that they foist insecurities on the population’ which become ‘all the more a problem when [one’s] livelihood is at stake’. Accordingly, it is the extent to which a programme of ‘emancipation’ from the market has been institutionalized within a given state, I argue, that is the principal political determinant of subjective well-being. Put differently, life satisfaction should increase as we move from less to more generous welfare states. More generally, life satisfaction should vary positively with the dominance in government of political parties committed to the social democratic programme of limiting human dependence on the market. The argument is straightforward. Even the most ardent defenders of capitalism, while rejecting the normative implications of the prison metaphor, concede that insecurity is endemic in market economies. We know, further, that economic insecurity translates into chronic psychological stress, which in turn is clearly inimical to life satisfaction (e.g. Brenner, 1977; Dooley and Catalano, 1980; Korpi, 1997). To the extent that society insulates its citizens against market dependence, it simultaneously insulates them from the stress and anxiety of market insecurity. Furthermore, markets, as economists since Adam Smith have recognized, contribute to inequality. The social democratic project again facilitates well-being by mitigating the worst effects of inequality through the idea of social citizenship: the state guarantees a basic, and ideally middle-class, level of need satisfaction through income maintenance, family allowances and other forms of public welfare. In sum, the political programme of the left attempts to ‘marginalize the market as the principal agent of distri- The Political Economy of Human Happiness 37 bution and the chief determinant of peoples’ life chances’ (Esping-Andersen, 1985: 245). Satisfaction with life, then, will increase as security becomes increasingly a matter of social provision. Hence, if, as argued, market independence contributes to subjective well-being, then the expansive ‘emancipating’ welfare state associated with the social democratic vision, or, more generally, the electoral dominance of parties devoted to that vision, the greater should be the general level of life satisfaction.

Analysis

The empirical analysis focuses on satisfaction in the industrial democracies. Two data collections are utilized. The ﬁrst is the time series available for the member states of the European Union. The second is cross-sectional data on a wider selection of 16 industrial democracies. Following the convention of the literature, the dependent variable is the mean level of national life satisfaction. The method is to regress that variable upon terms designed to measure the market-versus-politics distinction, along with appropriate control variables; a full discussion of the econometric procedures and various modelling strategies are provided in Radcliff et al. (2000).

Time Series analysis: the European Community The most extensive set of comparable time serial data on subjective well-being are from the Eurobarometer. It contains the question commonly used to assess life satisfaction: ‘On the whole, are you very satisfied, fairly satisfied, not very satisfied, or not at all satisfied with the life you lead?’ I utilize the national mean on this indicator, with the response categories coded one to four in ascending order. The data are from 1973 to 1990.5 In specifying the appropriate model, three sets of independent variables are included. The first are economic, given the large body of evidence that economic development, unemployment and perhaps short-term economic performance all affect aggregate levels of happiness. These are operationalized, respectively, as real per capita GDP in purchasing power parity, the unemployment rate (in per cent) and the annual percentage change in GDP. I also include the year, given Lane’s (1996) vigorous argument that the evolving logic of modernization has driven a consistent, linear decline in subjective well-being in recent decades. The second set of variables concerns political culture. Given that the basic proposition of the cultural school is that national levels of satisfaction are determined by the nation’s distinctive culture, the simplest way – substan- tively and econometrically – is to include dummy variables for every country, excepting a reference category. The effect of the dummies is of course to fit 38 Benjamin Radcliff separate intercepts for each country, thus accounting for the large and sustained differences in satisfaction that the cultural interpretation suggests. By accounting for level differences in the dependent variable, these terms also remove the possibility that unit effects could bias the resulting parameter estimates. This leaves political variables to be accounted for. The ideal variable would be one that captures the extent to which social rights have replaced markets, but to my knowledge no such data exist in time serial form. As a consequence, I of necessity look to two reasonable surrogates (used independently): (a) the common expedient of welfare expenditures, measured as annual real per capita welfare spending (expressed in 1985 purchasing power parity) and (b) the extent to which parties of the left, as advocates of social rights, have dominated politics, using a measure described presently. The utility of considering welfare spending is entirely obvious: spending represents the best summary measure of the state’s commitment to substitut- ing politics for markets that is available over time for a large number of countries.6 The rationale for focusing on the relative strength of parties is equally clear: the balance of party control of government is the best way to measure the ideology of a nation’s policy regime. Simply put, I presume that the ideology of a government is a function of the ideology of the parties that comprise it. By extension, the policy regime that citizens face at any given time is approximated by the cumulative governmental tenure of parties in office to that point. Given that parties of the left generally have advocated, and parties of the right have typically opposed, using the state to grant citizens universal social rights, then the relative balance between right and left accumulated over time should reflect the extent of provision of social rights. In other words, leftist parties share a commitment to ensuring that basic human needs are not, as the right would prefer, relegated to the market economy.7 The predominance of the left’s approach should produce greater security and thus greater life satisfaction. I measure this concept in the obvious fashion: the cumulative portion of left party cabinet seats less the cumulative portion of right party seats (to the given year). The expectation is that the greater this difference, the higher will be subjective appreciation of life. The impact of centre parties is less clear. Centre (and especially Christian Democratic) parties have often been argued to be less hostile to state provision of welfare than the right, although the policies they favour tend to be far more market conforming than those of the left. Their effect on quality of life would therefore tend to be minimal. To account for whatever effect the centre parties might have, though, I include in the party control model their cumulative cabinet seats as a separate variable.8 Initial results for welfare spending are consistent with expectations: greater spending implies greater happiness, other things being equal. In assessing the The Political Economy of Human Happiness 39 magnitude of this relationship, it is perhaps most instructive to compare the impact of spending to that of unemployment, the factor that is widely agreed to have especially dramatic effects on well-being at both the individual and aggregate levels (see Oswald, 1997). The difference between the maximum and minimum observed values on unemployment in the sample would predict a difference of just over one standard deviation on the dependent variable. The equivalent change over the range of welfare spending is fully twice that. Similar conclusions emerge when considering party control: the greater the time in power of the left less the time in power of the right, the happier are citizens. The magnitude of the relationship is similar to that for welfare spending. Furthermore we found the centre party variable lacks statistical significance, suggesting that centre governments have no impact on subjective quality of life.

Cross-sectional analysis: OECD countries While the pooled time series approach has obvious advantages, it also suffers from two limitations. One is the sample: data are available only for nations comprising the European Community. Moving to a cross-sectional design utilizing the second wave of the World Values Study allows the sample to include the United States, Canada, Japan, Finland, Sweden and Switzerland. I utilize the mean value of the life satisfaction question: ‘All things considered, how satisfied or dissatisfied are you with your life now?’ The cross-sectional approach has an additional advantage: it allows the use of more refined indicators of the extent to which states protect citizens against market dependence. Perhaps the best, and certainly the best known, of these are provided by Esping-Andersen (1990). He provides, first, three interval level variables tapping how much a nation’s welfare system embodies elements of his three ideal types of welfare regimes: liberal, conservative and socialist. The liberal ‘effectively contains the realm of social rights’ with modest and typically means-tested programmes. The conservative may provide more generous benefits, but in the end has a ‘negligible’ redistributive impact that reflects its emphasis on ‘the preservation of status differentials’. The socialist is, of course, predicated explicitly upon the substitution of citizenship rights for markets. As Esping-Andersen (1990: 27) puts it, this vision of the welfare state is one that ‘would promote an equality of the highest standards . . . by guar- anteeing . . . [everyone] full participation in the quality of rights enjoyed by the better-off’. This approach ‘crowds out the market’ in favour of state- provided social rights. As these are ideal types, countries naturally do not fit neatly into any one single category. Actually existing welfare states combine elements from each. For that reason, Esping-Andersen (1990) provides, and I utilize, three separate indices accessing the amount of socialist, conservative and liberal elements in 40 Benjamin Radcliff a nation’s policy regime. As social democratic attributes are clearly the ones most conducive to ending market dependence, I expect that subjective quality of life will vary directly with the extent to which social democratic elements are present. Esping-Andersen (1990) provides an alternative operationalization of market independence in the form of his index of the decommodification of labour. Labour is decommodified to ‘the degree to which individuals, or families, can uphold a socially acceptable standard of living independently of market participation’ (Esping- Andersen, 1990: 37). For reasons developed previously, the extent of decommodification should obviously contribute to the quality of life that citizens experience.9 The national level of satisfaction is modelled as a function of the three economic variables from the prior analysis, a measure of national culture offered by Triandis (1989) that has sometimes been argued to affect aggregate levels of happiness (e.g. Diener et al., 1995; Veenhoven, 1997b)10 and either the decommodification variable or the three welfare regime variables discussed above. The evidence suggests the more socialist a welfare regime, or, similarly, the greater the decommodification accomplished by state policy, the more satisfied are people with their lives. The magnitude of these effects is pronounced.

Democracy and life satisfaction

The results of democratic competition have dramatic effects on national levels of life satisfaction. Subjective appreciation of life is positively affected by governments of the left. It is also heavily influenced by the generosity and quality of the welfare state. Per capita spending, the extent of social democratic attributes and the degree of decommodification all contribute to more people leading more satisfied lives. Several implications follow. The most obvious are the practical implications for electoral politics and social policy. Simply stated, the political programme of the left, whatever other failings may be attributed to it, does appear to improve a nation’s subjective quality of life. While no empirical evidence is ever capable of adjudi- cating between whether left or right governments are ‘better’, the evidence presented here does suggest that citizens should look toward parties of the left to the extent that they wish to build a society in which people are, on average, more satisfied with life. That conclusion depends, of course, upon the left maintaining its traditional ideological principles. The 1990s have witnessed a moderation in or retreat from such commitments, in varying degrees, across the industrial democracies. There is as yet no consensus on the magnitude or importance of such retrenchment. Nor is it by any means certain that the present trajectory The Political Economy of Human Happiness 41 will or can be maintained. What is clear from the present analysis is that the ideological dilution of the core values of progressive parties may diminish quality of life. The present analysis also speaks to the perennial debate over the desirability of the welfare state. For social democrats, the welfare state has long been conceived as a mechanism for empowering the working class in its presumed struggle with capital. For workers to successfully engage in collective action, they first establish a basic level of security outside the market. Whether viewed as a step in the transformation of society away from capitalism or merely as a way of improving the relative position of wage-earners, the welfare state is in this tradition an absolute necessity for the well-being of the great mass of citizens. Non-socialist progressives and various liberal reformers have also looked to the welfare state as a method for correcting ‘market failures’. Whether driven by a sense of social justice or the simple desire to order capitalism in a way that makes it more self-sustaining, they too look to the welfare state as a mechanism for improving people’s lives. Conversely, many conservatives and classical liberals have decried the dele- terious effects of the welfare state on the positive and empowering institution of the market. The effort to protect people from ‘cradle to grave’ may be well intended, it is argued, but the net effect is to breed dependency and compla- cency, with predictable consequences for people’s self-respect and autonomy. Individuals similarly find themselves alienated and powerless in face of the huge and intrusive governmental bureaucracy that accompanies the welfare state. Further, it is said, the enormous financial burdens of social spending are simply passed on to taxpayers and businesses, with, again, predictable effects on investment, profits and prosperity. The welfare state, then, will ultimately contribute to worse rather than better lives. It would be both naive and presumptuous to claim that the evidence presented comes remotely close to resolving the question of whether the welfare state – or the more general programme of the left – actually enhances the human condition. Such an assessment is ultimately both normative and ideological; as such, it is unlikely to be settled by any empirical evidence. Still, the findings reported above do add to the body of evidence in favour of the desirability of an expansive regime of social security and social rights. The welfare state, whatever else it might do, makes people, on average, more satisfied with their lives. While that hardly settles the argument over the wisdom of the welfare state, it would seem of some moment. Elsewhere (Radcliff et al., 2000) I consider the implications of the present findings for our theoretical understanding of life satisfaction. For present purposes, it is sufficient to note that whether happiness is relative or not, or whether it is affected by cultural norms or not, it is affected by democratic politics. 42 Benjamin Radcliff

This has profound implications for our empirical and normative understanding of liberal democracy. Subjective well-being is affected meaningfully by the political choices made by voters and the governments they elect. That conclusion should certainly be welcomed by democrats of all ideological per- suasions. It may also offer solace to all who would look to the ballot box as a tool to help build a better world.

Notes

1. The author gratefully acknowledges the assistance with this project afforded by Ruut Veenhoven. Thoughtful comments and suggestions were also made by Arend Lijphart, Suzanne Coshow, Ed Diener, Keith Dowding, Amy Gille, Andrew Gould, Michael Krassa, Brian Krueger, Robert Lane, Anthony Messina, Jack Nagel, Jan Ott and Duane Swank. 2. The major efforts within political science in the last decade to model cross-national levels of satisfaction are Inglehart (1990) and Clarke, Dutt and Kornberg (1993). While both are valuable and illuminating, they say little about politics per se. Inglehart treats satisfaction as a cultural phenomenon. While he does identify one political variable associated with satisfaction – years of continuous democracy – he argues that causality ﬂows predominantly from satisfaction to democracy. Subjective well-being thus becomes an independent variable affecting politics rather than a dependent variable determined by politics. Clarke, Dutt and Kornberg (1993) examine a variety of public attitudes in western Europe, including satisfaction with life, satisfaction with democracy, support for governing parties and desire for social change. They model life satisfaction as a function of economic conditions alone, purposely omitting the political variables they include in the equations estimating the other attitudes. They go on, too, to use satisfaction as an independent variable that affects politics but is not directly affected by it. 3. Measures of subjective well-being are generally agreed to be reliable and valid; for a discussion, see Myers and Diener (1997). After an extensive review, Veenhoven (1996: 4) concludes that any misgivings on this score ‘can be discarded’. This is especially so when using national averages rather than individual level data, given that whatever imprecision is involved will tend to balance out in large samples. 4. It is sometimes objected that linguistic or cultural barriers prohibit meaningful cross-national comparison. Detailed examination of these claims by Veenhoven (1993, 1996, 1997a, 1997b) and Inglehart (1990) ﬁnd no support for them. 5. The countries included are Belgium, Denmark, West Germany, Greece, Spain, France, Ireland, Italy, Luxembourg, the Netherlands and Great Britain. Portugal is effectively excluded due to missing data on other variables. The time series for Greece and Spain are shorter, given their later entry into the EC; removing these countries from the sample does not affect the reported results. Sample size is about 1000 or more per country (the Luxembourg sample is smaller). Multiple observations per year are averaged. There are no data for 1974. Other data utilized in the time series analysis are from Huber, Ragin and Stephens (1997), the Penn World Table (version 5.6), Lane et al. (1991) and the World Bank (World Data 1995). 6. The most common method of measuring welfare state generosity has been as a proportion of GDP. While measuring welfare spending relative to the size of the The Political Economy of Human Happiness 43

economy is perfectly sensible when considering welfare ‘effort’, per capita spending is more appropriate for considering the actual impact of state provision on people’s lives. A country that spent a large share of its (per capita) GDP might indeed receive high marks for effort, but if its (per capita) economy were small, the net consequences for quality of life would also remain small. Note too that while the concept of ‘generosity’ that spending captures is surely critical in assessing welfare states, it is sometimes argued that the quality of social rights cannot be completely reduced to spending levels (e.g. Esping-Andersen, 1990). I do not mean to suggest otherwise. It may well be that not all welfare dollars buy the same degree of market independence. Nonetheless, the correlation between the extensiveness of social rights and the amount of spending necessary to provide those rights is certain to be very high. Spending thus seems to be a perfectly plausible if not entirely perfect surrogate for market independence. In any case, the differences between left, centre and right approaches to public welfare may be captured in a more qualitative way when considering the indices of party control discussed below. 7. While party ideologies are always in flux, party identities – or, at least, relative ideological distances – have remained reasonably stable during the time-frame in question. To be sure, there were some instances of significant moderation among left-wing parties (e.g. the French Socialists in the 1980s) as well as any number of more subtle programmatic shifts insightfully detailed by Kitschelt (1994), but such developments hardly challenge the basic identification of the left with the programme of protecting citizens against the indifference of the market. In any case, Huber and Inglehart (1995: 79) report that the correlation between expert ratings of party ideology in the early 1980s with that of the early 1990s is 0.94. It is arguable that some later developments (such as the advent of ‘New Labour’ in Britain, or perhaps even the emerging though still contested centrism of the German SPD) have produced more troublesome questions, but these are fortunately beyond the temporal limits of the present analysis. 8. I have relied on the ideological categorization (and concomitant cumulative cabinet data) offered by Huber, Ragin and Stephens (1997). They divide parties into left, centre secular, centre Christian, centre Catholic, secular right, right Christian and right Catholic. I create left, centre, and right by collapsing the centre and right categories in the obvious fashion. I also tested for a possible Christian Democratic effect per se by (a) treating the centre group as only the Catholic and Christian centre, or (b) pooling the four religious party types (with or without the secular centred added as well) and removing the two religious right parties from the total right category. Neither proved significant in the analysis that follows. Cabinet data for Spain and Greece (not included within Huber, Ragin and Stephens (1997)) are from Lane et al. (1991); as noted previously, results are unchanged when removing these countries from the sample. 9. The three welfare regime characteristic variables are from Esping-Andersen (1990: table 3.3). The details of the operationalization are complex (1990: 55–78). Suffice it to say that the variables are cumulative index scores on multiple indicators of each dimension; higher values indicate a greater prevalence of the particular type of ideal regime type. The decommodification variable is from (1990: table 2.1). Briefly, it is computed by multiplying a multi-item index of the generosity of unemployment insurance by the percentage of the labour force eligible (for a description, see 1990: 49–54); higher values indicate greater decommodification. 10. The Triandis (1989) index is an attempt to measure how ‘individualistic’, as opposed to ‘collectivist’, a given culture is. Individualistic cultures have been argued 44 Benjamin Radcliff

to encourage subjective well-being, in that they provide greater independence to choose and achieve one’s own life-goals.

References

Brenner, M. Harvey (1977) ‘Personal Stability and Economic Security’, Social Policy, 9: 2–14. Clarke, Harold, Nitish Dutt and Allan Korneberg (1993) ‘The Political Economy of Attitudes Toward Politics and Society in Western Europe’, Journal of Politics, 55: 998–1021. Diener, Ed, Marissa Diener and Carol Diener (1995) ‘Factors Predicting the Subjective Well-Being of Nations’, Journal of Personality and Social Psychology, 5: 851–64. Dooley, David and Ralph Catalano (1980) ‘Economic Causes of Behavior Disorder’, Psychological Bulletin, 41: 81–93. Easterlin, Richard (1974) ‘Does Economic Growth Improve the Human Lot?’, in P. A. David and M. W. Reder (eds), Nations and Households in Economic Growth. New York: Academic Press. Easterlin, Richard (1995) ‘Will Raising the Incomes of All Increase the Happiness of All?’, Journal of Economic Behavior and Organization, 27: 35–48. Esping-Andersen, Gosta (1985) Politics Against Markets. Princeton, NJ: Princeton University Press. Esping-Andersen, Gosta (1990) The Three Worlds of Welfare Capitalism. Princeton, NJ: Princeton University Press. Huber, Eveylne, Charles Ragin and John D. Stephens (1997) Comparative Welfare States Dataset, Northwest University and University of North Carolina. Available freely on line at http://lissy.ceps.lu/compwsp.htm Huber, John and Ronald Inglehart (1995) ‘Expert Interpretations of Party Space and Party Locations in 42 Societies’ Party Politics, 1: 73–111. Inglehart, Ronald (1990) Culture Shift in Advanced Industrial Society. Princeton, NJ: Princeton University Press. Inkeles, Alex (1997) National Character. New Brunswick, NJ: Transaction Publishers. Kitschelt, Herbert (1994) The Transformation of European Social Democracy. New York: Cambridge University Press. Korpi, T. (1997) ‘Is Utility Related to Employment Status?’, Labour Economics, 4: 125–48. Kuttner, Robert (1997) Everything for Sale. New York: Alfred A. Knopf. Lane, Jan-Erik, David McKay and Kenneth Newton (1991) Political Data Handbook OECD Countries. Oxford: Oxford University Press. Lane, Robert (1978) ‘Autonomy, Felicity, Futility’, Journal of Politics, 40: 1–24. Lane, Robert (1996) The Joyless Market Economy. Paper delivered to the Conference on Economics, Values, and Organization, Yale University, 1996. Lindblom, Charles (1977) Politics and Markets. New York: Basic Books. Maslow, A. (1970) Motivation and Personality. New York: Harper & Row. Myers, David and Ed Diener (1995) ‘Who Is Happy?’, Psychological Science, 6: 10–19. Oswald, Andrew (1997) ‘Happiness and Economic Performance’, Economic Journal, 107: 1815–31. Polanyi, Karl (1944) The Great Transformation. New York: Rinehart. Radcliff, Benjamin with Suzanne Coshow, Brian Krueger and Samuel Best (2000) Politics, Markets and Life Satisfaction: The Political Economy of Human Happiness in the Industrial Democracies. Paper presented at the Political Studies Association Conference, London. The Political Economy of Human Happiness 45

Schyns, Peggy (1998) ‘Crossnational Differences in Happiness’, Social Indicators Research, 43: 3–26. Triandis, H.C. (1989) ‘The Self and Social Behavior in Differing Cultural Contexts’, Psychological Review, 96: 506–20. Veenhoven, Ruut (1993) Happiness in Nations. Rotterdam: Risbo. Veenhoven, Ruut (1995) ‘The Cross-National Pattern of Happiness’, Social Indicators Research, 34: 33–68. Veenhoven, Ruut (1996) ‘Developments in Satisfaction Research’, Social Indicators Research, 37: 1–46. Veenhoven, Ruut (1997a) ‘Advances in Understanding Happiness’, Revue Quebecoise de Psychologie, 18: 29–74. (Article published in French; English translation kindly provided by Ruut Veenhoven, Erasmus University, The Netherlands.) Veenhoven, Ruut (1997b) ‘Quality of Life in Individualistic Society’, in The Gift of Society. Nijker, the Netherlands: Enzo Press. Weber, Max (1958) The Protestant Ethic and the Spirit of Capitalism. New York: Scribner’s. 4 Freedom and the Achievement of Happiness Jan Ott

Introduction

According to Veenhoven (1992) ‘happiness’ is an adequate indicator for the quality of life or the ‘livability’ of nations, the unobserved fit between provisions and requirements of society and the needs and capacities of citizens. This ‘output-indicator’ directly measures the flourishing of citizens as apparent in the observable appraisal of life. ‘Input indicators’ like GNP, equality, literacy and so on are based on assumptions about the contribution of these factors to the quality of life. The selection of these indicators and the weights they get in comprehensive indicators, like the Human Development Index, are more arbitrary. Opinion about the impact of freedom on happiness and the ‘livability’ of nations is mixed. Different philosophies stress different effects. Individualistic social philosophy stresses the positive effects. It is assumed that people themselves know best what will make them happy, and hence that they will enjoy life more if they can follow their own preferences. Conflicts of interest are seen to be solved by the invisible hand of the market, which is believed to yield optimal solutions more than prescription by king or custom. Conservative thought tends to emphasize the negative consequences of freedom. Conservatives doubt that people really know what is best for them. The wisdom of tradition and the benefits of solidarity are seen to bring a better life than short-sighted egoism. Some schools see different effects of different variants of freedom. The New Right is positive about economic freedom but critical about freedom in the private sphere of life. A leading view in South-East Asia is that economic but not political freedom will improve the human lot. According to Amartya Sen, economist and Nobel prize winner, freedom is the most important condition for development. Two questions are addressed in this paper: first, are the effects of freedom on happiness positive on balance? and second, what kind of freedom is most conducive to happiness?

K. Dowding et al. (eds.), Challenges to Democracy © Political Studies Association 2001 Freedom and the Achievement of Happiness 47

Earlier studies

Veenhoven (1999) assessed average happiness in 48 nations in the early 1990s. These nations are not a representative sample of all nations in the world, since these countries are relatively rich and developed. The scores are systematically higher in nations that are most afﬂuent, free, equal, educated and harmonious. Happiness was not signiﬁcantly related to unemployment, state welfare, income equality, religiousness and trust in institutions. In rich nations the additional value for happiness of extra wealth seems to decrease while the additional value of rights seems to be stable. Veenhoven made a distinction between economic, political and personal freedom. The correlation of happiness and economic freedom is rather general; the correlation of happiness and political freedom is limited to rich nations. Correlations between happiness and personal freedom are different for different types of personal freedom. Since the relation between happiness and economic freedom seems to be rather general, this relation deserves more attention. This paper partially repeats Veenhoven’s empirical test with data for happiness in 1995, economic freedom in 1995/1996 and political and press freedom in 1994/1995. The data and sources are given in Appendix 4.1.

Comprehensive economic freedom, political freedom and freedom of the press

Table 4.1 correlates happiness and different types of freedom in 1994. Since higher scores for political freedom, press freedom and economic freedom as measured by the Heritage Foundation indicate less freedom the + and – for correlations with these factors have been reversed to obtain more consistency: positive correlations always indicate a positive effect on happiness and vice versa. If we split the sample into rich and poor nations (see Table 4.2), we see that zero order correlations are high for economic and political freedom and

Table 4.1 Economic and political freedom, correlations with happiness and controlling for wealth

Zero order Controlled for wealth Economic freedom (Fraser) 0.694** 0.5028** Economic freedom (Heritage) 0.591** 0.3025 Political freedom 0.404** 0.0309 Freedom of the press 0.482** 0.0712

** = signiﬁcant at 0.01% * = signiﬁcant at 0.05% 48 Jan Ott

Table 4.2 Economic and political freedom in rich and poor nations

Rich nations Poor nations Economic freedom (Fraser) 0.617** 0.381 Economic freedom (Heritage) 0.328 0.248 Political freedom 0.647** –0.241 Freedom of the press 0.701** –0.138

*/** see Table 4.1.

Table 4.3 Components of freedom rather exclusively related to the behaviour of governments (Fraser Institute)

Zero order Controlled for wealth 1. Government consumption 0.096 0.4223* 2. Transfers and subsidies –0.137 0.2706 3. Government enterprises 0.746** 0.6560** 4. Deposits in private banks 0.581** 0.4410* 5. Negative interest (by regulation) 0.558** 0.3586*

*/** see Table 4.1. freedom of press: more freedom goes together with more happiness. These correlations are much lower if the effects of differences in wealth (real GDP per capita in 1994; source: Human Development index) are neutralized. Only correlation with economic freedom, as measured by the Fraser Institute, remains high (see Table 4.3). Splitting the nations into relatively rich and relatively poor (split at $10 000 a year per capita in 1994) is informative. Correlations between happiness and economic and political freedom and press freedom are still high in rich nations. In poor nations correlations between political and press freedom are the opposite: there seems to be greater happiness in nations with lower political and press freedom (though this relationship is not signiﬁcant).

Components of economic freedom: zero order and wealth controlled

Tables 4.4 to 4.6 show components of economic freedom, general monetary and institutional development within the nation, and general monetary and institutional development within international relations. Seven components of economic freedom show a high correlation with happiness. After controlling for wealth only freedom resulting from a low level of government enterprises and investment, more deposits in private banks and the absence of negative interest situations still have a high positive correlation with happiness. A remarkable result is the high correlation between happiness and Freedom and the Achievement of Happiness 49

Table 4.4 ‘A-components’ of freedom rather exclusively related to the behaviour of governments (Fraser Institute)

Zero order Controlled for wealth 1 Government consumption 0.096 0.4223* 2. Transfers and subsidies –0.137 0.2706 3. Government enterprises 0.746** 0.6560** 4. Deposits in private banks 0.581** 0.4410* 5. Negative interest (by regulation) 0.558** 0.3586*

*/** see Table 4.1.

Table 4.5 ‘B-components’ of freedom more related to general monetary and institutional development within the nation (Fraser Institute)

Zero order Controlled for wealth 1. Price controls 0.366* 0.1476 2. Recent inﬂation 0.518** 0.2429 3. Viability contracts 0.459** 0.1120 4. Rule of law 0.283 –0.2167

*/** see Table 4.1.

Table 4.6 ‘C-components’ of freedom more related to monetary institutional development in international relations (Fraser Institute)

Zero order Controlled for wealth 1. Non-tariff trade restraints 0.054 –0.1309 2. Capital transactions with foreigners 0.392* 0.0161

*/** see Table 4.1. freedom as a result of a low level of government consumption, if wealth is controlled.

Components of economic freedom: rich and poor nations

Table 4.7 shows components of freedom related exclusively to the behaviour of governments, and Table 4.8 shows them related to general monetary and institutional development within the nation. In Table 4.7 only freedom resulting from less government enterprises and investment shows a high positive correlation in 1995 with happiness in rich and poor nations; correlation is higher in poor nations. Freedom as a result of low government consumption and more deposits in private banks correlates with more happiness only in poor nations. Freedom by low levels of transfers and subsidies shows more 50 Jan Ott

Table 4.7 ‘A-components’ of freedom rather exclusively related to the behaviour of governments (for four components comparable data for 1990 were available; these data are presented in brackets) (Fraser Institute)

Rich nations Poor nations 1. Government consumption –0.148 0.533* 2. Transfers and subsidies 0.184 (–0.122) 0.334 (0.696**) 3. Government enterprises 0.499* (0.162) 0.645* (0.609**) 4. Deposits in private banks 0.164 0.471* 5. Negative interest (by regulation) 0.055 0.322

*/** see Table 4.1.

Table 4.8 ‘B-components’ of freedom more related to general monetary and institutional development within the nation (Fraser Institute)

Rich nations Poor nations 1. Price controls 0.372 0.077 2. Recent inﬂation 0.496* 0.295 3. Viability contracts 0.253 (0.558**) –0.027 (0.291) 4. Rule of law 0.597** (0.550**) –0.088 (–0.536*)

*/** see Table 4.1.

correlation with happiness in poor nations than in rich nations (but not significantly so). For 1995 freedom by low recent inflation and more rule of law only shows significant correlation with happiness in rich nations. Freedom by absence of price controls and more viability of contracts correlates more with happiness in rich nations than in poor nations (but not significantly so). In poor nations these correlations, and the correlation for the rule of law, are negative: low scores for these freedom components go together with low scores for happiness. The ‘B-components’ of the Heritage Foundation show a similar pattern: higher correlations with happiness in rich nations (monetary policy, wage and price controls, banking, property rights). Freedom by less non-tariff trade restraints and less restrictions on capital transactions with foreigners goes together with positive scores for happiness in rich nations and negative scores in poor nations, but these correlations are low. The ‘C-component’ of the Heritage Foundation (capital flows and foreign investment policy) shows a similar pattern: a positive correlation with happiness in rich nations and a negative one for poor nations, but not significantly so. Freedom and the Achievement of Happiness 51

Discussion

Before concluding anything some reservations have to be made.

1. The 52 nations in this empirical test (see Appendix 4.2) are not representative of all nations in the world (about 174). These 52 countries are relatively rich and developed. The nations in the test are nevertheless very different in terms of development, political regime, international position (isolated or integrated) and religion. The possibilities of reaching conclusions with this heterogeneous sample, and with cross-sectional analysis and relatively simple statistical techniques, are limited. Only ‘rough’ conclusions are justified. With a more homogeneous sample and more sophisticated techniques more ‘refined’ conclusions are possible. Benjamin Radcliff has demonstrated this by establishing that the results of democratic competition have dramatic effects upon national levels of life satisfaction. Specifically he demonstrates that subjective appreciation of life is positively affected by governments of the left and by the level (and kind) of welfare state development (Radcliff et al., 2000; Radcliff, this volume). 2. Some of the selected freedom components may have a direct short-term effect on individual freedom. Others may have a more indirect long-term effect via organizations like corporations (including multinationals), churches and trade unions that can play a role in the economic and political system. More political rights and press freedom can change a political system; in due time they can be a starting point, also in poor countries, for more individual freedom for minorities. This can create happiness in the future. 3. Freedom can also enhance economic growth. Ayal and Karras, also using information from the Fraser Institute, have specified this effect of aggregate economic freedom and separate components. In this way too, freedom can create happiness in the future. Time serial data could be used to analyse these long-term interactions (Ayal and Karras, 1999). 4. ‘Freedom’ is a very difficult concept. Freedom can be defined as the possibility of choosing but this can only be measured by ‘circumstantial evidence’. This is complicated since it takes a lot to be free. It requires absence of obstacles created by governments or otherwise (‘negative freedom’ as defined by Berlin (1969)) and it requires opportunities and information about options. It requires capabilities, literacy and knowledge. Individual freedom also requires some psychological inclination to go one’s own way (Bay, 1970). A lack of freedom can be caused by very different problems. It is therefore difficult to choose an adequate mix of observable manifesta- tions to measure freedom. Perhaps it is realistic to accept that we will never find such a mix and every choice has its limitations. Perhaps, because of 52 Jan Ott

that, every choice has some ideological background. But freedom is nevertheless a concept that can be used fruitfully in scientific research. Obviously the focus of attention for the Fraser Institute and the Heritage Foundation is negative freedom, or more specifically: freedom from government interventions. But freedom also requires capabilities to utilize opportunities. This might require investment by governments in education and this requires taxation. The Fraser Institute index is technically based on statistics about economic activities, but is this a valid measurement of freedom or is it a possible result of freedom? The indexes of the Heritage Foundation and Freedom House are based on expert ratings. But then the frame of reference of the experts can become rather dominant. In surveys, like the World Value Survey, the knowledge, consciousness and frankness of respondents set limits. The definition and measurement of freedom obviously requires more research. Since in our statistical exercise information from the Fraser Institute and the Heritage Foundation is used, we must realize that the role of governments must be the focus in our conclusions about the relation between economic freedom and happiness.

Conclusions

Correlations between happiness and comprehensive indicators for economic, political and press freedom are significant. Correlation with economic freedom is rather international but correlation with political freedom and press freedom is limited to relatively rich nations. Distinguishing separate components of economic freedom specifies the relation between economic freedom and happiness. Correlation with components more exclusively related to the behaviour of governments is limited to relatively poor nations. Among relatively poor nations average happiness is higher in nations with low government consumption, a lower level of transfers and subsidies, less government enterprises and investment, more deposits in privately owned banks and fewer interest rate controls and regulations that lead to negative interest rates. Correlation with components related to the monetary and juridical institutional development is, like political freedom, limited to rich nations. Among relatively rich nations average happiness is higher in nations with fewer price controls, lower inflation, better viability of contracts and greater rule of law. Correlation with components more related to monetary institutional development in international relations is positive among rich nations and negative among poor nations, but all these correlations are low and not significant. Given the reservations to be made, only a provisional conclusion can be formulated. In poor nations government interventions in the economy seem to have more short-term impact on happiness compared to government inter- Freedom and the Achievement of Happiness 53 vention in rich nations. Governments of relatively poor nations should therefore be very careful when proposing interference which reduces economic freedom. Governments in relatively poor countries should respect and promote economic freedom by enhancing monetary and juridical institutional development and stability, for example by supporting and respecting the independence of the central bank and the judiciary. This is probably relatively cheap and has positive effects for economic development and happiness in the future. Limiting freedom by expensive policies requiring high government consumption, transfers, subsidies and taxes can demotivate citizens, and perhaps foreign investors, from starting economic activities. Perhaps Montesquieu should have added the central bank to his trias politica.

Appendix 4.1 data and sources

Happiness Average happiness in 52 nations in 1994 on a 0–4 scale: World Value Survey 1997, Cumulative ﬁle, ICPSR, Michigan, USA. See appendix with scores for average happiness in 1995. Scores are obtained by the question: ‘Taking all things together, would you say you are: very happy, quite happy, not very happy or not at all happy.’ Higher scores indicate more happiness. There has been a lot of discussion about the methodological aspects of this measurement but in general this measurement is considered to be sufﬁciently valid.

Political freedom Political freedom 1995: Freedom in the World – The Annual Survey of Political Rights and Civil Liberties 1998–1999, Freedom House (1972–73 to 1999–2000). Scores for the two components, political rights and civil liberties, range from 1 to 7; higher scores indicate less political rights or civil liberties. Scores are based on expert ratings (judgements of the Survey team at Freedom House). The correlation between scores for political rights and civil liberties is high. For comprehensive ‘political freedom’ the sum is taken of the separate scores for the two components (scores from 2 to 14: a higher score indicates less political freedom).

Press freedom Press Freedom Survey, Freedom House (1999–2000). Scores for the two components ‘broadcast’ and ‘print’ range for 1994/1995 from 0 to 50; higher scores indicate less freedom. Correlation between these components is high. Scores are based on expert ratings (f.e. correspondents). For ‘press freedom’ the sum is taken of the separate scores for the two components (scores from 0 to 100: a higher score indicates less press freedom). 54 Jan Ott

Economic freedom: Fraser Institute Economic freedom in nations (general index and speciﬁc components) in 1995: ‘Economic Freedom of the World 1975–1995’, James Gwartney, Robert Lawson and Walter Block, the Fraser Institute. The index consists of scores from 0.0 to 10.0: a higher score indicates more economic freedom. The index comprises 23 components designed to identify how consistent institutional arrangements and policies are with economic freedom: higher scores indicate more freedom. All scores are based on statistical information. Components cover seven areas. Each area has a certain weight within the index and within each area components have a certain weight. The weight of each component in the index can be calculated by multiplication. The seven areas covered by the index are:

1. size of government: consumption, transfers and subsidies (2 components; weight in index 11.0 per cent); 2. structure of the economy and use of markets (4; 14.2 per cent); 3. monetary policy and price stability (3; 9.2 per cent); 4. freedom to use alternative currencies (2; 14.6 per cent); 5. legal structure and property rights (security of property rights and viability of contracts) (3; 16.6 per cent); 6. international exchange: freedom to trade with foreigners (5; 17.1 per cent); 7. freedom of exchange in capital and ﬁnancial markets (4; 17.2 per cent).

The 52 nations with scores for happiness are relatively rich and developed. The variety in scores for the components of economic freedom is therefore in many cases too limited to calculate correlations with happiness. To assess the relation with happiness 11 components are selected if diversity in scores is sufﬁcient for such a calculation. It appears then that these 11 components can be roughly divided into three groups (in brackets the weights of the components in the index):

A. Components of freedom rather exclusively related to the behaviour of governments:

1. General government consumption expenditures as a percentage of total consumption (5.5 per cent); 2. Transfers and subsidies as a percentage of GDP (5.5 per cent); 3. Government enterprises and investment as a share of the economy (4.6 per cent); 4. Ownership of banks: percentage of deposits held in privately owned banks (4.7 per cent) (= absence of state owned banks); Freedom and the Achievement of Happiness 55

5. Interest rate controls and regulations (by the government) that lead to negative interest rates (4.2 per cent).

B. Components of freedom more related to general monetary and institutional development within the nation:

1. Price controls: extent to which businesses are free to set their own prices (4.8 per cent); 2. Annual inﬂation rate during the most recent year (3.0 per cent); 3. Viability of contracts (risk of contract repudiation by the government) (5.6 per cent); 4. Rule of law: legal institutions supportive of rule of law and access to a nondiscriminatory judiciary (5.3 per cent).

C. Components of freedom more related to monetary institutional development in international relations:

1. Percentage of international trade covered by non-tariff trade restraints (3.3 per cent); 2. Restrictions on the freedom of citizens to engage in capital transactions with foreigners (4.7 per cent).

Obviously the distinctions between A, B and C are matters of degree. The role of governments is rather important in all components. But in the ‘B-’ and ‘C-components’ other factors play a more substantial role, like the behaviour of the judiciary and the central bank in the ‘B-components’ and the behaviour of other governments, international institutions like the IMF and the World Bank, the size of the economy and foreign investors in the ‘C- components’.

Economic freedom: Heritage Foundation ‘Index of Economic Freedom’, Johnson, B. T. and Sheeny T. P., Heritage Foundation, Washington DC (1995). A comparable index for economic freedom has been developed by the Heritage Foundation. Like the index of the Fraser Institute this index is primarily directed at – the absence of – obstacles created by governments. The Heritage Foundation deﬁnes economic freedom as the absence of government coercion or constraint on the production, distribution or consumption of goods and services. However, while the index of the Fraser Institute is based on statistics about daily economic practice or output, the index of the Heritage Foundation is based on estimates by experts of institutional conditions or input for economic activity. Ten factors, 56 Jan Ott equally weighted, contribute to the index of the Heritage Foundation. These factors can be divided in the same way in three groups.

A. Components of freedom rather exclusively related to the behaviour of governments:

1. Trade policy; 2. Taxation; 3. Government intervention in the economy; 4. Regulation; 5. Black market.

B. Components of freedom more related to general monetary and institutional development within the nation:

1. Monetary policy; 2. Wage and price controls; 3. Banking; 4. Property rights.

C. Components of freedom more related to monetary institutional development in international relations:

1. Capital ﬂows and foreign investment policy.

Appendix 4.2 Happiness and freedom in 52 nations (mid-1990s)

Nation Code Name Happiness Economic Economic Political Press freedom freedom freedom freedom 1994/95 1995 1995/96 1995 1995 World Value Fraser Heritage Freedom Freedom Survey Institute Foundation House House Scale 0–4 Scale 0–10 Scale 0–3 Scale 2–14 Scale 0–100 (not happy (not free – (free – (free – (free – – happy) free) not free) not free) not free) GB Britain 3.30 8.80 1.95 3 22 BG Bulgaria 2.33 5.20 3.50 4 39 CND Canada 3.18 8.40 2.00 2 18 RCH Chile 3.04 8.20 2.50 4 30 CN China 2.98 5.60 3.80 14 83 CR Croatia 2.69 4.30 3.70 8 56 CZ Czech Rep. 2.75 6.80 2.10 3 21 DK Denmark 3.30 8.10 1.95 2 9 DO Dominica 3.05 6.30 3.40 6 35 EW Estonia 2.61 6.10 2.25 5 25 SF Finland 3.11 8.10 2.30 2 15 F France 3.14 8.10 2.30 3 27 GE Georgia 2.68 3.85 10 70 D Germany 3.00 8.20 2.00 3 18 H Hungary 2.79 6.80 2.80 3 38 IS Iceland 3.40 8.10 2.50 2 12 IND India 2.89 5.00 3.70 8 49 IRL Ireland 3.36 8.60 2.20 3 15 I Italy 2.92 7.50 2.50 4 30 J Japan 3.08 8.20 1.95 4 20 LR Latvia 2.62 5.70 3.05 6 29 LT Lithuania 2.55 5.70 3.50 4 29 MEX Mexico 3.03 7.40 3.05 8 54 NL Netherlands 3.34 8.40 1.85 2 18 WAN Nigeria 3.11 3.40 3.15 12 69 N Norway 3.22 8.10 2.45 2 8 RP Philippines 3.32 7.70 3.30 7 46 PL Poland 2.99 6.40 3.25 4 29 P Portugal 2.84 7.80 2.80 2 16 RO Romania 2.63 4.50 3.55 8 50 SU Russia 2.51 5.00 3.50 7 55 SA South Africa 2.97 6.60 3.00 9 30 ROK South Korea 2.93 7.70 2.15 4 28 SLK Slovakia 2.51 6.20 2.75 7 55 SLO Slovenia 2.73 5.60 3.35 3 37 SP Spain 3.03 7.90 2.60 3 23 S Sweden 3.30 7.90 2.65 2 10 CH Switzerland 3.31 8.50 1.80 2 10 TR Turkey 3.21 6.30 3.00 8 73 UK Ukraine 2.99 3.1 3.90 8 42 US USA 3.30 8.90 1.90 2 12 UR Uruguay 2.99 7.2 2.90 4 25 58 Jan Ott

References

Ayal, E. B. and Karras, G. (1998) ‘Components of Economic Freedom and Growth’, Journal of Developing Areas, vol. 32, Spring 1998, pp. 327–38. Bay, C. (1970) The Structure of Freedom. Stanford University Press, Stanford, CA. Berlin, I. (1969) ‘Two concepts of liberty’ in Four Essays on Liberty. New York and Oxford: Oxford University Press. Freedom House (1999–2000) Press Freedom Survey. New York: Freedom House. Gwartney, J., Lawson, R. and Block, W. Economic Freedom of the World 1975–1995. The Fraser Institute. Johnson, B. T. and Sheeny, T. P. (1995) Index of Economic Freedom. Washington, DC: Heritage Foundation. Karatnycky, A., Cavanugh, C. and Finn, J. (1995 and 1998/1999) Freedom in the World: The Annual Survey of Political Rights and Civil Liberties. New York: Freedom House. Radcliff, B, ‘The Political Economy of Human Happiness’, this volume. United Nations Development Programme, New York: Oxford University Press. UNPD (1995, 1999) ‘Human Development Report’. Veenhoven, R. (1992) Happiness in nations. Subjective appreciation of life in 56 nations 1946–1992. Rotterdam: RISBO Rotterdam (see also World Database of Happiness). Veenhoven, R. (1999) ‘Freedom and happiness: a comparative study in 44 nations in the early 1990’s’, in E. Diener and E. Suh (eds), Subjective Well-being Across Cultures. Cambridge, MA: MIT Press. World Value Survey 2 (1995, 1996). World Value Survey 2 1990–92, Cumulative ﬁle ICPSR, ﬁle 6160, Ann Arbor, Michigan. Part II Challenges from Society: Changing Patterns of Political Involvement 5 Is There a Crisis of Democracy in Great Britain? Turnout at General Elections Reconsidered Ron Johnston and Charles Pattie

There is currently something of a crisis in democracy, we are told, with many countries experiencing an ‘erosion of confidence in the institutions of representative democracy’ (Norris, 1999, p. 257). One of the consequences of this growing cynicism, she suggests, is a decline in ‘conventional participation: discouraging electoral turnout, political activism, and civic engagement’. The extent of the decline in turnout can be exaggerated, however: some studies have identified no overall trend since 1945 in the ‘established democracies’ (p. 258), although an IDEA (1997) study did find a ‘modest dip’ averaging around 6 percentage points since the 1970s. The United Kingdom only partially fits this description. Average turnout at general elections fell from highs of over 82 per cent at the two elections of 1950 and 1951 to a range between 71 and 78 per cent thereafter, but with no evidence of a secular trend – although the lowest postwar turnout rate did occur in the last of the twentieth century’s general elections (71.2 per cent in 1997). Irrespective of whether the low turnout in 1997 indicates a crisis in British democracy, it is clearly of concern, especially when considered in conjunction with other low turnout rates at elections in the late 1990s: 23 per cent in the 1999 European Parliament elections; 29 per cent in the 1999 English local government elections; 40 per cent in the 1999 Welsh Assembly elections; although 59 per cent in the 1999 Scottish Parliament elections. In 1998 the House of Commons Home Affairs Committee investigated electoral law and administration, arguing that (Home Affairs Committee, 1998a, p. vi):

… participation rates in British elections are low and efforts should be made to increase them … [they] are either lowest or equal lowest among the larger West European democracies.

K. Dowding et al. (eds.), Challenges to Democracy © Political Studies Association 2001 62 Ron Johnston and Charles Pattie

Increasing turnout levels was important because

… we regard it as important for the health of the democratic and political system that participation in the electoral process be as high as possible … [their emphasis].

This paper explores why people vote at general elections, in the context of rational voter models, especially Downs’s (1957) classic work. His chapter on ‘The causes and effects of rational abstention’ set out several propositions, including:

1. When voting is costless, every citizen who is indifferent abstains and every citizen who has any preferences whatsoever votes; 2. If voting is costly, it is rational for some indifferent citizens to vote and for some citizens with preferences to abstain; … 4. The cost of information acts in effect to disenfranchise low-income groups relative to high-income groups when voting is costly; … and 6. It is sometimes rational for a citizen to vote even when his short-run costs exceed his short-run returns, because social responsibility produces a long-run return.

Evaluating these relies on measures of some of the important concepts embraced, such as the cost of voting. Downs notes (1957, p. 265) that

… every act takes time. In fact, time is the principal cost of voting: time to register, to discover what parties are running, to deliberate, to go to the polls, and to mark the ballot. Since time is a scarce resource, voting is inherently costly.

In deciding whether to commit time to voting, therefore, electors have to evaluate the benefits that might accrue relative to the costs. Downs (1957, p. 267) argues that this produces a ‘maze of conjectural variation’; for each elector ‘the importance of his own vote depends upon how important others think their votes are, which in turn depends on how important he thinks his vote is’. Individuals may determine that their vote has so little value (i.e. is so unlikely to affect the outcome) that voting is not worthwhile – in which case nobody turns out and democracy collapses. Alternatively, they might argue that if others reach that conclusion too, few will vote and those that do will have a significant impact on the outcome; in that case everybody who is not indifferent to the outcome will vote, and will then discover that their vote has little weight, and so will adopt the first position next time – and eventually democracy collapses! Turnout at General Elections Reconsidered 63

This paradox is only resolved because (Downs, 1957, pp. 267–8):

1. Rational men in a democracy are motivated to some extent by a sense of social responsibility relatively independent of their own short-run gains or losses. 2. If we view such responsibility as one part of the return from voting, it is possible that the cost of voting is outweighed by its returns for some but not all rational men.

In other words, some vote because they see it as a civic duty – or, as Dunleavy (1991, p. 85) puts it, a ‘highly valued resource … [that] simply goes to waste when unused’. They are most likely to value it if they believe that they will receive greater benefits when one party is in power rather than another, what Dunleavy (1991, p. 85) terms a ‘party identity’, which he distinguishes from the conventional political science concept of ‘party identification’. In consequence, people are more likely to vote when they see a clear differential between the parties and a close relationship between their desires and what one of the parties is offering – although evidence that the party is going to have a landslide victory ‘may convince some supporters that they no longer need to vote to be assured of victory’ (Dunleavy, 1991, p. 86: the same presumably could be argued regarding the supporters of those parties which appear destined to a landslide defeat). Turnout rates thus reflect a calculus incorporating both short- and long-run costs and benefits. The volume of the latter for each voter (Downs, 1957, p. 274)

… depends upon (1) the beneﬁts he gets from democracy, (2) how much he wants a particular party to win, (3) how close he thinks the election will be, and (4) how many other citizens he thinks will vote. These variables insure a relatively wide range of possible returns similar to the range of voting costs. Thus when citizens balance their costs and returns, some vote and others abstain.

Such costs are relative, however, so abstention rates should be higher among lower-income citizens who ‘have a harder time paying the cost of voting, [and so] it takes higher returns to get them to vote. And since they can less easily bear the cost of information, they have fewer data and are more uncertain; therefore they discount the returns from voting more heavily’. Downs (1957, p. 221) also addressed information acquisition with a number of propositions, including:

1. Society’s free information stream systematically provides some citizens with more politically useful information than it provides others; and 64 Ron Johnston and Charles Pattie

2. Certain specialists in the division of labor act automatically to reduce data costs drastically and to focus citizens’ attention on the areas most relevant to their political decision-making.

The latter includes a number of agencies – government and the governing party, other political parties, the media, interest groups and private citizens: the information that they transmit to voters may be either accidentally received or sought. If people vote because of a perceived difference between the parties in the likely returns from their election, then they need prior information with which to formulate those perceptions, which they either invest in (i.e. take time to obtain) or which they are provided with freely. Thus it is rational – an argument not developed by Downs – for political parties to focus provision of that free information on those voters whose support is most important to them and most winnable. This targeted dissemination is costly for the parties but effectively free for the voters, but the potential returns to the parties are high.

Testing these propositions empirically involves analyses at a number of scales. Downs’s propositions are phrased at the level of the individual voter operating in an, assumed, undivided electoral system, but the UK comprises 659 separate constituencies, each a separate milieu within which individuals activate their voting calculi. Thus when determining whether their vote will make any difference, electors may take into account both the national situation and that within their own constituency. It could be, as Butler suggested with regard to the low turnout in the 1997 general election (Home Affairs Committee, 1998b, p. 3), that:

… it was fairly obvious there was going to be a Labour victory, not on the scale there was but everyone was assuming, it did not seem an exciting election and people did not turn out.

Close-run elections, according to the information available to voters (from opinion polls published by the media and others on which parties base their campaign strategies and tactics), should attract higher turnouts than those where one party is expected to win by a substantial majority – hence an ‘explanation’ for why turnout was some seven percentage points higher in 1992 than 1997, though not why it was higher in 1987 than 1997.1 Turnout may be affected by the voter’s constituency situation rather than (or in addition to) the national position. Referring to local rather than national elections, Denver told the Home Affairs Committee (1998b, p. 215) that:

Politicians and others who beat their breasts over low turnout should ask themselves what exactly is the point of voting (other than out of a sense of Turnout at General Elections Reconsidered 65

duty) in a ward which is always won by the same party in an authority which is always controlled by the same party. Electors are not fools. They vote in satisfactory numbers when it is worthwhile doing so.

Nor are the parties fooled: they focus resources on providing information that will stimulate higher turnout where it is potentially most important to them:

My own research has shown that turnout is higher where candidates and parties campaign harder – but under the present system relatively lively campaigning is conﬁned to marginal areas.

Together, these arguments suggest that studies of turnout variations need investigations at:

1. the national scale, to enquire whether the turnout reﬂects perceptions of the closeness of the contest in addition to the voters’ own preferences and their local situations; 2. the constituency scale, to enquire whether (as Denver and Hands, 1974, 1985, demonstrated) turnout is higher where the potential of an individual’s vote mattering is greater, and where that potential is brought to voters’ attention by the parties, candidates and others; and 3. the individual scale, to identify those who vote because it is a perceived civic duty, whatever the electoral situation, plus those who vote because they identify with a particular party and believe that the outcome of an election could inﬂuence them personally.

The national pattern

Turnout at general elections since the Second World War saw a substantial increase between 1945 (an election held before the war ended) and 1950 (Figure 5.1). It dropped again after 1951, with an apparent secular decline until 1970 but the pattern thereafter suggesting trendless ﬂuctuation. Our analysis of the post-1955 period uses two sets of independent variables to represent the national and local expectations component of the rational voting model.2

1. Expectations of the national result based on the two main parties’ standing in the polls immediately before the election.3 The larger the difference between the percentages for those two parties (irrespective of sign), the greater the anticipation of a clear victory for one party, and so the lower the expected turnout. In 1997, for example, the polls indicated 46 per cent for Labour, 19 points more than the Conservatives, suggesting a low turnout, especially so in comparison with 1992 when the difference was only 0.4 of a percentage point. 66 Ron Johnston and Charles Pattie

Turnout 75

Predicted

70 Actual 1940 1950 1960 1970 1980 1990 2000

Election

Figure 5.1 Variations in turnout at general elections in Great Britain 1945–97, and predicted turnout level 1959–97.

2. Expectations of the constituency results, based on the proportion of constituencies which were marginal after the previous general election. (Because of boundary changes, we were unable to derive this ﬁgure for the 1970 general election to be applied in February 1974; estimates for the 1979 and 1992 elections if they had been held in the new rather than old constituencies are available for 1983 and 1997, however.4) The larger the percentage of constituencies that were marginal, the higher the expected turnout overall. There was substantial variation in the percentage of seats that were marginal across the ten elections: around 30 per cent of constituencies had a gap of less than 10 percentage points between the ﬁrst- and second-placed parties in the 1950s and 1960s; this jumped to approximately 40 per cent at the two 1974 elections, fell back to 21–23 per cent during the 1980s and increased to 27 per cent in 1992.

Only the first variable is significantly related to turnout. Marginality at the constituency scale was entered in three different ways (percentage of the constituencies that were very marginal – a difference between the first two parties at the preceding election of less than 5 percentage points; percentage of the constituencies that were marginal – a difference between the first two parties at the preceding election of less than 10 percentage points; and percentage of the constituencies that were very safe – a difference between the first two parties at the preceding election of more than 20 percentage points) but none Turnout at General Elections Reconsidered 67 was significant alongside the national poll difference variable.5 Our equation for predicting turnout during the period 1959–97 (including February 1974 as no measure of marginality at the constituency scale was significant) was

TURNOUT = 77.36 – 0.27POLLDIFF ±0.11 r2 = 0.39

Plotting the predicted values alongside the actuals shows a very good ﬁt for several elections, and three main residuals (Figure 5.1): 1959, for which the turnout was substantially under-predicted; and 1970 and October 1974, for both of which it was substantially over-predicted. It is possible to suggest reasons for each of these (voter fatigue at the second 1974 election, for example), but the clear conclusion is that the only general inﬂuence on turnout at the national scale was the expected closeness of the elections.

The constituency scale

Given both the trendless ﬂuctuation in national turnout rates over most of the period and the importance of expectations of the result on those rates, it could be concluded that there is no crisis of democracy in Great Britain, or at least no increasing crisis; nothing fundamental has changed for at least three decades. Whatever the short-term trends, however, the absolute level of turnout may be considered low, and thus indicative of considerable disenchantment with, if not a crisis in, British representative democracy at the national scale. Furthermore, as Figure 5.2 shows, there is considerable variation in turnout at the second of our scales – the constituency – posing questions regarding why people in some places are more likely to vote than their contemporaries elsewhere, and suggesting that, if there is a crisis, it is localized. Figure 5.2 shows the mean and the standard deviation for turnout rates across the constituencies at all elections since 1950. Again, there is no indication of any major shift over the period – the one-standard-deviation band around the mean is approximately the same width across all 14 elections. The variation is substantial, however, with more than one-third of the constituencies outside that band (e.g. having a turnout either above 77 or below 66 per cent in 1997). Why is this? The conventional wisdom has it that turnout is highest in the more marginal seats – thus we would expect a close relationship between the two variables at each of the 14 contests.6 We test this, ﬁrst, for the 1997 election, and then for the whole sequence.

The pattern of turnout in 1997 Turnout at the 1997 general election varied substantially across the 639 constituencies in Great Britain (i.e. excluding Northern Ireland plus West 68 Ron Johnston and Charles Pattie

Turnout 70 Mean + 1sd

Mean

60 Mean – 1sd 1950 1960 1970 1980 1990 2000 Election Figure 5.2 Variations in turnout by constituency at general elections, 1950–97: means and standard deviations.

Bromwich West – contested by the incumbent Speaker – and Tatton – where the incumbent Conservative was challenged by a non-party anti-sleaze candidate). The mean was 71.5 per cent, with a standard deviation of 5.5: the highest turnout was 82.2 and the lowest 51.9 (in Cardiff North and Liverpool Riverside respectively). But there was only a very weak relationship between turnout and margin (measured as the distance, in percentage of the votes cast at the previous election, between the ﬁrst- and second-placed parties then). There was the expected negative regression coefﬁcient (the greater the margin the lower the turnout), but with a correlation of only 0.24 just 0.06 (i.e. 0.242) of the variation in turnout could be accounted for by marginality (Model I in Table 5.1).

Table 5.1 Regressions of constituency variations in turnout, 1997

Model: I II III Margin: –0.10 –0.09 –0.06 Components: White collar 1.57 1.15 Deprivation 3.08 3.20 Rural 0.57 0.32 Pensioners –0.28 n.s. Energy industries 1.67 1.55 Total campaign spending 0.02 Scotland/Wales dummy 1.68 R2 0.06 0.64 0.68 Turnout at General Elections Reconsidered 69

This result provides very little evidence of voters reacting differently in their propensity to vote according to the local electoral situation. What of their individual circumstances, given Downs’ arguments regarding the cost of obtaining information, processing it and then taking time to vote, which he suggested may reflect their incomes? Unfortunately we have no data on incomes at the constituency scale, but we have taken 12 separate variables representing various aspects of each constituency’s population, and reduced these to five separate dimensions using principal components analysis (the results of which are given in full in Johnston, Rossiter and Pattie, 1997). The five factors were labelled as: 1. White collar; 2. Deprivation; 3. Rural; 4. Pensioners; and 5. Energy industries. Regression analysis showed all five significantly related to turnout, holding constant marginality (Model II in Table 5.1). Higher turnout levels were associated with: greater proportions of white-collar workers, lower deprivation levels, more rural areas, lower proportions of pensioners, and higher proportions of the workforce employed in the energy industries. These are consistent with Denver and Hands’ findings for previous elections: turnout was highest in 1997 in the more affluent parts of the country. Finally, what of the provision of information by parties? Denver and Hands have shown that the more parties spend on their constituency campaigns (as an index of campaign intensity), the greater the turnout. The amount that can be spent varies by election and constituency so, following our standard practice (e.g. Pattie, Johnston and Fieldhouse, 1995), we have expressed spending by each party as a percentage of the constituency maximum. The percentages for Conservative, Labour and Liberal Democrat parties, plus the nationalist parties in Scotland and Wales, were summed to give an index of total campaign intensity, which is added to the other variables – it forms Model III (because the total spent can be higher in Scotland and Wales, we added a dummy variable for constituencies there). The average expenditure in English constituencies (where the maximum was 300) ranged from 56 to 294, with a mean of 185 and a standard deviation of 48: for Great Britain as a whole, the mean and standard deviation were little different at 189 and 53 (although the maximum possible was 400 and the maximum recorded was 373). The Model III results show that the more the parties spent, the greater the percentage of the electorate who voted (Table 5.1), although the additional ‘explanation’ provided is slight – as shown by the R2 figures.7 The overall message of the regression results in Table 5.1 is that although turnout was higher in 1997 where (a) there was a close contest (the margin of victory in 1992 was small) and (b) the parties campaigned hardest, together these elements of the rational voter model account for very little of the variation across the 639 constituencies. By far the most important determinants of the proportion who voted were the components representing the constituen- 70 Ron Johnston and Charles Pattie cies’ population characteristics and thus, assuming that the ecological fallacy is unlikely to be involved, the characteristics of the voters themselves. This suggests variations between people and places in the proportion who have clear preferences between the parties that they wish to express whatever the local electoral context; and/or see voting as a civic duty; or, in Downs’ terms, are prepared to meet the high costs of voting in the short-term because of the long-run benefits that democracy brings.

Turnout variations at the constituency scale, 1959–97 A problem involved in applying this model to a sequence of elections is that it is not easy to compile a comparable data set for the population characteristic variables over a sequence of elections – because of not only the need to get data from separate censuses (far from easy for the 1950s and 1960s) but also differences in the variables available. An applicable surrogate that can be applied across all elections is the party which won the constituency at the previous election to that being analysed. In 1997, there were signiﬁcant differences between constituencies according to victor in 1992 – turnout was 74.4 per cent in Conservative-held seats, 67.5 per cent in Labour-held, and 72.4 per cent in seats held by either the Liberal Democrats or one of the two nationalist parties. Conservative-held seats were also more white-collar than those held by Labour, for example, had lower deprivation levels, were more rural, had fewer pensioners and recorded substantially higher turnouts. The model was again ﬁtted in three stages (using ANOVA) – I, margin only; II, margin plus winner at previous election; III, margin plus winner at previous election plus total campaign expenditure. The R2 values are reported in Table 5.2 for each general election since 1955. Marginality is only weakly related to

Table 5.2 R2 values associated with ANOVAs of turnout variations across constituencies at the three stages of the analysis

Model Election I II III 1959 0.07 0.09 0.18 1964 0.06 0.12 0.20 1966 0.12 0.19 0.31 1970 0.21 0.25 0.34 1974 Oct. 0.11 0.20 0.30 1979 0.14 0.18 0.24 1983 0.08 0.19 0.27 1987 0.03 0.19 0.26 1992 0.03 0.31 0.41 1997 0.06 0.42 0.49

For key to model stages, see text. Turnout at General Elections Reconsidered 71 turnout – it accounted for more than 10 per cent of the variation at just four of the elections, between 1966 and 1979 inclusive, and was virtually unrelated to it in 1987 and 1992. Which party held the seat was increasingly associated with turnout variations after 1970, however, and especially after 1979. Finally, variations in campaign intensity also substantially increased the model’s performance – accounting for 6–12 per cent of the total on each occasion. At the constituency scale, therefore, there was little evidence to confirm the ‘conventional wisdom’ that turnout was highest in the more marginal constituencies. Instead there was a growing divide between Conservative- and Labour-held constituencies, with the amount of information provided to the voters, as measured by our campaign intensity index, also contributing significantly to the explanation. Voters vote where they are encouraged to do so by the parties, but more so where one of the main parties is the stronger of the two. This is shown in Figure 5.3 for the 1997 election in England only (excluding the small number of seats held by the Liberal Democrats): the best- fit, moving-average, line for all constituencies indicates a positive relationship between total expenditure and turnout, which was steepest up to a threshold of c. 200 per cent (i.e. each of the three parties spent on average about two- thirds of the maximum allowed), and a very clear difference between Conservative- and Labour-held seats of about 10 percentage points. This growing difference in turnout between Conservative- and Labour-held seats suggests either clear differences between socio-economic groups in their propensity to vote or an increasing difficulty for Labour in convincing voters

65 Turnout 1992 Winner

60 Labour

55 Conservative

50 All constituencies 0 100 200 300 Total Expenditure

Figure 5.3 Variations in turnout by constituency in England at the 1997 general election according to the total amount spent on the local campaign and which party won the seat in 1992. 72 Ron Johnston and Charles Pattie of the necessity of turning out in its strongholds, or both. Regarding the former, it may be that less afﬂuent individuals, who have been more likely to support Labour over recent decades, have become more disillusioned with the political situation, and less stimulated to vote as a consequence – for any party. Regarding the mobilization problems, the last two decades of the twentieth century saw substantial erosion of Labour’s organizational strengths – in the trade unions, for example – which were important vehicles for ‘getting out the vote’. Their joint impact is shown in the (most of them now former) coal- mining constituencies, where turnout was traditionally high despite them being very safe seats, but where levels are now much closer to the average.8 There was less incentive for Labour supporters to turn out by the end of the century, it seems, and less peer pressure – at work and in the local community – to do so.

Turnout at the individual scale

We turn ﬁnally to individual-scale data to appreciate why some people vote and others do not, in the context of both Downs’ argument that certain groups within society are more likely than others to treat voting as a civic duty and our ﬁndings above that constituency population characteristics are the main determinant of turnout variations at that scale. We start with an analysis of the 1997 British Election Study survey.9

Civic responsibility and turnout at the 1997 general election The 1997 BES asked a direct question about civic responsibility and voting:

Thinking now about voting in general elections. Generally speaking, do you think people need not vote unless they really care who wins or, is it everyone’s duty to vote?

Some 79 per cent said that voting is a civic duty, with 20 per cent saying that it is only worth voting if you care who wins; 86 per cent of the former voted in 1997, compared with 51 per cent of the latter. One key to why some of those who think that voting is a civic duty nevertheless abstained comes from the reasons they gave for not voting. The 1997 BES question allowed several answers: most offered only one, and we concentrate on those for grouping abstainers into two types: voluntary abstainers, for whom not voting was a positive decision; and involuntary abstainers, who were unable to vote despite wishing to (Johnston and Pattie, 1997; Pattie and Johnston, 1998a, 1998b). Nearly two-thirds of those who failed to vote were involuntary abstainers (Table 5.3), with a major difference between the two categories: less than one-third of the voluntary abstainers saw voting as a civic duty, whereas the majority of the involuntary abstainers did so. Turnout at General Elections Reconsidered 73

Table 5.3 Attitudes to voting by types of abstainer, 1997 (percentages of row totals; don’t knows excluded)

Reason for not voting: Care Duty N Voluntary abstainers Deliberately abstained 75.1 21.4 33 Couldn’t be bothered/not interested 63.0 29.3 76 Couldn’t decide between the parties 62.4 30.6 45 Not affected by who won 61.8 29.7 18 Never vote 86.5 13.5 12 Total 70.9 29.1 184 Involuntary abstainers Work prevented me 28.2 71.1 47 Sickness prevented me 24.9 68.5 32 Away on election day 30.4 67.1 95 Other commitments/no time 46.0 47.9 84 Polling card/station problems 39.3 59.3 67 Total 36.6 63.4 325

Key to columns: Care – those who think that you need only vote when you care who wins; Duty – those who see voting as a civic duty.

Most of those who see voting as a civic duty turned out in 1997, therefore, whereas most of those among them who failed to cast a ballot were prevented by (largely) unavoidable circumstances.10 Of those who take a pragmatic approach, only voting when it seems worthwhile, many fewer voted in 1997, and of those among them who abstained a majority did so deliberately. These findings are in line with Downs’ propositions. He also argued that those who see voting as a civic duty are more likely to have preferences among the parties, and thus are more likely to vote.11 The data sustain the first proposition; among the pragmatists, 17.6 per cent had no party identification, compared to only 3.9 per cent of the committed democrats (those for whom voting was a civic duty). Furthermore, among those with a party identification, many fewer of the pragmatists (7.3 per cent) than the committed democrats (17.6 per cent) were strongly identified with a party. Downs also suggested class differentials in the percentages of committed democrats and pragmatists, but this proposition is not confirmed by BES data. Differences between the classes – according to occupation, income, educa- tional qualifications and housing tenure – are generally slight.12 One of the few substantial exceptions is the large percentage of pragmatists among those who have never had a job, which undoubtedly relates to the only demographic variable for which there is a clear pattern – age: the older the respon- dent the greater the probability that he or she is a committed democrat. If, then, there is a crisis in representative democracy in Great Britain it appears to rest – as many have argued – with the younger generations, high proportions 74 Ron Johnston and Charles Pattie of whom, especially among the males, are pragmatists rather than committed democrats. But evidence from earlier BES surveys provide no indication that the current youngest generation are different from their predecessors: young people were less likely than their older counterparts to see voting as a civic duty at every election since 1964, suggesting that age brings with it a greater commitment to the political system.13 Party identification is a good identifier of not only which party people vote for but also whether they abstain.14 Less than half of those with no party identification voted, compared to 88 per cent among the very strong identifiers, 86 per cent among the fairly strong identifiers, and 72 per cent among those who classified themselves as not very strong identifiers (Table 5.4). Voluntary abstainers are more numerous than involuntary abstainers among those with no party identification, and vice versa. Pragmatists are less likely to identify with a party, therefore are less likely to vote, and if they abstain are more likely to do so as a deliberate act. Table 5.4 also shows inter-party differences. Of those who did not identify very strongly with a party, turnout was much lower among Labour identifiers. This could be associated with the ‘national marginality’ hypothesis discussed above. Labour was widely expected to win, and to win well, in 1997, generating much less pressure to vote on those who supported it only weakly than was the case with Conservative supporters – every vote counted for the latter

Table 5.4 Voting/abstaining by party identiﬁcation, 1997 (percentages of row totals)

Abstained Voted Vol. Invol. Other No party identification 47 32 15 6 Very strong identification 88 4 6 1 Fairly strong identification 86 4 7 3 Not very strong identification 72 10 13 5 Conservative identifiers Very strong identification 90 2 7 1 Fairly strong identification 89 2 6 3 Not very strong identification 75 7 13 5 Labour identifiers Very strong identification 88 5 7 1 Fairly strong identification 85 4 8 3 Not very strong identification 65 15 15 5 Liberal Democrat identifiers Very strong identification 98 0 0 2 Fairly strong identification 84 9 5 3 Not very strong identification 82 4 7 6

Vol. – Voluntary; Invol. – Involuntary. Turnout at General Elections Reconsidered 75 party in trying to stave off defeat (as it did for Liberal Democrats seeking to increase their Parliamentary representation).15 In addition, the pressure on Labour supporters to turn out was especially weak in many of its ‘safe seats’, where victory was expected and campaigning was far from intense. Was constituency marginality important therefore? Constituencies have been placed into eight categories according to their marginality for each party after the 1992 general election: a positive margin indicates that the party held the seat then (Table 5.5). Among Conservative identifiers, abstention rates were lowest in their party’s safest seats: they were higher in the more marginal seats, and highest where the party came more than 20 percentage points behind the victor in 1992 – the party faithful turned out in greatest numbers where victory was most assured. Subdivision according to strength of identification shows very low abstention rates save in the final category (where there was no hope of victory) in each group but little overall pattern otherwise. There was a very clear pattern among Labour identifiers: turnout was highest in the more marginal seats. This U-shaped relationship held for all

Table 5.5 Abstention (percentages who did not vote) by party identiﬁcation, attitudes to voting and constituency marginality, 1997

Margin: Held by Party Not held by Party +20 +10 +5 +0 –0 –5 –10 –20 Conservative identifiers All 15 12 18 22 19 16 18 31 Very strong 4 12 * * * * 14 33 Fairly strong 11 7 16 28 18 6 11 27 Not very strong 24 17 29 21 21 33 32 36 Labour identifiers All 24 17 21 9 15 22 22 19 Very strong 14 13 8 7 * 7 7 14 Fairly strong 19 15 12 7 8 9 15 14 Not very strong 42 26 36 13 * 42 42 30 No party identification Conservative margin 55 46 29 62 33 65 62 54 Labour margin 50 60 65 57 39 26 50 54 Liberal Democrat margin * * * 18 * * 50 53 Committed democrats Conservative identifiers 11 9 16 20 15 4 17 29 Labour identifiers 16 14 14 2 8 18 18 14 Pragmatists Conservative margin 44 44 38 52 46 60 38 62 Labour margin 64 38 60 39 45 41 49 44

*Number of respondents less than 20. 76 Ron Johnston and Charles Pattie groups of Labour identifiers, though the weaker the level of identification with the party the larger the proportion of abstainers – so much so that in Labour’s safest seats over 40 per cent of its not very strong identifiers failed to vote, presumably because there was little stimulus to do so given the party’s opinion poll lead and its relative lack of campaigning in such constituencies (Johnston, Pattie and MacAllister, 1999). Labour got the votes out where they were most needed.16 Similar differences in abstentions are shown by the ‘committed democrats’ and ‘pragmatists’ (Table 5.5). Among the former who are also Conservative identifiers, abstention was highest in the marginal constituencies (majorities of 5 percentage points or less) and in those where the Conservatives trailed the 1992 victor by at least 20 points. For Labour identifiers, however, the U-shaped curve is again very apparent, with very low abstention rates where turnout really mattered – in the marginal constituencies. Among the pragmatists, using Conservative margin there is slight evidence that abstention rates were lowest in the Conservative-held seats (suggesting that these voters were most prepared to turn out where they might help to remove a candidate of the incumbent government). The pattern with Labour margin is, not surprisingly, to a large extent the reverse of that: abstention rates were lowest where Labour was the challenger rather than the incumbent.

The longer view The implication from the above analyses of the 1997 result is that variations in turnout between general elections can be accounted for by one or more of:

1. differences in the proportion of electors who are either ‘committed democrats’ or ‘pragmatists’; 2. differences in the proportion of electors who are voluntary and involuntary abstainers; 3. differences in the proportion of electors who identify with one of the political parties; 4. differences in the proportion of those who identify with a party in the strength of that identiﬁcation; and 5. differences in the marginality of individual constituencies.

Not all of these are readily tested: we look at the third and fourth only, using BES data for 1964–97. The data confirm what has been generally appreciated – the proportion of the electorate who identified with no political party increased substantially, from 4.4 per cent in 1964 to 11.9 per cent in 1997. Abstention among them Turnout at General Elections Reconsidered 77 averaged 31 per cent; it was low at the two 1960s elections (28.3 per cent and 27.7 per cent) and in the 1980s, and highest in 1992 (39.1 per cent), however, so changes in the proportion of ‘non-identifiers’ were not related to variations in turnout since 1964. Among identifiers, abstention was always higher the weaker that identification. Regarding differential abstention rates according to both party identified with and strength of party identification, the data give little hint as to the reasons for the cross-election variations: Labour’s relative difficulties in turning out its not very strong identifiers in 1997 were, however, characteristic of some other elections – 1964, 1966, October 1974 and 1983 – the last two of which were low-turnout elections overall.

Conclusions

The low levels of turnout at British general elections may be indicative of a crisis of democracy, but our analyses suggest that the crisis did not deepen over the last four decades of the twentieth century: from 1959 on there was a trendless fluctuation around an average turnout of 75 per cent. Turnout was higher at elections which the voters thought would produce a close outcome between the two main parties, hence the lowest turnout in 1997 when the electorate expected a substantial victory for Labour. There was virtually no evidence that this rationality extended to the constituency scale, with very low correlations between marginality (i.e. the expected closeness of the outcome) and turnout. Instead, we found that turnout was generally higher in Conservative-held seats (implying that Labour had increasing difficulties stimulating turnout among its supporters). Indeed, the strongest relationships were with constituency types and with campaign intensity. Turnout was generally higher where the parties campaigned hardest – the latter providing further evidence in support of Downs’ propositions regarding the role of free information as a stimulant to turnout. The differences between constituencies according to which party won there suggests differential turnout among various groups within the electorate. Analysis of the 1997 British Election Study survey data did not show this, however, with one exception: turnout rates were much lower among young voters – which suggests that the ‘crisis of democracy’ may be no more than a generational effect. Other analyses of those data indicated differences in turnout according to people’s attitudes to voting, reflected in their identification with parties and the strength of that identification: together indicative of beliefs that they would benefit from the relevant party’s electoral success. 78 Ron Johnston and Charles Pattie

Notes

1. Although it was generally believed at the time that the 1987 result would be a certain Conservative victory, as polling results show, there were moments during the campaign when this certainty was queried (termed ‘wobbly Thursday’ by the Conservatives), which may have encouraged a slightly higher turnout than would otherwise have been the case. 2. We lacked needed data to conduct the analyses before 1955. 3. The data were obtained from Butler and Butler (1995) except for 1997; these were taken from the Gallup Political INDEX. 4. The 1979 estimates are derived from BBC/ITN (1983), and for 1992 using the methodology described in Rossiter, Johnston and Pattie (1997). 5. This is not strictly true. When both the first and the third constituency marginality variables were included along with the poll difference variable, all three were significant with an R2 of 0.69 (adjusted to 0.534 for degrees of freedom: there are only ten observations as we have no constituency marginality data for February 1994). Other variables (the length of the preceding Parliament, the outdatedness of the electoral register, a dummy variable for the extension of the franchise to those aged 18–21) were all tried and found to be insignificant. 6. Thus ‘the marginality or safeness of a constituency or ward for the various parties importantly influences turnout in elections’ and ‘the contextual factor which most strongly influences election turnout – constituency marginality’ (Denver et al., 2000, pp. 129 and 130). 7. As a check on the impact of higher potential spending levels in Scotland and Wales, where turnout was on average 1.68 points higher according to Model III (Table 5.1), the regressions have been rerun for England only. The results were not substantially different from those for Great Britain. 8. Denver and Hands (1974, 1985) show that some of the largest positive residuals from their regressions of turnout and marginality were in coal-mining constituencies. 9. The BES was conducted by CREST, based at Nuffield College, Oxford: the data have been obtained from the Data Archive at the University of Essex. 10. Interestingly, the smallest percentage giving civic duty as a reason for voting among the involuntary abstainers were those who said that they failed to vote because of ‘other commitments/no time’, suggesting that their commitment to the civic duty was less than those who gave other answers. 11. Similarly, Dunleavy (1991) argued that those who see benefits from voting will have a ‘party identity’; he gave no indication how this might be measured, hence we assume here that the conventional, party identification measure provides a useful first approximation. 12. Robertson (1984, p. 101) did show for the 1970s that those in the middle class with a ‘deviant ideology’ were more likely to abstain, suggesting the need for finer- grained analyses of socio-economic differentials. (See also Johnston et al., 2000.) 13. We have not included the data supporting this case here for reasons of space. Note that Park (1999), using data from the British Social Attitudes surveys, has found some evidence of a cohort effect over recent years – but only time can tell if this represents a permanent distancing from the political system by those socialized in the 1980s and 1990s. 14. Dunleavy (1991, p. 85) makes clear that ‘party identification’ as conventionally measured is not the same as his concept of ‘party identity’. Nevertheless, if people Turnout at General Elections Reconsidered 79

identify with a party in the way suggested by the conventional questions – in the BES data analysed here, ‘Generally speaking, do you [usually] think of yourself as a Conservative, Labour, Liberal or what?’, with the word usually used until 1970: Crewe, Day and Fox (1991, pp. 471–2) – they are likely to believe that they are more likely to benefit from that party’s policies than from those promoted by its competitors, especially if they also indicate that they care about the election outcome, which presumably implies that they think it will make a difference to them. 15. It could be argued that the expectation of defeat would lead many Conservative supporters to abstain as well. Conventional wisdom has it that the Conservative party has always been better than Labour at mobilizing its supporters, however, and in 1997 it could well have been that – apart from the seats where the party had no hope of victory (and where its vote held up best!) – Conservative supporters hoped that if they turned out in force they would at least stem the flow to Labour. 16. These findings are confirmed by logistic regression analysis. Among all respondents, there is a negative relationship between the probability of abstaining and age; those with a party identification are less likely to abstain than those with no identification; and abstention is significantly higher in the seats where the Conservatives were more than 20 percentage points behind the winner in 1992. Analysis of those with a party identification only has the same relationship with age and ‘hopeless’ Tory seats; there are significant differences according to strength of identification (the weaker it is, the greater the chance of abstaining), but no inter- party differences. For Conservative identifiers only, there are significant relationships with age and strength of identification, plus the difference between ‘hopeless’ and ‘very safe’ seats. But for Labour identifiers, as well as the age and strength of identification relationships, there is also a significantly lower rate of abstentions in the marginal than in the safe constituencies.

References

BBC/ITN (1983) The BBC/ITN Guide to the New Parliamentary Constituencies. Chichester: Parliamentary Research Services. Butler, D. and Butler, G. (1995) British Political Facts 1900–1994. London: Macmillan. Crewe, I., Day, N. and Fox, A. (1991) The British Electorate 1963–1987: A Compendium of Data from the British Election Studies. Cambridge: Cambridge University Press. Denver, D. and Hands, G. (1974) ‘Marginality and Turnout in British General Elections’, British Journal of Political Science, 4, 17–35. Denver, D. and Hands, G. (1985) ‘Marginality and Turnout in General Elections in the 1970s’, British Journal of Political Science, 15, 381–8. Denver, D. and Hands, G. (1997) Modern Constituency Electioneering: the 1997 General Election. London: Frank Cass. Denver, D., Mitchell, J., Pattie, C. and Bochel, H. (2000) Scotland Decides: The Devolution Issue and the Scottish Referendum. London: Frank Cass. Downs, A. (1957) An Economic Theory of Democracy. New York: Harper & Row. Dunleavy, P. (1991) Democracy, Bureaucracy and Public Choice. Hemel Hempstead: Harvester Wheatsheaf. Home Affairs Committee (1998a) Fourth Report: Electoral Law and Administration. Volume I Report and Proceedings of the Committee, Cm. 768–I. London: Stationery Ofﬁce. Home Affairs Committee (1998b) Fourth Report: Electoral Law and Administration. Volume II Minutes of Evidence and Appendices, Cm. 768–II. London: Stationery Ofﬁce. 80 Ron Johnston and Charles Pattie

IDEA (1997) Voter Turnout from 1945 to 1997: a Global Report, 2nd edn. Stockholm: IDEA. Johnston, R. J. and Pattie, C. J. (1997) ‘Towards an Understanding of Turnout at British General Elections: Voluntary and Involuntary Abstention in 1992’, Parliamentary Affairs, 49, 281–91. Johnston, R. J., Pattie, C. J., Dorling, D. F. L., MacAllister, I., Tunstall, H. and Rossiter, D. J. (2000) ‘The Neighbourhood Effect and Voting in England and Wales: Real or Imagined?’, in P. J. Cowley, D. Denver and J. Fisher (eds), British Elections and Parties Review 10. London: Frank Cass. Johnston, R. J., Pattie, C. J. and MacAllister, I. (1999) ‘The Funding of Constituency Party General Election Campaigns in Great Britain’, Environment and Planning C: Government and Policy, 17, 391–410. Johnston, R. J., Rossiter, D. J. and Pattie, C. J. (1997) Three Classiﬁcations of Great Britain’s New Parliamentary Constituencies, Essex Papers in Politics and Government No. 117. Colchester: University of Essex. Norris, P. (1999) ‘Conclusions: the Growth of Critical Citizens and its Consequences’, in P. Norris (ed.), Critical Citizens: Global Support for Democratic Governance. Oxford: Oxford University Press, pp. 257–72. Park, A. (1999) ‘Young People and Political Apathy’, in R. Jowell, J. Curtice, A. Park and K. Thomson (eds), British Social Attitudes, the 16th Report: Who Shares New Labour Values? Aldershot: Ashgate, pp. 23–44. Pattie, C. J. and Johnston, R. J. (1998a) ‘Voter Turnout at the British General Election of 1992: Rational Choice, Social Standing or Political Efﬁcacy?’, European Journal of Political Research, 33, 263–83. Pattie, C. J. and Johnston, R. J. (1998b) ‘Voter Turnout and Constituency Marginality: Geography and Rational Choice’, Area, 30, 38–48. Pattie, C. J., Johnston, R. J. and Fieldhouse, E. (1995) ‘Winning the Local Vote: the Effectiveness of Constituency Campaign Spending in Great Britain, 1983–1992’, American Political Science Review, 89, 969–83. Robertson, D. (1984) Class and the British Electorate. Oxford: Basil Blackwell. Rossiter, D. J., Johnston, R. J. and Pattie, C. J. (1997) ‘Estimating the Partisan Impact of Redistricting in Britain’, British Journal of Political Science, 27, 319–31. 6 Social Capital, Civic Literacy and Political Participation: Explaining Differences in Voter Turnout Henry Milner

Introduction

In his litany of the symptoms of shrinking social capital in America, Robert Putnam (1995) gives some prominence to the steady decline in voter turnout in national elections. Turnout is also one of the components of his ‘civic community index’, used to distinguish high from low social capital regions in Italy (1993). And yet Putnam expends little analytical effort on factors affecting voter turnout, emphasizing membership in voluntary organizations as indicators of civic engagement, something qualitatively different from voting. And this approach has largely been followed in the current burgeoning social capital literature. In my view, the result has been that the discussion of civic engagement gives inadequate attention to voting, the most basic form of political participation in democratic societies. No activity other than turning out to vote serves the objective of comparing long-term and broad trends of political participation. At a deeper level, moreover, in a globalizing, pluralistic world, it becomes inconceivable that civic engagement be dissociated from the most basic expression of democratic citizenship, the exercise of the franchise. Voting is a visible, readily identifiable and quantifiable act, for which we have objective as well as subjective measures. And – apart from that very small minority of anarchist-oriented activists who consciously reject all forms of electoral politics – voting is the sine qua non of political participation. Non- voters very seldom take part in more active forms of politics.1 Moreover, unless built into the institutional structure so as to be effectively passive (as with certain interest organizations in which members are automatically affiliated to a political party), more active forms of participation are, as a rule, limited to a minority of the population.

K. Dowding et al. (eds.), Challenges to Democracy © Political Studies Association 2001 82 Henry Milner

If civic engagement is not by its very definition to be restricted to a minority of the population, political participation must be something open to and, under normal circumstances, practised by a solid, consistent majority of the population. There remains only one candidate for an indicator of such participation, namely voter turnout. And, of course, turnout is an objective indicator than can be operationalized for the purposes of international comparison. Such comparison is fraught with difficulty with regard to the other, standard measures of levels of social capital: participation in voluntary organizations and interpersonal trust. Nevertheless, voter turnout is not a simple phenomenon. Though citizens in democratic countries vote to select representatives to a variety of legislative and executive offices, comparative studies of voter turnout typically consider only participation in elections for national legislatures. Thus a rich source of comparative data – especially if we are interested in voting as an expression of civic engagement – remains relatively unexplored. In limiting itself to elections to national office, comparative analysis investigates only political participation in relation to positions which, by definition, are relatively remote from people’s immediate experience. And while not all democratic countries elect intermediate or regional assemblies, elected local or municipal councils are effectively universal. In our search for a comparative indicator of civic engagement, we can do no better than focus on turnout in local elections. In this paper I present certain findings from my current research (Milner, 2001) in an effort to bring this missing dimension of civic engagement to the literature on social capital. Note that the data below compares national aggregates, though individual-level differences drawn from largely American data are brought to bear to better explain variations at the aggregate level.

The costs of non-voting

Let us step back for a moment from explaining why some people vote and others do not to ask why it matters. In recent years, some political scientists inﬂuenced by – or critical of (e.g. Green and Shapiro, 1994) – the standard economic model of behaviour have taken it to imply that there is something irrational or ‘paradoxical’ about the ‘effort’ put into voting. Yet, this ‘economic model of democracy’ turns out to be a straw man, the result of a faulty application of the model of analysis used by economists. Economists are not interested in the ‘effort’ people put into going to buy or sell, i.e. to participate in the market: this is simply taken for granted (Pellikaan, 1998). What is of interest is the effort they make to inform themselves so that they can make efﬁcient choices. It was in relation to his discussion of the costs of informing oneself that the father of the economic model of democracy, economist Social Capital, Civic Literacy and Political Participation 83

Anthony Downs, raised the question of ‘rational ignorance’ in relation to voting. Downs stressed the contradiction between, on the one hand, rational citizens wanting democracy to work well so as to gain its benefits and knowing it to work best when citizens are well-informed, and on the other hand (their knowing) the fact that the benefits of informing themselves are outweighed by the costs they must bear. According to Downs, what differentiates the ‘political market’ from product markets is its effect on the decision to invest in relevant information, since the existence of other voters reduces the incentive to pay the costs of that information. If producers and consumers in product markets based their choices on insufficient information, the market would never reach an efficient equilib- rium. But this is not normally the case, since individuals gain from investing in relevant information. It becomes the case when it comes to elections, since voters base their decision on incomplete information. This is what Downs referred to as the ‘voters’ paradox’. A similar distinction is made by Mancur Olson. Olson points to ‘knowledge of the real consequences of different public policies’ as a non-marketable form of human capital. Unlike the acquisition of marketable skills, such knowledge is a public good but cannot improve the individual’s relative position (Olson, 1996, p. 16). This is where social capital comes in. To the extent, then, that the economists’ framework is relevant to the overall discussion, it is in relation not to the choice over whether to vote, but of whether to acquire the requisite political information. When it comes to the acquisition of such information it is, according to Fishkin, the existence of social capital that resolves the voters’ paradox.

Citizens in a civic community, one with high social capital, have many reasons to participate in politics together and to stay informed. They are part of a dense network of civic associations, both political and nonpoliti- cal, which provides them lots of reasons to read newspapers, to stay informed, to participate in community activities. They internalize norms that motivate them to participate and to join with others – norms that give them satisfaction regardless of any calculation about the effects of their individual actions. And, as Tocqueville noted, these habits of association, the widespread acceptance of working together, make it far easier for each individual to participate, to combine with others for some cause of mutual interest. (Fishkin, 1995, p. 148).

Unfortunately, this idealized picture of rational political participation in a high social capital community doesn’t correspond to the United States of the 84 Henry Milner high social capital early 1960s, let alone contemporary America. The massive empirical data about the political knowledge of the American voter confirms Downs’ expectation of ‘rational ignorance’. A famous essay of that earlier period (Converse, 1964) concluded that the average American citizen lacked a stable orientation toward politics, and exhibited a low level of knowledge, lacked consistency between attitudes, showed attitude instability and provided vacuous answers to open-ended questions. Since that time, American students of politics have systematically retested his assertions, gauging citizens’ knowledge of American political institutions, processes and leaders, with the result that, as Bartels (1996, p. 184) put it: ‘The political ignorance of the American voter is one of the best documented data in political science.’ But there is much disagreement as to what, if anything, needs to be done. Fiorina, for one, simply takes rational ignorance as inevitable, indeed trivial. He argues that well-informed voters do not gather information in order to cast an informed vote; rather, they are able to cast an informed vote because they have accumulated information for other unrelated purposes (Fiorina, 1990, pp. 336–41). But what of the ‘paradox’ of rationally ignorant citizens losing out on the gains from an informed electorate? Here the cause is taken up by a number of American scholars (e.g. Lupia, 1994; Sniderman, Brody and Tetlock, 1991), who maintain that the voters’ paradox does not have the ill- effects attributed to it since politically uninformed people use heuristics by following social cues and rules of thumb. These, they argue, allow them to arrive at decisions – including the decision not to vote – resulting in overall outcomes similar to those that would have been attained through the participation of informed people. Consequently, the costs of rational ignorance are in fact negligible. Yet, though appealing, this dispensing with the concerns about suboptimal outcomes raised by economists Downs and Olson begs a key question. We need to ask if more knowledgeable voters vote differently from uninformed ones, for, if so, a more knowledgeable electorate would mean that the (larger number of) voters brought to the polls would affect outcomes. These are matters of controversy among political scientists, especially in the United States, given some of the findings emerging from recent research. For example, Dimock (1998) found that while partisanship in the US – as elsewhere – has been steadily declining over 40 years, it has occurred primarily among voters lacking knowledge about politics. And a detailed study by Delli Carpini and Keeter (1995) showed that the quartile of their sample who were most ignorant of political actors, institutions and procedures were only very slightly more likely than not to voice support for a candidate consistent with their expressed views on a given issue. Yet such trends do not in themselves prove that the absence of non-voters from the voting pool affects government policies. Some research shows that Social Capital, Civic Literacy and Political Participation 85 non-voters’ views, when tested, do not differ substantially from those of voters (Teixeira 1992, pp. 96–7). Lijphart, (1997) challenges this interpretation by raising an important objection. Findings of this kind based on polling data, he says, are suspect because the non-voters polled have not given the issues much thought, which they would have done had they been mobilized to vote. Because of this, asserts Lijphart, their votes could be expected to be different enough to affect outcomes. A number of studies provide support for this con- tention.2 Bartels (1996) and Delli Carpini and Keeter (1996), among others, suggest that in a significant number of cases, uninformed people in the United States would make different political judgements if they were well informed.3 In the end, individual-level analysis cannot resolve this issue, since we ultimately cannot know what uninformed non-voters would do if they became informed voters, and thus cannot prove that their absence translates into suboptimal political decisions. It is at the wider comparative level of analysis that this debate can most usefully be engaged. Lijphart (1997) marshals data from comparative studies showing a strong link between turnout and support for left-of-centre parties, citing, in particular, the findings of Pacek and Radcliff (1995). This study, which analysed all national elections in 19 industrial democracies from 1950 to 1990, showed that the left’s share of the total vote increased by almost one-third of a percentage point for every percentage point of added turnout. In light of such findings, it is unlikely – comparatively speaking – that the costs and benefits of low turnout will be equitably distributed. It is fair to conclude that turnout matters, with regard both to the quality of democracy and policy outcomes. Moreover, the individual-level data shows a discernible relationship between voting and non-voting and the level of political knowledge. I contend, therefore, that we can only assume that a community’s stock of social capital is reflected by its level of political participation if our conceptualization of social capital incorporates the political knowledge dimension. Simply put: the primary link between social capital and political participation is political knowledge. To elaborate this we return to the discussion of political participation, and voting in particular, as an indicator of social capital, looking first at the relationship between voter turnout and the standard indicators of social capital.

Interpersonal trust, associational density and voter turnout

We begin with comparative data on turnout in national elections. We now have reasonably reliable standardized turnout figures for elections in most countries, compiled by reputable international organizations such as IDEA, the Stockholm-based Institute for Democracy and Electoral Assistance. Official 86 Henry Milner voter turnout data of this kind frees us from relying on subjective self-reported assessments. We know that not only do expectations based on cultural norms affect responses to such commonly used tests of political involvement as ‘do you discuss politics at home?’ but also, ‘did you vote in the last election?’ (The American national election survey reports an average discrepancy of roughly 10 per cent between respondents who claim to have voted in the previous presidential election and the actual turnout figures.) The IDEA figures for national legislative elections (presidential elections in the United States) since 1945 (http://www.int-idea.se/Voter_turnout/index. html), unlike previous compilations, take into account differences in rates of registration by using as a base age-eligible citizens rather than registered voters. In Figure 6.1 we can see the average percentage of citizens of legal voting age (VAP) that turned out to vote in legislative elections between 1990 and 1996 for the mature democracies – excluding those using compulsory voting.4 (One advantage of using VAP as a base is that it brings the data from other countries into line with the way turnout is calculated for American national elections. This avenue is, however, not open to us when it comes to calculating local turnout since the different sources from which it is obtained, as a rule, use registered voters as a base.) As a visible and quantifiable act, voting compares favourably with standard indicators of civic engagement as a comparative indicator. Though useful for a longitudinal analysis of developments in a given country, interpersonal trust is too culturally embedded to adequately serve the purposes of comparative analysis, while the usefulness of associational density is marred by the fact that voluntary organizations play different roles in different institutional settings. Moreover interpersonal trust is necessarily subjective, while organiza-

70.0 NO SWE FI 60.0 DN NE US 50.0 IR UK SWI JA 40.0 GE Vertical trust 1991* Vertical

30.0 FR AU 20.0 Rsq=0.0637 30 40 50 60 70 80 90

Figure 6.1 VAP turnout in the 1990s (from IDEA). (* From World Values Survey, 1991.) Social Capital, Civic Literacy and Political Participation 87

300.0 NE

SWE 200.0 NO US FI CA DN GE UK AU 100.0 SWI IR FR

Organizational density* JA

0.0 Rsq=0.1713 40 50 60 70 80 90

Figure 6.2 Average national turnout 1945–97 (from IDEA). (* From World Values Survey, 1991.) tional density usually relies on self-reporting (with figures usually taken from the World Values Survey). Is turnout higher in democratic countries with higher levels of trust and organizational density? Using the numbers from IDEA for mature democracies participating in the 1991 World Values Survey (excluding those with compulsory voting), we test the statistical relationship between average levels of voter turnout and interpersonal or vertical trust at the beginning of the 1990s. Figure 6.1 shows there to be little basis for the presumption that people in high trust societies, that is those with a higher percentage responding that most people can be trusted, vote more than in low trust ones.5 And, in Figure 6.2, replacing level of interpersonal trust with that of organizational participation (using data from the World Values Survey), we find that countries with greater associational density6 do show somewhat higher turnout, though far from any kind of statistically significant relationship. This corresponds to individual-level findings in the American Citizen Participation Study (Verba et al., 1995) to the effect that there is a significant relationship between membership in voluntary groups and more active forms of political participation, but not with voting.

Comparing turnout in local elections

If trust and associational density are indicators of civic engagement, we would expect a stronger relationship with turnout in local than in national elections. What do we learn from the aggregate data? Note that data on turnout in local elections is harder to come by. Neither IDEA nor any other international organization assembles comparative local voting averages. Hence it has 88 Henry Milner required considerable effort to seek out this data from the relevant authorities and specialists in each country. This effort is more than justified, however, by the value of such data for the comparative analysis of civic engagement. For one thing, local elections are less prone to being distorted by the ‘noise’ of emotion-laden television campaigns designed to bring uninformed voters to the polls, since local elections seldom offer sufficient incentives for incurring the high costs of such campaigns. It is reasonable to expect that higher turnout in local elections will usually be reflected in higher levels of participation in non-political, local organizational activities. For example, Bean (1989) found a close interrelationship between turning out to vote, taking part in the campaign, and being involved in some effort to deal with a local problem. While voting in local elections does not match the standards of the active participation in New England towns described by Tocqueville, it does offer a greater potential of living up to the Rousseaunian ideal of political participation as an act of political learning. Mill put the case simply: local politics best affords opportunities for learning about democracy since holding local – as opposed to national – office is a reasonable aspiration for ordinary citizens. Moreover, though people participate to ensure that their interests are taken into account, in so doing they learn to effectively define their interests in the context of the interests of the wider community for which they also have responsibility (Mill, 1910, pp. 346–59). Figures 6.3 and 6.4 present the aggregate averages for municipal-level turnout for elections that took place between the late 1970s and the early 1990s for the main mature democracies for which such data is available. Countries with compulsory voting at the local level are excluded. Apart from the methodological reasons for doing so, there is something contradictory

70.0 NOSWE FI 60.0 DN NE CA US 50.0 IR UK SWI 40.0 GE Vertical trust 1991* Vertical

30.0 FR AU 20.0 Rsq=0.0046 30 40 50 60 70 80

Figure 6.3 Municipal turnout average. (* From World Values Survey, 1991.) Social Capital, Civic Literacy and Political Participation 89

300.0 NE

SWE 200.0 NO US FI CA DN GE UK AU IR 100.0 SWI FR Organizational density*

0.0 Rsq=0.0155 30 40 50 60 70 80

Figure 6.4 Municipal turnout average. (* From World Values Survey, 1991.) about treating the action of individuals who vote to avoid paying a fine as an expression of civic engagement. Moreover, compulsory voting, though probably the most significant, is only one of many institutional factors affecting local turnout. In some cases, there is as much variation within countries as among them. Moreover, the relevant regulations change fairly frequently since it is typically far easier to enact such changes at the local level. Among relevant rules are: the extent to which the elections are simultaneous across the region or country, or with elections at another level; whether non-citizens or non-resident property owners are permitted to vote; the use of postal ballots; the day of the vote and length of voting hours; and the ease of registering to vote.7 There is, however, no reason to think that, apart from compulsory voting, any of these factors systematically distorts outcomes to the extent of making overall aggregate relationships suspect. Indeed, differences on certain of these features within the more pluralistic countries to some extent cancel out differences among those countries. Significant variation in important features of local elections and politics have been noted in Germany, Australia and, especially, Canada (see Goldsmith, 1992).8 For the unitary countries where figures are provided by central government agencies, incomplete data are fairly rare. Among them, the French data are the least comprehensive, with results only for the 1983 election and only for the 386 municipalities with over 20 000 in population.9 For the UK and New Zealand we are limited to elections that took place in the 1990s, though the British figures correspond closely to an estimate for an earlier decade by Goldsmith and Newton (1986, p. 146). The main problems arise with the federal states – Canada, the United States, Australia, Germany, Austria and Switzerland – since they leave it to the 90 Henry Milner states/provinces to collect municipal data and, in some cases, the states or provinces leave it up to the municipalities themselves to collect the data, though this has become less common in recent years. Thus, the data that we do have for these countries tends to be partial, with more complete data available only for recent elections. Typical perhaps is Australia where such data are now collected by the states, but several have only recently begun to do so. Three such states are Tasmania, South Australia and Western Australia, which are of particular interest in that they are the only ones where municipal voting is voluntary. Australian data used in this study are thus limited to those three states. The 34 per cent average arrived at is low by international standards, but corresponds to an earlier estimate by Goldsmith and Newton (1986, p. 146) of 25 per cent, once we take into account the turnout boost in elections during the 1990s due to the introduction of postal ballots and other administrative reforms designed to encourage participation. The situation is somewhat similar in Canada, though compulsory voting does not complicate matters. We use the mean turnout in contested municipal elections in the two largest provinces, Ontario and Quebec, as well as the fourth largest, Alberta.10 For Austria, I have only the four elections up to 1996 for Vienna, the largest province, and the 1997 figures for the second largest province, Upper Austria. The German results are somewhat more complete. Of the ten former West German Länder, I have been able to collect data from electoral statistics offices in seven, representing over 75 per cent of the population of the former West Germany.11 When it comes to acquiring information from regional or national authorities, the most difficult cases are the United States and Switzerland. The former is especially disappointing, given the immense quantity of electoral data collected by American researchers. It has proven impossible to find any comprehensive municipal turnout data in the various American databases; nor have I been able to identify specific state government agencies that collect such data on a state-wide basis. Hence there is almost no literature based on systematic analyses of turnout in American municipal elections. There are only two partial and dated sources in the published literature (Morlan, 1984; Alford and Lee, 1968). In both cases the data were acquired through a mailed survey of officials of municipalities with over 20 000 population. Fortunately, I have gained access to unpublished data from the results of a recent request for such information gathered by Schuckman (1998).12 The 23.5 per cent average that I derive is based on his results and complements the 31.5 per cent average cited by Morlan, given the overall decline in voter turnout in the US in recent decades. Nevertheless using these data places the US at a comparative disadvantage as they are based on potential rather than registered voters and leave out the small municipalities.13 Hence I have attempted to correct for these disadvan- Social Capital, Civic Literacy and Political Participation 91 tages by adjusting the average upwards in an amount corresponding to the estimated difference between registered and potential voters by reducing the base by 25 per cent of potential voters. Hence the US figure goes up from 27.5 (the average of Morlan and Schuckman’s figures) to 36.5 per cent.14 Switzerland presented an equally difficult problem, since there are no official records assembling the results of all local elections in any Canton, let alone for Switzerland’s over 3000 municipalities (Kommunes) as a whole. Fortunately, a research project conducted at the Institute of Sociology at the University of Zurich surveyed the secretaries of all Swiss municipalities (see Ladner and Milner, 1999) to which 81.6 per cent of the Kommunes responded, providing turnout and other valuable data for the most recent local election held up to 1988. Because of these limitations of the data, several of the numbers in Figures 6.3 and 6.4 are just approximations. Hence the correlations in the figures are also approximations, provided to allow us to visualize overall comparative aggregates. As we can see in Figure 6.3, there is, as with national level turnout, no significant relationship at the aggregate level between average local turnout and vertical (interpersonal) trust. As far as organizational density is concerned, we can see in Figure 6.4 that, unlike in the case of trust, there is no sign in local elections of the somewhat higher turnout in more trusting societies that we saw in Figure 6.2 for national elections – quite the opposite of what we might have expected.

Civic literacy: the missing link between social capital and political participation

If, as is now clear, a country’s average level of interpersonal trust and organizational participation does not predict the level at which its citizens will turn out to vote in the elections closest to their ‘civic’ lives, what does this tell us about these indicators when used for comparative purposes, and, more profoundly, what does this suggest to us about the nature of civic engagement? In my forthcoming book, I present an alternative approach to civic engagement, developing the argument that the connection between social capital and political participation passes through political knowledge – what I term ‘civic literacy’. Here I focus on one element of that argument, the institutional underpinnings of civic literacy, to explain the absence of a relationship between local voting turnout and the standard indicators of social capital. Rosenstone and Hansen (1993) refer to the ‘more intense exertions’ by European political actors to promote citizen participation. In investigating cross-national variation in what they term ‘political sophistication’, Gordon and Segura usefully identify such exertions as reﬂecting not individual ‘capabilities’, but institutions. 92 Henry Milner

Lack of sophistication need not indicate an inherent weakness of mass publics; this shortcoming may, instead, be a product of the systems – and the individual choices structured by those systems within which it emerges . . . Capable people, in environments where information is available may choose to be uninformed if that information is expensive and difﬁcult to accumulate, or if the use of that information is of limited utility. (Gordon and Segura, 1997, p. 129).

Competitive elections and multi-party systems are institutions they associate with greater political sophistication. I pursue this insight, arguing that the relationship between political knowledge and political participation tran- scends the characteristics of the potential voters. Critical are the institutional incentives upon political actors’ efforts to get voters to the polls, especially those efforts to inform citizens of the issues at stake in the election. Consensual political institutions (Lijphart, 1999), and proportional electoral systems in particular, have an impact on the accessibility, intelligibility and usefulness of political information. In various ways, PR-based political institutions place greater incentives on political actors to inform citizens, and on citizens to seek political information. Using a range of existing – admittedly incomplete – comparative measures of political knowledge, I show voter turnout at all levels to be significantly higher in high civic literacy countries, that is countries with high average levels of political knowledge. The term ‘civic literacy’ reflects the approach taken. The word ‘civic’ is chosen, as in civic engagement, because it combines in one word the notion of exercising one’s role as citizen with one’s membership in a local community; the word ‘literacy’ is selected because it implies that there is a known quantity that is attainable by each individual, that, unlike, say, knowledge, cannot be ‘stocked’.15 Since local political decisions are those closest to the citizen’s everyday activities, local voter turnout can be regarded as a measurable expression of informed citizenship. Hence the degree to which citizens regularly participate politically by voting in local elections provides a kind of barometer of the level of the civic component of civic literacy. The literacy, or cognitive capacity, component is reflected by the proportion of members of that society demonstrating the knowledge and skills to act as competent citizens. Although there is no single international scale comparing societies’ average level of such knowledge or capacity, the various incomplete indicators allow for a quite consistent ranking of countries, one conforming quite well to average local turnout as well as to newspaper circulation, adult education availability and so on. Policies encouraging newspaper reading and adult education can be described as affecting the ‘demand’ side of civic literacy; the choice of PR electoral systems and related consensual political institutions work which, Social Capital, Civic Literacy and Political Participation 93 through the incentives placed on political actors, affect the accessibility, intelligibility and usefulness of political information can be described as impacting the ‘supply’ side. Together, these supply and demand factors account for differences in civic literacy. To illustrate these explanations, I end here with figures portraying the relationship of average turnout in local elections with three indicators. The first two compare countries on the basis of indicators of the literacy or demand side of civic literacy. Figure 6.5 uses results from the OECD-sponsored International Adult Literacy Survey (IALS). The study tested the level of comprehension of three types of written materials: narrative prose; documents, such as maps, train schedules, etc.; and questions applying basic arithmetic skills (OECD, 1997). Figure 6.5 plots the relationship between average local turnout and the average per cent of a large sample of the adult population in 11 countries16 scoring in the lowest (subliterate) category in the three tests.17 Figure 6.6 uses average daily newspaper circulation per thousand. Figure 6.7 breaks countries down based on the electoral system used. As we can see, average voter turnout in (non-compulsory) local elections is far higher in countries using proportional representation than in those using majority systems. While not the only factor accounting for this difference, Figures 6.5 to 6.7 combined show a clear three-way relationship between political participation, political institutions and civic literacy. There is much more information to be marshalled to do justice to the complexity of the relationships these figures depict. I have only touched upon those factors directly relevant to the purpose of this paper, that is demonstrating that the link between social capital and political participation passes through political knowledge. Clearly, in comparing advanced democratic societies, we cannot deny there is a knowledge – civic literacy – component to

IR US UK NZ 20 AUL CA SWI

NE GE 10 SWE Functional illiteracy rate* Functional illiteracy

0 Rsq=0.5331 30 40 50 60 70 80

Figure 6.5 Municipal turnout average. (* From OECD-IALS, 1997.) 94 Henry Milner

700

600 NO

500 FI SWE

400 SWI UK NE DNGE 300 AU AUL NZ US FR 200 CA IR Daliy newspapers per thousand 1995* Daliy newspapers 100 Rsq=0.2853 30 40 50 60 70 80

Figure 6.6 Municipal turnout average. (* From UNDP, 1999.)

30 Municipal turnout per cent average 20 N= 6 8 1.00 2.00

Figure 6.7 Electoral system: 2 = PR; 1 = other.

political participation that is different from (and not captured by indicators of) interpersonal trust. This component, moreover, contributes to the level of voter turnout but not necessarily to the density of voluntary associations. If political participation is an expression of social capital, and voting in local elections an essential element of civic engagement, then the civic literacy dimension must be integrated into the conceptualization of social capital – especially if that conceptualization is to serve the purposes of comparative analysis. Social Capital, Civic Literacy and Political Participation 95

Notes

1. For example, McCann (1998) not only confirms that more informed voters are more likely to vote, but finds American voters to be significantly more socially and politically engaged than non-voters, even when controlling for age, education, etc. 2. Althaus (1998) finds that the low levels and uneven social distribution of political knowledge in the US results of opinion surveys often misrepresent the mix of voices in a society. First, those who are poorly informed about politics tend to give ‘don’t know’ and ‘no opinion’ responses at much higher rates than more knowledgeable people. In addition, information asymmetries affect the ‘quality of opinions’. The well informed are more likely to express policy preferences consistent with their political interests, which are then more accurately reflected in measures of collective opinion. Overall, ‘one in five policy questions might have a different collective preference if everyone were equally well informed about politics’ (p. 555). 3. As Kuklinski et al. (1998, pp. 144–6) put it, it is less that people are uninformed than that they are systematically misinformed. Thus their policy preferences are distorted, and they resist accepting the correct facts even if these are made available. Their study of a sample of Illinois residents ‘found that many people had systematically inaccurate factual beliefs (for example, they grossly overestimated the payments received by a typical family) [with an] overall bias toward beliefs that one would expect to be associated with negative views about welfare . . . People held their beliefs confidently, . . . [and] these beliefs are strongly linked to policy preferences in the expected ways.’ The authors relate their conclusions to the strategy of deliberative polls sponsored by James Fishkin (Luskin and Fishkin, 1998). ‘Both in several days of briefings and discussion and through written materials provided in advance, Fishkin exposes a representative sample of citizens to a massive amount of competing advocacy . . . There is evidence that deliberative polls improve people’s judgments. At least, they often alter people’s opinions.’ 4. I do not include countries with compulsory voting. A rare longitudinal analysis comparing the turnout in Australian states and for the Commonwealth in the election before and after the introduction of compulsory voting found an increase averaging 23 per cent, varying from a low of 12.4 to a high of 37.8 (McAllister and Mackerras, 1998, p. 2). The difference in Australia is significantly higher than in the Netherlands, which changed in the opposite direction, eliminating compulsory voting. The average difference in Dutch elections before and after the change was 10 per cent (Goldsmith and Newton, 1986, p. 146; Hirczy, 1994). 5. The absence of any significant relationship at the aggregate parallels the findings of studies using individual data: while people who trust their political institutions are more likely to participate in politics, ‘studies have found that trusting citizens are not more likely to vote, engage in campaign activities, or be interested in politics’ (Norris, 1996, p. 475). 6. This number reflects the average number of associational memberships to which (a sample of) individuals claim membership. 7. Other rules that could be relevant concern the selection of the mayor – elected directly, indirectly elected or appointed (as a town manager). Here there is much variation. For example, while the mayor is now directly elected in all German municipalities, this is not the case in Austria where mayors are directly elected in only five of the Länder. Relevant also could be whether, as in France, the mayors and councillors are eligible to hold higher-level office. 96 Henry Milner

8. There is one exception: that is when local elections occur simultaneously with national elections, which, we must assume, bring a certain number of people to the polls who would otherwise not have voted. Only one country among those considered in this study uses such a system, namely Sweden. Sweden is included in the data set on turnout, but, to take this factor into account, the Swedish turnout results have been reduced by 10 per cent, an amount derived by looking at differences in local and national turnout in Sweden’s Nordic neighbours. 9. New Zealand data are for 1992 and 1995, and are taken from the Local Authority Election Statistics of the Department of Internal Affairs. For the UK, results are from 1991 and 1995 (Rallings and Thrasher, 1998). 10. The three together account for more than 80 per cent of the population; I was unable to acquire data for the third largest, British Columbia. However, it is only recently that the two English-speaking provinces began to assemble the data, while Quebec provides only the raw data, city by city. (For 1997 and 1998, I arrived at an average of 59 per cent for those in which the position of mayor was contested. This compares to a 51 per cent average for the eight large cities that elected mayors in those years, and to the 54 per cent for the 28 of the 53 Quebec municipalities with over 20 000 population that had a contested election for mayor in 1993 or 1994.) 11. In the cases of North Rhineland-Westphalia, Bavaria, Hesse, Schleswig-Holstein, Baden-Würtemburg and Hamburg I was able to obtain the results for all local elections since 1980; for the Rhineland-Palatinate only 1994 was available, and for Saarland, only 1989 and 1994. As free elections are still recent arrivals in the other five Länder, I exclude those from the former East Germany. 12. Schuckman wrote to the city halls of the 79 cities with over 200 000 population, requesting information about the results of the last mayoral election held before spring 1998. He then calculated totals (from the 65 cities that responded) against the estimated number of potential voters. 13. The apparent effect of size can be estimated from Ontario results, where the 21 cities with more than 100 000 averaged approximately 10 per cent less than Ontario municipalities as a whole (Kushner et al., 1997). Note that it is also to some extent unfair not to take into account the fact that US voters are, on average, called upon to vote more frequently, just as they still, though less than before (Highton and Wolfinger, 1998), face obstacles to registration not faced by voters elsewhere. 14. Using available newspaper reports, I was able to note the results of fall 1997 local contests in Connecticut as well as in a number of counties in central New York State. Based on a commonly made – and probably still conservative – assumption that 25 per cent of eligible voters do not register in the United States, the 43 per cent average of eligible voters that turned out in both cases corresponds to 31 per cent of potential voters, 8 to 9 per cent above the big city totals provided by Schuckman. 15. This distinction becomes clear in the discussion of social capital. One can build up a stock of physical and, by implication, social capital. Yet, clearly, an individual who is a member of five voluntary associations does not contribute as much to the ‘stock’ of social capital as do five individuals, each members of one. The term literacy does not permit such an interpretation: though one person may read much more than another, both contribute equally to the overall literacy rate of their community. 16. For Belgium, only Flanders was surveyed, and I have excluded Poland. 17. We must be careful in drawing conclusions since there are only 11 countries involved. The IALS is currently conducting or planning tests in a number of other Social Capital, Civic Literacy and Political Participation 97

countries, the scores of which, when compiled and included here, should bring the number of countries to a number high enough that the values of single cases will not have the same powerful effect.

References

Alford, R. and Lee, E. (1968) ‘Voting Turnout in American Cities’, American Political Science Review, 62: 796–813. Althaus, Scott L. (1998) ‘Information Effects in Collective Preferences’, American Political Science Review, 92: 545–58. Bartels, Larry M. (1996) ‘Uninformed Votes: Informational Effects in Presidential Elections’, American Journal of Political Science, 40: 194–230. Bean, Clive (1989) ‘Orthodox Political Participation in Australia’, Australia and New Zealand Journal of Sociology, 25, 3: 451–79. Converse, Philip E. (1964) ‘The Nature of Belief Systems in Mass Publics’, in David Apter (ed.), Ideology and Discontent. New York: Free Press. Delli Carpini, Michael and Keeter, Scott (1995) Political Knowledge and Enlightened Group Interests’, paper presented at the 1995 Congress of the American Political Science Association, Chicago, IL. Delli Carpini, Michael and Keeter, S. (1996) What Americans Know about Politics and Why it Matters. New Haven, CT: Yale University Press, pp. 1–397. Dimock, Michael A. (1998) Political Knowledge and the Political Environment: Reassessing Trends in Partisanship, 1960–1996’, paper presented at the American Political Science Association Annual Meeting, Boston, MA. Downs, Anthony J. (1957) An Economic Theory of Democracy. New York: Harper. Fiorina, Morris P. (1990) ‘Information and Rationality in Elections’, in John A. Ferejohn and J. A. Kuklinski (ed.), Information and Democratic Processes. Urbana: University of Illinois Press. Fishkin, James (1995) The Voice of the People: Public Opinion and Democracy, New Haven, CT: Yale University Press. Goldsmith, M. (1992) Options for the Future – Local Democracy Abroad. London LGMB (Local Government Management Board). Goldsmith M. and Newton, K. (1986) ‘Aspects of Local Democracy’, in Report of Committee of Inquiry into the Conduct of Local Authority Business, Vol 4. London: HMSO. Gordon, Stacy B. and Segura, Gary M. (1997) ‘Cross-National Variation in the Political Sophistication of Individuals: Capability or Choice?’, Journal of Politics, 59: 126–47. Green, Donald P. and Shapiro, Ian (1994) Pathologies of Rational Choice Theory: A Critique of Applications in Political Science. New Haven, CT: Yale University Press. Highton B. and Wolﬁnger, R. E. (1998) ‘Estimating the Effects of the National Voter Registration Act 1993’. Political Behavior, 20, 2: 79–104. Hirczy, Wolfgang (1994) ‘The Impact of Mandatory Voting Laws on Turnout: A Quasi- Experimental Approach’, Electoral Studies, 13 (March): 64–76. Kuklinski, James H., Paul J. Quirk, David W., Schwieder, Robert F. Rich (1998) ‘Just the facts, ma’am’: political facts and public opinion’. Annals of the American Academy of Political and Social Science; 560: 143–54. Kushner, J., D. Seigel and H. Stanwich (1997). ‘Ontario Municipal Elections: Voting Trends and Determinants of Electoral Success in a Canadian Province’. Canadian Journal of Political Science. 20, 3. 98 Henry Milner

Ladner, Andreas and Milner, Henry (1999) ‘Politicization, Electoral Institutions, and Voting Turnout: the Evidence from Swiss Communal Elections in Comparative Context’, Electoral Studies, 18, 2: 235–50. Lijphart, Arend (1997) ‘Unequal Participation: Democracy’s Unresolved Dilemma’, American Political Science Review, 91, 1: 1–14. Lijphart, Arend (1999) Patterns of Democracy: government forms and performance in thirty- six countries. New Haven, CT: Yale University Press. Lupia, Arthur (1994) ‘Shortcuts versus Encyclopedias: Information and Voting Behavior in California Insurance Reform Elections’, American Political Science Review, 88: 63–76. Luskin Robert C. and James S. Fishkin (1998) ‘Deliberative Polling, Public Opinion and Democracy: the Case of the National Issues Convention’. Presented at the annual meeting of the American Political Science Association, Boston, MA, September 1998. McAllister, Ian and Mackerras, Malcolm (1998) ‘Compulsory Voting, Party Stability and Electoral Bias in Australia’, paper presented at the ECPR joint sessions, Warwick, UK, March. McCann, James A. (1998) ‘Electoral Participation and Local Community Activism: Spillover Effects, 1992–1996’, paper presented at the annual meeting of the American Political Science Association, Boston, MA. Mill, J. S. (1910) Representative Government. London: Dent & Sons. Milner, Henry (2001) Civic Literacy: How Informed Citizens Can Make Democracy Work. However, NH: University of New England Press. Morlan, Robert L. (1984) ‘Municipal vs. National Election Voter Turnout: Europe and the United States’. Political Science Quarterly, 99: 457–70. Norris, Pippa (1996) ‘Does Television Erode Social Capital?’, PS: Political Science and Politics, September, 474–80. OECD-IALS (1997) Literacy Skills for the Knowledge Society: Further Results from the International Adult Literacy Survey. Paris: OECD. Olson, Mancur (1996) ‘The Varieties of Eurosclerosis: the Rise and Decline of Nations since 1982’, in N. Craft and G. Toniolo (eds), Economic Growth in Europe. Cambridge: Cambridge University Press. Pacek, Alexander and Radcliff, Benjamin (1995) ‘Turnout and the Vote for Left-of- Centre Parties: A Cross-National Analysis’, British Journal of Political Science, 25: 137–43. Pellikaan, Huib (1998) Rationality of Voter Turnout, paper presented at the annual meeting of the Society for the Advancement of Socioeconomics (SASE), Vienna, July. Putnam, Robert (1993) Making Democracy Work. Princeton, NJ: Princeton University Press. Putnam, Robert (1995) ‘Bowling Alone: America’s Declining Social Capital’, Journal of Democracy, 6: 65–78. Rallings, Colin and Thrasher, Michael (1998) The Slow Death of a Governing Party: the Conservatives in Local Government 1979–97, paper presented at the ECPR joint sessions, Warwick, UK. Rosenstone, Steven and Hansen, John M. (1993) Mobilization, Participation and Democracy in America. New York: Macmillan. Schuckman, Harvey P. (1998) ‘What Motivates Civic Engagement in Local Politics: Deconstructing the Role of Local Political Institutions: An Examination of Mayoral Electoral Turnout’, paper presented at the American Political Science Association Annual Meeting, Boston, September. Sniderman, Paul M., Brody, Richard A. and Tetlock, Philip E. (1991) Reasoning and Choice. New York: Cambridge University Press. Social Capital, Civic Literacy and Political Participation 99

Teixeira, Ruy A. (1992) The Disappearing Voter. Washington, DC: Brookings Institution. UNDP (United Nations Development Programme) (1999) Human Development Report. New York: Oxford. Verba, Sidney, Schlozman, Kay Lehman and Brady, Henry E. (1995) Voice and Equality: Civic Voluntarism in American Politics. Cambridge, MA: Harvard University Press. 7 A Virtuous Circle? The Impact of Political Communications in Post-Industrial Democracies Pippa Norris

Recent years have seen growing tensions between the ideals and the perceived performance of democratic institutions. While there is no ‘crisis of democracy’, many believe that all is not well with the body politic. Concern in the United States has focused on widespread cynicism about political institutions and leaders, fuelling fears about civic disengagement and a half-empty ballot box (Nye et al., 1997; Putnam, 2000). The common view is that the American public turns off, knows little, cares less and stays home. Similar worries echo in Europe. Commentators have noted a crisis of legitimacy following the steady expansion in the power and scope of the European Union despite public disengagement from critical policy choices (Hayward, 1995; Anderson and Eliason, 1996, Pharr and Putnam, 2000). The growth of critical citizens is open to many explanations, explored in a previous study (Norris, 1999). One of the most popular accounts attributes public disengagement to political communications. The political science literature on ‘media malaise’ or ‘videomalaise’ originated in the 1960s, developed in a series of scholarly articles in the post-Watergate 1970s, and rippled out to become the conventional wisdom in the popular culture of journalism and politics following a ﬂood of books in the 1990s. The chorus of critics is loudest in the United States but similar echoes can be heard in Europe. These accounts claim that common practices by the news media and by party campaigns hinder civic engagement, meaning learning about public affairs, trust in government and political activism. Media malaise theories share two core assumptions: (a) that the process of political communications has a signiﬁcant impact upon civic engagement; and (b) that this impact is in a negative direction. There is nothing particularly novel about these claims. Many critics expressed concern about the effects of the popular press on moral decline throughout the nineteenth century as newspapers became more widely avail-

100

K. Dowding et al. (eds.), Challenges to Democracy © Political Studies Association 2001 Political Communications in Post-Industrial Democracies 101 able (Curran and Seaton, 1991). The phenomenon of the ‘yellow press’ in the 1890s caused worry about its possible dangers for public affairs. In the 1920s and 1930s, the earliest theories of mass propaganda were based on the assumption that authoritarian regimes could dupe and choreograph the public by manipulating radio bulletins and newsreels (Lowery and DeFleur, 1995). Recent decades have seen multiple crusades against the supposed pernicious inﬂuence of the mass media, whether directed against violence in movies, the ‘wasteland’ of television, the impact on civic engagement of watching TV entertainment, the dangers of tobacco advertising or the supposedly pernicious effects of pop music (Starker, 1991). While hardly new, what is different today is the widespread orthodoxy that has developed around this theory. Let us ﬁrst outline the American and European accounts of media malaise and then consider some evidence surrounding this thesis.

Theories of media malaise

American theories of ‘media malaise’ emerged in the political science literature in the 1960s. Kurt and Gladys Lang (1966) were the first to connect the rise of network news with broader feelings of disenchantment with American politics. TV broadcasts, they argued, fuelled public cynicism by overemphasiz- ing political conflict and downplaying routine policymaking in DC. This process, they suggested, had most impact on the ‘inadvertent audience’, who encountered politics because they happened to be watching TV when the news was shown, but who lacked much interest in, or prior knowledge about, public affairs. The Langs proved an isolated voice at the time, in large part because the consensus in political communications stressed the minimal effects of the mass media on public opinion. The idea gained currency in the mid-1970s since it seemed to provide a plausible reason for growing public alienation in the post-Vietnam and post- Watergate era. Robinson (1976) first popularized the term ‘videomalaise’ to describe the link between reliance upon American television journalism and feelings of political cynicism, social mistrust and lack of political efficacy. Greater exposure to television news, he argued, with its high ‘negativism’, conflictual frames and anti-institutional themes, generated political disaffec- tion, frustration, cynicism, self-doubt and malaise. This process was seen as most critical during election campaigns, where viewers were turned off, he argued, by TV’s focus on the ‘horse-race’ at the expense of issues, analysis rather than factual information and excessive ‘bad news’ about the candidates (Robinson and Sheeney, 1983). Many others echoed these claims over the years. According to a widely influential report for the Trilateral Commission, the news media had eroded 102 Pippa Norris respect for government authority in many post-industrial societies, contributing towards a widespread ‘crisis’ of democracy evident on the streets of Washington DC, Paris and Tokyo (Crozier, Huntington and Watanuki, 1975). During the 1990s the trickle of complaints about the news media became a popular deluge (Entman, 1989; Postman, 1985; Fallows, 1996; Hart, 1994, 1996; Gabler, 1998; Sabato, 1988; Patterson, 1993, 1996; Cappella and Jamieson, 1996, 1997; Dautrich and Hartley, 1999). The list of complaints goes on and on and on. Criticisms have moved well beyond the halls of academe: many US journalists share the belief that something is badly wrong with their profession (Pew, 1999) and the Committee of Concerned Journalists, led by Bill Kovach and Tom Rosensteil (1999), has debated potential reforms to the profession. In Europe similar voices can be heard although these accounts emphasize structural developments in the news industry and in party campaigning (Blumler and Gurevitch, 1995; Achille and Bueno, 1994; Dahlgren, 1995; Schulz, 1997, 1998; Kaase, 2000). During the 1980s, the public sector experienced a massive programme of privatization throughout Western Europe. During the same era, the growth of alternative commercial channels, breaking down the monopoly of public service broadcasting, undermined the rationale for subsidizing television through state resources. Schulz (1997, 1998) argues that in Germany the decline of public service broadcasting and the rise of commercial channels, the latter emphasizing the more sensational and negative aspects of political news, may have increased public cynicism. Kaase (2000) fears that these developments may produce audiences segmented according to the amount of political information to which they are exposed, possibly reinforcing a ‘knowledge gap’. In the print sector, there is widespread concern that increased competition for readers has increased the pressure on traditional standards of news, leading to ‘tabloidization’ or ‘infotainment’. ‘Yellow journalism’ in the 1890s rou- tinely highlighted the moral peccadilloes and sexual proclivities of the rich and famous. Sensationalism, crime and scandal in newspapers are hardly new, providing a popular alternative to the dull business of politics (Gabler, 1999). But today routine and daily front-page news about government scandals appears greater than in previous decades – whether sleaze in Britain, Tagentopoli in Italy, Recruit and Sagawa in Japan, or l’affaire Lewinsky in America (Lull and Hinerman, 1997). This coverage is believed to corrode the forms of trust underpinning social relations and political authority. The process of ‘tabloidization’ may have gone further in Europe than in the American or Japanese press, with papers like The Sun or Der Bild leading the pack, each with many millions of readers. But similar phenomena are evident in the chase for ratings in local TV news and ‘all talk, all the time’ cable news magazines in the US. Political Communications in Post-Industrial Democracies 103

Many hope that the Internet can escape these problems, but others fear that this medium may reinforce political inequality (for a full discussion see Norris, 2001). Owen and Davis (1998) conclude that the Internet provides new sources of information for the politically interested, but given uneven levels of access there are good grounds to be sceptical about its transformative potential for democratic participation. Murdock and Golding (1989) argue that the new medium may merely reproduce, or even exacerbate, existing social biases in conventional political participation. Hill and Hughes (1998) believe that the Internet does not change people; it simply allows them to do the same things in a different way. A related stream of commentators attributes the problems of political communications primarily to the practice of professional marketing. One of the most striking developments in many countries has been the declining importance of the ‘pre-modern’ campaign involving local party meetings, door-to-door canvassing and direct voter–candidate contact. The rise of the ‘modern’ campaign is characterized by the widespread adoption of the techniques of political marketing (Swanson and Mancini, 1996). Strategic communications is part of the ‘professionalization’ of campaigning, giving a greater role to technical experts in public relations, news management, advertising, speech-writing and market research (Norris, 1997a; Norris et al., 1999). The rise of political marketing has been widely blamed for growing public cynicism about political leaders and institutions. The central concern is that the techniques of ‘spin’, selling and persuasion may have undermined the credibility of political leaders (Jones, 1995; Rosenbaum, 1997). If everything in politics is designed for popular appeal then it may become harder to trust the messages or messenger. Although lacking direct evidence of public opinion, Franklin (1994) provides one of the clearest statements of this thesis, decrying the ‘packaging of politics’. Many others have expressed concern about the ‘Americanization’ of election campaigning, in Britain, Germany and Scandinavia, and the possible impact this may have had upon public conﬁdence in political parties (Pfetsch, 1996). The use of ‘negative’ or attack advertising by parties and candidates has also raised anxieties that this practice may demobilize the electorate (Ansolabehere and Iyengar, 1995). Therefore, to summarize, American and European accounts differ in the reasons given for media malaise.

• Structural perspectives emphasize institutional developments common to many post-industrial societies, such as economic pressures moving the news industry downmarket, the erosion of public service broadcasting, and the emergence of a more fragmented, multi-channel television environment. 104 Pippa Norris

• Cultural accounts stress historical events speciﬁc to journalism in the United States, notably the growth of a more adversarial news culture following Vietnam and Watergate. • Campaign accounts focus on the growth of political marketing with its attendant coterie of spin-doctors, advertising consultants and pollsters, reducing the personal connections between citizens and representatives.

Of course there are counterclaims in the literature and the number of sceptics questioning the thesis has been growing in recent years. Earlier studies by the author found that, contrary to media malaise, although TV watching was related to some signs of apathy, attention to the news media was associated with positive indicators of civic engagement, in the United States and Britain, as well as other countries (Norris, 1996, 1997b, 2000). Kenneth Newton (1997) showed that reading a broadsheet newspaper in Britain, and watching a lot of television news was associated with greater political knowledge, interest and understanding of politics. Holtz-Bacha (1990) demonstrated similar patterns associated with attention to the news media in Germany, while Curtice, Schmitt-Beck and Schrott (1998) reported similarly positive findings in a five- nation study from elections in the early 1990s. The most recent examination of the American NES evidence, by Earl Bennett et al. (1999), found that trust in politics and trust in the news media went hand in hand, with no evidence that use of the news media was related to political cynicism. So far, however, counterclaims have been published in scattered scholarly studies, and have thereby been drowned out by the Greek chorus of popular lament for the state of modern journalism. Before we all jump on the media malaise bandwagon, what is the solid evidence supporting this thesis? Here we can briefly outline two sources of data that throw sceptical light on some of the core claims, namely aggregate indicators of the major structural trends affecting the news media’s political coverage in the postwar era, and survey evidence about the individual-level impact of attention to the news media on civic engagement.

Trends in the news industry

In examining the evidence for media malaise we need to distinguish between the production, contents and effects of political communications. While the production process has undoubtedly been transformed during the last ﬁfty years, the impact of this upon the contents has not been well established, still less the inﬂuence upon the general public. The news industry has certainly changed in response to major technological, socio-economic and political developments in the postwar era. Since the 1950s, the printed press has seen greater concentration of ownership and a Political Communications in Post-Industrial Democracies 105

600

500 481

Number of TVs 419 400

334 300 294 290 271 295 262 Newspaper circulation 263

200 187 Number per 1000 population

100 58 8 0 1950 1960 1970 1980 1990 1996 Figure 7.1 TV viewing and newspaper circulation in the postwar era. reduction in the number of available independent outlets. At the same time, however, many media malaise accounts fear that newspaper sales have declined in postindustrial societies and this is not the case. As shown in Figure 7.1, in the postwar era TV viewing surged but at the same time newspaper sales across OECD countries have remained stable. During the 1980s public television, which had enjoyed a state monopoly throughout much of Western Europe, faced increased competition from the proliferation of new terrestrial, cable, satellite, digital and broadband television channels. Since the mid- 1990s, the explosion of the Internet has challenged the predominance of television, a pattern most advanced in Scandinavia and North America. The net result of these developments is greater fragmentation and diversiﬁcation of formats, levels and audiences in the available news outlets. The available comparative evidence suggests ﬁve important trends, each with important implications for structural claims of media malaise. First, overall news consumption is up. During the last three decades the proportion of Europeans reading a newspaper every day almost doubled, and the proportion watching television news every day rose from one half in 1970 to almost three-quarters in 1999 (see Table 7.1). Social trends, including patterns 106 Pippa Norris

Table 7.1 The growth in the size of the news audience, EU5

% everyday 1970 1999 Difference Read newspaper 27 45 +18 Watch TV news 49 72 +23 Listen to radio news 44 46 +2

Note: For consistent comparison over time media use is compared only in Belgium, France, Italy, the Netherlands and Germany. Media use in all EU15 member states in 1999 was about 5–8 percentage points higher than these ﬁgures. Source: Eurobarometer surveys, 1970, 1999. of higher literacy, afﬂuence and leisure, have probably contributed towards these developments. Second, the structure of the news industry varies widely across OECD states and TV has not necessarily displaced newspapers as an important source of news in many societies. We often generalize based on the American literature but compared with other post-industrial societies, the US proves exceptionally low in consumption of newspapers and TV news (see Figure 7.2 and Table 7.2). Other countries like Sweden, Austria and Germany are far heavier users of the

80 High news users

70 Finl Ger Neth 60 Swe Den Austria Lux 50 UK Ire

40 US Belg 30 Italy Fr Spain % read paper every day % read paper every 20 Port Gre

10 Low news users

0 50 60 70 80 90

% watch TV news every day

Figure 7.2 News consumption in EU15 and US (1999). Political Communications in Post-Industrial Democracies 107

Table 7.2 Variations in regular sources of news, Europe and the US, 1999

Country Regular Regular TV Regular radio Online users newspaper (% news news (% (% with ‘read (% ‘watch listen access) every day’)‘ every day’) ‘every day’) Austria 54 63 67 11 Belgium 30 66 42 11 Denmark 56 76 65 44 Finland 69 82 49 39 France 26 58 37 9 Germany 63 68 56 8 Greece 17 80 19 7 Ireland 44 66 64 14 Italy 29 82 23 14 Luxembourg 53 71 60 22 Netherlands 61 76 56 32 Portugal 16 62 27 5 Spain 27 70 32 8 Sweden 58 63 47 61 UK 49 71 45 22 US 34 53 29 49 Northern Europe 60 71 57 48 Western Europe 48 70 52 17 Southern Europe 22 74 25 9 EU15 45 71 47 20

Notes: Regular newspaper readers: reads the news in daily papers ‘every day’. Regular television news: watches the news on television ‘every day’. Regular radio news: listens to the news on the radio ‘every day’. Northern Europe: Denmark, Finland and Sweden. Western Europe: Austria, Belgium, Germany, France, Ireland, Luxembourg, The Netherlands, and the UK. Southern Europe: Italy, Greece, Portugal and Spain. Sources: Eurobarometer, 51.0, Spring 1999; American National Election Study, 1998. press while there are far higher users of both newspapers and TV news in Finland, the Netherlands and to a lesser extent the UK. Moreover, news formats and outlets have diversified. In the 1960s viewers normally got the news from the standard flagship evening news programmes and current affairs programmes. Today these have been supplemented by 24- hour rolling news, on-the-hour radio headlines, TV magazines and talk shows, as well as the panoply of online news sources. Access to the Internet has been exploding in many post-industrial societies. By the late-1990s, about a fifth of all Europeans and half of all Americans and Scandinavians surf online. Getting news is one of the most popular uses of the Internet in the US and Europe. As a result of all these developments in the news environment it has become easier to bump into the news, almost accidentally, than ever before. In part as a result, recent decades have broadened the social background of the news audience, especially for the press. Tables 7.3 and 7.4 show regression 108 Pippa Norris

Table 7.3 Models predicting readership of newspapers in 1970, 1980 and 1999, EU5

Predictors of Sig. Predictors of Sig. Predictors of Sig. newspaper newspaper newspaper readership readership readership 1970 1980 1999 Demographics: Education 0.16 ** 0.16 ** 0.04 * Gender: male 0.25 ** 0.15 ** 0.08 ** Age (years) 0.16 ** 0.13 ** 0.15 ** Left–right ideology –0.04 ** –0.04 ** 0.01 SES 0.08 ** 0.04 ** 0.08 ** Household income 0.09 ** 0.10 ** 0.12 ** Urbanization 0.02 0.10 ** 0.01 Use of news: TV news use 0.11 ** 0.19 ** 0.18 ** Radio news use 0.15 ** 0.12 ** 0.16 ** Nation: Belgium –0.17 ** –0.07 ** –0.21 ** France –0.12 ** –0.25 ** –0.23 ** Italy –0.14 ** –0.27 –0.16 ** Netherlands –0.01 ** –0.01 ** –0.05 * Constant 0.56 0.63 0.74 R2 0.22 0.24 0.25 N 8567 6521 6218

Notes: The table reports the standardized beta coefficients predicting frequency of reading newspapers based on ordinary least squared regression models. The dependent variables are the 5-point scales measuring frequency of use of newspaper and television news, where 5 = ‘every day use’ and 1 = ‘never use’. Sig. P. **>0.01 *>0.05 The German dummy variable is excluded as a predictor in these models. Education: age finished full- time education. L–R ideology scale: coded from left (1) to right (10). SES: manual (0) or non-manual head of household. Urbanization rural (1), small town (2), large town/city (3). TV news and radio News: frequency of use on 5-point scales. Sources: European Community Study, 1970; Eurobarometer, 13.0, April 1980, weighted for EU6; Eurobarometer, 50.1, March–April 1999, weighted for EU6. models predicting the social background of regular newspaper readers and TV news viewers using Eurobarometer surveys in five countries in 1970 and 1999. The results of the standardized coefficients show that readership has widened in terms of education, gender and class, with no shift in the age profile of readers. Lastly, the new information environment has greatly expanded the opportunities to learn about public affairs in different channels, programmes, formats and levels. Since the 1970s, the amount of news and current affairs broadcast on public service television in OECD countries more than tripled (see Table 7.5). And of course this does not count the development of new commercial 24- hour news services like Sky and CNN. Political Communications in Post-Industrial Democracies 109

Table 7.4 Models predicting TV news viewership in 1970, 1980 and 1999, EU5

Predictors of Sig. Predictors of Sig. Predictors of Sig. TV news TV news TV news viewership viewership viewership 1970 1980 1999 Demographics Education –0.07 ** –0.03 ** –0.03 Gender 0.04 ** 0.01 * 0.00 Age 0.16 ** 0.09 ** 0.14 ** Left–right ideology –0.01 0.05 ** 0.04 * SES 0.01 –0.05 ** 0.01 Household income 0.08 ** 0.01 ** 0.01 Use of media Newspaper use 0.13 ** 0.21 ** 0.21 ** Radio news use 0.01 0.16 ** 0.08 ** Nation Belgium –0.13 ** –0.02 * 0.01 France –0.11 ** –0.06 ** –0.04 * Italy –0.18 ** 0.15 ** 0.12 ** Netherlands 0.01 –0.03 * 0.05 * Constant 3.3 3.25 3.65 R2 0.08 0.12 0.11 N 8567 8827 6218

But have structural trends eroded traditional standards of political coverage? Many commentators have expressed concern about a decline in long- term ‘hard’ news, such as coverage of international affairs, public policy issues and parliamentary debates. In its place, many suggest, news has ‘dumbed down’ to become ‘infotainment’, focusing on human-interest stories about scandal, celebrities and sex. ‘Tabloid’ papers in Britain, the ‘boulevard press’ in Germany and local television news in the US share many common characteristics. Rather than an inexorable downwards erosion in the standards of serious journalism, it seems more accurate to understand trends during the 1980s and 1990s as representing a diversiﬁcation of the marketplace in terms of levels, formats and topics. Soft news and ‘infotainment’ has undoubtedly grown in some sectors of the market, but serious coverage of political events, inter- 110 Pippa Norris

Table 7.5 The expansion in the news broadcast on public TV, 1971–96

Country Change in the number of hours of Change in the number of hours of news and current affairs entertainment broadcasting broadcasting 1971–96 1971–96 Australia +931 +42 Austria +2489 +5105 Belgium –507 +2321 Czech Rep. +2648 +5848 Denmark +21 +1391 Finland +1051 +2474 France –464 +6448 Greece +2709 +5324 Hungary +3412 +2296 Ireland +592 +4655 Italy +7300 +12945 Korea, S. +2751 +5195 Netherlands +963 +2243 Norway –115 +1342 Poland +4195 +3698 Portugal +2634 +12051 Spain –238 +2469 Sweden –1069 +992 Switzerland +4251 +8315 Turkey +7259 +14699 EU15 +1290 +4868 OECD total +2041 +4992

Note: For the full range of categories see Norris (2000). Source: Calculated from UNESCO Statistical Yearbooks (Paris: UNESCO), 1971–1998.

national affairs and financial news has also steadily expanded in availability elsewhere. Endless Senate debates shown on C-Span coexist today with endless debates about sex and personal relationships on the Jerry Springer Show. The Sun sits on the same newsstands as The Economist. News.bbc.co.uk is as easily available as Amsterdam pornography sites. Diversification does not mean that the whole of society is being progres- sively ‘dumbed down’ by trends in the news media. By focusing only on excesses at the popular end of the market, such as the wasteland of endless punditry on American cable TV talk shows or ‘if it bleeds it leads’ on local American TV news, we overlook dramatic changes such as the ability to watch live legislative debates, to witness natural disasters like the Mozambique floods in real time, or to find online information about local government services. Potentially diversification may lead to another danger, namely greater divisions between the information haves and have-nots. But as we have seen, the Political Communications in Post-Industrial Democracies 111 audience for news has greatly expanded in size and broadened socially during the last quarter-century, not narrowed. The evidence for other assumed long-term changes in the news culture remains limited. We need more systematic data to establish whether, for example, there actually has been a growth in negative coverage of politicians during election campaigns, or whether a more adversarial relationship has developed between journalists and governments. The available studies, however, strongly suggest that developments in political coverage observed in particular countries are often highly particularistic and contextual, rather than representing trends common across post-industrial societies. For example, the most comprehensive comparison of news cultures in 21 countries, by Weaver (1998) based on surveys of journalists, found almost no consensus about professional roles, ethical values and journalistic norms. Rather than the emergence of a single prevalent model of journalism, based on American norms, this suggests considerable diversity worldwide. In the same way, without being able to discuss this in detail in this paper, there is little doubt that political campaigns have been transformed by the diversification in the news industry and also by the widespread adoption of political marketing techniques. Countries have not simply imported American campaigning practices lock, stock and barrel, but politicians in states like Israel, Argentina and Britain seem to be paying more attention to formal feedback mechanisms like polls and focus groups, with an expanding role for campaign professionals from marketing and public relations. Comparative surveys have found that in a ‘shopping’ model, parties adopt whatever techniques seem well suited for their particular environment, supplementing but not dis- carding older forms of electioneering (Plassner et al., 1999). Even in America, traditional forms of grassroots voter contact have been maintained, for example in New Hampshire, alongside newer forms of campaign communications like websites. Rather than decrying the ‘black arts of spin doctors’, the professionalization of political communications can be regarded as an extension of the democratic process if these techniques bind parties more closely with the concerns of the electorate. The key issue is less the increased deployment of marketing techniques per se, which is not in dispute, than their effects upon politicians and voters, which is.

The impact on civic engagement

This brings us to the issue at the heart of the debate: whether there is solid evidence that changes in political communications have contributed towards civic disengagement. Theories of media malaise argue that exposure to the news media discourages learning about politics, erodes trust in political 112 Pippa Norris leaders and government institutions, and dampens political mobilization. The net result, it is argued by proponents, has been a decline in active democratic citizenship. Extensive evidence cannot be presented within the space of a brief paper, but other work by the author (Norris, 2000) demonstrates that there is extensive evidence from a battery of surveys in Europe and the United States, as well as experiments in Britain, that cast strong doubt upon these claims. The results of the analysis show that, contrary to the media malaise hypothesis, use of the news media is positively associated with a wide range of indicators of political knowledge, trust and mobilization. People who watch more TV news, read more newspapers, surf the net and pay attention to campaigns are consistently more knowledgeable, trusting of government and participatory. This relationship remains signiﬁcant even after introducing a battery of controls in multivariate regression models. For example, Table 7.6 shows the model predicting campaign activism in the US, based on the 1998 NES. The

Table 7.6 Predictors of campaign activism, US 1998

Campaign activism Sig. Campaign activism Sig. (i) (ii) Structural Education 0.13 ** 0.04 Gender: male 0.09 ** 0.04 Age 0.08 * 0.03 Household income 0.08 * 0.15 ** Attitudinal Political discussion 0.12 ** 0.11 ** Lib-Con ideology 0.01 0.06 Use of news media Media news use 0.13 ** Newspaper 0.08 * National TV news 0.11 * Local TV news –0.01 Radio news 0.05 Net campaign news 0.12 * Constant –0.82 –1.01 R2 0.10 0.08

Notes: Columns report the standardized beta coefficients predicting campaign activism based on ordinary least squared regression models. The participation variable is the 6-point scale measuring attending a candidate meeting, working for a candidate or party, donating money to a candidate or party, displaying a campaign button, and talking to others for or against a candidate. Use of news sources are measured using 7-point scales. The overall Media Use index is a 29-point scale based on use of TV news + paper + radio news. Sig. P. **>0.01 *>0.05 For other details see Norris (2000), Table 13.5. Source: American NES, 1998, N.1,281 Political Communications in Post-Industrial Democracies 113 results confirm that attention to newspapers, network TV news and campaign news on the Internet is significantly associated with campaign activism even after controlling for social background. Similar positive relationships are evident in Europe and the US using multiple indicators of civic knowledge and political trust (for full details see Norris, 2000). Far from a case of ‘American exceptionalism’, this pattern is found in Europe and the United States. Repeated tests using different datasets in different countries across different time periods during the last half-century, confirm this positive relationship, even after controlling for factors that char- acterize the news audience like their education and prior political interest. The evidence strongly suggests that the public is not simply passively responding to political communications being presented to them in a naive ‘stimulus–response’ model; instead they are critically and actively sifting, dis- carding and interpreting the available information. A more educated and liter- ate public is capable of using the more complex range of news sources and party messages to find the information they need to make practical political choices. The survey evidence shows that news exposure was not associated with civic disengagement in America and Europe.

Conclusions: a virtuous circle?

Why should we find a positive link between civic engagement and attention to the news media? There are three possible answers, which cannot be resolved here. One interpretation is selection effects. In this explanation, those who are most predisposed to participate politically (for whatever reason) could well be more interested in keeping up with current affairs in the news, so the direction of causation could be one-way, from prior attitudes to use of the news media. This view is consistent with the ‘uses and gratification’ literature (Blumler and Katz, 1974) which suggests that mass media habits reflect prior predispositions in the audience: people who love football turn to the sports results, people who invest in Wall Street check the business pages, and people interested in politics read editorials about government and public policy. But if we assume a purely one-way selection effect, this implies that despite repeatedly turning to the news about public affairs, we learn nothing whatever from the process, a proposition that seems inherently implausible. Another answer could be media effects. In this explanation, the process of watching or reading about public affairs (for whatever reason) can be expected to increase our interest in, and knowledge about, government and politics, thereby facilitating political participation. The more we watch or read, in this interpretation, the more we learn. News habits can be caused by many factors such as leisure patterns and broadcasting schedules: people may catch the 114 Pippa Norris news because it comes on after a popular sit-com, or because radio stations air headline news between music clips, or because the household subscribes to home delivery of a newspaper. In this view, the direction of causality would again be one-way, but in this case running from prior news habits to our subsequent political attitudes. Both these views could logically make sense of the associations we establish. One or the other could be true. It is not possible for us, any more than for others, to resolve the direction of causality from cross-sectional polls of public opinion taken at one point in time. But it seems more plausible and convincing to assume a two way- interactive process or a virtuous circle. In the long term through repeated exposure, like the socialization process in the family or workplace, there may well be a ‘virtuous circle’ where the news media and party campaigns serve to activate the active. Those most interested and knowledgeable pay most attention to political news. Learning more about public affairs (the policy stances of the candidates and parties, the record of the government, the severity of social and economic problems facing the nation) reduces the barriers to further civic engagement. In this interpretation, the ratchet of reinforcement thereby moves in a direction that is healthy for democratic participation. In contrast, the news media has far less power to reinforce the disengagement of the disengaged, because, given the easy availability of the multiple alternatives now on offer, and minimal political interest, when presented with news about politics and current affairs this group is habitually more likely to turn over, turn off or surf to another web page. If the disengaged do catch the news, they are likely to pay little attention. And if they do pay attention, they are more likely to mistrust media sources of information. Repeatedly tuning out political messages inoculates against their potential impact. This theory cannot be proved conclusively from the available cross-sectional survey evidence, any more than can theories of media malaise, but it does provide a plausible and coherent interpretation of the different pieces of the puzzle found in this study. Claims of media malaise are methodologically flawed so that they are at best unproven, to use the Scottish verdict, or at worst false. As a result too often we are ‘blaming the messenger’ for more deep-rooted ills of the body politic. This matters, not just because we need to understand the real causes of civic disengagement to advance our knowledge, but also because the correct diagnosis has serious implications for public policy choices. This is especially important in newer democracies struggling to institutionalize a free press in the transition from authoritarian rule. ‘Blaming the messenger’ can prove a deeply conservative strategy, blocking effective institutional reforms, especially in cultures that idealize protection of the press from public regulation. Political Communications in Post-Industrial Democracies 115

This paper does not seek to claim that all is for the best in the best of all possible political worlds. If not ‘broken’, there are many deep-rooted ﬂaws embedded in the core institutions of representative democracy; we are not seeking to present a Panglossian view. The important point for this argument is that many failings have deep-seated structural causes, whether the ﬂood of dollars and lack of viable third parties in American elections, the wasteland of corruption and malfeasance in Russia, or the lack of transparency and accountability in Brussels. If we stopped blaming the news media’s coverage of politics and directed attention to the problems themselves, perhaps effective remedies would be more forthcoming.

References

Achille, Y. and Bueno, J. I. (1994) Les télévisions publiques en quête d’avenir. Grenoble: Presses Universitaires de Grenoble. Andersen, Svein S. and Eliassen, Kjell A. (1996) The European Union: How Democratic is It? London: Sage. Ansolabehere, Stephen and Iyengar, Shanto (1995) Going Negative: How Political Advertisments Shrink and Polarize the Electorate. New York: Free Press. Bennett, Stephen Earl, Rhine, Staci L., Flickinger, Richard S. and Bennett, Linda L. M. (1999) ‘Videomalaise Revisited: Reconsidering the Relation between the Public’s View of the Media and Trust in Government’, Harvard International Journal of Press/Politics, 4(4): 8–23. Blumler, Jay G. and Gurevitch, Michael (1995) The Crisis of Public Communication. London: Longman. Blumler, Jay G. and Katz, Elihu (eds) (1974) The Uses of Mass Communications: Current Perspectives on Gratiﬁcations Research. Beverly Hills, CA: Sage. Cappella, Joseph N. and Jamieson, Kathleen H. (1996) ‘News Frames, Political Cynicism and Media Cynicism’, in The Media and Politics, ed. Kathleen Hall Jamieson, Annals of the American Academy of Political and Social Science, vol. 546. Cappella, Joseph N. and Jamieson, Kathleen H. (1997) Spiral of Cynicism: The Press and the Public Good. New York: Oxford University Press. Crozier, Michael, Huntington, Samuel P. and Watanuki, Joji (1975) The Crisis of Democracy. New York: New York University Press. Curran, James and Seaton, Jean (1991) Power Without Responsibility: The Press and Broadcasting in Britain. London: Routledge. Curtice, John, Schmitt-Beck, Rüdiger and Schrott, Peter (1998) ‘Do the Media Matter?’ Paper presented at the Annual Meeting of the Mid-West Political Science Association, Chicago. Dahlgren, Peter (1995) Television and the Public Sphere. London: Sage. Dautrich, Kenneth and Hartley, Thomas H. (1999) How the News Media Fail American Voters: Causes, Consequences and Remedies. New York: Columbia University Press. Entman, Robert (1989). Democracy without Citizens: Media and the Decay of American Politics. Oxford: Oxford University Press. Fallows, James (1996) Breaking the News. New York: Pantheon Books. Franklin, Bob (1994) Packaging Politics. London: Edward Arnold. Gabler, Neil (1998) Life the Movie. New York: Alfred A. Knopf. 116 Pippa Norris

Hart, Roderick (1994) Seducing America. New York: Oxford University Press. Hart, Roderick (1996) ‘Easy Citizenship: Television’s Curious Legacy’, in The Media and Politics, ed. Kathleen Hall Jamieson, Annals of the American Academy of Political and Social Science, vol. 546. Hayward, Jack (1995) The Crisis of Representation in Europe. London: Frank Cass. Hill, Kevin A. and Hughes, John E. (1998) Cyberpolitics. New York: Rowman & Littleﬁeld, p. 44. Holtz-Bacha, Christina (1990) ‘Videomalaise Revisited: Media Exposure and Political Alienation in West Germany’, European Journal of Communication, 5: 73–85. Jones, Nicholas (1995) Soundbites and Spin Doctors. London: Cassell. Kaase, Max (2000) ‘Germany’, in Democracy and the Media: A Comparative Perspective, eds Richard Gunther and Anthony Mughan. New York: Cambridge University Press. Kovach, Bill and Rosenstiel, Tom (1999) Warp Speed. New York: Century Foundation Press. Lang, Kurt and Lang, Gladys (1966) ‘The Mass Media and Voting’, in Reader in Public Opinion and Communication, eds Bernard Berelson and M. Janowitz. New York: Free Press. Lowery, Shearon A. and DeFleur, Melvin L. (1995) Milestones in Mass Communication Research. New York: Longman. Lull, James and Hinerman, Stephen (1997) Media Scandals. Oxford: Polity Press. Murdock, Graham and Golding, Peter (1989) ‘Information Poverty and Political Inequality: Citizenship in the Age of Privatised Communications’, Journal of Communication, 39: 180–93. Newton, Kenneth (1997) ‘Politics and the News Media: Mobilisation or Videomalaise?’, in British Social Attitudes: the 14th Report, 1997/8, eds Roger Jowell, John Curtice, Alison Park, Katarina Thomson and Lindsay Brook. Aldershot: Ashgate. Norris, Pippa (1996) ‘Does Television Erode Social Capital? A Reply to Putnam’, PS: Political Science and Politics, 29(3). Norris, Pippa (1997a) ‘Political Communications’, in Developments in British Politics 5, eds Patrick Dunleavy, Andrew Gamble, Ian Holliday and Gillian Peele. London: Macmillan. Norris, Pippa (1997b) Electoral Change since 1945. Oxford: Blackwell. Norris, Pippa (1999) Critical Citizens: Global Support for Democratic Governance. Oxford: Oxford University Press. Norris, Pippa (2000) ‘Television and Civic Malaise’, in What’s Troubling the Trilateral Democracies, eds Susan J. Pharr and Robert D. Putnam. Princeton, NJ: Princeton University Press. Norris, Pippa (2001) Digital Divide: Civic Engagement, Information Poverty and the Internet in Democratic Societies. New York: Cambridge University Press. Norris, Pippa, Curtice, John, Sanders, David, Scammell, Margaret and Semetko, Holli A. (1999) On Message: Communicating the Campaign. London: Sage. Nye, Jr, Joseph, Zelikow, Philip and King, David (1997) Why People Don’t Trust Government. Cambridge, MA: Harvard University Press. Owen, Diane and Davis, Richard (1998) New Media and American Politics. New York: Oxford University Press, p. 185. Patterson, Thomas E. (1993) Out of Order. New York: Vintage. Patterson, Thomas E. (1996) ‘Bad News, Bad Governance’, in The Media and Politics, ed. Kathleen Hall Jamieson, Annals of the American Academy of Political and Social Science, vol. 546. Political Communications in Post-Industrial Democracies 117

Pew (1999) Striking the Balance: Audience Interests, Business Pressures and Journalists’ Values. Washington, DC: Pew Research Center for the People and the Press. Pfetsch, Barbara (1996) ‘Convergence through Privatization? Changing Media Environments and Televised Politics in Germany’, European Journal of Communication, 8(3): 425–50. Pharr, Susan J. and Putnam, Robert D. (eds) (2000) Disaffected Democrats: What’s Troubling the Trilateral Countries. Princeton, NJ: Princeton University Press. Plassner, Fritz, Scheucher, Christian and Senft, Christian (1999) ‘Is There a European Style of Political Marketing?’ in The Handbook of Political Marketing, ed. Bruce I. Newman. Thousand Oaks, CA: Sage. Postman, Neil (1985) Entertaining Ourselves to Death. New York: Viking. Putnam, Robert D. (2000) Bowling Alone. New York: Simon & Schuster. Robinson, Michael J. (1976) ‘Public Affairs Television and the Growth of Political Malaise: The Case of “the Selling of the President” ’, American Political Science Review, 70(3): 409–32. Robinson, Michael J. and Sheehan, Margaret A. (1983) Over the Wire and on TV: CBS and UPI in Campaign ‘80. New York: Russell Sage Foundation. Rosenbaum, Martin (1997) From Soapbox to Soundbite: Party Political Campaigning since 1945. London: Macmillan. Sabato, Larry (1988) Feeding Frenzy: How Attack Journalism has Transformed American Politics. New York: Free Press. Schulz, Winfried (1996) ‘Changes of Mass Media and the Public Sphere’, Javost – The Public, Theme Issue on Virtual Democracy, 3(1): 57–90. Schulz, Winfried (1998) ‘Media Change and the Political Effects of Television: Americanization of the Political Culture?’, Communications, 23(4): 527–43. Starker, Steven (1991) Evil Empires: Crusading Against the Mass Media. London: Transaction. Swanson, David and Mancini, Paolo (1996) Politics, Media and Modern Democracy. New York: Praeger. 8 Hacktivism: Direct Action on the Electronic Flows of Information Societies Tim Jordan1

Direct action in the streets is now mirrored by direct action in the networks. As the World Trade Organization (WTO) met in Seattle and there was tear-gas in the streets, thousands of people across the world attempted to bring computers serving the meeting to a halt through an electronic sit-in. Virtual and real activism have been joined in a direct-action hacking movement called hacktivism that is, perhaps, emblematic of the rise of information societies. While terms such as the ‘information society’ remain controversial, it is clear that information and communication technologies are a central and increasingly important part of current societies and it is within these technologies that hacktivists protest (Webster, 1995). The combination of popular protest with technologically sophisticated attacks on the electronic nervous system of global societies has significant implications for the development of both democracy and popular protest. There are then two questions posed by the rise of hacktivism: ‘what is the nature of hacktivism?’ and ‘what are hacktivism’s implications?’ It is the former of these that this chapter attempts to answer, with only brief comments possible in the conclusion that address the latter. Hacking has existed as long as there have been computer networks. Although hacking is most widely understood as illicit computer intrusion, it also has a wider meaning as the utilization of technology in novel or unfore- seen ways (Taylor, 1999; Jordan and Taylor, 1998). Hacktivism draws on this pre-existing hacking community and on links to 1990s protest movements. Popular mobilizations, particularly around ecological, animal rights and anti- capitalist political agendas, are the most recent social movements to have emerged since the 1960s and have begun to be called ‘DiY cultures’. The established community of hackers has, for the first time, become connected to the social movements that have dominated popular protest since the 1960s (McKay, 1998; Jordan and Lent, 1999; Anonymous, 1999). This paper maps the main outlines of this new, electronic direct action movement (Wray, 1998a). The argument will be developed by first briefly out-

118

K. Dowding et al. (eds.), Challenges to Democracy © Political Studies Association 2001 Hacktivism 119 lining the nature of the hacking community. This will provide a basis on which to understand the innovations hacktivism brings and the resources it relies on. It will also clarify what is meant by ‘a hack’ and the politics of hacking. The outline of the hacking community will make clear two types of hacks that have been claimed for and reinterpreted within hacktivism: the exploration of computer security and the defence of virtual cultures. Following this, hacktivism’s roots will be outlined and two new hacks will be defined in the politically motivated hack and electronic civil disobedience. Taken together these four different hacks constitute political actions typical of hacktivism. With the nature of hacktivism clearer, its connections to DiY culture will be outlined, before concluding by briefly introducing some general implications of hacktivism for democracy in the twenty-first century.

Hacking: communities, hacks and informational politics

Hackers form a community. Hackers run conferences, magazines, media campaigns, create places to meet (virtual and non-virtual), retell their history and argue over their politics. The hacking community is created through discussions and actions between its members and in a process of boundary formation between hackers and the computer security industry. Within the hacking community a collective identity is constantly constructed though ongoing actions around six themes: technology, secrecy, anonymity, membership fluidity, male dominance and motivations (Jordan and Taylor, 1998; Taylor, 1999). In each of these areas a particular ‘hacker’ understanding is developed. For example, hackers wish to remain both secret and anonymous to avoid ret- ribution for their hacks, but at the same time wish to publicize their achievements. They resolve this dilemma in a number of ways, for example by developing pseudonyms (or handles) to ‘sign’ hacks or by taking ‘trophies’ to prove they really did break into a difficult site. Across all six areas mentioned above, there is an ongoing process of identity construction that allows hackers to recognize each other as hackers (Jordan and Taylor, 1998, pp. 762–70). Hackers also identify themselves against an ‘outside’ that is largely constituted by the community which hackers are most in contact with, the computer security industry. The complexity of this relationship results from the difficulty of distinguishing the two communities on the basis of either typical actions or membership. Hackers often seek to become industry professionals and many professionals have a ‘hacker’ past; in addition many industry professionals admit to undertaking typically illicit hacker actions as part of their ‘legitimate’ business. Yet these are two often vociferously opposed communities. Differentiation is largely achieved by both communities developing metaphors for the hack and, crucially, imbuing these metaphors with an ethical component. Typically a computer security professional will describe 120 Tim Jordan hacking as a crime, like burglary, while hackers will describe it as an intellectual game, like chess. There is a wide ethical gulf between chess and burglary and it is through such metaphors that hackers and security professionals differentiate their communities (Jordan and Taylor, 1998, pp. 770–5). At the centre of the hacking community’s structures and activities is the act of hacking. Understood in its broadest sense a hack is putting technology to previously unimagined uses. For example, Count Zero from hacking group Cult of the Dead Cow (CDC) said, ‘I was always interested in how the innards of things worked (mechanical and electrical). Hacking was just a way of expressing that curiosity’ (Count Zero, 1999). This definition grasps the potentially broad nature of the hack which is currently interpreted in two main ways. First, the hack is most commonly and widely understood as the act of illicitly breaking into a computer via telephone or computer network, a usually illegal activity also called cracking. Examples of such hacks are numerous, from the rigging of a radio phone-in to ensure two hackers won red Porsches in successive contests to the defacing of the USA’s Central Intelligence Agency website so that it was entitled ‘Central Stupidity Agency’. Here high status hacks are defined by the difficulty of the target – it is a test of expertise (Jordan and Taylor, 1998; Littman, 1997; for websites see http://www.hackernews.com/defaced). The second and often less discussed sense of a hack is designing new software tools that are then distributed, usually also free and open-source. The development of such tools that include automated attacks for some complex assaults on computers and brute force attacks such as phone diallers (that program a computer to dial number after number seeking out a modem) has moved to the centre of the discussion of hacking in recent years. Howard demonstrates there is an increasing complexity in types of hacks over time and there is concern that automating difficult hacks is both ‘dumbing down’ the hacking community while increasing levels of hacking (Howard, 1998) A related way of distinguishing hacks is by noting there are ‘lone wolf’ or individualistic hacks and tool-based collective hacks. In the first case, there are hacks that could be undertaken by individuals. These include hacks such as the attack on the British Labour Party website that altered titles and pictures to lampoon the Party (for example, the link ‘Road to the Manifesto’ was changed to ‘Road to Nowhere’) and the reported cracking of Indian government systems in protest at nuclear tests by India (Miller, 1996; Glave, 1998). Such attacks are rarely actually taken by individuals – often hackers either hack physically together or are in constant touch online through chat-rooms or messaging. However, the essence of these hacks is that they could be conducted by an individual who, acting as a classic science fiction cyberpunk, takes action to disturb or affect a computer network. The second type of hack are those that cannot be undertaken unless there is mass participation or a collective effort. Typical here are hacks that rely on the distribution of a software tool that Hacktivism 121 enables people unskilled in hacking to take action. Hacking programs such as Back Orifice (discussed below) create their effects through both the collective effort to write and distribute the software and its wide use. The second type of hack cannot be undertaken individually and consists of the conjunction of expertly created tools with widespread distribution and use. Before grasping changes hacktivism has brought, it is also important to examine the nature of hacking politics prior to hacktivism. There are three main axes to hacking politics up to the late 1990s. First, hackers often discuss the service they are performing for other computer users when they identify security problems. Second, hackers defend their community and its cultures. Third, hackers have a distinctive ideology, blending anarchist and libertarian rhetoric and ideas with a focus on the world of information. All three of these are carried forward into and are changed by hacktivism. Hackers often see themselves as improving computer security, rather than compromising it, by identifying flaws in both particular sites and general systems. This view is reinforced by the practice of hiring teams of hackers to probe sites for security problems, seemingly legitimating what are otherwise criminal practices (McIntosh, 2000; Hencke, 1998). Actions exploring net security can be taken either collectively or individualistically. The main example of a collective hack in this style is Cult of the Dead Cow’s (CDC) Windows hacking software Back Orifice.2 This program allows any user of a computer attached to a Windows-based network to secretly access any other user’s machine. Someone using Back Orifice can read someone else’s e-mail, examine files or track someone’s use (either by watching that use or by recording it to be examined later) (Virus Bulletin, 1998a, 1998b; CERT, 1998). Back Orifice fundamentally compromises privacy and security on any computer attached to a Windows-based network and Back Orifice 2000 had been downloaded 128 776 times by late February 2000. The justification for the writing and release of this software is that it forces Microsoft and users of Microsoft’s operating system to face up to security issues that are normally hidden. CDC also point out that Microsoft attacks the stealth features of Back Orifice but that Microsoft includes as part of its Systems Management Server software tools that allow secret, remote access to computers. The difference between the two software packages is not in what they do but that Microsoft’s stealth features are highly likely to remain in the hands of those who administer a network, whereas Back Orifice is available to anyone. Distinguishing Microsoft from CDC here revolves around whether ‘ordinary’ users feel reassured or threatened that their systems administrators can secretly examine their computer activities and they cannot return the compliment. Whichever way one feels as a user, it is clear CDC is uncovering privacy and security issues CDC, 1999. Back Orifice is one of the clearest examples of a collective effort to uncover security flaws in networks and expose them to the world, and it is by no means the only example of such action. Countless individualistic hackers have also 122 Tim Jordan justified breaking into systems by claiming they have provided a service when they identify security flaws. Whether pursued individualistically by hacking computer after computer or collectively by writing and releasing software, the identification of security flaws and the meaning those flaws have for the privacy and security of individual users has been a central theme of hacking politics. Moreover, it is one that lasts to the present day and has been integrated within hacktivism. Back Orifice is an important example not only because of its nature but because it is currently active, with CDC explicitly proclaiming the launch of Back Orifice 2000 in 1999 as part of the hacktivist struggle. The second main axis of hacking politics prior to hacktivism is defence of the hacking community and its cultures. This defence is often understood and enacted with a very broad view. Hackers may take it on themselves to defend general Internet cultures as their own or, perhaps more accurately, hackers often understand general Internet cultures as hacker cultures. Incidents such as attacks on junk e-mailers or ‘spammers’, both in automated attacks that bulk e-mail the spammer or individual attacks on the spammers’ computers, have occurred (Taylor, 1999, pp. 59–64). A further example is the extended toywar in 1999, when hacktivists sought to support the long-standing online artists’ group etoy against an attempt by toy company etoys to take their domain name. The domain name was ‘www.etoy.com’ which the arts group uses for its main website. The toy company etoys tried to have the name taken away to try and retain control of their customers, who were only an ‘s’ away from the arts. Here hacktivists defended a fundamental online right, a virtual name which is equivalent to a virtual place (http://www.toywar.com). It is in these sorts of contexts that hackers are often treated as, or argue themselves to be, the guardians of ‘real’ virtual cultures. The third main political direction for hackers is the adoption of a number of political tropes and metaphors that articulate their beliefs. Hackers typically reach for anarchism and libertarianism to support claims that everybody has the right to cyberspace’s information and that cyberspace should be self- governed. This is most simply expressed in the almost universal hacker slogan ‘information wants to be free’ (Jordan, 1999, pp. 193–5) and in virulent opposition to governments, corporations or anything that is part of an often vaguely defined ‘system’. It is a rare hacker meeting place that does not contain files or discussions about anarchy or ring with sentiments like those of hacker Mentor:

We seek after knowledge … and you call us criminals. We exist without skin colour, without nationality, without religious bias. . and you call us criminals. You build atomic bombs, you wage wars, you murder, cheat and lie to us to make us believe that it’s for our own good, yet we’re the criminals. Yes, I’m a criminal. My crime is that of curiosity. (Cited in Sterling, 1992, p. 86) Hacktivism 123

Hackers articulate what might be called an informational anarchism that draws mainly on symbols and slogans from the Western traditions of anarchism and libertarianism, rather than a deep reading of either literature (liter- atures that are incompatible for many anarchists and libertarians). Count Zero from CDC expressed this informational politics this way: ‘Freedom of expression is very important to hackers. Anything that seriously stifles that is wrong’ (Count Zero, 1999). This informational anarchism/libertarianism sometimes descends into ominous but extremely vague attacks on an enemy of governments, security forces and corporations, but it also consistently uses anarchist/libertarian tropes to argue for the individual’s right to explore all information in cyberspace. An ideology of informational anarchism/libertarianism, a belief in the necessity of certain virtual cultures and a deep-rooted concern for the development of computer networks secure for the whole virtual world, largely form hacker politics up to the late 1990s. It is notable that this politics is focused on the virtual world; discussions of the state or corporations tend to occur only when these bodies interfere in online life. It might seem obvious that a community based on virtual activities would not see far beyond the virtual, but it is also striking that a division between online and offline runs through hacker politics up to the late 1990s locating its activism firmly within cyberspace. It is with the advent of hacktivism that such a focus on virtuality changes.

Hacktivism and new hacks

In 1998 the Electronic Disturbance Theatre (EDT) initiated a number of protests in support of the Zapatista movement, attempting to halt some government sites. Soon after:

On September 9, 1998, EDT exhibited its SWARM project … where it launched a three-pronged FloodNet disturbance against web sites of the Mexican presidency, the Frankfurt Stock Exchange, and the Pentagon, to demonstrate international support for the Zapatistas, against the Mexican government, against the U.S. military, and against a symbol of international capital. (Wray, 1998a, 1998b)

Prior to these actions politicized hacking, beyond Internet security or defence of virtual cultures, had been largely limited to isolated incidents or to the developments of ideas around hacking (CAE, 1994, 1996). Since these early times hacktivism has, first, brought the politically motivated hack to centre stage and, second, generated what appears to be a new form of hacking in electronic civil disobedience. Both these innovations rely on the pre-existing hacker community and build on pre-existing skills, but they also represent 124 Tim Jordan novel forms of politically charged hacking. When added to the already discussed hacks, we can see the full range of hacktivist actions. Individualistic hacking has gained new ideological purpose within hacktivism. For example, and though hard evidence is difficult to establish, attacks that announce a political message on websites seem to have increased. For a movement with an anarchist-inspired ideology, however fragmentary, it might seem surprising that politically motivated attacks are an innovation. Yet prior to hacktivism such ideological hacks were rare, whereas hacks that demonstrated someone’s abilities as a hacker were of primary importance. From this perspective what mattered was the difficulty of hacking a site and not the politics of a site’s owner. The numerous attempts to hack US military sites in the 1990s should be understood in this light as targeting what were assumed to be the best secured sites, not as protests against US militarism. What hackers have begun to do with hacktivism is to connect an organization that is online to its politics and sometimes to attack the online presence. At the 1999 launch of Back Orifice 2000, a CDC spokesman claimed, to general laughter, ‘I’m not going to tell you that web-site hacks for political point are wrong, but pick the cause before you pick the site you’re going to hack; like make it a little bit relevant’ (http://www.defcon.org/html/defcon-media- archives-defcon.html). Here the impact of hacktivism is registered along with the continuing focus of many hackers on the difficulty of hacking a site rather than its political character. One of the earliest examples of a politically motivated hack is the 1989 hacking of NASA in protest at the launch of nuclear powered space probes. Here NASA scientists had to sit and watch screens that told them their files were being deleted accompanied by the message that they had been ‘wanked’ (worm against nuclear killers). The program spread from computer to computer (making it technically a ‘worm’) paralysing systems as it went, though its claims to be delet- ing files were found to be false (Dreyfus, 1997, pp. 1–44). Individualistic hacks obviously have the potential for political use, but what is surprising is that hacks such as WANK were relatively uncommon. However, hacktivism is beginning to create hacks that are more about the politics of the target than the difficulty of the break-in. For example, in 1999, DoctorNuker of the Pakistan Hacker Club altered the US Marine Corps University page to call for the freedom of Kashmir, the Ku Klux Klan site was hacked to begin with the words ‘Your Webpage And All Accounts Associated With It Have Been Compromised, And Deleted, For Crimes Against The Human Race, by S C R E A M of the OLM (OnLine Mafia) and H.A.R.P (Hackers Against Racist Parties)’, and H.A.R.P. also attacked a ‘whitepride’ site.3 In 1998, British Hacker ‘jf’ reportedly added anti-nuclear protests to over 300 websites (Wray, 1998a). However, a brief review of archives of hacked sites conducted for this chapter found that clearly hacktivist hacks, that is web pages altered explicitly for political reasons, were in a minority with most hacks extolling hacking abilities. Archives of hacks demonstrate that while Hacktivism 125 hacktivism brings innovations it is also developing within the existing hacking community, and many in that community remain more concerned with demonstrating their expertise than with politics. While politically motivated hacks, particularly website hacks, have begun to appear in increasing numbers, a more collective effort has emerged, often called electronic civil disobedience. The Critical Arts Ensemble argues:

The strategy and tactics of ECD [electronic civil disobedience] should not be a mystery to any activists. They are the same as traditional CD [civil disobedience] … As in CD, the primary tactics in ECD are trespass and blockage. Exits, entrances, conduits, and other key spaces must be occupied by the contestational force in order to bring pressure on legitimized institutions engaged in unethical or criminal actions. Blocking information conduits is analogous to blocking physical locations; however, electronic blockage can cause ﬁnancial stress that physical blockage cannot, and it can be used beyond the local level. ECD is CD reinvigorated. What CD was, ECD is now. (CAE, 1996, pp. 11 and 18)

A brief example of this has already been given in the virtual sit-in against the WTO meeting in Seattle. A group called the ‘electrohippies’ created a web page whose function was to automatically and repeatedly load key index pages from the WTO conference website. Effectively each person who opened up the electrohippies page downloaded to their own computer a copy of the programme that repeatedly called up the WTO site. The electrohippies claim they interrupted the WTO servers on two days and significantly slowed the WTO network down on two other days. They also claim 450 000 people participated over five days (Electrohippies Collective, 2000). Another such action, requiring less technical support than a floodnet, was the ‘Global Jam Echelon Day’ on 21 October 1999. Echelon is a worldwide lis- tening system run by US security agencies in partnership with a small number of other governments. It attempts to record all telecommunications traffic and, it is believed, subjects this traffic to keyword searches. A call went out for everyone on 21 October to send as many e-mails as possible that included keywords that Echelon picks up and a list of 50 keywords, such as revolution and hacktivism, was thoughtfully provided. It was hoped this would flood and disable the Echelon system (Wright, 1998; http://www.echelon.wiretapped. net/original-release.html). Another example is a web page that generates a fax protesting the conviction and death sentence for journalist and former Black Panther Mumia Abu-Jamal (http://www.bckweb.com/mummia/). There have also been a series of attacks on Chinese Internet connections protesting the denial of freedom of speech. At one stage CDC announced a piece of e-mail software that would allow web pages to be requested and automatically e-mailed to the recipient. The motive was to subvert the Chinese govern- 126 Tim Jordan ment’s blocking of access for all China-based Internet users to certain web pages. (McKay, 1998; Hesseldahl, 1998). In all these, and other, examples the structure of protest is the provision of tools by a small group of ‘experts’ that reduce barriers to mass participation. The expertise may be as minimal as guessing the 50 words most likely to trigger Echelon’s surveillance or as complex as writing new software tools, but some provision of expertise consolidated into a tool is essential. Hacktivism’s most innovative protests join expertise in information technologies to mass action, creating a powerful conjunction of forces. Having outlined the typical actions hacktivists undertake it is now important to place these developments into the wider field of 1990s and twenty-first century popular protest.

Ideologies of hacktivism: hacking culture and DiY culture

Hacktivism did not emerge into a political vacuum. There are two ideologies whose combination has been crucial in forming hacktivism: the already discussed informational anarchism of hacking and the rise in the 1990s of what in the UK is called DiY culture. Enough has been said of hacking’s ideology to see two broad elements that are taken forward into hacktivism, alongside the two typical hacks that have been integrated within hacktivism. First, there is a belief in virtual or informational politics and, second, there is a commitment to anarchism and/or libertarianism. These two elements have themselves been transformed by a developing relationship to DiY culture that creates a loose, at times self-contradictory, set of ideas through which hacktivists can discuss which sites should be attacked or what actions are hacktivist actions. DiY culture can be broadly understood as social movements and popular protest that emerged in the late 1980s and particularly in the 1990s. Following from earlier social movements, such as feminism or black liberation, protests emerged that celebrate environmental activism, animal rights and anti-capitalism, that use a form of organization called disorganization and that integrate pleasure and symbolic politics (McKay, 1996, 1998; Jordan and Lent, 1999; Wall, 1999). DiY culture is now a global movement. Like all social movements, DiY culture does not consist of a single belief, organization or type of action. We must be careful to provide a characterization of it (and of hacktivism) that is not homogenized. However, a number of key themes that are clearly part of DiY culture have affected hacktivism and these will be briefly outlined. First, certain core political principles seem to be crossing between the two movements. From DiY culture there are two interrelated ethics that should be noted. First, and perhaps most important, is a broad set of issues that can best be called environmental. These cover as diverse topics as the destruction of rainforests, the building of roads in overdeveloped countries, the need for public transport, pollution, the ozone layer and so on. Taken together they all Hacktivism 127 in some way attempt to put the earth first. Second, increasingly anti-capitalist and anti-globalization rhetoric has developed (Jordan, 1998). This became abundantly clear in 1999 with the J18 day of worldwide action that saw linked protests around the world against the effects of globalizing capitalism and with the protests in November 1999 against the WTO in Seattle (Anonymous, 1999; Jordan, 1998). While hackers have not necessarily taken such ideologies as central, hacktivism has forged connections to DiY cultures that result in such ethics being used to generate political positions. The roots of hacktivism in pro-Zapatista campaigns, that have often targeted global neo-liberalism, solidifies this connection, as does the use in DiY cultures of similar anarchist language to hackers. Common metaphors and tropes of language enable common understandings. Second, DiY culture celebrates a type of organization that is best described by its name, do-it-yourself, though many activists now amend this to do-it-ourselves. This is the challenge to become active now! and is linked to a decentred and open form of organization usually called disorganization. This combination of an expectation that anyone concerned can be active and that the way to be active is to be open, inclusive and to avoid hierarchy forms a strong organizational ethic within DiY culture (McKay, 1998; Jordan and Lent, 1999). Hackers have for a long time subscribed to a similar ethic within their community, with the rationale of the hack being to take action to use technology in novel ways. Hackers have also created a community that is dispersed and networked rather than organized and hierarchical. At this second level of organizational culture, hacker and DiY cultures find they have much in common.4 Finally, DiY culture integrates pleasure into popular protest. Many of the forms of pleasure found within DiY cultures are derived from the dance floors and clubs where a form of party called ‘raving’, that developed in the late 1980s, spread through youth cultures generally (Jordan, 1998; McKay, 1996; Jordan, 1995). There is not the room here to rehearse the nature of raving and its contributions to dance cultures and popular protest – all that can be done is to note a series of pleasures of the dance floor that have been integrated into DiY cultures and are clearly well-known to hackers. For example, CDC’s announcement of Back Orifice 2000 at hacker conference DefCon7 began with visuals and music straight from the dance floor (http://www.defcon.org/ html/defcon-media-archives-defcon.html). Hacktivism is not simply unified; it does not base itself on any single manifesto or subscribe to any one theorist. What has been identified in this section is not so much a coherent set of beliefs as a series of ideas and commitments that circulate within hacktivism and DiY culture, and are taking on the shape of a movement’s ideology. The exchange in early 2000 between CDC and the electrohippies over the legitimacy of some forms of electronic civil disobedience, attacked by CDC for limiting even the enemies’ right to free flows of information, is a moment of development and division within hacktivism 128 Tim Jordan

(Electrohippies Collective, 2000; Ruffin, 2000). We must also be careful in assuming that the tools of hacktivism are only available to those subscribing to hacktivist ideologies. For example, there have been reports of anti-abortion hackers utilising hacktivist tools (Count Zero, 1999). However, by the early 2000s the following positions or values were characteristic of hacktivism: a concern for both Internet security and virtual cultures that results from a belief in free flows of information; a growing environmental consciousness and anti-capitalist politics; a commitment to taking direct action and organizing in fluid networks; and a commitment to certain cultures of pleasure.

Hacktivism: the movement

In times when our lives are increasingly dominated by electronically based exchanges, hackers can seem the most powerful villains of all. Hackers’ abilities to alter electronic resources that most of us simply access and use can make them into folk devils and objects of fear. Hacktivism, by adding a radical political message to this fear, faces even greater problems of exaggeration. Yet, even though hacktivism is still in formation, its fusion of hacking expertise and the politics of DiY cultures creates the basis of a potentially powerful movement in informational times, perhaps even as powerful as the fear of it might suggest. In understanding hacktivism we must be clear that it has arisen with one eye looking at the hacking community it came from and is still part of, and one eye looking at the radical politics of the 1990s. For example, Count Zero’s explanation of hacktivism focuses almost entirely on informational flows. He argues that ‘focusing on empowering the people in those places [he is referring to China and Iraq] with the TOOLS of hacktivism … making the WORLD know about the injustices and human rights abuses … in other words, getting the FLOW of INFORMATION pumpin’ around the globe … UNIMPEDED and UNCENSORED … THAT’S hacktivism . . .!’ (Count Zero, 1999 – … are included by Count Zero). In contrast, others such as the Critical Arts Ensemble or the Electronic Disturbance Theatre have stressed the role of hacktivism in supporting the Zapatistas or, more generally, in attacking globalizing capitalism. What is important is that these two sides can be both contradictory of each other and constitutive of hacktivism. For example, Wray notes one possible contradiction in that some hacktivists oppose floodnets because they take up large amounts of bandwidth and have the potential to slow the Internet for everyone. Wray calls this the ‘digitally correct position’ within hacktivism (Wray, 1998a). Out of these creative conjunctions has come a movement whose potential is enormous, because it plays in the electronic flows of information deeply embedded in millennial socio-economies. The implications for democracy in the twenty-first century of electronic direct action can only now be touched on, as what is meant by such action is only clear from within an understanding of hacktivism. Future work will have Hacktivism 129 to take account of the power of electronic action to subvert key elements of informational socio-economies. The democratic need to allow popular protest, both as a fundamental political right and as one mechanism for expressing and identifying social conflicts, is here potentially contradicted by the technical requirements for security and reliability of over-developed societies’ informational infrastructures (Melucci, 1996; Everard, 2000). How the needs of e-commerce for reliable transactions or of intellectual property for ownership of infinitely reproducible digital properties are going to be balanced against the ability of hacktivists to take electronic direct action are examples of some of the dilemmas democracy faces in the twenty-first century that are posed by hacktivism. The fear generated by those who play in the electronic nervous system of informational socio-economies combined with the fear eco and anti-capitalist activists generate produces a powerful nightmare for digital societies. Not only has hacktivism arisen, and appears to be here to stay, it also produces fundamental questions for democracy and popular protest.

Notes

1. Thanks to Christine Michael, Alan White, Helen Margetts, Roger Moore and discus- sants in BS for help with this chapter. 2. I am using the term Back Oriﬁce to refer to all versions of BO. 3. All hacked sites were viewed at www.hackernews.com/defaced. 4. Though, of course, this should not be overemphasized. Hackers have also developed a highly competitive culture based on expertise and such macho competitiveness is something DiY cultures tend to reject, at least rhetorically.

Bibliography

Anonymous (1999) ‘Friday June 18th 1999: confronting capital and smashing the state’, Do or Die, 8, 1–12. CAE (Critical Art Ensemble) (1994) The Electronic Disturbance. New York: Autonomedia. CAE (Critical Art Ensemble) (1996) Electronic Civil Disobedience and Other Unpopular Ideas. New York: Autonomedia. CDC (1999) Don’t Worry Windows Users, Everything Will Be BO2K. Press release from Cult of the Dead Cow, available at http://www.cultdeadcow.com/news CERT (1998) CERT Vulnerability Note VN–98.07. Available at http://www.cert.org Count Zero (1999) Interview with Count Zero, member of Cult of the Dead Cow, ABCNEWS.com. Available at http://www.abcnews.go.com/sections/tech/DailyNews/ chat_countzero.html Dreyfus, S. (1997) Underground: Tales of Hacking, Madness and Obsession on the Electronic Frontier. Kew: Mandarin. Electrohippies Collective (2000) Client-side Distributed Denial-of-Service: Valid Campaign Tactic or Terrorist Act?, Occasional Paper No.1. Available at http://www.gn.apc.org/ pmhp/ehippies Everard, J. (2000) Virtual State: the Internet and Boundaries of the Nation-state. London: Routledge. Glave, J. (1998) ‘Crackers: We Stole Nuke Data’, Wired News, available at http://www.wired.com/news 130 Tim Jordan

Hencke, D. (1998) ‘Whitehall Attempts to Foil Net Hackers’, Guardian Weekly, 26 April. Hesseldahl, A. (1998) ‘Hacking for Human Rights?’, Wired News, available at http://www.wired.com/news. Howard, F. (1999) ‘Out of the Mouths of Babes and Hackers’, Virus Bulletin, December, p. 2. Howard, J. (1998) Security Incidents on the Internet. Paper given at Inet98, Geneva, available at http://www.isoc.org/inet. Jordan, T. (1995) ‘Collective Bodies: Raving and the Politics of Gilles Deleuze and Felix Guattari’, The Body and Society, 1(1), 125–144. Jordan, T. (1998) ‘The Art of Necessity: the Subversive Imagination of Anti-road Protest and Reclaim the Streets’, in McKay, G. (ed.), DiY Culture: Party and Protest in Nineties Britain. London: Verso, pp. 129–51. Jordan, T. (1999) Cyberpower: the Culture and Politics of Cyberspace and the Internet. London: Routledge. Jordan, T. and Lent, A. (eds) (1999) Storming the Millennium: the New Politics of Change. London: Lawrence & Wishart. Jordan, T. and Taylor, P. (1998) ‘A Sociology of Hackers’, Sociological Review, 46(4), 757–80. Kahney, L. (1999a) ‘RealNetworks Probe Begins’, Wired News, 1/11/99, available at www.wired.com/news. Kahney, L. (1999b) ‘The Internet’s “Living Treasure” ’, Wired News, 2/11/99, available at www.wired.com/news. Levy, S. (1984) Hackers: Heroes of the Computer Revolution. London: Penguin. Littman, J. (1997) The Watchman: the Twisted Life and Crimes of Serial Hacker Kevin Poulsen. New York: Little, Brown. McIntosh, N. (2000) ‘Could you pass the tiger test?’, Guardian, 8 March. McKay, G. (1996) Senseless Acts of Beauty: Cultures of Resistance Since the Sixties. London: Verso. McKay, G. (ed.) (1998) DiY Culture: Party and Protest in Nineties Britain. London: Verso. McKay, N. (1998) China: the Great Firewall, available at http://www.wired.com/news. Melucci, A. (1996) Challenging Codes: Collective Action in the Information Age. Cambridge: Cambridge University Press. Miller, S. (1996) ‘Hacker takes over Labour’s cyberspace’, Guardian, 10 December. Rufﬁn, O. (2000) Valid Campaign Tactic or Terrorist Act? The Cult of the Dead Cow’s Response to Client-Side Distributed Denial-of-Service’, available at http://www.gn.acp.org/ pmhp/ehippies. Sterling, B. (1992) The Hacker Crackdown: Law and Disorder on the Electronic Frontier. New York: Bantam Books. Taylor, P. (1999) Hackers: Crime in the Digital Sublime. London: Routledge. Virus Bulletin (1998a) Virus Bulletin, September, p. 4. Virus Bulletin (1998b) Virus Bulletin, October, p. 4. Wall, D. (1999) Earth First! and the Anti-Roads Movement. London: Routledge. Webster, F. (1995) Theories of the Information Society. London: Routledge. Wray, S. (1998a) Electronic Civil Disobedience and the World Wide Web of Hacktivism. A Mapping of Extraparliamentarian Direct Action Net Politics. Paper presented at The World Wide Web and Contemporary Cultural Theory Conference, Drake University. Wray, S. (1998b) The Electronic Disturbance Theatre and Electronic Civil Disobedience, available at www.nyu.edu/projects/wray/EDTECD.html. Wright, S. (1998) An Appraisal of Technologies of Political Control: European Parliament Scientiﬁc and Technological Options Assessment, Working Document PE 166 499, European Union Directorate General for Research. 9 Deliberative Democracy and Referendums John Parkinson1

A deliberative ideal

The discovery that I can offer no persuasive reasons on behalf of a proposal of mine may transform the preferences that motivate the proposal. (Cohen, 1989, p. 144)

It has become almost commonplace to argue that direct democratic institutions, particularly referendums, embody the principles of popular control and political equality most closely. Bogdanor (1981) claims that ‘in the final analysis, the arguments against referendums are arguments against democracy’, while Saward (1998, p. 83) asks, ‘What better way to maximize responsiveness of rulers to the ruled than by fostering a system in which the ruled themselves make the decisions?’ Beyond these appeals to democratic ideals, some writers appeal to specifically deliberative ideals to support direct democracy. The argument is developed in two ways: the first route is to claim that referendums have a rationalizing effect on policy by allowing ordinary citizens to raise problems and enact solutions directly rather than filtering them through ‘distorting’ representative and bureaucratic structures (Butler and Ranney, 1994, p. 12); the second route is the argument that public debate has rationalizing effects on citizens. This claim is about more than democratic rights and legitimacy: it asserts that participation in public debate educates the people, transforming crude, individual preferences into considered opinions about the common good (Barber, 1984, pp. 235–6; Cronin, 1989, pp. 10–11; Pateman, 1970, p. 23; Schmidt, 1989, pp. 25–6). In this paper I question whether the deliberative defence of referendums stands up by assessing them in practice against a set of four ideal conditions derived primarily from Joshua Cohen (1989, pp. 146–7). These ideal condi-

131

K. Dowding et al. (eds.), Challenges to Democracy © Political Studies Association 2001 132 John Parkinson tions are: open participation; communicative competence, such that people have the motivation, skills and information to be able to listen to, weigh and make reasoned arguments; equality of resources, status and respect between participants; and the requirement to reach consensus.2 Taking these four conditions, I begin with a brief comparison of the deliberative ideal with Swiss direct democratic institutions, since Switzerland is most often held up as the paragon of deliberative virtue (Beedham, 1996, p. S4, Saward, 1998, pp. 95–100). I then examine a series of 27 immigration referendums and petitions in Switzerland since the late 1960s. I suggest that while referendums may appear deliberative in isolation, their social and institutional setting introduces fundamental distortions which make genuine deliberation highly unlikely.

How open is participation?

For a decision-making procedure to come close to the ideal, participation must be open to all. How open is participation in a Swiss referendum? The first major exclusion is ‘foreigners’. While most countries limit political rights to citizens and exclude other residents, it is difficult to gain citizenship in Switzerland – there is a 12-year residence requirement, with complex registration rules which vary by canton and by commune. There are further naturalization difficulties for the children of immigrants, difficulties which the government attempted to lessen in 1983 and again in 1994, both attempts being defeated by facultative referendums (for a full table of immigration votes and petitions, see the appendix to this paper). So, participation is limited to 60 per cent of the total population, compared with best-case 72 per cent eligibility in New Zealand which gives voting rights to all adult permanent residents. Along with restrictions based on criminal convictions and mental capacity, these are the only formal limits to participation. However, there are informal barriers as well. While one of the common justifications for direct democracy is that it will improve participation rates, the reverse seems to be the case in practice. Switzerland has seen a steady decline in participation rates over the decades, particularly in multiple-question ballots, with the average dropping from a peak of 64.6 per cent in the 1930s to 41 per cent in the 1990s. This means that, on average, a proposition requires not a majority of the Swiss people to carry the day, but just 12 per cent of the total population (Linder, 1994). Beyond the low overall participation rates, there are differences in participation levels between demographic groups. As in most Western democracies, voter participation is higher among older, more conservative, German- speaking, better educated and wealthier citizens, while younger, liberal, lower Deliberative Democracy and Referendums 133 income and less educated members of society participate less (Kobach, 1993, p. 80). While participation is quite open in theory, there is a clear bias in practice to those in privileged positions and those with conservative views (Kobach, 1993, p. 80; Linder, 1994, p. 95; Magleby, 1984, p. 121; Wilson, 1999). The door to direct participation is open, but it’s not open wide.

Do all participants have equal resources, status and respect?

In this section we are concerned with two related questions: beyond the inherent inequalities of the social structure, are ‘the people’ equally able to form groups and get propositions on the ballot; and is there equal power between the two sides once a proposition is accepted? One dimension is money. We lack Swiss expenditure figures because of financial secrecy laws, but can draw inferences from California where disclosure of the amounts and sources of funding is mandatory. First, it costs around US$1m to qualify a proposition for the ballot in California, because of the high number of signatures (419 260 for statute initiatives in 2000 and 2002) which must be collected in a short space of time, 150 days. The logistics needed to gather these 2795 signatures per day has meant that only well-financed groups with large networks, often using paid signature-gathering firms, are able to qualify initiatives for the ballot. In Switzerland, the thresholds are much lower: initiative proponents only need to gather 100 000 signatures and have 18 months to do so (180 signatures per day). It might be imagined that therefore the ‘average citizen’ is much more able to place an initiative on the ballot in Switzerland and yet experience shows otherwise. To take the example of immigration politics, every anti-immigration initiative has been promoted by a well-established political party. While a number of ‘initiative entrepreneurs’ (as Kobach calls them) have used the device, they have done so using pre-existing, national organizations often with larger parliamentary goals in mind. The influence of money in the success of actual campaigns has been hotly debated, but the consensus is that while money cannot necessarily buy a win, it can be very successful at buying a defeat. As Figure 9.1 shows, when the proponents of a measure in California outspend opponents by two-thirds to one- third, the likelihood of success is about even. When spending is more even, only a quarter of propositions succeed. When opponents outspend proponents by two-thirds to one-third, they defeat 87 per cent of the measures on the ballot. In Switzerland, Linder (1994, p. 144) estimates that well-financed sides win 80 to 90 per cent of the campaigns. The situation is further complicated by what is known as the Neinsager (‘nay-sayer’) effect in Switzerland (Kobach, 1993, pp. 48–9; 1997). This is a core of about 20 per cent of Swiss voters who will oppose the government view on any issue; the example is often given of the April 1990 ballot which 134 John Parkinson

90% 80% 70% 60% 50% Adopted 40% Failed 30% 20% 10% 0% Total parity Rough >67% by >67% by opponents proponents

Figure 9.1 Effects of spending on California proposition outcomes, 1954–82. (Source: Magleby, 1984, p. 148.) saw the rejection of six uncontroversial measures as a means of venting anger over a privacy-invasion scandal. Kobach suggests that the Neinsager phenomenon is partly the result of the ‘magic formula’ which fixes the allocation of seats on the Federal Council to the four main parties, making it almost impossible for Swiss voters to express dissatisfaction with the government by voting them out of office. Its effect is to strengthen the ability of moneyed or conservative interests to torpedo a measure. In this context it is interesting to note that, despite the apparent enthusiasm for initiatives in Switzerland, they have a dismal success rate. Only 10 per cent of initiatives succeed, compared with 63 per cent of government-initiated measures. The combined demands of extensive networks, campaign finance and the generally ‘conservative and obstructionist’ features of the referendum (Kobach, 1993, p. 27) make it unlikely that power is equalized between participants.

Are participants communicatively competent?

The next criterion is that, to reach a good decision, all participants need to be well informed, and fully capable of making and questioning arguments. Again, this is an area with multiple dimensions, but I will restrict my comments to two aspects only: the availability and usage of a range of information. There is some debate about how deeply observations about voter ignorance cut (Catt, 1996, p. 35): it may be that certain things which political Deliberative Democracy and Referendums 135 scientists hold dear are simply not salient. However, this still leaves open the question of whether, if an issue is salient, voters are well-informed. In the context of a referendum debate, the ideal is that people are well informed about the pros and cons of a ‘yes’ or ‘no’ vote, and have both the ability and the opportunity to question the arguments of others and put forward their own opinions. This is combined with the general assertion that the responsibility granted to ordinary voters transforms unthinking ‘opinion’ into informed ‘choice’ (Cronin, 1989, pp. 60–1). In Switzerland the Federal Council publishes a voter handbook for every referendum and makes it available on the Internet, but it does not give fiscal impact calculations (unlike the Californian model), gives only limited, government-authored summaries of the partisans’ cases, and puts the Federal Council’s and parliament’s own recommendations forward strongly. The Swiss electronic media is small and largely state-owned, comprising eight television and a dozen radio stations all with an obligation to try to achieve ‘balance’ in their coverage, which tends to repeat the handbooks’ lines (Internet Press, 1999; Linder, 1994, p. 109). The newspapers serve up a much more varied diet of political coverage. Historically, however, newspapers were affiliated with political parties, and while these ties have loosened significantly in view of a need to attract broader advertising support, newspapers still pursue a partisan editorial line in a similar fashion to the British press. While in the larger towns like Zurich and Geneva this presents less of a problem, there being a range of newspapers and views to choose from, smaller localities are often served by just one journal ‘which often prevents pluralism by suppressing views that do not correspond to its own’ (Linder, 1994, p. 110). This is to say nothing of the more general problem that readers are often unable to identify the political views of their paper, famously exposed in the UK in 1979 when a third of the readers of The Sun thought the paper supported Labour, despite its Conservative Party affiliations (Newton, 1986, p. 324). So, Swiss voters can gain access to a mix of expert and popular opinion, and partisan and independent analysis, if they make the effort. And yet, despite this information, only a quarter of voters in a 1977–80 study could remember the specifics of an issue shortly after the vote (Linder, 1994, p. 111). Further, the reliance on broadcast media, particularly by lower socio-economic group voters, indicates that they are relying on one-way sources of information: rather than actively creating a dialogue, the average voter in a referendum is more a passive receiver of information than communicatively competent. However, the one-way nature of the information is only part of the story, and not necessarily the most important part. As will be seen shortly, mediated information is distorted in ways which make it unlikely that even a two-way information flow would ‘enlighten’ voters. 136 John Parkinson

Do referendums deliver agreements approaching consensus?

The referendum is an inherently majoritarian device: it splits the range of opinion on an issue into two camps, favouring conﬂict over compromise. But there may be cases in which a referendum delivers a result somewhat approaching consensus. In Figure 9.2 I have counted how often initiatives in both California and Switzerland achieve different winning margins, ranging from the sharply divided (winner receives 50–55 per cent of the vote) to broad consensus (winner receives 80 per cent or more, which means even some of the Neinsagers have agreed). The graph shows how rare near-consensus decisions are in California. Almost a quarter of decisions (12 out of 49) can be characterized as sharply divided, with more than half (25) failing to reach 60 per cent support. Only nine polls achieved a vote of 70 per cent or more for one side or the other. However, there are some clear differences in Switzerland, with more than a third of the votes (11 out of 29) resulting in winner’s shares greater than 70 per cent, only one in the 50–55 per cent range and four in the 55–60 per cent bracket. Part of the explanation for the difference can be found in Switzerland’s consensual structures which force the Federal Council to work with the different parties, cantons and ofﬁcially recognized pressure groups (the Verbände) to mould actions that already have broad support, often deliberately to head off initiative action (Kobach, 1993, p. 60). By the time an initiative is organized

California (n=49) 6 Switzerland (n=29)

0 50–55% 55–60% 60–65% 65–70% 70–75% 75–80% 80%+

Figure 9.2 Initiative results by winner’s share of vote, 1992–98 (Sources: California Secretary of State, 1999; Centre d’étude et de documentation sur la démocratie directe, 1999). Deliberative Democracy and Referendums 137 and a vote put to the people, the government has already acted and the people reject the initiative (as in the immigration initiatives 220, 305, 320 and 432). Reflecting the dismal success rate noted above, almost all of the Swiss votes are defeats (only 3 successes out of 29), where the Californian initiative success rate is closer to 33 per cent. So, while the Swiss referendums appear to be more consensual, this has more to do with the federal structures than the devices of direct democracy. Majoritarian California, lacking such consensus-building institutions, shows voting patterns which reinforce the impression that the referendum is more likely to be a divisive tool than a uniting one. In terms then of a purely institutional analysis, referendums in Switzerland do not seem to match Cohen’s deliberative criteria at all well. Despite this, could it be that in actual debates over many years, voter preferences are transformed in public debate in ways which have not yet been captured by this analysis? I now turn to a brief summary of anti-immigration referendums to see whether such transformation has taken place and, if so, for what reasons. Along with much of the rest of Western Europe, Switzerland has seen a steady rise in the proportion of its population classed as ‘foreigners’, thanks to the deliberate attraction of migrant workers in the two decades following the Second World War, and to the welcoming of refugees during and after the Cold War. The proportion of foreigners now stands at 19 per cent of the total population, which is claimed to be the highest in Western Europe by a considerable margin (Layton-Henry, 1988; Swiss Federal Statistical Office, 1999). These foreigners have not been entirely welcome in Switzerland. Various groups have launched 17 different initiative petitions trying to limit their numbers and rights, and only two trying to expand them. While five of the ‘illiberal’ petitions failed to gather enough signatures, and two have yet to be put to the vote, ten anti-foreigner measures have been successfully placed on the ballot in national referendums. Yet of those ten, not one has been successful, and it is this statistic which is often cited in defence of the Swiss people’s deliberative abilities in direct democracy (Budge, 1996, pp. 96–7; Cronin, 1989, p. 123; Linder, 1994, p. 143). On this issue, it seems that Swiss ‘ordinary people are often surprisingly (to politicians) shrewd in their decisions’ (Beedham, 1996, p. S4).

The social construction of ‘Swiss’ and ‘foreign’

Swiss concepts of the ‘the Swiss people’ are partly the result of deliberative attempts to forge a unified national identity following long-standing rifts and a brief civil war in the mid-1800s. The Swiss self-image is based on ideals of fierce independence and individual will, focusing on the Alps and a rural lifestyle to symbolize both the strength of character of those who live among 138 John Parkinson the mountains and the strength of the barrier that the mountains present to, perhaps envious, outsiders (Arnaout, 1999; Besson, 1999; Linder, 1994, p. 17). The Swiss also have an image of themselves as guarantors of human rights throughout the world, and look back with pride on their role in allowing tens of thousands of refugees from the former communist Eastern Europe to enter and settle in Switzerland in safety (Tschopp, 1998). Over many years the Federal Council has affirmed a set of principles which guarantee asylum against persecution; active participation in aid programmes following war or natural disaster; provision of a temporary safe haven for foreigners if local conditions make direct intervention impossible; and various programmes to minimize the social, political and economic causes of involuntary migration (Federal Department of Justice and Police, 1999a). Another important element in the Swiss self-image is the concept of citizenship based on descent, ius sanguinis (Schmitter Heisler, 1988, p. 685). This contrasts with ius solis countries like the United Kingdom, where citizenship is claimed by place of birth and more immediate parentage. When it comes to citizenship and political rights, the child of Turkish parents born in London is British, not Turkish. This is an important distinction, because it calls into question the claim noted earlier that Switzerland has the highest proportion of foreigners in Europe: its figure is high, and the UK’s is low (3 per cent in 1988), partly because of Switzerland’s exclusive definitions of ‘Swiss’ and the UK’s inclusive definition of ‘British’ (Layton-Henry, 1988, p. 590; Swiss Federal Statistical Office, 1999).3 There is thus a tension between the parts of the Swiss self-concept that are exclusive and conservative, and the parts that celebrate humanitarian and democratic values. This tension is played out in the relationship between the Swiss people and their primary ‘other’: foreign residents. There are at least three ways in which the Swiss categorize foreigners, all linked to distinct historical moments.4 The first is the group of guest workers encouraged to come from Italy in particular between 1948 and 1960 as Switzerland experienced an economic boom as its neighbours rebuilt after the war. As time went by, many migrant workers brought their families to Switzerland and shifted out of low-skilled work into higher paid employment, competing more directly with ‘native’ Swiss (Schmitter Heisler, 1988, pp. 687–90). It was these migrant workers who were the target of the five initiatives ‘against foreign influence’ (see the appendix). The focal event which sparked initiative action was a 1964 treaty which loosened restrictions on Italian workers and made mention of the need to help new immigrants better integrate into Swiss life. This was the first official acknowledgement that guest workers were not ‘temporary’, an admission which sparked widespread unrest in German Switzerland (Smith, 1976, p. 16). Taking advantage of this focal event a local nationalist party, the Democratic Party of the Canton of Zurich, launched a petition to reduce the Deliberative Democracy and Referendums 139 number of foreign residents; the petition was withdrawn because it won assur- ances that the Federal Council would work to reduce the number of foreigners in Switzerland (Schmitter Heisler, 1988, pp. 690–1). These measures were seen as insufficient, and within two months of the first petition’s withdrawal the second initiative against foreign influence (no. 220) was launched by another Zurich-based group, Nationale Aktion gegen die Überfremdung von Volk und Heimat (National Action against the Over-foreignization of People and Homeland – note the distinction between ‘people’ and ‘foreigner’ in the title). The Schwarzenbach initiative, as 220 became known after National Action’s founder, demanded that the proportion of foreign workers be reduced to 10 per cent of the total population within four years, and despite opposition from ‘almost every conceivable Swiss organization’ (Schmitter Heisler, 1988, pp. 691–2) it was only narrowly defeated when it came to the vote in June 1970, gaining 46 per cent of the total vote and a majority in eight cantons on a very high 74.7 per cent turnout. Despite its apparent failure at the polls, however, it is crucial to note that 220 did not need to win a majority in order to change government policy. Faced with the threat of continuing initiatives from National Action, a threat made credible by the closeness of the result, the government introduced several measures to ‘stabilize’ the number of foreign workers (Smith, 1976, p. 16). So, it is misleading simply to say that the anti-foreigner initiatives of the early 1970s failed (Beedham, 1996, p. S4); they may have lost the vote, but because of the susceptibility of central government to interest group pressures they led indirectly to restrictive action. The other three initiatives ‘against foreign influence’ were defeated by significant margins because of the length of time it took to bring them to a vote. While the three petitions were launched in 1971, 1972 and 1973, the votes were not held until 1974 (for 242) and 1977 (for 265 and 266) and all were soundly defeated for reasons which are not entirely clear: the limited academic evidence credits the mid-1970s oil crisis and the new government immigration restrictions following 220 with creating economic conditions which encouraged many migrant workers to return home (Schmitter Heisler, 1988, p. 692), but local social movement activity provides more persuasive explanations for the rise and fall of anti-immigration politics than economic accounts (Karapin, 1999, pp. 423–30). Regardless, by 1974 and 1977 the Swiss electorate was less charmed with National Action and the crude xenophobia which had been the driving force more than ten years earlier. Interpreting the defeat of 242 this way, ten days later a coalition of churches and left-wing organizations launched the Mitenand (‘together’) initiative (305), which would have extended a range of rights to foreigners and cancelled some of the more restrictive laws regarding seasonal workers. The government agreed with the broad thrust of 305 and delayed setting a vote while it developed its own counter-proposals which aimed to harmonize Swiss 140 John Parkinson law with international immigration conventions and norms. By the time the government proposal was ready, six and a half years had gone by: 305 wasn’t put to a vote until April 1981, was deliberately paired with the government’s counter-proposal and thus, having lost momentum and its raison d’être, also lost the vote by a massive 84 to 16 per cent, with every canton voting against. The government counter-proposal was then successfully challenged by facultative referendum 309, the law failing to win by the slimmest margin, 49.6 per cent in favour, 50.4 per cent against (Federal Authorities of the Swiss Confederation, 1999). Given these two significant failures to liberalize immigration law, there are limits to any claims that the Swiss electorate had had its preferences transformed by the debates over the migrant workers of the 1970s.

The category reconstructed

The second category of foreign residents was perhaps the most welcome, precisely because as a category they were consonant with the Swiss self-image. These were the large numbers of refugees allowed into the country following the Soviet invasions of Hungary in 1956 and Czechoslovakia in 1968. These refugees fitted perfectly with the Swiss humanitarian ideal: European, intellec- tuals, civilized elites fleeing a repressive regime, they matched stereotypes of Nazi era refugees like Albert Einstein. They were relatively easily integrated into Swiss society and thus caused few conflicts with the Swiss sense of ‘proper behaviour’ (Besson, 1999). After that period though, a third group of immigrants started arriving, fleeing civil wars in Vietnam, then Angola and Sri Lanka, and more recently Iraq and the former Yugoslavia. This group in particular is talked about in very different terms: they are not ‘immigrants’, as are the Italians and other guest workers; they are not ‘refugees’ fleeing political oppression as were the second group; they are ‘asylum seekers’ and in the language of European immigration politics that label carries with it the slur of ‘economic’ refugee rather than ‘true’ political refugee status (Schmitter Heisler, 1988, p. 695; Steiner, 1999b, p. 47):

The German term Flüchtling (refugee) swells up sympathy within us and motivates us to help. Therefore, supporters of tighter asylum tend to avoid this term and instead use the pejorative Asylant (asylee) that inspires little sympathy. (Steiner, 1999a, p. 354)

Between 1989 and 1991 the Swiss government worked hard to accommodate over 100 000 asylum seekers – 41 629 in 1991 alone (Federal Department of Justice and Police, 1999b). The reaction came ﬁrst from another far-right party, the Swiss Democrats (SD), which launched restrictive petitions in Deliberative Democracy and Referendums 141

February 1990 and January 1991. The first failed to attract enough signatures, but the second was the subject of delays and an eventual invalidation by the Federal Council on the grounds that it was contrary to international law and Switzerland’s humanitarian duties and traditions (Steiner, 1999a, p. 358). The cancellation caused uproar in German-speaking Switzerland, but by then the SD had achieved two things. The first was further government-ordered restrictions to reduce the attractiveness of Switzerland as a destination by giving sweeping new powers to cantonal authorities to detain and deport illegal immigrants, and by requiring refugees to claim asylum in the first safe country they arrived in. These restrictions dramatically lowered the numbers of asylum applications until the collapse of Albania and the Kosovo crisis sent applications climbing again in 1997 and 1998. The second SD achievement was the major language shift in immigration politics noted above, from Flüchtling to Asylant, attaching to Asylant the concept of ‘economic’ rather than ‘political’ refugee, the idea of ‘abuse’ of Switzerland’s hospitality, and the idea of the Asylant as ‘criminal’. The latter is a particularly interesting feature of the immigration debate. The German-Swiss use the word ‘criminality’ to describe a very wide range of actions which include the English and French ideas of merely anti-social behaviour. Where the French-Swiss distinguish between criminalité and delinquance, the German- Swiss use just one word: Kriminalität (Menusier, 1999). There is some truth in the crime claims. In Geneva alone, 78 per cent of known offenders are foreigners (of whom half are non-residents), and 23.5 per cent are asylum seekers; a quarter of the population of Geneva’s largest prison are young Kosovars (Menusier, 1999). This, it should be noted, is in a canton which does not automatically regard every petty misdemeanour as a ‘crime’. This kind of activity is seen as abusing Swiss hospitality, giving ‘genuine’ refugees and the Swiss themselves a bad name. Indeed, criticism of recent asylum law is deeply resented by some Swiss, especially Federal Councillors who see themselves as trying to offer aid to those in need without opening the door to cynical criminals (Mamarbachi and Béguin, 1999; Steiner, 1999b, p. 61). On the other hand, given the exclusive concept of ‘Swiss-ness’ and the Asylant–Kriminalität linkage, it is likely that young Kosovars and other immigrants are targeted by the various branches of the Swiss police in ways that young ‘Swiss’ men are not; just as Hall noted the way that the British police deliberately targeted young black men during Britain’s anti-immigrant crack- downs of the late 1960s and early 1970s (Hall et al., 1978, p. 43). Foreigners may be grossly overrepresented in Swiss crime figures not because they are inherently more criminal, but because they are easily categorized and targeted. While it is important not to minimize the seriousness of some of these crimes, the problem has been framed in terms which lumps all unusual behaviour under the ‘crime’ umbrella; thus, the public debate in German-speaking 142 John Parkinson

Switzerland has discussed the ‘foreigner problem’ as if it were one of serious crime, and not mere failure to adapt to the norms expected of them by native Swiss. While Husbands (1988, p. 715) notes that Swiss anti-foreigner parties have always been seen as more about the protection of traditional values and ‘far less identified in the public mind with the extreme right’, this reframing of the immigration debate has served to make the anti-immigration position more mainstream. Most recently, the issue has been taken up for the first time by one of the four parties permanently represented in the Federal Council. This is the party that has one permanent seat, the right-leaning Schweizerische Volkspartei/Union Democratique du Centre (Swiss People’s Party in German, Centrist Democratic Union in French: SVP/UDC). Since 1992, the SVP/UDC and its leader, Christoph Blocher, have used immigration to advance their interests in Switzerland by promoting a number of initiatives and indirect law changes to toughen restrictions on asylum seekers. These have included the somewhat extreme initiative ‘against clandestine immigration’ (432) which was rejected partly because of the cost of the bureaucracy it would set up, partly because it would have centralized functions that were cantonal preroga- tives, and partly because the government had pushed through two law changes which implemented many of the SVP/UDC demands (Mamarbachi and Béguin, 1999). The new asylum laws were challenged by refugee agencies in facultatives 454 and 455, but the SVP/UDC successfully campaigned for a vote to keep the laws in place. These measures were certainly popular: in the 1999 federal elections, the SVP/UDC became the second largest party in the Swiss parliament on the back of its tough asylum stance. Its strength is now such that it is demanding a second seat on the Federal Council, the first time the power-sharing formula has come under serious pressure for change since it was created in 1959 (Federal Assembly, 1999; Linder, 1994). In this environment where the immigration issue has been reframed as being about guarding against ‘abuse’ and ‘crime’, opponents have had very little traction. The refugee agencies’ efforts to challenge the Asylant–abuse–crime linkage look like they are defending drug traffickers instead of ‘genuine’ refugees. Using the new construct, the Swiss can claim that they are not anti-immigrant, just anti-abuse. Thus it is not at all clear that the immigration debate has transformed opinions; rather, the debate has transformed the political language, allowing anti-immigrant measures to dress themselves in new, more acceptable clothes.

Conclusion

Ten times the Swiss have gone to the polls on CIR which would have limited the rights of immigrants, and ten times they have rejected those limitations. Deliberative Democracy and Referendums 143

However, the reasons for those rejections are hardly what a deliberative democrat might hope. They were not rejected because public debate had a rationalizing effect: sometimes it was because central government had already acted; sometimes because of fears of the further centralization of cantonal power, cost and the impact on taxes; sometimes because the language was transformed and they were not seen as anti-immigrant issues at all. Furthermore, there have been nine occasions on which the voters have either approved government restrictions or reversed federal laws liberalizing conditions for immigrants; there is only one case where a government-initiated liberalization has been approved (314). In all cases, the issues were based on a limited and politically powerful construction of what it means to be ‘Swiss’ and what it means to be ‘foreign’. Switzerland possesses unique consensual institutions leaving the centre in a much less powerful position to defend its attitudes. In that environment, even unsuccessful referendums can force change in government policy. CIR and federal institutions in Switzerland therefore present highly ‘favourable opportunity structures’ for anti-immigrant feeling (Karapin, 1999; Tilly, 1978) because the likelihood of effecting policy change is high regardless of whether one wins a vote or not, and because reframing the issue made it more acceptable, more mainstream, closer to the centre of power, more consonant with fundamental elements of the Swiss self-image. The device also provided ideal opportunities for ambitious politicians, from James Schwarzenbach and National Action to Christoph Blocher and the SVP/UDC, while opponent groups have had pitiful results. Two liberalizing initiatives have failed dramatically, one through deliberate government delaying and counter-proposal, the other through simple lack of signatures. There is not much in a Swiss referendum matching Cohen’s conditions for deliberation. Issues come pre-framed, wielding cultural categories in ways not readily open to challenge. Contra Cohen, preferences are not transformed by ‘the discovery that I can offer no persuasive reasons on behalf of a proposal’ but by pre-existing and unconscious values and emotions (Schon and Rein, 1994), the selective interplay of part-fact and half-truth that makes up mediated public debate. Therefore, while referendums may be promoted and defended on all sorts of grounds, the deliberative defence doesn’t stand up to close scrutiny.

Appendix: Swiss immigration petitions and referendums

Table 9.1 summarizes the key elements of all 31 Swiss federal initiatives and referendums on immigration and asylum, including initiatives which failed to reach the signature requirement or were withdrawn. Where possible, I have indicated whether the proposal was an extension or restriction of immigrants’ 144 0 cantons favour 5/6 cantons 17/19 & 8 cantons 0 cantons 16/19 & 5/6 cantons 0 cantons 0 cantons 46% 38.1% Sponsoring organization and aimsSponsoring organization Turnout % in all foreigner residents. measures. unknown). amendment (details As above, but amending the As above, but amending Unknown proponents, to removeUnknown proponents, 46% 15.9% amendment of the naturalization rules (details unknown). Democratic Party of the Canton of named after the leader of National 20/3/68 after government promised to increase efforts. Government-initiated constitutionalGovernment-initiated 68% 62.2%; Action, to reduce the proportion of foreigners to 10% in 4 years. Government initiative constitutionalGovernment initiative 45% 70.7%; number of foreigners; withdrawn National Action initiative to limit 70%National Action initiative to annul 34.2% 45% 29.5% The ‘Schwarzenbach initiative’, 75% 46% of foreigners nationality and expulsion and expel naturalization procedures length of stay and nationality and expulsionof foreigners constitution’s national security settlement of foreigners naturalization and the overpopulationof Switzerland (3rd) the total number of foreigners Switzerland (4th againstforeign influence) resident to 500 000. guest worker agreement with Italy. penetration (1st againstforeign influence) to reduce the Zurich petition (2nd) Initiative Against foreign 20/5/28 105 Obligatory Federal decree on 25/10/25 100 Obligatory Federal decree on the to vote Key elements of Swiss federal initiatives and referendums on immigration and asylum on immigration and referendums of Swiss federal initiatives Key elements Launch Vote No. Type Title Table 9.1 28/6/19 11/6/22 90 Initiative of Swiss Acquisition 28/6/19 11/6/22 89 Initiative of Swiss Acquisition 15/12/64 Not put 15/5/68 7/6/70 220 Initiative Against foreign influence 27/3/71 20/10/74 24230/6/72 Initiative 13/3/77 foreign influence Against 265 Initiative For the protection of 145 favour 0 cantons 0 cantons 40% 16.2% 35.2% 49.6% 36% 60.8% ) continued or ‘Together Mitenand immigrant and left organizations to immigrant and left organizations rights, guarantee basic human migration, choice of work, family right to social security coverage, renew work and residence permits, and remove laws on seasonal work. Government promised law change; implemented but successfully challenged by 309 below. Successful challenge to a government law which would have harmonized Swiss law with international treaties, spelled out legal rights and protection, allowed for family migration, and some social integration. Minor liberalization of rules (details removed from official database following the complete constitutional revision of 1999). Sponsoring organization and aimsSponsoring organization Turnout % in classifying Swiss nationality The National Action proposal to limitNational Action proposal 45% 33.8% foreigners new policy on foreigners Initiative’ supported by churches, nationality law (5th against foreigninfluence) more than 40 000 per year. number of naturalizationsnumber of naturalizations to no the 6/6/82 309 Facultative Revision of law on 4/12/83 314 Facultative Revision of the Key elements of Swiss federal initiatives and referendums on immigration and asylum ( on immigration and referendums of Swiss federal initiatives Key elements Launch Vote No. Type Title Table 9.1 30/10/74 5/4/81 305 Initiative Together in favour of a 15/3/73 13/3/77 266 Initiative For a limit to the annual 146 favour cantons 7/20 & 3/6 0 cantons 36% 44.8% 42.2% 65.7% ) , from continued Lex Friedrich Sponsoring organization and aimsSponsoring Turnout % in Unsuccessful challenge to two lawstightening previous conditions, such 42.4% 67.3% Successful challenge to a law Successful challenge and allowing simpler procedures for fewer barriers to naturalization and refugees, stateless persons children of immigrants. annual cap on real estate sales to annual cap on real estate enacted a foreigners, eventually version as the which National Action initiative signatures, failed to gather enough and declared invalid 4 August 1987. that asylum seekers must register as they cross the border, not in the canton of residence; police granted power to intern seekers refused asylum; but financial aid given to those deported. As above. immigration by not permitting more pejorative term for German tourists. pejorative term for German National Action initiative to limit 53% 32.7% National Action proposal to put anNational Action proposal 42.5% 48.9%, immigrants to enter than had emigrated the previous year, and to limit the numbers of refugees and regular cross-border commuters. contre ’ asylum laws naturalizations national regarding foreigners immigration the nation’s land ‘ le bradage du sol overpopulation Initiative Against foreign 4/12/83 315 Facultative Law allowing easier to vote 5/4/87 344 Facultative Revision of political 5/4/87 345 Facultative Revision of law Key elements of Swiss federal initiatives and referendums on immigration and asylum ( on immigration and referendums of Swiss federal initiatives Key elements Launch Vote No. Type Title Table 9.1 14/11/78 20/5/84 320 Initiative Against the selling off of 11/10/83 4/12/88 355 Initiative For limits on 14/1/86 Not put 147 -cantons 2 / 1 favour failed double majority: 2/6 9/20 full, 44% 72.9% ) continued Movement . seekers and immigrants, in order to limit the number of foreigners giving false identities or going into hiding to prevent expulsion, by giving cantons the power to detain and expel illegal immigrants or rejected asylum seekers, and to limit the movements of foreigners without proper permits. Sponsoring organization and aimsSponsoring Turnout % in years in Swiss schools. Challenge to a federal law declared invalid 5 December 1988. declared invalid 5 December Humaniste gather enough signatures, and gather enough signatures, and declared invalid 21 March 1989. and declared invalid Failed to Proponents unknown. procedures simpler for the children of immigrants, aged 15 to 24, who have been ‘integrated’ into Swiss life, and who have spent at least 5 amendment to make naturalization Government-initiated constitutionalGovernment-initiated 47% 52.8%, but regarding the law onforeigners tightening conditions for asylum of requests for asylum and gather enough signatures, values inherent in the Universal Declaration onUniversal Declaration Human Rights enough signatures, Failed to gather immigration of foreigners and asylumseekers declared invalid 21 August 1991. young foreigners Initiative limit to acceptance For a Failed to Proponents unknown. Initiative on the For education Initiative by the Initiative Against massive 4/12/94 417 Facultative of constraint Measures to vote to vote to vote 12/6/94 411 Obligatory Easier naturalization for Key elements of Swiss federal initiatives and referendums on immigration and asylum ( on immigration and referendums of Swiss federal initiatives Key elements Launch Vote No. Type Title Table 9.1 2/6/87 Not put 1/9/87 Not put 20/2/90 Not put 148 favour 10/20 & 2/6 canton 47% 46.3%; ) in continued Lex Friedrich Sponsoring organization and aimsSponsoring Turnout % in Challenge by the far-right SwissChallenge by the far-right liberalizing the –40% –46% minor ways, raising the limit on minor ways, raising the from property sales to foreigners simplifying 1500 to 4000pa and gains and procedures, for economic to buy land to allow overseas Swiss back home. 16/3/96, Council recommendation on the grounds that it contravened human rights and international norms. Schweizerische Volkspartei/Union Democratique du Centre (SVP/UDC)-sponsored initiative to prevent illegal immigrants from seeking asylum, restricting their means of appeal, and managing their income. Initiative by the SD to limit the Parliament following Federal Parliament following number of foreign immigrants in any one year to no more than the number of foreign emigrants in the previous year. Failed to gather enough signatures, and declared invalid 13 March 1997. by people overseas Democrats (SD) to a federal law immigration asylum immigration Initiative sensible policy on For a SD petition, declared invalid by Initiative For restraint in 25/6/95 424 Facultative Acquisition of property to vote to vote Key elements of Swiss federal initiatives and referendums on immigration and asylum ( on immigration and referendums of Swiss federal initiatives Key elements Launch Vote No. Type Title Table 9.1 15/1/91 Not put 21/4/92 1/12/96 432 Initiative Against clandestine 12/9/95 Not put 149 favour 43.8% 70.9% ) continued , or AMU) creating new arrêté fédéral sur les mesures refugees must seek asylum in the SVP/UDC initiative to limit SVP/UDC initiative specifying that first safe country they reach, still in Sponsoring organization and aimsSponsoring organization Turnout % in signature-gathering stage. federal law and an urgent decree federal law and an urgent ( d’urgence asylum; grounds for granting cooperation tightening identity and a rapid requirements; and creating expulsion system. As above. foreigners to 18% of the total population; Federal Council has recommended rejection, but a date for the vote has not yet been set. Challenge by refugee agencies to aChallenge by refugee 43.8% 70.5% right to asylum foreigners immigration 13/6/99 455 Facultative on asylum and AMU signatures 13/6/99 454 Facultative Asylum law to be set Key elements of Swiss federal initiatives and referendums on immigration and asylum ( federal initiatives and referendums on immigration Key elements of Swiss Launch Vote No. Type Title Table 9.1 1/3/94 Vote date25/5/99 Initiative Gathering For the regulation of Initiative Against abuses of the 150 John Parkinson rights and whether the proposal became or was confirmed as law. I have relied on official Swiss sources (Federal Authorities of the Swiss Confederation, 1999) which, before 1994, often do not list the sponsoring organization, membership of the referendum committees or details of the law being proposed or challenged. I have filled in some details from other sources, notably Schmitter Heisler (1988) and Kobach (1993), but in some cases the lists are contradictory: for example, Kobach’s numbering system differs from the official Swiss government list from the 1950s onwards, while Schmitter Heisler’s dates are sometimes at variance with both Kobach and the government list. In many cases further information is not available outside Switzerland, or in languages other than German. The first date is when the initiative petitions were launched, while the second is the date of vote; I have not included launch dates for the facultatives because they are invariably between three and five months prior to the vote date. The titles are author translations of the official French title rather than the German.

Notes

1. This paper draws together ideas developed while working with Helena Catt. I received valuable advice and assistance from John Dryzek, Claudia Heierli and Niklaus Steiner. To them, Michael Saward, the participants in the PSA 2000 conference, and everyone at the Political Studies Department, University of Auckland, I am profoundly indebted. 2. It may be that the consensus condition is a little old-fashioned: Dryzek (2000) notes most deliberative theorists have abandoned pure consensus in favour of more plausible ‘workable agreements’. I stick with Cohen’s word but use it more in Dryzek’s sense throughout. 3. Finding current, comparable statistics on foreign-national UK residents is itself difficult. The Office of National Statistics holds data on ethnic origin and residency but not citizenship, while the Immigration and Nationality Directorate of the Home Office does not maintain citizenship figures on its website (www.ind.org.uk). Based on 1998 ethnicity data, the non-European population of Great Britain is just over 6 per cent of the total, but of course many of these people are UK nationals, while others will be nationals of other European countries. 4. There is a fourth category focused on the somewhat tense relationship with Germany. The German-Swiss in particular distrust the economic and cultural power of their northern neighbours.

References

Arnaout, Ghassan (1999) ‘Notre conception de la Suisse est l’enjeu de la votation sur l’asile’, Le Temps, 26 May, asylum Internet dossier, http://www.letemps.ch Barber, Benjamin (1984) Strong Democracy: Participatory Politics for a New Age. Berkeley and Los Angeles: University of California Press. Beedham, Brian (1996) ‘A survey of democracy: “Happy 21st Century voters” ’, The Economist, 21 December. Deliberative Democracy and Referendums 151151

Besson, Sylvain (1999) ‘Les demandeurs d’asile sont devenus la cible privilégiée de la xénophobie suisse’, Le Temps, 31 May, asylum Internet dossier, http://www.letemps.ch Bognador, Vernon (1981) The People and the Party System: the referendum and electoral reform in British politics. Cambridge: Cambridge University Press. Budge, Ian (1996) The New Challenge of Direct Democracy, Chicago: Polity Press. Butler, David and Ranney Austin, (eds) (1994) Referendums around the World: the Growing Use of Direct Democracy. Washington, DC: AEI Press. California Secretary of State (1999) Internet site http://www.ss.ca.gov Catt, Helena (1996) Voting Behavior: a Radical Critique. New York: Leicester University Press. Centre d’études et de documentation sur la démocratie directe (1999) Base de données des votations fédérales. University of Geneva, Internet site http://129.194.47.76/LL/ Example/FSearch.html Cohen, Joshua (1989) ‘Deliberation and democratic legitimacy’, in R. E. Goodin and P. Pettit (eds), Contemporary Political Philosophy: an Anthology. Oxford: Blackwell. Cronin, Thomas (1989) Direct Democracy: the Politics of Initiative, Referendum and Recall. Cambridge, MA: Harvard University Press. Dryzek, John (2000) ‘Deliberative Economy and Discursive Legitimacy’. Paper read at ‘Deliberating about Deliberative Democracy’, 4–6 February 2000, at University of Texas at Austin. Federal Assembly (1999) The Swiss Parliament, Internet site http://www.parlement.ch/ Poly/Framesets/E/Frame-E.HTM. Federal Authorities of the Swiss Confederation (1999) Droits politiques dans le plan popu- laire, Internet site http://www.admin.ch/index.html Federal Department of Justice and Police (1999a) L’asile en Suisse, Internet site http://www.ejpd.admin.ch/Franz/themen/aktuelles/Asyl/asyl3_r.htm Federal Department of Justice and Police (1999b) L’asile en Suisse: statistiques, Internet site, 4 October 1999, http://www.ejdp.admin.ch/Franz/themen/aktuelles/Asyl/ Statistik/statistik.htm Hall, Stuart, Critcher, Chas, Jefferson, Tony, Clarke, John and Roberts, Brian (1978) Policing the Crisis: Mugging, the State, and Law and Order. London: Macmillan. Husbands, Christopher (1988) ‘The dynamics of racial exclusion and expulsion: racist politics in Western Europe’, European Journal of Political Research, 16: 701–20. Internet Press (1999) Switzerland, Internet site http://www.gallery.uunet.be/internet- press/swiss.htm Karapin, Roger (1999) ‘The politics of immigration control in Britain and Germany: subnational politicians and social movements’, Comparative Politics, 31 (4): 423–44. Kobach, Kris (1993) The Referendum: Direct Democracy in Switzerland. Aldershot: Dartmouth. Kobach, Kris (1997) ‘Spurn thy neighbour: direct democracy and Swiss isolationism’, West European Politics, 20 (3): 185–211. Layton-Henry, Zig (1988) ‘The political challenge of migration for West European states’, European Journal of Political Research, 16: 587–95. Linder, Wolf (1994) Swiss Democracy: Possible Solutions to Conﬂict in Multicultural Societies. New York: St. Martin’s Press – now Palgrave. Magleby, David (1984) Direct legislation: Voting on Ballot Propositions in the United States. Baltimore, MD: Johns Hopkins University Press. Mamarbachi, Esther and Béguin, Jean-Marc (1999) ‘Politique suisse d’asile: le débat’, Le Temps, 31 May, asylum Internet dossier http://www.letemps.ch 152 John Parkinson

Menusier, Antoine (1999) ‘ “Un requérant criminel n’est pas un requérant délinquant” ’, Le Temps, 5 June, asylum Internet dossier http://www.letemps.ch Newton, Kenneth (1986) ‘Mass Media’, in Henry Drucker, Patrick Dunleavy, Andrew Gamble and Gillian Peele (eds), Developments in British Politics 2. London: Macmillan. Pateman, Carole (1970) Participation and Democratic Theory. Cambridge: Cambridge University Press. Saward, Michael (1998) The Terms of Democracy. Cambridge: Polity Press. Schmidt, David D. (1989) Citizen Lawmakers: the Ballot Initiative Revolution. Philadelphia: Temple University Press. Schmitter Heisler, Barbara (1988) ‘From conflict to accommodation: the “Foreigners Question” in Switzerland’, European Journal of Political Research, 16: 683–700. Schon, Donald, and Rein, Martin (1994) Frame Reflection: Toward the Resolution of Intractable Policy Controversies. New York: Basic Books. Smith, Gordon (1976) ‘The functional properties of the referendum’, European Journal of Political Research, 4: 1–23. Steiner, Niklaus (1999a) ‘Arguing about international legal norms in parliamentary asylum debates’, Global Society, 13 (3): 349–66. Steiner, Niklaus (1999b) ‘Moral arguments in asylum debates’, International Politics, 36: 45–65. Swiss Federal Statistical Office (1999) Swiss Statistics, Internet site, 28 October, http://www.statistik.admin.ch/eindex.htm Tilly, Charles (1978) From Mobilization to Revolution. New York: Random House. Tschopp, Peter (1998) ‘Asile et minorités’, Tribune, Geneva, http://www.tribune.ch Wilson, David (1999) ‘Exploring the limits of public participation in local government’, Parliamentary Affairs, 52 (2): 246. 10 Creating Spaces of Deliberation in Barcelona: It’s Good to Talk? Georgina Blakeley

This chapter analyses three weaknesses of deliberative democratic theory,1 illustrated by empirical examples taken from a participatory experiment carried out in Barcelona in the 1980s and 1990s. This experiment occurred within a wider context of democratization of the Spanish polity and against a background of a mobilized and radical civil society with strong historical roots. As a result of this associational history, democratization at the local level in Barcelona went beyond the establishment of constitutional procedures which could guarantee elections and party competition: the key demand of social movement and associational activists was to render this constitutional democracy more participatory. This demand was adopted by the local political elites who gained power in the ﬁrst democratic local elections in 1979 and translated into government action. Among the many mechanisms of participation which the local council established, the most frequently used are a series of sectoral and territorial advisory councils which represent spaces of dialogue and interaction between the associational fabric of the city and the local council. These advisory councils question some of the key premises of deliberative democracy such as the idealized separation of public and private and the ability to bracket, or at least to neutralize, social and economic inequalities within deliberative forums. Moreover, these advisory councils highlight the importance of identifying the social foundations on which the deliberative ideal can potentially be realized. Deliberative democracy’s abstract framework of analysis, however, hinders this linkage between theory and praxis. Finally, the empirical evidence from Barcelona questions deliberative democracy’s validity as a model of democracy, rather than simply another mechanism among others that can help to extend and deepen democracy, given that it does not have the means to address the key question of how people can be encouraged to deliberate in the ﬁrst place.

153

The boundary problem in theory

Boundaries are constitutive of deliberative democracy in three ways: ﬁrst, as exempliﬁed in Habermas’s Theory of Communicative Action, deliberative democracy is based upon the separation of society into the lifeworld and the systems of the economy and the state, whose coordinating mechanisms are rational discourse, money and power respectively. Habermas’s account of the separation of spheres is an improvement on dichotomous approaches to the public and private given that it posits a relationship among four spheres – the family, the economy, the state and the public sphere – as well as high- lighting the overlap and interaction between them (Fraser, 1989, p. 41). Nevertheless despite the sophistication of Habermas’s account, the overriding concern is to prevent the encroachment of one domain onto another. The blurring of the boundaries between these domains represents a negative development which results in the ‘inner colonization of the lifeworld’ where the media of money and power literally invade the lifeworld (Habermas, 1989a, pp. 301–73). Habermas’s colonization thesis posits a kind of golden age in which the public sphere, free from state or market interference, was characterized by rational-critical discourse (Habermas, 1989b). Of course, such a thesis ignores the extent to which the public sphere was always characterized as much by negotiation as by rational-critical discourse as well as the extent to which these different domains have always overlapped. Nevertheless, Habermas upholds the liberal assumption that the separation of spheres is necessary for each to function adequately and for the adequate functioning of the polity as a whole. Habermas makes this point clearly:

Notwithstanding this discursive rationalization, only the administrative system itself can ‘act.’ The administration is a subsystem specialized for collectively binding decisions, whereas the communicative structures of the public sphere comprise a far-flung network of sensors that in the first place react to the pressure of societywide problematics and stimulate influential opinions. The public opinion that is worked up via democratic procedures into communicative power cannot ‘rule’ of itself, but can only point the use of administrative power in specific directions. (Habermas, 1996b, p. 29)

Habermas’s account of state/society relations ignores the extent to which power relations are as much a part of civil society as they are of the state and the economy. Moreover, by assuming that civil society can and should be defended against the encroachment of the state and the economy, Habermas fails to adequately understand the extent to which these spheres overlap as well as the nature of the interactions between them, which may be character- Creating Spaces of Deliberation in Barcelona 155 ized as much by conflict and contestation as by deliberation and discussion. In this respect, it may be appropriate to draw on Gramsci’s concept of hege- mony which could serve to highlight the power relations that are constitutive of civil society as well as the dual nature of deliberative arenas, which can be used both to contest the system as well as to secure its legitimacy. A Gramscian interpretation could also serve to highlight another significant omission within deliberative democracy, namely a lack of any explicit analysis of the state. While deliberative theorists assume that the role of the state is important in both the establishment and the maintenance of arenas of deliberation, this role is not adequately conceptualized and is generally regarded as unproblematic. Habermas sees the constitutional state as necessary to guarantee deliberative arenas. He takes from the republican model of democracy the importance of political opinion and will-formation, while from the liberal model he takes the centrality of the constitutional state, which he regards as ‘a consistent answer to the question of how the demanding communicative forms of a democratic opinion- and will-formation can be institutionalized’ (Habermas, 1996b, p. 27). Cohen likewise emphasizes the importance of institutional design for deliberative theory. He claims that ‘The institutions themselves must provide the framework for the formation of the will; they determine whether there is equality, whether deliberation is free and reasoned, whether there is autonomy, and so on’ (Cohen, 1991, p. 26). Neither Habermas nor Cohen, however, give any clues as to how and in what circumstances the state can fulfil these roles, nor the problems that may arise from doing so. The second boundary problem concerns the belief that participants in a deliberative process are able to bracket economic and social inequalities to allow the best argument, rather than status and power, to determine policy. This claim fits nicely with the liberal assumption of the autonomy of the political and stresses the high degree of overlap between political liberalism and deliberative democracy. However, not only is this claim empirically dubious in itself as both Marxists and feminists have demonstrated, it also ignores the extent to which ‘power sometimes enters speech itself’ (Young, 1996, p. 123). Foucault has also suggested that there is no such thing as a neutral discourse and that power relations of some kind are constitutive of all discourses (Foucault, 1990, pp. 92–3). In other words, the very practice of deliberation itself cannot presume to be neutral and value-free. Deliberative practices are generally culturally specific and often reflect underlying power configurations within society. Finally, in its bid to strike a balance between the interest-based politics of liberalism and the ethics-based politics of communitarianism, deliberative democracy is based on a separation between norms and procedures. Habermas claims that: ‘Discourse theory has the success of deliberative politics depend 156 Georgina Blakeley not on a collectively acting citizenry but on the institutionalization of the corresponding procedures and conditions of communication’ (Habermas, 1996b, p. 27). This stress on procedures acknowledges the diversity and plurality characteristic of modern societies which implies that as agreements are less likely at the level of substantive beliefs, they can only be achieved at the level of procedures. In other words: ‘Proceduralism is a rational answer to persisting value conflicts at the substantive level’ (Benhabib, 1996, p. 73). This separation is even more noticeable in liberal theorizing on deliberative democracy. In contrast to the relatively open agenda of public debate which Habermas’s model allows, Rawls (1997) is concerned to specify limits to the use of his concept of public reason which he applies only to political questions involving ‘constitutional essentials’ and ‘questions of basic justice’. This purposely excludes principles regulating matters of distributive justice for example. Ackerman (1989) also offers a restrictive idea of deliberative democracy in his model of ‘conversational restraint’, which excludes what he calls ‘value judgements’ from public debate. The extent to which norms and procedures can actually be separated in this way is, however, debatable. McCarthy cautions that: ‘The separation of formal procedure from substantive content is never absolute: we cannot agree on what is just without achieving some measure of agreement on what is good’ (McCarthy, 1992, p. 62). Just as deliberation itself cannot be regarded as neutral and value-free, procedures too will be a reflection of the predominance of some norms over others.

The boundary problem in practice

The advisory councils in Barcelona illustrate quite clearly these points: they are not based upon an ideal separation of civil society and state but rather on a relationship of mutual interdependence and collaboration, which recognizes that associations and local government need each other. Indeed, the majority of advisory councils in Barcelona have been created by the city council itself. As such they are all essentially ‘municipal organs’ and are largely financed and managed by the city council.2 Moreover, it is no coincidence that the first advisory council created was in the field of social welfare policy as this was precisely where the need to enlist the cooperation and support of associations was considered paramount. The collaborative relationship the city council has tried to promote between associations and the local administration is viewed as an essential mechanism for maintaining or advancing the standards of social welfare within the city. Without the cooperation of associations in terms of co-managing projects and services, it would be difficult to maintain, let alone advance, in social welfare provision. The establishment of this kind of reciprocal relationship between Creating Spaces of Deliberation in Barcelona 157 local administration and associations is also viewed as part of a long-term political strategy of maintaining Barcelona as a city with certain standards of social welfare whether or not the Right or the Left is in government. This growing interdependence between state and civil society, and the increasingly blurred boundaries between them, supports Fraser’s argument that Habermas’s division of society into what she calls ‘weak publics’ which are limited to opinion formation and ‘strong publics’ whose deliberation can result not just in opinion formation but also in decision-making is ultimately a negative and restrictive division (Fraser, 1992, p. 134). Such a division does not allow us to envisage other, perhaps equally valid, modes of coordination between the state and civil society. As the advisory councils have developed, new experiences have arisen which suggest that associational activists have both the capacity and the desire to go beyond a discursive role of opinion formation to roles of co-management and decision-making within certain spheres of municipal action.3 Pace Habermas, those active in deliberative arenas do not necessarily wish to be spared ‘the burden of decision-making’ nor do they share Habermas’s view that ‘the capacity of the public sphere to solve problems on its own is limited’ (Habermas, 1996a, pp. 362 and 359, italics in the original). The importance of the state for both establishing and maintaining public spheres is also evident in the Barcelona model as are the difficulties inherent in this state role given that the majority of the advisory councils are heavily dependent on the financial support of the state to survive. While this is not problematic per se, it can be so if associations become concerned with shad- owing government projects at the expense of promoting their own alternative projects. If associations become preoccupied with managing or delivering services, they may neglect their alternative roles of holding the city council accountable, of offering alternatives or of promoting their own autonomous projects, not least because associations have neither the time nor the resources to do everything! Pindado maintains: ‘With its own autonomous project, an association can collaborate on equal terms with the administration or even be one step ahead of municipal policy. Without that autonomous project, there is a danger that the association will end up being swallowed up by the administration’s project.’4 The Barcelona model also casts doubt on the deliberative democracy claim that participants within deliberative forums are able to bracket social and economic inequalities in order to allow reasoned argument to provide the only criteria for policy-making. Empirical evidence suggests that this is not only a careless treatment of power but also a simplistic interpretation of it. ‘Informal’ barriers to participation may be present within organizational structures, protocols of style and expected behaviour as well as within the practice of deliberation itself. It is necessary to find mechanisms to deal with these 158 Georgina Blakeley

‘informal’ or ‘concealed’ power configurations, rather than simply believing that we can bracket them. The participation of women in the advisory councils shows quite clearly to what extent inequalities of all kinds can hinder deliberation even when all formal barriers and exclusions have been removed. During the first Congress of the Women of Barcelona in January 1999 a workshop on participation took place which showed that some of the protocol and language used in the councils were barriers to women’s participation. Some felt that the administration treated the women’s groups as ‘guests’ and therefore it was they as the ‘invi- tees’ who had to adapt to the former’s way of doing things rather than vice versa. In addition, many women experienced lack of time as an obstacle to their participation exacerbated by the council’s excessive dependency on meetings and on written communication. This last point highlights the artificial distinction between norms and procedures referred to earlier. Procedures too are a reflection of underlying norms and structures of power favouring some interests at the expense of others. As one interviewee commented, ‘women are still not able to participate on an equal footing with men. For example, the women who tend to participate more are either generally older women who no longer have family responsibilities, or professional women with the economic resources to permit their participation.’5 The fact that neither formal nor informal inequalities can be either brack- eted or neutralized also points to the importance of creating counterpublics (Fraser, 1992, p. 123). While shared spaces of dialogue and interaction are needed between associations and administration, it is equally important for associations to have their own space to interact among themselves within which their own proposals and projects can be articulated. For associational activists to be able to have a real chance to influence the administration rather than to simply respond to whatever the administration puts forward, i.e. to be proactive rather than reactive, associations need their own space, free from all the administrative machinery, in which to articulate their ideas prior to engaging with the council.

The abstraction problem in theory

The second key weakness of deliberative democratic theory is its abstract framework of analysis, which all too often seems far removed from the practice of daily life. Of significance here is the extent to which Habermas revised his earlier theory in The Structural Transformation of the Public Sphere, which was far more historically and socially specific, to produce a communicative theory of action which is not tied to any particular historical moment. Calhoun notes that in his later work, ‘Habermas shifts his attention from the institutional construction of a public sphere as the basis of democratic will Creating Spaces of Deliberation in Barcelona 159 formation to the validity claims universally implicit in all speech’ (Calhoun, 1992, p. 31). While Habermas’s later theoretical framework undoubtedly gains in terms of universalism and general validity, it nonetheless loses in terms of linkage to praxis. Historical references and arguments are replaced by an abstract, philosophical argument. What is also lost in this shift in Habermas’s writing from the location of practical reason within historically specific social institutions in the public sphere to the location of practical reason within the universal, transhistorical communicative practices in the lifeworld, is the focus on agency. Any notion of a ‘macrosocial subject’ is supplanted with the idea of ‘subjectless communications’ (Leet, 1998, p. 91). This denotes, in particular, an implicit recognition of the loss of faith in the working class as the historical subject. The idea of ‘subjectless communications’, however, does not provide any satisfactory basis to guide political praxis which, by contrast, requires that political agency be identified and accounted for. If we require theory to help to change practice, rather than simply to describe it, a more historically and contextually specific analysis is required. The increasing pessimism in Habermas’s thinking is also reflected in a nar- rowing of the aims of his democratic project. Habermas rejects his earlier conceptualization in The Structural Transformation of the Public Sphere suggesting that both the state and the economy could be internally democratized. In the light of the failures of state socialism he no longer believes it possible to democratize from within the systemically integrated action fields of either the state or the economy without damaging their ability to function. Nor does he believe any longer in the possibility of democratizing formally organized parties and interest groups, instead all his faith is placed in the sponta- neously formed, informal public spheres whose ‘wild’ and ‘anarchic nature’ he celebrates (Habermas, 1996a, p. 307). Rather than transformation we are now looking at a more defensive, protective mechanism of boundary maintenance, to prevent or limit the encroachment of these systematically integrated action fields on the lifeworld. Habermas’s new vision of the task of radical democracy is ‘no longer to supersede an economic system having a capitalist life of its own and a system of domination having a bureaucratic life of its own but to erect a democratic dam against the colonializing encroachment of system imperatives on areas of the lifeworld’ (Habermas, 1992, p. 444). However, pace Habermas, I believe that his concept of the public sphere as elaborated in The Structural Transformation of the Public Sphere is a far more useful starting point for democratic theory than his later theory, precisely because the former combines normative ideals with a grounding in a particular historical and social setting. Without such a concrete historical grounding, we run the risk of being left with a normative ideal that tells us nothing about 160 Georgina Blakeley the social conditions which can facilitate its realization. Habermas writes that ‘Informal public-opinion formation generates “influence”; influence is transformed into “communicative power” through the channels of political elections and communicative power is again transformed into “administrative power” through legislation’ (Habermas, 1996b, p. 28). But at no point in this account is there an explanation of how this happens, in what circumstances this happens and under what conditions. Yet these are precisely the kinds of questions that need to be investigated by empirical research. Deliberative democracy requires this concrete historical grounding if it is not to become an example of the worst kind of abstract theorizing constructed from the top down. Habermas’s abstract framework of communicative action, however, does not allow us to ask how and in what circumstances the public sphere is capable of detecting social problems, sharpening them into a ‘consciousness of crisis’ and introducing them into the parliamentary context ‘in a way that disrupts the latter’s routines’: this can only be done via an empirical examination of a specific, historically grounded civil society (Habermas, 1996a, pp. 357–8, italics in the original). Yet Habermas eschews such an empirical examination stating:

In the present context, of course, there can be no question of a conclusive empirical evaluation of the mutual inﬂuence that politics and public have on each other. For our purposes, it sufﬁces to make it plausible that in a perceived crisis situation, the actors in civil society thus far neglected in our scenario can assume a surprisingly active and momentous role. (Habermas, 1996a, p. 380, italics in the original)

Other authors seem similarly disinclined to tackle empirically the way in which the ‘indirect’ influence of associations within civil society can actually influence the state apparatus by means of changing attitudes and values. McCarthy (1992, p. 63), for example, states ‘I won’t raise here the important question of how effective voluntary associations and social movements can be in monitoring and influencing the formal decision-making processes of a systematically integrated economy and state’. Benhabib (1996, p. 83, italics added) also skates over the issue of exactly how Fraser’s idea of counterpublics to renegotiate the public and private distinction do so in reality, writing: ‘It is nonetheless a long step from the cultural and social rethinking and reformula- tion of such distinctions as between the public and the private to their implementation in legislation and government regulation.’ It is precisely this ‘long step’, however, which deliberative theorists should be concerned with if they are to answer the question of exactly what influence deliberation has and why we should consider it important in the first place. Creating Spaces of Deliberation in Barcelona 161

The abstraction problem in practice

An empirical examination of the extent to which deliberative arenas can influence the sphere of the administration is necessary to understand both the possibilities and limitations of deliberation within a liberal democratic context. In the case of Barcelona, it is clear that there are both advantages and disadvantages arising from deliberation in the advisory councils. The key benefit for the city council is that by implicating different associations in the elaboration of its policies, it gains valid interlocutors and legitimators for its activities. Advisory councils can serve to generate an ample consensus on the city council’s policies and activities, thereby increasing what the council refers to as the ‘social profitability’ of its actions. For the associational fabric, the chief importance of the advisory councils lies in creating ‘public spheres’: spaces for dialogue and interaction, not just between themselves and the administration, but among themselves. The importance of facilitating dialogue between diverse social actors can help increase comprehension of discrete viewpoints even if it does not change them. As one trade union representative commented with regard to their participation within the Economic and Social Council:

It is a privileged framework which represents a daily contract with the bosses which is useful in many different ways. For example, by explaining your reasons for solidarity with Cuba, you can stop the employers from boycotting your solidarity campaign even if they don’t actually participate in it. You can therefore avoid zero-sum positions.6

Many participants also stressed the educative experience gained from working together in the advisory councils not just in the narrower sense of personal enrichment, but also in the wider sense of learning to appreciate the wider picture, rather than their own particular viewpoint. With time, participants came to regard the advisory councils, not as a space or mechanism with which to promote their own particular interests, but rather as one with which to promote the common good for the city as a whole. Finally, it is undoubtedly true that unless participants felt that their recommendations and proposals were incorporated, at least to some degree, into municipal policy, their continued participation would be difficult to explain, not least because such participation entails devoting considerable time and energy on a voluntary basis. Nevertheless, the limitations of such influence are also obvious. Key problems arise once the proposals and recommendations made by the advisory councils touch on areas outside the jurisdiction of the city council. Although participants were generally positive about their ability to influence municipal policy, it was acknowledged that it was considerably more difficult, if not 162 Georgina Blakeley impossible, to influence policy made at either the level of regional or national government. Even within the city itself there is a real difference between the influence of city-level and district-level advisory councils. The impact of the latter on municipal policy is far less than that of the former given that there is less decision-making capacity at the district level. Yet the advisory councils at district level are precisely where ordinary citizens have the greatest chance to participate given the closer proximity of the district to the citizen. City-level advisory councils, on the other hand, while more influential, are where experts or professionals take part, rather than the ordinary citizen. Women too have questioned their ability to influence decision- making. Despite significant successes women generally felt that the proposals and suggestions arising out of the women’s advisory councils did not have a real influence on municipal policy. ‘What stands out in terms of municipal policy is the sheer invisibility of women’s lives, experiences and demands’ (Primer Congres de les Dones de Barcelona, 1998, pp. 19–21). Pindado (1997, p. 95) concurs:

The real centres of political debate about the city and its problems and solutions and decisions on strategy, continue to be centralised and are not very open to citizens’ interventions. Proximity between administration and associations exists for concrete or small issues, but there is a distancing with respect to big plans and strategic initiatives.

Such difficulties are related to what Fraser (Fraser, 1992, p. 131) labels the ‘rhetoric of domestic privacy’ and the ‘rhetoric of economic privacy’: the boundaries which are drawn around certain issues, thereby excluding them from public debate, under the guise that they are matters for the individual, the market or private ownership. Various women’s associations, for example, claimed that issues of concern to them were frequently confined to their own public spheres, rather than being transversal. Likewise, decisions concerning controversial issues, particularly large infrastructural projects likely to have significant environmental impact such as expanding Barcelona’s airport, are excluded from the agenda of these deliberative arenas. Such decisions are taken elsewhere in non-public forums and although they do not receive the legitimization of the advisory councils, such projects tend to go ahead regardless given the economic power of those involved. The celebration of the Olympic Games in Barcelona in 1992 represents a key example of the limitations of deliberation when real-estate interests are at stake. Notable defeats for the neighbourhood associations in this respect included the Hotel Plan which requalified areas originally intended for public facilities and green zones (1989) and the failure to gain social housing in the Olympic Village (1990) (Huertas and Andreu, 1996, p. 36). In other words, whereas spaces of delibera- Creating Spaces of Deliberation in Barcelona 163 tion may be of crucial significance for subordinate groups, they are not essential for the advantaged: a complementary mechanism which they may, or may not, avail themselves of, but not an essential tool. Bohman’s (1997, pp. 338–9) caution is pertinent:

Powerful groups attain a greater extent of social freedom by excluding many topics from public debate by implied threats and other non-deliberative means. [. . .] In these cases, public functioning is not required for the groups whose interests carry the day for structural reasons, since their credible threats circumvent the need to convince others of the reasons for their policies.

The model of democracy problem in theory and in practice

The final area of weakness within deliberative democratic theory is the confu- sion surrounding the ultimate pretensions of deliberative theory: is deliberative democracy a model of democracy as many theorists purport, or is it simply an element within a wider theoretical framework? In many respects, the various labels which precede democracy such as ‘participatory’, ‘aggregative’ or ‘deliberative’ are axiomatic in that participation, aggregation and deliberation are all elements of democracy. Of course such labels serve to give importance to one element over others within any given model of democracy, but whether they can actually be used to delineate one particular model of democracy in respect to another is questionable. Saward (1998, p. 64) states emphatically: ‘There is no such thing as a “deliberative model of democracy”, despite efforts like Cohen’s to construct one. What there is is an effort to increase public deliberation on policies within a larger aggregative framework of constitutional democratic provisions.’ In other words, deliberative democracy is one mechanism among others that can help to extend and deepen democracy rather than a discrete model of democracy. This is not just a question of semantics: the problem that arises from the deliberationists’ aim to construct a model of democracy is that they neglect other forms of political activity which may be just as legitimate as deliberation. Mansbridge argues that just as deliberation may be more or less legitimate depending on the context and conditions within which it takes place, coercion and competition among actors can also be more or less legitimate and as such have their place within a normative understanding of democracy (Mansbridge, 1996, p. 51). Evidence from Barcelona shows that while participants are happy to engage in deliberation, wherever disagreement appears to be insurmountable participants tend to resort to all kinds of other, more effective, mechanisms to influence the public administration, not 164 Georgina Blakeley least those based on coercion. Political participation can be enacted through far more media than that of deliberation. The principal failure of deliberative democracy as a model of democracy, however, is that it does not have the means to address the key question of how people can be encouraged to deliberate in the first place, particularly when, in the context of an unequal society, opportunities to participate within deliberative forums may be seen by many, particularly the disadvantaged, as a burden rather than an opportunity. Habermas’s model of democracy involves the reciprocal recognition of the basic rights and duties which citizens require to engage in rational discourse, but without the strong enforcement of those duties that a communitarian model might imply. According to Leet (1998, p. 90) therefore, ‘The prospects for democracy depend, in the last instance, on citizens choosing to exercise their communicative capacities.’ The experience of Barcelona shows quite clearly that even in a context where both the political will to facilitate spaces of deliberation exists, as well as a historical tradition of associational activity, this is no guarantee that citizens can, want to, and indeed will participate. The same minority of people tend to attend the advisory councils. Moreover, evidence suggests that, particularly within the city-level advisory councils, what takes place is a type of ‘participation of the notables’ whereby experts or professionals in the field take part rather than the ordinary citizen. Women’s associations in particular remarked upon what they felt was the yawning divide between their ability to participate at district level compared to city level. As Elster has argued, the problem of establishing deliberative arenas which by their very existence can create a sense of obligation for citizens to participate in political discussion, may well, in an unequal society, lead to a ‘self-selected elite’ (Elster, 1997, p. 13). In this sense, the deliberative model of democracy does not go much further than a Schumpeterian model of democracy in its failure to tackle the fact that some citizens will enjoy a far greater capacity to choose to participate in arenas of rational discourse than others. If the deliberative theorists are serious in their aim of establishing rational critical argument as the linchpin of democratic systems, it will require a far stronger basis than simply that of the individual’s freedom to choose a course of deliberative action, not least because democracies in the real world are structured by inequalities of class, gender and race that restrict the individual’s ability to make that choice. Without a stronger basis we are left with the ultimate conclusion that deliberative democracy is for the chattering classes alone: no pun intended!

Notes

1. Despite problems evident within Habermas’s theory I concentrate mainly on his version of deliberative democracy because it is theoretically the most developed. It also presents the soundest basis for further theorizing and empirical research, pri- Creating Spaces of Deliberation in Barcelona 165

marily because it contains within it the most developed conceptualization of the public sphere. 2. The following advisory councils currently exist: the Municipal Council of Social Welfare, the Senior Citizens’ Advisory Council, Barcelona Women’s Council, the Council of Health and Safety at Work, the Municipal Sports Council, the Municipal Council of Associations and Volunteers, the Economic and Social Council, the Council of 100 Young People, the Advisory Council of Foreign Immigrants, the Municipal Schools’ Council and the Barcelona Youth Council. The latter is the only example of a non-municipal advisory council. 3. Examples of co-management include the ‘Barcelona Youth Project’ elaborated in 1985 by the city council with the co-responsibility of the Barcelona Youth Council; the Integral Plan for the Development of Social Services elaborated in 1995 by the city council with more than 130 social entities; the Municipal Institute for the Handicapped where 50 per cent of the governing body are elected from amongst the associations themselves, while the other 50 per cent are made up of political representatives from the city council; and, the Secretariat of Associations of Sants which is an example of the co-management of public services between the associational fabric (involving approximately 200 associations) and the administration. 4. Interview with F. Pindado, Lawyer and Activist in Nou Barris Neighbourhood Association, 10.10.97. 5. Interview with Imma Moraleda, Socialist Councillor for Barcelona City Council in charge of the Youth and Women Department, 26.11.97. 6. Interview with Vicens Tarats, Trade Unionist, Barcelona Workers’ Commissions, 28.11.97.

References

Ackerman, B. (1989) ‘Why Dialogue?’, Journal of Philosophy, vol. 86, no. 8, pp. 5–22. Benhabib, S. (1996) ‘Toward a Deliberative Model of Democratic Legitimacy’, in Benhabib, S. (ed.), Democracy and Difference. Princeton, NJ: Princeton University Press. Bohman, J. (1997) ‘Deliberative Democracy and Effective Social Freedom: Capabilities, Resources, and Opportunities’, in Bohman, J. and Rehg, W. (eds), Deliberative Democracy. Cambridge, MA: MIT Press. Calhoun, C. (1992) ‘Introduction’, in Calhoun, C. (ed.), Habermas and the Public Sphere. Cambridge, MA: MIT Press. Cohen, J. (1991) ‘Deliberation and Democratic Legitimacy’, in Hamlin, A. and Pettit, P. (eds), The Good Polity. Oxford: Basil Blackwell. Elster, J. (1997) ‘The Market and the Forum: Three Varieties of Political Theory’ in Bohman, J. and Rehg, W. (eds), Deliberative Democracy. Cambridge, MA: MIT Press. Foucault, M. (1990) The History of Sexuality, London: Penguin. Fraser, N. (1989) ‘What’s Critical about Critical Theory? The Case of Habermas and Gender’, in Unruly Practices: Power, Discourse and Gender in Contemporary Social Theory. Minnesota: University of Minnesota Press. Fraser, N. (1992) ‘Rethinking the Public Sphere: A Contribution to the Critique of Actually Existing Democracy’, in Calhoun, C. (ed.), Habermas and the Public Sphere. Cambridge, MA: MIT Press. Habermas, J. (1989a) The Theory of Communicative Action Two, trans. T. McCarthy. Cambridge: Polity Press. Habermas, J. (1989b) The Structural Transformation of the Public Sphere: An Inquiry into a Category of Bourgeois Society, trans. T. Burger with F. Lawrence. Cambridge, MA: MIT Press. 166 Georgina Blakeley

Habermas, J. (1992) ‘Further Reﬂections on the Public Sphere’, in Calhoun, C. (ed.), Habermas and the Public Sphere. Cambridge, MA: MIT Press. Habermas, J. (1996a) Between Facts and Norms, trans. W. Rehg. Cambridge: Polity Press. Habermas, J. (1996b) ‘Three Normative Models of Democracy’, in Benhabib, S. (ed.), Democracy and Difference. Princeton, NJ: Princeton University Press. Huertas, J. and Andreu, M. (1996) Barcelona en Lluita. Barcelona: Federació de Associacións de Veïns de Barcelona. Leet, M. (1998) ‘Jürgen Habermas and Deliberative Democracy’, in Carter, A. and Stokes, G. (eds), Liberal Democracy and its Critics. Cambridge: Polity Press. McCarthy, T. (1992) ‘Practical Discourse: On the Relation of Morality to Politics’, in Calhoun, C. (ed.), Habermas and the Public Sphere. Cambridge, MA: MIT Press. Mansbridge, J. (1996) ‘Using Power/Fighting Power: The Polity’, in Benhabib, S. (ed.), Democracy and Difference. Princeton, NJ: Princeton University Press. Pindado, F. (1997) ‘La Participació Ciutadana A La Vida de Les Ciutats’, unpublished manuscript, Barcelona. Primer Congres de les Dones de Barcelona (1998) Transformem la ciutat donant valor a la participació de les dones. Barcelona. Rawls, J. (1997) ‘The Idea of Public Reason’, in Bohman, J. and Rehg, W. (eds), Deliberative Democracy. Cambridge, MA: MIT Press. Saward, M. (1998) The Terms of Democracy. Cambridge: Polity Press. Young, I. (1996) ‘Communication and the Other: Beyond Deliberative Democracy’, in Benhabib, S. (ed.), Democracy and Difference. Princeton, NJ: Princeton University Press. Part III Challenges from the State: Changing Institutions 11 External Inﬂuences on Party Development and Transnational Party Cooperation: the Case of Post-Communist Europe Geoffrey Pridham

Introduction

The importance of international factors in regime change in Central and Eastern Europe (CEE) needs no introduction as a theme among transitologists, who now readily accept that these predominated when the Communist regimes collapsed. Subsequently, international factors have remained important, and at times decisive, in the construction and establishment of new democracies in the region. Of these, the European Union has emerged as by far the most influential external actor with its different mechanisms for facilitating this transformation. However, the impact has not been uniform. The influence of European integration on post-Communist democratizations has been substantial in areas like economic transition and political choice (especially through the EU requirements of democratic conditionality) and rather less with respect to civil society and nation-building (Pridham, 2000, ch. 9). In general, however, Brussels has considerable pull on these post-Communist democracies through the promise of eventual membership of the EU. There is significant scope for exploring international and especially European influences on party development in the new post-Communist democracies, all the more as emerging party systems are likely to be central to their operation, though the area has hardly been academically explored. Indeed, it has for long been a standard assumption that party development is essentially a concern of domestic political arenas and that, at best, external influences – if they are present – can only be a dependent variable. This paper seeks to challenge this assumption, which, not least in the European environment, has become increasingly unrealistic. It will look

169

K. Dowding et al. (eds.), Challenges to Democracy © Political Studies Association 2001 170 Geoffrey Pridham briefly at the possibilities for exploring external influences in party development in general, and then turn to how far European influences are present in democratization in post-Communist Europe with respect to party system patterns there. Finally, the paper will examine more precisely the impact of European integration through the mechanism of transnational party cooperation, such as whether this has offered a special channel for influencing party system development in these applicant states. Examples will be drawn especially from the countries of East-Central Europe which have so far developed the most intensive links with the EU. These questions will be explored by applying the transnational concepts of ‘convergence’ and ‘conditionality’ used in democratization studies. ‘Convergence’ is characterized by gradual movement in terms of system conformity with a grouping of established democratic states that has the power and institutional mechanisms to attract transiting regimes and to help secure democratic outcomes. The concept of ‘conditionality’ is more suggestive of deliberate efforts to determine from outside the course and outcome of regime change. It usually involves specifying (pre-) conditions for support and political monitoring of domestic developments in the countries under discussion.

External inﬂuences in party system development: formulating an approach

Conventional studies of political parties and party systems in Europe usually focus on such dimensions as party typologies and families, party structures and functions including political roles, ideology and programmes as well as political competition, electoral support and presence in society. At the same time, there is an established practice in comparative studies of looking cross- nationally at party development. But, rarely is this European perspective taken a stage further to examine transnational influences over programmatic formulation (emulation) and forms of political competition and electoral appeal, let alone deeper effects. The main exception is work on the special arena of transnational parties in the European Union and particularly their activity in the European Parliament (EP). A familiar starting point is the importance of external policy directions and commitments. Political parties are required to formulate full-scale programmes for electoral purposes, including statements covering foreign, defence and European policy. The salience of external issues varies between campaigns, although some countries are more prone to their mobilizing effects than others. Taking different historical examples, such salience is most frequently linked to policy divisiveness and a potential for enhancing political competition. The major conflict between governing and opposition parties over European integration and rearmament in the first decade of the German Party Development and Party Cooperation 171

Federal Republic revolved around the new democracy’s international future after the Second World War. A similar debate between left and right occurred in post-Franco Spain over Spain’s membership of NATO. Party ideology played a major part in the latter case especially on the left, although historical over- tones due to anti-American resentment concerning Washington’s support for the reviled dictator were also present. Among established European democracies in Western Europe, European integration has been the most commonly divisive issue between and within political parties. In the UK, this issue has over several decades persisted as the sharpest policy divide but not consistently between left and right. The same was once true of states like Denmark, not to mention Norway where the repeated rejection of EU membership by referendum left deep marks on the party system. Undoubtedly, the current crisis between Austria and the EU is a special case where the party factor – here, the ideological leanings and significance of the Freedom Party – is prominent as a systemically relevant issue in European policy, given the common perception of Haider as flouting the EU’s liberal democratic conditions for membership. As a whole, the phenomenon of ‘Euroscepticism’ in the 1990s has opened up new possibilities for party political engagement over EU affairs. In some countries one may speak of an external policy cleavage, where the divide endures and is rooted in public attitudes over time. A pertinent example is found in postwar Italy, where the chasm between left (in particular the PCI) and right over their respective allegiances to the USSR and USA lasted for over two decades. This was hardly surprising since the main parties’ links with the two superpowers involved different forms of active support from the latter. At various crucial moments in postwar Italian politics, outside influence – especially that of the USA – was quite decisive in the move towards new coalition arrangements involving the participation of one or other parties on the left (Pridham, 1998, ch. 11). Accordingly, the main Italian parties became identified to varying degrees with their respective international patrons. Admittedly, some parties are more inclined to external allegiances. Communist parties are the best instance, in many cases owing their founding to the inspiration of the Bolshevik Revolution. This special link with the USSR later became institutionalized with the ruling Communist parties in Eastern Europe after the Second World War (Dawisha, 1990, ch. 4). During the Fascist period, the PCI depended on Moscow at the low point of its suppression, although it eventually surged into becoming a mass party once the Resistance from 1943 gave it the opportunity to mobilize support (Urban, 1986). In this and other cases of West European Communist parties, the shift towards autonomy from Moscow, later labelled ‘Euro-Communism’, loosened the link from ideological dictate to sentiment but once again showed the importance of external factors in their development as parties. Furthermore, the subse- 172 Geoffrey Pridham quent conversion in the early 1990s of many Communist parties to centre-left tendencies in both parts of Europe illustrates once again how much major international developments – here, the fall of Communist rule – may have fundamental effects on party development. In summary there is an interplay between national traditions or domestic politics on the one hand and a variety of external influences on the other. It is also possible to differentiate between direct and indirect effects of such influences. Direct influences are those between a national party and outside support. This relationship is usually bilateral and can be quite close, and may become part of a particular party’s own history and tradition. However, EU transnational parties represent a multilateral form of direct influence. Indirect influences on party development derive from the general and more diffuse impact on a country’s politics from world or regional events which have a special potential for activating opinion. War is a paramount example, but in more normal times international economic downturns can also have a powerful influence on national policy performance, and hence on the fortunes of parties. Other issues may also arouse national concern or interest or internally divide parties. One may hypothesize that there are special conditions encouraging outside interest in new democracies. Authoritarian collapse usually brings a reconsid- eration of external policy allegiances, engaging the concern of foreign powers. Since political parties are central actors the balance of strength between them will affect policy choice. The type of regime transition matters when it comes to external influences. The postwar transitions and those in Southern Europe three decades later illustrate these patterns but in different ways. West Germany falls into a special category since outside influence on early party development was official through the licensing system that vetted and permitted the activity of newly formed parties. This discriminated mostly against neo-Nazi or hard right parties, thus allowing the CDU/CSU a head start in integrating the right of the political spectrum. But most cases of direct external influence in transition are best described as informal through bilateral cross-national links or even personal connections. The German influence in favour of the Iberian Socialist parties in the later 1970s is an instance. The Portuguese party is unusual because it was founded abroad, at Bad Munstereifel near Bonn in 1973, while the dictatorship was still in place. This owed much to the close personal links cultivated by Mario Soares with figures in the Socialist International during the period before democracy returned to Portugal (Janitschek, 1985). Subsequently, the SPD and Brandt in particular (who became President of the SI in 1975) played a key part in bolstering the chances of both the Portuguese PS and the Spanish PSOE through political training and material support (Pridham, 1991, pp. 241–2). It has commonly Party Development and Party Cooperation 173 been supposed that German support remained purely at the party level although new evidence suggests this was buttressed by government sources.1 Behind this activity was not merely ideological preference but also a systemic concern to stabilize new democracies in the Iberian peninsula. In general, external influences can have a formative effect although this may vary cross- nationally according to the type of transition. Greek party development was rather less open to European influences than that in Iberia for cultural reasons and anti-German feeling. Over time, European integration has had an ever more penetrating effect on the domestic politics of member states through the close interrelationship between European policy and traditional areas of domestic concern. Indirectly the scope for diffuse effects on party development has increased helping to explain the pressures to which party politics has been subjected over not only the issue of accession but also controversial matters like monetary integration. Directly, effects on national parties have come from involvement in transnational party organizations. In the past their activity has been secondary to mainstream integration concerns, but the increasing institutional and political weight of the EP has enhanced their influence. These patterns have emerged more strongly in the past decade during which post-Communist states have democratized and sought entry to the EU as well as to other European or international organizations.

Integration and convergence: context to party development

The special form of external influence where European integration affects party development in new democracies may be examined both indirectly and directly employing the concepts of ‘convergence’ and ‘conditionality’. The effects of accession pressures on regime change may be effectively explored by these concepts, for they point towards democratic consolidation and also provide a framework for exploring interactions between external effects and domestic developments. Furthermore, they highlight the need to focus on the regional context (Schmitter, 1996, p. 40). Integration gives the EU a direction and purpose towards convergence and reinforces the power to attract transiting regimes. It provides the EU with a leverage over elites and a channel of influence over the publics in new European democracies. Convergence affects policy choice and of course economic interests, but it especially influences elite mentalities as they emerge from international isolation. Such influence deriving from ever closer contacts in established democracies may well be system-reinforcing. At the same time, prospective entrant countries have to satisfy some basic requirements, particularly meeting a range of democratic criteria. 174 Geoffrey Pridham

The concept of ‘conditionality’ takes this up, for it speciﬁes conditions for support, involving either promise of material aid or political opportunities. Its method of political monitoring of domestic developments is being increasingly adopted by several international and European organizations. The EU is now most associated with democratic conditionality since this is linked to the prize of eventual membership (Pridham, 1999a, section 3). Entry to the EU allows ample time to observe the practice of democratic conditions in fragile new regimes. Over time, the EU has come to be more insistent and comprehensive in specifying democratic criteria concerning the countries from Central and Eastern Europe. Increasingly, these criteria have moved from mainly procedural conditions of formal democracy (e.g. rule of law, separation of institutional powers, free elections, freedom of expression) to include criteria of substantive democracy, such as the role of political parties, pluralism of the media, the importance of local government and an involved civil society (Kaldor and Vejvoda, 1997, esp. pp. 62–7). The EU now follows up the implementation of democratic criteria through detailed annual progress reports on applicant countries. Through conditionality the EU exercises more immediate pressures than supposed by those who argue that European integration can only have long- term effects on the consolidation process. At the same time, this pressure from Brussels has been complemented by support mechanisms for democracy- building. Of the various aid programmes, the most pertinent is the PHARE Democracy Programme administered by the Human Rights Foundation in Brussels. Its brief is to ‘support the activities and efforts of non-governmental bodies promoting a stable open society and good governance’, and it focuses support on ‘political reform and democratic practice, where local advocacy bodies are weak and professional expertise is particularly lacking’. In doing so, it seeks to further many of the criteria of both procedural substantive democracy (Pridham, 1999a, p. 67). Conditionality derives its strength from convergence but is inevitably constrained by the limitations of convergence. But conditionality has a wider meaning, for there are practical reasons why applicant countries need to gear up their democratic acts. Applicant countries are entering a form of political system in which the rule of law is respected (and the European Court of Justice asserts its authority in the legislative process), and in which the principles of representative government are tested at least in terms of participation in the Council of Ministers (CM) and the EP (Pinder, 1997, p. 124). There are different indirect and direct ways in which conditionality and convergence might affect party development, such as through EU policy pressures and transnational elite socialization. But transnational party organizations also implement their own form of conditionality for new members just as they provide a useful mechanism for democracy-building. These two con- Party Development and Party Cooperation 175 cepts will now be applied in turn to party development in post-Communist countries.

Convergence and diffuse effects: patterns of party development in post-Communist Europe in the context of EU accession

Convergence may be explored at the level of political parties in diverse ways. Parties are required to respond to EU policy concerns, all the more as accession negotiations proceed. It is also likely that the more difficult phase here, yet to come, will force parties to define more precisely their positions over contentious issues as has become evident in Poland (Blazyca and Kolkiewicz, 1999). For our purposes, though, attention turns here to the main party political tendencies over the past decade. This is a largely descriptive exercise and may be treated briefly as a prelude to examining more active and direct effects in the next section. By and large, four decades of Communist rule left greater discontinuities in party development than was the case in the earlier two experiences of European democratizations. Historical parties like the Social Democrats did not perform well in the first free elections in 1990, leaving the way open for new parties to dominate the political stage although, in the second phase of party development, many former regime parties successfully converted themselves into democratic Socialist or Social Democratic parties. Hence, party identification has taken longer to emerge. Furthermore, the initial dominance of anti-Communist umbrella movements inhibited the ready occurrence of political pluralism, as expressed through the medium of organized parties along recognizable ideological lines. The continuing importance of Solidarity is one reason for the belated flowering of full competitive politics in Poland, and similar problems were encountered with Civic Forum and Public Against Violence in both parts of Czechoslovakia. Party development in the CEECs has also been marked by a focus on parliamentary activity with a strong propensity for elite control. At the same time, a pattern of splits and recombi- nations has persisted in party formation (Olson, 1998). As a whole, parties in these new democracies have often demonstrated weak organizational links with the public (Pridham and Lewis, 1996, pp. 17–18). These problems, especially that relating to party identification, open up real prospects for Europeanization patterns in party development. Against the background of EU promises and pressures political elites and their parties in the new democracies may well approximate to Western European ideological tendencies. Such a pattern has been most noticeable on the Left with the Social Democratization of former regime parties in the CEECs. Reform Communists looked consciously to European models and the domestic motivation was evident in their desire for legitimation, with recognition by the Socialist 176 Geoffrey Pridham

International (SI) being the most obvious external accolade. This influence of European models on the left was most pronounced in the case of the newly proclaimed Hungarian Socialist Party, whose leaders and activists made frequent visits to Western Europe in the early transition years (Agh, 1995, p. 493). But it was later evident in other former Communist parties. The Slovak party changed its name in 1990 to Party of the Democratic Left (SDL) with Peter Weiss, its new chairman, deliberately aiming to create a modern European left party. In particular, it was the Italian Democratic Party of the Left (PDS) that acted as a national model for the European left. It had a strong influence on the SDL in both strategic and organizational ways (Pridham, 1999b, p. 1231). What is shown more broadly by comparisons with models of party development in Western Europe? Using conventional ideological criteria, we look at party families and how far these are replicated in the CEECs. In a survey of party development a few years ago, the following patterns were identified (Segert, 1996, pp. 226–7):

• the scarcity of modern conservative parties, with a notable exception in the Czech ODS; • a pattern of centre-right parties settling in the Liberal camp; with traditional authoritarian parties in the right in certain countries (Hungary, Poland and Slovakia); • strong left parties as modernized successors to the former Communist regime parties or, alternatively, in the form of Social Democratic parties (notably, the CSSD in the Czech Republic); • the failure of Green parties to establish themselves despite their initial impact during the fall of Communism; and • the absence of agrarian parties except in Poland and Hungary.

By and large, these patterns remained through the later 1990s. It should be noted, however, that this survey underrates the importance of centre-right parties, including those of the Christian Democratic tendency. But there are limitations to this exercise. It offers some descriptive value with restricted mileage for cross-national comparison. Labels alone do not automatically tell us much of substance, although admittedly individual party self-identification with a West European type of party family may denote an element of conviction, or otherwise opportunism. The two types of left parties are the clearest case of transnationally identifiable forces, as are the Christian Democrats on the centre-right in some countries, such as the Czech and Slovak republics. But there are many examples in the CEECs of parties that do not adopt transnational nomenclature, although in terms of Party Development and Party Cooperation 177 ideological precepts they may approximate to one party family or the other. This is true of other parties on the centre-right in particular, such as the Civic Democratic Party (ODS) in the Czech Republic and the Hungarian Democratic Forum (MDF) and their allegiance with conservative or Christian Democratic parties. Taking together self-identified and de facto members of European party families does nevertheless allow us some scope for cross- national comparison. However, there are still deficiencies in that conventional ideological labels may not be meaningfully transferable to the CEECs. According to Agh (1996, p. 22) this is because their local connotations may be idiosyncratic and often have obscure profiles. Similar to this exercise is the categorization of parties into ‘standard’ and ‘non-standard’, terms initially applied to the Slovak party scene (Meseznikov, 1994, pp. 105ff). ‘Standard’ parties are those related without too much difficulty to the Left/Right spectrum in European politics, while ‘non- standard’ parties fail to do so because of their national and social populism, authoritarianism, radicalism and extremism as well as possibly a confronta- tional and charismatic approach – the archetype of the latter being Meciar’s Movement for a Democratic Slovakia (HZDS). This categorization of ‘standard’ parties furthermore stresses ‘compatibility with international party structures’ thus overlapping with the exercise of aligning party families (Meseznikov, 1995, p. 106). Whatever approach is adopted there are different types of parties in CEECs not usually found in Western Europe. This is because of cleavage fault lines which are particular to CEECs or individual states in the region. The most obvious category is that of ethnic parties like the Hungarian parties in Slovakia (now united in the SMK) as well as the similar Hungarian Democratic Federation of Romania (HDFR) in Hungary and the Turkish Movement for Rights and Freedoms (MRF) in Bulgaria. Olson (1986, pp. 447, 461–2) calls into question the resemblance of party systems in East-Central Europe to their counterparts in Western Europe precisely because of issue alignments being different. It is clear that familiar comparative concepts may only be applied to the new democracies in the Eastern half of Europe to a restricted degree. In particular, they encounter not merely differences of historical experience but also problems in trying to compare unsettled party systems with long-established ones. With one or two exceptions, the CEECs have suffered from the debilitat- ing effects on civil society from long Communist rule, and it would be surprising if that inheritance had not inhibited party development. Notwithstanding these difficulties, they have developed into key actors in the new democracies in the CEECs. 178 Geoffrey Pridham

Opportunities for and constraints on transnational party cooperation between East and West

There are different levels at which the Europeanization of party development can take place. These may be categorized as identity and ideology, programme, organization, electoral politics and personnel. In all these respects transnational party cooperation (TPC) may impact on party development in new democracies providing a more concrete vehicle for approximation to West European ideological tendencies. It exerts real pressures through inter- elite socialization between countries in both halves of Europe and through a variety of organizational channels. They include the party groups in the EP as well as in the Parliamentary Assembly of the Council of Europe, the different EU party federations related to these party groups and the traditional party internationals. Not to be omitted, bilateral links between parties in different countries are growing. There are recognizable limitations on the impact of TPC. TPC links operate primarily via party leaders and organizations including local or regional branches as well as international offices in national party headquarters. Activity has involved election training as well as policy and organizational advice and sometimes includes financial support. Whatever its limitations, transnational party cooperation provides a convenient and pertinent mechanism for assessing how far party development in the CEECs has been determined by Western European models. In this sense, it is unique and should be utilized as a method in the comparative analysis of political parties. In one obvious way, TPC accords strongly with the categorization into party families. Parties from new democracies interested in joining transnational party formations have to declare an adherence to one or other official ideological tendency. This is not merely nominal as even this decision may entail some ideological adjustment or certainly a sharpening of party identity. The process of being accepted has been tightened up in response to the wave of new parties seeking membership in the 1990s. On the left this policy was motivated by difficulties in dealing with applications from former regime parties but also reflected general problems that emerged early on of identifying acceptable partner parties in the CEECs. Finally, once parties have joined, they are subject to various pressures to conform to TPC programmes and policy positions. These are, admittedly, largely platonic given that the transnational party formations do not as such transmit policy stands into policy action. Party identification proved a major problem for TPC especially in the first few years. Similar ideological labels sometimes turned out to be delusive; transnational actors occasionally discovered more ideological sympathy within rival parties than with their formal fraternal partners (Pridham, 1996, Party Development and Party Cooperation 179 p. 199). The greatest problems were located on the centre-right of the political spectrum. Among the Christian Democrats, the problem was not so much ascertaining which parties were de facto CD while not using this nomenclature but trying to exclude those holding hard right views. According to the adviser on Eastern Europe to the chairman of the EPP group in the EP, the European Union of Christian Democrats (EUCD) had become more careful about scrutinizing party programmes before admitting new parties into the first stage of formal links. This followed some dilemma over contacts with Christian parties in Poland which were right-wing and anti-Semitic.2 In one case the Hungarian Independent Smallholders’ Party was suspended from membership of the EUCD because of its increasingly nationalist-populist standpoint.3 The ISP returned to TPC organizations in 1999–2000 – with full membership of the EDU and observer status in the EPP – but only after strengthening its support for democratic rules and adopting more standard conservative positions. This process was overseen and influenced by the Bavarian CSU and the Hanns Seidel Stiftung which put pressure on the ISP and trained its functionaries, organized seminars and invited top figures to Bavaria.4 The same problem was faced by conservative parties looking for reliable partners in the CEECs. Party terminology was particularly difficult since ‘conservative’ was a rare name in these countries and in some was associated with reactionary tendencies. Instead conservative parties displayed myriad and sometimes idiosyncratic names. The usual way of sifting them out was to study party programmes to test their conformity with standard centre-right views.5 There was the same problem of self-demarcation from the hard right, the main difficulty concerning parties with nationalist tendencies. These were tested for their views on patriotism and soundings could produce a negative response, as in a reluctance to pursue further talks with the Croatian Democratic Union (HDZ) – the party of Franjo Tudjman. Human rights and treatment of minorities were generally major issues in keeping certain parties at a distance from transnational organizations. On the positive side, however, there were many cases on the centre-right of especially Christian Democratic parties in CEECs that found no difficulty in aligning themselves with these transnational formations. This was, for instance, true of Christian Democratic parties in the Czech Republic, Slovakia, Hungary, Poland, Romania and Slovenia. Some like the Slovak Christian Democratic Movement (KDH) took slightly divergent lines, such as a funda- mentalist standpoint on religious matters, reflecting a relatively strong religious presence in Slovak society, but this did not embarrass it transnationally, for the KDH’s record on human and minority rights was clear-cut. These Christian Democratic parties from the CEECs went on to play a fairly active part in the transnational organizations, this being facilitated by a strong 180 Geoffrey Pridham support for the values of European integration – their most salient commonal- ity with standard Christian Democratic parties in Western Europe (Pridham, 1999b, pp. 1229–30). The European Liberals encountered severe problems of identifying partner parties since the tradition of Liberalism was generally not strong in the CEECs. Eventually, this provided transnational actors with a special scope for helping to establish if not mould new parties. One can see this in the Slovak case. The Slovak Democratic Union (DU) was established late in transition, shortly before the parliamentary elections of 1994. Electoral pressure was therefore a factor in explaining the DU’s strong reliance on advice from European Liberals. To this extent, the DU’s early identity formation was significantly determined by standard European Liberal ideas.6 It also reflected the general pattern in Slovakia during the 1994–98 parliament when opposition parties placed that much more importance on transnational linkages as a defence mechanism against authoritarian pressures from the then Meciar government (Pridham, 1999b, p. 1232). While the anti-Communist umbrella movements delayed party formation, some of their successor organizations conformed to ‘standard parties’ eventually. In Czechoslovakia, Civic Forum divided in 1991 into the Civic Democratic Party led by Klaus and the Civic Movement led by Dienstbier. While the former allied with European Conservatives in the EDU, the latter, renamed Free Democrats, joined the Liberal International. In Slovakia, Public Against Violence split into a couple of small parties, but the bulk of its support went with Meciar’s Movement for a Democratic Slovakia (HZDS). But the HZDS as a populist-nationalist force contained diverse ideological elements, and this prevented it from establishing viable transnational links with Western Europe. However, European doubts about Meciar’s dubious democratic credentials also played a part, as did intransigent attitudes within the HZDS itself (Pridham, 1999b, pp. 1231–2). A fairly clear pattern emerged during the first half of the 1990s whereby ‘standard parties’ in the CEECs declared themselves in relation to transnational membership. It was also apparent that a line of demarcation was drawn on the right to exclude parties that were nationalist, racialist or ideo- logically extremist. On the political Left, there was much less of a problem in identifying parties. The main difficulty was how to respond to former Communist regime parties that had now converted themselves into parties called Socialist (as in Hungary and Bulgaria), Social Democratic (as in Poland) or Democratic Left (as in Slovakia). Their new programmes tended to adopt centre-left policies and suitable policy lines, notably support for EU entry, as well as economic reform, albeit in some cases offering a milder version of this. This was often combined with leadership change involving generational turnover. Continuities nevertheless remained at the level of party activists, Party Development and Party Cooperation 181 electoral support and strong organization. Only very gradually, however, did the SI – and later the PES – come round to considering former regime parties as members. This required a long process of vetting candidate parties, a practice applied by all transnational organizations and one that employed a fairly strict form of democratic conditionality at the party political level. It included examining party programmes and public statements by leaders, using confidential reports by West European embassies in the country of the candidate party, sending ‘missions’ by transnational organizations to national party headquarters and often invitations to party delegations to come and visit the transnational party offices, usually in Brussels, for detailed discussions of policy matters. This procedure demonstrated the degree of influence that could be exerted by transnational actors. Former Communists sought links with the SI and it soon became clear their motivation was international legitimation. These parties were closely scrutinized over several years, and it helped that several of them were elected to national government during 1992–94. They lobbied hard for a formal link and hardly disclosed the high priority they accorded this. By the mid-1990s, the parties in Hungary and Slovakia (HSP and SDL) were recommended for membership and observer status was offered the two Polish parties – the Social Democrats (SLD) and Union of Labour (UP). But some cross-national differentiation was evident in the SI approach. Doubts were expressed over the Slovak party for a while because of its internal divisions and, more seriously, over the Bulgarian Socialist Party (BSP). The BSP proved rather difficult to handle as it was deeply divided but also as the Bulgarian transition was for a while in doubt, not least as a result of the BSP’s own questionable policies while in government during 1995–97. Even in the mid-1990s, SI circles in Brussels were not sure as ‘we still don’t know if it is a reformed party or if it still has very strong tendencies of a Communist party; but there was a willingness to keep open contact with the BSP and to involve it in policy discussions’.7 Problems were already present in differences between the factions over SI membership together with reservations among activists, due in part to lack of knowledge about the International but also suspicion over criticisms of the BSP by some West European parties.8 Finally, in 1997 the BSP split with the Social Democrats leaving it to form the Euro-Left, thereby allowing them more freedom to pursue transnational links. Other factors were influential and these related to emerging political cultures in these countries. In Hungary, for instance, there was no trace of wari- ness towards transnational involvements as was evident not only in Bulgaria but also in Slovakia, where such links of the opposition parties were attacked by the ruling HZDS for being ‘anti-Slovak’. This difference undoubtedly helped account for Hungarian parties of different tendencies being to the fore among those in the CEECs in initiating and developing transnational links. In 182 Geoffrey Pridham general, several interview respondents among transnational actors commented on some differences of understanding and mentality in dialogue with party representatives from the CEECs, especially in the first years after transition began. In part this was due to unfamiliarity with democratic techniques and competitive politics, but the problems could be deeper. The Czech Social Democratic Party (CSSD) had no ideological reservations about affiliation with the SI but complications arising from experience under Communist rule sur- faced, particularly over dislike for returned exiles and former dissidents. However, in the course of time European links – such as attending party con- gresses abroad and involvement in transnational policy projects – helped to instil more confidence among this party’s elite groups, all the more once the party’s popular support rose.9 Cultural problems tended to diminish with time. Political and electoral experience was acquired first-hand as well as vicariously through transnational actors, while increasingly routine contact with the latter fostered elite socialization that also helped reduce the pull of inheritances from the Communist period. There was, as noted by interview respondents, some differentiation between different regions and countries. This was particularly marked between the CEECs and republics from the former Soviet Union, with the exception of the Baltic states due to their own European background and close party links with Scandinavia and Sweden in particular. But there was another factor that impinged – and increasingly so – on TPC and its influence on party development in the new democracies, and that was the process of EU enlargement. This second way in which TPC evolved became more evident during the course of the later 1990s. Increasingly now, party personnel from CEECs are involved in policy fora in transnational organizations as one way of preparing their parties for eventual entry to the EU. Transnational organizations and contacts came increasingly to be viewed by party leaders from the CEECs as a non-official channel for networking in favour of EU accession. It could generally be said that the greater the possibility of eventual EU membership, the more likely it was that parties from CEECs would be ready to conform with European party political patterns, subject to obvious constraints in domestic politics. Clearly, one cannot view this relationship in isolation from other developments – of which the most important was the state of regime transition, which in turn affected both party systems and decisions in Brussels about opening entry negotiations. Many interview respondents remarked on significant differences between East-Central European and Balkan countries when it came to TPC. In some cases, these responses appeared to convey cultural preconceptions, but, all the same, links with parties in Balkan countries proved difficult for a number of reasons. They were basically non-existent with republics in the former Yugoslavia until the Dayton peace agreement of late 1995, with the one excep- Party Development and Party Cooperation 183 tion of Slovenia. Since then, the SI and its member parties in Western Europe have begun to show an interest in these republics but there has been little progress with countries like Serbia and Croatia because of their authoritarian regimes. This is now beginning to change in Croatia’s case following Tudjman’s death in December 1999 and the victory of the democratic opposition early in 2000 in both parliamentary and presidential elections. Milosevic’s Socialist Party of Serbia (SPS) is distinctly a pariah party in transnational circles, but links have developed with parties in Macedonia and Montenegro.10 Among centre-right parties, it was considered much easier to find Christian Democratic parties in East-Central Europe than in the Balkans for confessional reasons (Slovenia here counted among the former), save for a few small parties like the Hungarian Christian Democrats in Romania. This was a clear admission that the cultural divide between Catholic and Orthodox countries imprinted itself on transnational party developments. Nevertheless, transnational links were most developed in the Balkans with Romania and particularly Bulgaria judging by the membership of the SI, LI and EUCD by the mid-1990s (see tables of membership in Pridham, 1996, pp. 217–19). In Albania, the Democratic Party had formal links with the EDU and EUCD, becoming a full member of the latter in 1995, but these were not established without difficulty. Some parties in the EUCD were opposed to accepting it because of ethnic tensions, complaints having been received from the Greek minority, and because of doubts about its democratic commitment for it was the party of President Berisha, whose authoritarian practices were becoming controversial. But the Democratic Party’s desire for international acceptance was such that outside pressures on Albania to negotiate with Athens were successful. Overall, transnational links with Balkan countries were less intensive than with those in East-Central Europe. Geographical distance may have influenced this but more important was the problem in matching West European party formations with viable partner parties in countries that were undergoing, in many cases, difficult and uncertain transitions. The greater incidence of ethnic problems in the Balkan countries also presents serious problems. Behind these regional differences lies the issue of EU entry. Differences in prospects here influenced TPC to the extent that transnational actors in Brussels were more prepared to invest resources as well as time in countries with better chances of membership. The future balance of party group representation in the EP following enlargement was one major consideration. This differentiation was also influenced by progress with democratization. Transnational actors were as a whole more cautious about developing close links with countries where the outcome of regime change was difficult to predict, although in some cases they did choose to make a special effort to support allied parties to help democratization. 184 Geoffrey Pridham

In conclusion, TPC, while influential, was undoubtedly secondary to domestic factors in party development. This is shown by the experience of umbrella movements, which later succumbed to the dynamics of political competition in their own countries, for it was only then that transnational organizations started to relate effectively to their successors. TPC was also constrained by cultural reservations towards outside influences, including nationalism, which proved a firm deterrent to transnational activity. The Hungarian ISP was more the exception than the rule, although the more important case of the Slovak HZDS may or may not succeed in its repeated effort to enter a transnational organization. Meciar openly declared his party’s aim to join the EDU at its congress in March 2000, but that is dependent on the HZDS reforming itself and probably also on Meciar’s final retirement from political life. At the same time, there were different ways in which TPC could help influence party development. In some individual cases, transnational support had a fairly decisive effect in the formative stage of parties. Otherwise, influence was particularly evident at elite levels. It was party leaders and senior officials who were most directly concerned with TPC, but they themselves were in a position to mould their own parties given these were often top-down.11 Policy training and advice on election campaigns could also have wider effects within parties. Interview respondents remarked frequently on the sense of belonging to international groupings, and there was undoubtedly a feel-good factor emanating from involvement in transnational meetings. These were not simply regarded as tame events, since they had a European and particularly EU focus so that serious political messages were transmitted via these organizations. Recognition proved a valuable encouragement to party development, and in the case of the political left the legitimation this accorded to former regime parties was given a top priority.

Conclusion

Focusing on transnational party cooperation between EU member states and the new democracies in the CEECs adds significantly to identifying party families there since the transnational organizations strengthen ideological attach- ments and, at the same time, apply conditions for acceptance that can have a secondary influence on member parties’ development. With some exceptions, parties in the CEECs have overwhelmingly sought links with parties – whether transnational or national – in Western Europe rather than elsewhere in the CEECs, although regional groupings of parties especially from the Visegrad are beginning to develop. Thus, overall, the pattern in the 1990s has been for West European party political habits and attitudes to imprint themselves on parties from the CEECs. In the course of time, distinct national influences from new member parties in the CEECs are likely to influence transnational Party Development and Party Cooperation 185 organizations, all the more if they come from successful entrants to the EU and once they join the EP. One may therefore speak of a pattern of Europeanization, but it has been one-sided in its effects up till now. Europeanization through TPC has concentrated on the major political forces in European politics to the exclusion of certain categories. It has been mainly parties that are Christian Democratic, Conservative, Liberal and Socialist that have benefited from this activity. Greens have a weak form of TPC and they are not strong in the CEECs. However, parties that are nationalist or populist or ethnically hostile have been shunned, just as have parties with authoritarian leanings. In this way TPC has tended to draw a fairly clear distinction between standard and non-standard parties and so deepen the divide between them. It is difficult to argue that TPC has actually weakened the position of non-standard parties, especially if they draw on political cultures that are, at least in part, not friendly towards outside influences. In different ways, TPC has been strengthened by factors less present in previous phases of democratic transition in postwar Western Europe. First, parties that emerged in post-Communist countries had much weaker historical roots than those in post-Fascist and Southern European states. To that degree transnational organizations had more opportunity to have a formative influence on party development in the CEECs. Second, the impact of the international environment has been so much greater in the transitions of the 1990s compared with those of the 1970s in Southern Europe, and this has been present in the desire to join the EU. This link between EU entry and TPC will increase in the future, with continuing effects on party development. This very connection between EU entry and TPC illustrates the importance of assessing the latter’s activity as part of the overall process of convergence in Europe. Transnational party cooperation also offers unique scope for measuring the way in which convergence actually occurs at unofficial levels. Altogether, the institutionalized framework provided by European integration emerges as an important channel for studying external influences in party development in the new post-Communist democracies.

Notes

1. According to The Times, 2 February 2000, large sums from the German espionage budget were devoted to strengthening moderate parties in Iberia and thus promoting liberal democracy there. These revelations came as part of the major scandal over secret funding of the CDU during Kohl’s long chairmanship of the party. 2. Interview with Stephen Biller, EPP group, European Parliament, Brussels, January 1996. 3. Interview with Klaus Welle, Secretary-General of the EPP and EUCD, Brussels, January 1996. 186 Geoffrey Pridham

4. Interviews with Hans-Friedrich Freiherr von Solemacher, head of Hanns Seidel Stiftung in Hungary, and with Katalin Kiszely, European affairs spokesperson of the ISP and member of the Hungarian Parliament, both in Budapest, April 2000. 5. Interview with Alexis Wintoniak, Executive Secretary of the European Democratic Union (EDU), Vienna, November 1995. 6. Interview with Eduard Kukan, then Chairman of the DU, Bratislava, September 1998. According to him, the party’s formation shortly before the 1994 election meant it was extra reliant on European assistance. 7. Interview with Bo Toresson, Secretary-General of the SI Forum for Democracy and Solidarity, Brussels, January 1996. 8. Interview with Elena Poptodorova, BSP national deputy, Sofia, September 1995. 9. Interview with Svetlana Navarova, head of international office, CSSD, Prague, November 1995. 10. SI European Forum for Democracy and Solidarity, country updates, for Macedonia September 1997, and for Federal Republic of Yugoslavia January 1998. 11. According to a leader of the Hungarian Free Democrats, TPC had an influence on elite mentalities involving a small part of party elites, especially those that had a knowledge of foreign languages, but this was an influential component of party elites in Hungary (interview with Istvan Szent-Ivanyi, Chairman of the Foreign Affairs Committee, Hungarian National Assembly, Budapest, April 2000.

Bibliography

Agh, A. (1995) ‘Partial consolidation of the East-Central European parties: the case of the Hungarian Socialist Party’, Party Politics, October. Agh, A. (1996) The End of the Beginning: the Partial Consolidation of East-Central European Parties and Party Systems, Budapest Papers on Democratic Transition, No. 156. Budapest University of Economics. Blazyca, G. and Kolkiewicz, M. (1999) ‘Poland and the EU: internal disputes, domestic politics and accession’, Journal of Communist Studies and Transition Politics, December, pp. 131–43. Dawisha, K. (1990) Eastern Europe, Gorbachev and Reform: the Great Challenge. Cambridge: Cambridge University Press. Janitschek, H. (1985) Mario Soares: Portrait of a Hero. London: Weidenfeld & Nicolson. Kaldor, M. and Vejvoda, I. (1997) ‘Democratisation in Eastern and Central European countries’, International Affairs, January. Meseznikov, G. (1995) ‘The parliamentary election 1994: a conﬁrmation of the party system in Slovakia’, in S. Szomolanyi and G. Meseznikov (eds), Slovakia: Parliamentary Elections 1994. Bratislava: Slovak Political Science Association, pp. 105–6. Olson, D. (1998) ‘Party formation and party system consolidation in the new democracies of Central Europe’, Political Studies, vol. 46, no. 3, pp. 433–35. Pinder, J. (1997) ‘The European Community and democracy in Central and Eastern Europe’, in G. Pridham, E. Herring and G. Sanford (eds), Building Europe? The International Dimension of Democratisation in Eastern Europe. London: Leicester University Press. Pridham, G. (1988) Political Parties and Coalitional Behaviour in Italy. London: Routledge. Pridham, G. (ed.) (1991) Encouraging Democracy: the International Context of Regime Transition in Southern Europe. Leicester: Leicester University Press. Party Development and Party Cooperation 187

Pridham, G. (1996) ‘Transnational party links and transition to democracy: Eastern Europe in comparative perspective’, in P. G. Lewis (ed.), Party Structure and Organisation in East-Central Europe. Cheltenham: Edward Elgar. Pridham, G. (1999a) ‘The European Union, democratic conditionality and transnational party linkages: the case of Eastern Europe’, in J. Grugel (ed.), Democracy Without Borders: Transnationalisation and Conditionality in New Democracies. London: Routledge. Pridham, G. (1999b) ‘Complying with the European Union’s democratic conditionality: transnational party linkages and regime change in Slovakia, 1993–98’, Europe-Asia Studies, November. Pridham, G. (2000) The Dynamics of Democratisation: a Comparative Approach. London: Continuum. Pridham, G. and Lewis, P. (eds) (1996) Stabilising Fragile Democracies: Comparing New Party Systems in Southern and Eastern Europe. London: Routledge. Schmitter, P. (1996) ‘The inﬂuence of the international context upon the choice of national institutions and policies in neo-democracies’, in L. Whitehead (ed.), The International Dimensions of Democratization: Europe and the Americas. Oxford: Oxford University Press. Segert, D. (1996) ‘Party politics in the process of Europeanisation – is there a special way for party development in Central Eastern Europe?’, in M. Szabo (ed.), The Challenge of Democratisation in the Region: East Central Europe. Budapest: Hungarian Political Science Association. Urban, J. B. (1986) Moscow and the Italian Communist Party: from Togliatti to Berlinguer. London: I. B. Tauris. 12 Promoting Parties and Party Systems in New Democracies: Is There Anything the ‘International Community’ Can Do? Peter Burnell

Introduction

A consensus exists among political scientists that parties and competitive party systems are central to democracy and essential agents of democratization. But following early statements of the challenges of democratic transition we have come to appreciate how difﬁcult is the crafting of durable parties and appropriate forms of plural politics in new and emerging democracies. Thus Mainwaring’s (1998) survey concluded that many ‘third wave’ democracies have only weakly institutionalized party systems: the parties have shallow roots and tenuous links to organized interests; they enjoy fragile legitimacy; many parties’ popularity is short-lived. The quality of democracy and the prospects for democratic consolidation both suffer. Others identify the draw- backs of excessive pluralism, claiming that although political factions can help new parties to emerge, persistent factionalism frustrates democratic stability and consolidation (Gillespie et al., 1995). In contrast in Africa the problem is more the persistence of dominant party systems and ineffective opposition (Bogaards, 2000). The 1990s witnessed a remarkable expansion of assistance from the ‘international community’ to political reform and better governance in liberalizing and democratizing countries (Burnell, 2000). Help with party-building and electoral competition is one aspect. It is impossible to gauge the exact amount, or, indeed, to choose one best unit of measurement. Just as scholars are divided over whether the greater threat lies in the emergence of highly fragmented or very concentrated party systems, so disagreements exist about the wisdom of direct international involvement with parties in new democracies – even among commentators who support international cooperation for

188

K. Dowding et al. (eds.), Challenges to Democracy © Political Studies Association 2001 Party Systems in New Democracies 189 democratization more generally. For example, in Africa, van Cranenburgh (1999, pp. 101–2) maintains that their essentially clientelistic nature makes Africa’s parties unsuitable objects of democracy assistance; Mair (2000, pp. 142–5), in contrast, argues for increased international support. Meanwhile, organizations like the International Institute for Democracy and Electoral Assistance are exploring the practical issues, including modalities of party funding. There are fundamental objections to engagement by western liberal democracies in democracy promotion. They see involvement with parties as especially worrying, and not only in Africa. First, it could be said the established democracies have too little knowledge and understanding of the indigenous political realities to interfere in constructive, non-harmful ways. Second, the established democracies have deteriorated so much in respect of their own party politics that they should address their own weaknesses ﬁrst. Third, the erosion of de facto sovereignty that ‘party assistance’ represents (and, possibly, accelerates) is worrying. Fourth, the West intends to encourage only those forms of party politics that accommodate international and transnational capital, neglecting the political requirements of social justice. Fifth, parties are outmoded or inherently deﬁcient vehicles for genuine democratization. True democracy involves far more than replicating representative government as in the established democracies. It should emphasize popular empowerment, through encouraging social movements, civil society organizations and direct action. These fundamental objections are not addressed here. Instead the chapter considers party assistance within the framework of international democracy promotion as a fact. While noting that democracy in this context means a largely political construct, informed more by individualistic than communitarian notions of society, it does not challenge the ambition to disseminate western-style liberal democracy. International democracy promotion makes few concessions to radical models of participatory politics – the kind implied by ‘popular democracy’, ‘grass-roots democracy’, let alone social and economic democracy.

Object of the enquiry

‘Party assistance’ denotes assistance to parties severally or on an individual basis, and also in a more generic sense of engagement with party systems as a whole. Party systems are deﬁned tautologically as ‘the set of all signiﬁcant parties in the country, their interactions, and (sometimes) the electoral system and voter loyalties that produce it’ (McLean, 1996, p. 364). In theory a de jure one-party state might be classed as a type of party system, but democratic systems allow the right to mount organized political opposition and to compete for power against the ruling party. 190 Peter Burnell

The chapter reviews the international community’s efforts to assist the development of parties and party systems and assesses the options in the light of available evidence. It argues that the circumstances in the countries and the nature of the political reform process under way should be factored into a strategic approach to deciding what kinds of party assistance, if any, are appropriate. Contrary to how party assistance has evolved, a strategic approach would distinguish the several functions that parties and party systems can perform for democratic politics and government. It would identify the speciﬁc requirements and establish priorities among ‘needs’ in relation to each country individually. The account concludes that measures to strengthen parties are one of the more speculative ways of trying to promote democratization. The potential returns are uncertain. Steps to foster plural politics indirectly, while also having limitations, could be more constructive. The development of strong linkages between parties and civil society, often held to be vital for democracy, is especially problematic terrain for external intervention.

Democracy assistance, the ‘international community’ and intervention

Democracy assistance can be styled as a form of international political intervention. Political intervention is a contested term. It has acquired elastic quali- ties particularly in the light of the rise of ‘humanitarian intervention’. It may be coercive or non-coercive; it can take consensual and non-consensual forms. It embraces a continuum of modalities, from ‘hard’ to ‘soft’. Democracy assistance, of which party assistance is a subset, comprises just one of several possible approaches to international intervention on behalf of democracy. It is but one form of democracy promotion tout court. For the most part it lies at the ‘soft’ end, comprising various kinds of transfers of resources on concessionary, or non-commercial, terms. It is a positive form of engagement, spanning practical and symbolic support, incentives, inducements and rewards. Although it differs from other forms of intervention such as attaching political conditionalities to economic aid, the forms may be linked in practice, as when economic aid’s political conditionalities require a government to remove obstacles against political aid to reformers. Precisely because the European Union toughened its political requirements for accession, Pridham argues (1999, p. 73), transnational party cooperation has advanced democratization at the level of party development in Central/Eastern Europe since 1989. The international (donor) community is a loose term, not a refined social science concept. The referents vary according to context. In refugee situations the United Nations High Commissioner for Refugees is central; in respect of the international financial architecture, the Bretton Woods institutions dominate. In democracy assistance the international community denotes a plural- Party Systems in New Democracies 191 ity of actors in the established western democracies – chiefly the United States and Germany and, to a lesser degree, countries like Canada, Britain, the Netherlands and Scandinavia, plus certain international organizations where they are predominant. It includes different types of organization – government departments and intergovernmental bodies, formally autonomous (but government-funded) agencies, political parties and party networks, party foundations and other private foundations and non-governmental organizations. Some primarily provide finance, others are conduits, and several combine different roles including adding value like know-how and skills. Only some of the organizations were created specifically and exclusively to promote democracy abroad. For Germany’s Stiftungen, whose many field offices abroad date back over 30 years in some cases, democracy promotion has come to the fore only in the 1990s. Most actors are engaged in several sectors, such as government institutions and civil society, like America’s National Endowment for Democracy (NED) – founded in 1983 and with an annual income of around $32 million – and Britain’s Westminster Foundation for Democracy, set up in 1982 and receiving £4 million from the Foreign and Commonwealth Office (1999/2000). The NED, like USAID, disburses party assistance through such core grantees as the International Republican Institute (IRI) and the National Democratic Institute for International Affairs (NDI). By the mid-1990s USAID was devoting more than $10 million annually to party-related projects through these party institutions (Carothers, 1999, p. 139). For yet other bodies democracy promotion (and even more so party assistance) is a minor concern in terms of their formal objectives, operational goals and expenditure. This is true of USAID as well as Britain’s Department for International Development. They have a parallel interest in promoting economic liberalization and marketization. For some agents of democracy promotion, party assistance is no part of their agenda. The Commonwealth, whose 1991 Harare Declaration added principles of democracy, human rights and good governance to its mission, has concentrated on elections observation in member states, although the Commonwealth Parliamentary Association’s activities are clearly relevant to parties. There are many other actors whose specialized competence, such as in legal technical assistance to judicial reform, precludes direct assistance to parties. Definitional and methodological conundrums frustrate an exact estimate of democracy assistance, but the total financial value is probably around US$2 billion annually. United States government spending on political aid of around US$700 million – of which the Agency for International Development (USAID) budget for ‘building sustainable democracies’ took $515 million in 1999 – and the German government’s annual expenditure on democracy promotion (mainly the 120 million euros budgeted for the Stiftungen or party foundations) account for the largest shares. 192 Peter Burnell

Approaches to party assistance: the systemic level and the party level

In regard to party assistance we must be clear about the aims, the basis and whether the methods are appropriate, as well as the identity of the principal actors. Three objects of assistance can be distinguished. Parties form one of these: assistance may be provided on a partisan, non-partisan (all-party) or multi-partisan basis. A second is the assortment of afﬁliated organizations like labour groups and associated or ancillary organizations such as policy research institutes and think-tanks that often cluster around parties. Third, there is the wider political and regulatory environment that structures the terms of political engagement – participation and contestation – in formal and informal ways. Interventions at this level can aim to further a permissive environment for competition and fair rules of the game. For analytical purposes the plethora of activities denoted by party assistance can be distinguished in terms of the principal aims and objectives directed at two levels: the systemic level and the parties, individual or several. In practice the aims may be confused. Observers on different sides of the relationship may understand identical endeavours quite differently. Moreover even the declared purpose of an intervention may change once the activities unfold, or are adjusted by intermediaries and as the results – successes and failures – emerge. Furthermore, informal agendas may prevail over formally stated ones.

The systemic level At the system level the objective is presumed to be the establishment of a competitive political system offering the electorate meaningful choice, and inclusive of all pertinent social groups. No significant minority or, indeed, majority group feels excluded from effective representation. Political pluralism demands there be a minimum of two parties preferably of more or less equal weight. At the same time it is reasonable to infer that the requirements of political stability and effective government must be accommodated somehow, which means avoiding hyper-pluralism and atomized multipartism. Even if some individual parties rise and fall, or change their identity (their name even), over time, there must be some continuity if there is to be institutionalization of the system as a whole. Democracy promoters actually face a great variety of situations, with various shortcomings from a democratic perspective. No single approach will fit all cases. Although limited, quasi- and semi-democracies and the like cannot be defined solely in terms of the measure of political competition, they are echoed in a range of limited, quasi- and semi-competitive party systems. Each case war- rants individual consideration and, probably, merits its own response. For example, there is orderly no-partyism, as in Uganda’s formal ‘no-party demo- Party Systems in New Democracies 193 cracy’ (the official dogma claims all Ugandans belong to the ruling National Resistance Movement). This differs from a constantly changing kalaeidoscope of legally recognized but weak, poorly grounded and fragile political groups, parties and factions. A third possibility is a dominant party system in which organized political opposition to the ruling elite is either non-existent or weak, perhaps chronically divided among small groups, proto- and pre-parties. In a hegemonic party system opposition parties, while allowed, are not permitted to pose serious competition. There are further variants: the ruling party may or not be highly institutionalized or prone to internal factionalism. There might or might not be a long history of mass political organization of some form: some of today’s parties have a substantial history; others are novices. The new democracy might or might not qualify as a liberal democracy; the parties might or might not all accept the democratic ‘rules of the game’, or accept that the rules operate freely and fairly. Even where a viable and broadly representative party system is taking shape the parties may lack organizational skills and managerial competence, and have limited abilities to initiate new policy options or promote their message. They might simply inspire less confidence than did the parties in the more recently consolidated democracies of southern Europe at a similar stage of democratic transition, leaving a question mark hanging over the new party system’s sustainability. When faced with situations like these the democracy promoters rarely give explicit or precise formulation to the approximate number of parties they would prefer to see, let alone the relative weights. Perhaps they lack a clear view, or believe it would be politically inappropriate to express one. The belief that, of all the determinants of party systems, advice on electoral rules gives the greatest purchase might also be present, though it is highly debatable (see Taagepera, 1998). It seems that party assistance has not been consciously organized to maximize the chances of a democratically viable party system. In some places there have been efforts to encourage like-minded parties to unite, but the purpose has usually been to boost a particular ideological tendency, or to orchestrate a stronger challenge to a predominant party, not to influence the system per se. No one exercises responsibility for ensuring a well-balanced spectrum in ideological terms. Instead, the party system that does emerge is left to be a consequence of interventions at the level of particular parties and other influences, such as changes in the political culture, social and economic developments like modernization, urbanization, (de-)industrialization, growing prosperity (poverty) and income (in)equality, and the implications all these have for social cleavages.

The party level Much party assistance has gone in the first instance to parties, unmediated by a clear sense of purpose with respect to the party system. Sometimes they are 194 Peter Burnell chosen on a partisan basis, in others on a multi-party or non-partisan basis (or claimed as such). Occasionally there are cross-party ventures. In the early 1990s much of the assistance from parties, party institutes and foundations to new parties had proximate aims – helping them capture legislative seats and win executive power. That meant performing well in elections, most notably ‘transition elections’. The parties’ leaders, not the party as an organization or its programme, formed the centre of attention. Such a narrow and immediate focus can be defended. After all, the issue of who wins on such occasions can be critical to the longer-run chances of the country completing a successful transition to democracy (e.g. by defeating a reactionary party, or aiding progressives to take over a party). The fact that a transition election takes place successfully (peacefully, accepted as being fair, and the winners form a government) is what matters most for democratic transition, not the number of parties and their respective weights. Even a very one-sided contest can be a triumph for democratic transition. But in the longer term, assistance that, for example, confirms a dominant party’s position cannot serve to institutionalize a competitive party system. And the experience of ‘mature democracies’ suggests that party systems endure once they become embedded and are not easily changed (Ware, 1996, p. 213). Thus the early years can be crucial to democratic sustainability and to the eventual shape of the party system. And after democratic transition a longer-term agenda must surely take over – one that assists politicians and supporters to maintain a durable commitment to democracy between elections, notwithstanding occasional electoral defeats. This means developing party organization. It means ensuring that all parties cultivate an identity separate from the state, especially the party in power, although even that party might need strengthening, especially where the burdens of office distract it from consolidating party organization. A ruling party and the electorate may be slow to make these adjustments, especially in countries long accustomed to a de jure one-party state that is followed by a dominant party system, more so where the dominant party claims superiority as the party of national liberation (as in Zimbabwe).

A strategic approach The tasks in hand can be interpreted in different ways, depending on the choice of emphasis as between the different functions that parties can provide (but not every party will provide them all, or provide them well):

• providing political leadership and candidates to ﬁll government ofﬁce; • mobilizing political support for government and facilitating its legitimacy; • providing checks against the executive and potential abuse of power; • serving as mediums for political communication; Party Systems in New Democracies 195

• aggregating interests; • integrating social groups; • furnishing alternative policy options that address the citizen’s wants and needs; • articulating a vision of the public good.

While contemporary politics renders none of these functions obsolete, their interpretation, rank order and balance tend to be culture-specific. The record of the established democracies offers no uniform or constant example. The precise meaning, role and function of political opposition by ‘opposition parties’ are contested. All this complicates the giving of advice to new parties and new democracies. But it should not obscure the fundamentals of a strategic approach to offering support. A strategic approach starts by engaging with questions about what parties are and what they are for. It keeps in view that they are valuable not so much for themselves but for the contribution they make to democracy’s defining attributes: political accountability of the governors; opportunity for public participation in the policy process; and so on. It relates different forms of assistance to achieving these goals. Thus it would diagnose the hierarchies of need as they relate to the specificities of pre-democratic transition, transition and post-transition situations, taking into account the individual cultural, social and historical circumstances of the country and their bearing on its likely political trajectory, especially its parties.

The reality The theory is one thing; the reality is different. It is not obvious from the ad hoc and unsystematic growth of international party assistance – much of it opportunistic and short-term in nature – that careful thought has been given to relating form and content to a strategic assessment of functional requirements of the system overall. Perhaps it was inevitable that historical connections, accidents of personal contact, the need to seize the moment and tactical considerations would take precedence in the early days of political change. The absence of a theory of when to phase out party assistance – where an emerging democracy has ‘graduated’, or is deemed a hopeless case, or because other instruments become more appropriate – is symptomatic. It befits the uncertainty that studies of democratization show over the signs of completed democratic transition and indicators of consolidation. The analytical ground-clearing seems to have been left behind in the rush to execute the political and bureaucratic imperatives of party assistance. But the two should be connected. To illustrate, an objective that dwells on parties as vote-gathering, power-seeking machines will differ from a design to help parties replicate democratic norms in their internal arrangements, impose 196 Peter Burnell accountability requirements on their leadership and respond to wider opinion. There are interesting possibilities for tension between donor enthusiasm for ‘democratic decentralization’ (a USAID policy objective) and the typical wish of party central organs to retain control at provincial and local levels. Different again would be an objective to help parties organize around principles and programmes, not personalities, or to become the kind of party that reaches out to voters by offering reasoned positions on public policy, not by dispensing patronage and emotional appeals. Interestingly, attempts to move parties on from being personal (and often transient) vehicles, conceived primarily for the benefit of the leader’s political ambitions, to becoming more like structured organizations that reflect fundamental social cleavages, could appear to reverse the sequence undergone in several older democracies. Finally there are the unstated aims. Even the western parties engaged in interventions have self-regarding agendas; their ‘partners’ too have their particular agendas, in some instances personal financial enrichment.

The different forms of party assistance

An analytical distinction is between conveying the abstract ideas and values of democracy and its more practical mores, and on the other side disseminating particular belief systems drawn from the range of political ideologies compat- ible with democracy (socialism, liberalism, conservatism and so on) and aiding partisan political forces. The second often includes substantive policy advice to parties especially on central issues of economic and social policy. This sharing of intellectual property has extended even to help with drafting election manifestos. In practice the transfer of the two sets of ideas – some expressly about/supportive of the democratic way, and other, more speciﬁc belief systems or ideologies – may be thoroughly intertwined, reciprocally reinforcing.

Party identities Helping parties form their beliefs, policies and programmes begs questions about what dimensions should govern the ideological branding and the criteria to employ in selecting partners. The familiar left (socialism) – right (conservative) continuum, sometimes said to designate ‘standard parties’, appears uppermost. But quite apart from the fact that our present understanding of this continuum and its bandwidth may not be appropriate to all new democracies, especially post-communist democracies, its standing in the West is increasingly questioned. There we see ‘catch-all’ parties, largely stripped of ideological baggage, and ‘cartel parties’ denoted by ‘collusion and co-operation between ostensible competitors’ (Katz and Mair, 1995, p. 17). Traditional redistributive issues might be losing salience in the advanced industrial societies, where truly ‘left-wing’ parties are an endangered species, but extensive Party Systems in New Democracies 197 absolute poverty and gross inequality characterizes many emerging democracies. There is a presumption that eligibility for party assistance requires a basic commitment to the norms and values of liberal democracy. But is that sufﬁcient? Should the parties be all-inclusive, what does that mean in practice, and how realistic is it anyway? It is easy to say that parties avowing racism, sexism or xenophobia should be off-limits, but in reality this leaves a signiﬁcant margin for latitude. It might be useful to distinguish a party from its more colourful leaders; the electorate might ascribe an identity to a party that is at odds with its self-image and which it tries hard to disown. Many emerging democracies are plural societies with ‘non-standard’ parties. And according to Lijphart (see Mainwaring, 1998, p. 223), where sharp cultural, ethnic, religious or linguistic cleavages exist, a multi-partyism that mirrors this pattern can serve both stability and democracy. After all, nationalist parties, religious-based parties and feminist political groups are not unknown in western Europe. In countries like South Africa ethno-nationalist and region- ally-based parties might offer the best hope for consolidating political pluralism for the foreseeable future (Friedman, 1999, p. 8). Meanwhile, no one seriously questions the entitlement of Irish-Americans to sponsor (to the tune of several million dollars annually) Ulster’s republican party, Sinn Fein, which aspires to remove territorial sovereignty from Britain.

Practical assistance Advice, information and practical skills have been offered through various mechanisms including fact-ﬁnding visits, seminars/workshops and training programmes. These cover matters like party structure and organization, election campaign techniques, fund-raising methods, media management, opinion research and leadership skills. There have been donations of material as well as some cash gifts to help parties carry out essential activities. While some actors like the Stiftungen are forbidden to make all-purpose payments to parties, unions or campaign funds, observers claim that such restrictions are circumvented. There are also projects that aim to enhance the parliamentary skills of elected representatives. This includes the capacities to contribute to the legislative process and exercise scrutiny and oversight of the executive. In contrast, efforts to promote directly ‘bottom-up’ involvement and accountability by party leaders and ofﬁcials to the grass-roots are under-represented, although some projects claim to promote dialogue between representatives and electorate and to help women’s political advancement.

Political support In certain situations the symbolic beneﬁts, the chance to acquire enhanced credibility and legitimation from external contacts and recognition are most 198 Peter Burnell valuable, beneﬁting a political group’s claim to democratic credentials or giving extra conviction to its declared ideological positioning or both. Where the government harasses opponents a measure of international protection that raises the diplomatic costs can be most important. In extremis safe havens are offered outside their own country to leading opponents. In this way Chilean exiles had an opportunity to observe European socialism and communism after Chile’s 1973 military coup.

Does it work?

How would we know if party assistance is performing well? The criteria of success vary according to the objectives – partisan electoral assistance, longer- run institution-building, the advancement of systemic properties like a representative party system and so on. The interpretation of the consequences will vary along with the actors and their perspectives. Whose objectives should count – the ‘providers’ or the ‘recipients’, the ‘winners’ or the ‘losers’ and non-beneficiaries? The results will depend also on the kind of evidence that is judged relevant – the respective merits of qualitative and quantitative information – and the time frame used. In some interventions the possibility of registering an effect is virtually instantaneous. For others there will be lags, depending on the programme. It is essential to differentiate results defined in terms of the individual, or combined, objectives of projects/programmes and impact more broadly conceived. The effects on parties and party system are less significant than the consequences for the functions they are supposed to perform for healthy democracy. The answer to the question ‘does assistance help parties?’ could differ from whether it strengthens democracy or promotes democratization. Moreover, we can never know the counterfactual. Social scientists are notorious for being unable to agree on causality, even in the presence of high levels of correlation. For all these reasons, analysts of democracy assistance like Carothers (1999) have modest expectations about its influence. After all, democratization itself is a complex process; the key milestones and their signifiers are contested. The primary forces that determine political stasis/change are no less disputed. A common view is that domestic forces almost invariably predominate anyway, even more so in the formative years of party building and during democratic consolidation. And among all the possible external influences, many are much more powerful and more pervasive than democracy assistance, let alone party assistance.

The burden of evidence Democracy promoters’ public relations materials often cite politicians proclaiming how valuable was the support they received. We can also easily Party Systems in New Democracies 199 detect many examples of assistance that failed to achieve the stated objectives, or where a successful outcome seems unlikely. In itself this need not be dis- turbing. More unsatisfactory, for partisan supporters, are the parties that dis- appeared, or abandoned their ideological pretensions, or performed badly in office, following external support. Assistance that was not intentionally partisan has occasionally provoked the ruling party to work harder to maintain power by fair means and foul, construing the external intervention in a hostile light. In the choppy and unfamiliar waters of some emerging democracies choosing the ‘right’ partners has necessitated inspired guesswork by external actors (Phillips, 1999). More damning still is the evidence of parties and politicians who have weakened or shed their pro-democracy credentials, after having benefited from external assistance or following interventions which prompted the institutional reforms that made their accession to power possible. Critics cite Zambia’s Movement for Multi-Party Democracy, in office since 1991, as an example. All in all, it is difficult to prove on the basis of present evidence that the international support to parties and party-building is a crucial part of the explanation of the most successful recent transitions to democracy. Also it would be hard to demonstrate that the parties, party system and condition of democracy in, say, Zambia or Russia would now be very different if the international community had not attempted directly to strengthen their political pluralism.

So why persist?

The available evidence is not overwhelmingly positive, but this partly reflects that it is relatively easier to conceptualize, identify and collect firm evidence about failure. The most weighty argument possible against party assistance as a form of democracy promotion would be that it is counterproductive. At present we simply do not know enough about the consequences, intended and unintended, to confirm that. USAID apart, most democracy promoters are shy of making evaluations and impact assessments; they shun independent evaluations and are reluctant to publish any results. A rationalization is that the methodologies are still a youthful art form, not a science, and can be controversial, but this is not a justification. So, should the international community persist with party assistance? There are several reasons for saying yes. First, there is the belief that strong parties can owe much to elite behaviour and fortuitous history; they are not solely the result of deep social cleavages (Lipset, 2000, p. 54). Hence international involvement can possibly make a difference. Second, the arguments against non-intervention are nowhere stronger than where external actors helped topple the ancien régime and speed democratic breakthrough. A moral obligation to help democratic transition succeed is one inference. In situations where interventions enable ruling 200 Peter Burnell parties/leaders to acquire a cloak of democratic legitimacy by submitting themselves to elections, yet the rulers still abuse power, the international community, if it is consistent, should try to move the democratization further forwards, such as by strengthening opposition parties. Third, placing restrictions on intervention is non-neutral: it signals a decision to not try to offset the many forces impeding the development of appropriate parties and party systems. Many such forces begin in the international environment, and may include financial/economic support to governments and the bribery of politicians by international corporations. The negative social consequences of globalization can lead vulnerable groups to view political participation as an unaffordable luxury, irrelevant to their needs. Although such influences should be tackled directly, for now party assistance offers a more practicable route than, say, radically transforming the international economic system. Fourth, the risk of an unbalanced pattern of political development is made greater by approaches to international cooperation that include strengthening the governance capabilities of states and enhancing the vitality of economic markets but ignore political society – of which parties are a central component. Governance and state-building measures that increase the power at the disposal of governing parties necessitate parallel steps to ensure the party system furnishes parties that are up to the task of governing and a vigorous countervailing check (strong opposition). Of course party competition cannot be equated to democracy; there must be additional opportunities for the people to participate. But although civil society aid programmes are very attractive to donors they cannot supply a perfect substitute. Rather than civic associations offering an alternative, parties and their leaders provide ‘a crucial element in any discussion of the relationship between civil society and the impetus towards democratization and of the broader triadic relationship between democracy, civil society and the developmental capacity of the state’ (White, 1994, p. 382). Assistance to opposition parties may also be indicated where governments impose restrictions on non-governmental organizations’ access to foreign support for political activities like lobbying and advocacy. Fifth, even where party assistance does not advance political pluralism its potential to serve other objectives like human rights, more effective governance and poverty-reduction merit further exploration. The gaps in our knowledge here make such exploration that much more needful. Finally, cooperation in party building has often been demand-led, from new and emerging democracies. This demand may have been partly induced by western influence, but the relationship is one of freely consenting adults, often drawn from the private sector. For the West to seek somehow to curtail such cooperation could look imperialistic and necessitate an invasion of liberty at home, especially if it prohibited the use of private resources honestly Party Systems in New Democracies 201 acquired. For the sake of consistency and effectiveness a bar would have to be placed on other means whereby ideas and people cross borders – improbable in this, the communications age. The important thing, then, is to ensure there is a plurality of suitable advice and democracy assistance. In this context, although party funding in new democracies is especially problematic (Burnell and Ware, 1998), international financial transfers are particularly hard to justify, being easy to conceal and very open to abuse. The issues cannot be discussed here. But similar risks, for example becoming excessively oriented towards international ‘partners’ at the expense of cultivating ties with domestic constituencies, can also bedevil international transfers of ideologies and policy briefs.

Alternatives

Parties and party systems are influenced by many variables including electoral rules, social heterogeneity, political culture and informal institutional environment. The amount and pattern of resources of all kinds, including finance, might be the most susceptible to external influence but not necessarily the dominant force. Attention should be redirected to other ways in which the international community might strengthen party systems, apart from assisting individual parties directly, especially in cases where democratic transition has stalled or gone into reverse (see Burnell, 1998). Again countries should be studied individually. The nature of the political regime determines what the international community can reasonably hope to do. Naturally, fewer options exist where pluralism is suppressed, but even there, a worthwhile strategy might be possible. International elections monitoring and observing are well-tried, and should now focus more on the wider electoral infrastructure and the period before the election campaigns (Elklit, 1999). Civic education to inform voter participation (or reasoned abstention) is another indirect form of party assistance. The emphasis could shift from voter education to ‘the less glamorous task of teaching people to value democratic institutions and processes more for their own sake than for what they may deliver in terms of immediate and tangible benefits’ (Mattes and Thiel, 1998, p. 108). Assistance to civic associations is now very fashionable, following growing elections-support fatigue. Here the challenge is to help civil society give indirect support to multi-party politics, by pressing for an improved legal and institutional environment for political competition. This includes, for example, regulations affecting the registration of parties and their freedoms, party finance (both funding and expenditure), and the media. It also means encouraging civic actors to make appropriate demands on all the politicians, such as insisting on responsible behaviour and integrity. The more parties are recognized and ‘supported’ by relevant regula- 202 Peter Burnell tory measures the more the party system can be said to be institutionalized (Randall and Svåsand, 2001). Once they settle down, party systems ‘tend to be preserved by the institutional context in which they were formed’ (Ware, 1996, p. 220). Hence international attention early on to shaping the institutional framework, working through and in conjunction with civic actors, promises a longer reach than direct attempts to influence the parties. This is nowhere more needful than where the political executive shows a predisposi- tion to make constitutional initiatives that would emasculate the legislature, thereby rendering much party activity redundant.

Limits of inteventions Widner (1997, p. 102) argues for encouraging a closer liaison between politicians and civil society organizations. However, the encouragement of a positive direct relationship is especially problematic terrain for international democracy promotion. The new democracies each draw on their own recent experience. Furthermore, there are competing prototypes of the ideal relationship. And, like the Labour Party’s relationship with the trade unions in Britain, there is flux and change even in established democracies. To develop a web of interconnecting supports between political and civil society there must be a positive attitude towards constructive engagement on all sides. Sometimes, actors on at least one side are wary or contemptuous or indifferent. The prevailing perception may be one of mutual competition for influence and resources (including international democracy assistance and development aid). Politicians can resent the political ambitions of civic leaders, who may be strangers to the electoral process; they query their claims of representativeness. Sometimes they portray them as creatures of the donor community. It is difficult to see how foreign intermediaries can pose as matchmakers in such situations. To be sustainable, political–civic interactions must be authored and developed locally. If parties are captured by particularist social forces that are unaccountable and have objectives against democracy, the prospects for democratization will be compromised. The same is true where parties heavily impinge on the independence of important actors like the private media. The prospect of foreign actors endeavouring to close the spaces between parties and civic associations, as if they could, and should, weave the warp and weft of entire political communities, invites many pitfalls. A similar point could be made about attempts to overcome fragmented and ineffectual opposition among opposition parties by pressing them to cooper- ate, collaborate or unite. It is precisely in situations like this – where the opposition parties’ own political shortcomings prevent them making serious headway – that compensating efforts by the international community can be most difficult to justify. Interventions of this sort can look more like an inter- Party Systems in New Democracies 203 ference with the workings of democracy than an exercise that will strengthen democracy. Foreign intervention is no substitute for domestic credibility and popular support. The institutionalization of a competitive party system requires the development of trust in the parties as institutions, and that too is no easy-to-import talisman: it has to be earned. Nevertheless, assuming it is rational to plan party assistance in conjunction with different forms of democracy assistance (legal technical assistance, aid to civil society and so on) and democracy promotion more generally, then the other interventions should be considered as well. That means, for instance, the international pressures to reform economic and social policies and institutions. A comprehensive approach to trying to help build democracy would aim for coherence and consistency across the board, and reduce the conflicts between the different streams of developmental assistance and political advice. For example, if governments were allowed greater leeway in applying neo-liberalism to their socio-economic affairs, then the parties would be able to offer more varied programmatic alternatives to the electorate. That would counter, not reinforce, the trend towards greater external direction of domestic affairs. It would give greater meaning to the act of voting, by increasing the scope for parties to differentiate themselves. Only when people can see that making choices between parties matters (or can matter), in the sense of making a difference to what concerns them most (bread and butter issues of welfare and security from violent crime often come first), can we expect a lively political pluralism to take hold. The Bretton Woods institutions’ attitudes towards conditional lending and debt relief probably exert far more influence on the prospects for stable competitive party politics than any amount of direct party assistance. Finally, just as some of the most effective ways of advancing the possibility of stronger parties and plural politics could be through indirect measures of support, so even these do not guarantee success. Current experience indicates that all actors presently engaged in promoting democracy by whatever means should be modest about their capability to shape the future, even where they do think strategically.

References

Bogaards, Matthijs (2000) ‘Crafting Competitive Party Systems: Electoral Laws and the Opposition in Africa’, Democratization, 7: forthcoming. Burnell, Peter (1998) ‘Arrivals and Departures: A Preliminary Classiﬁcation of Democratic Failures and Their Explanation’, Commonwealth and Comparative Politics, 36: 1–29. Burnell, Peter (ed.) (2000) Democracy Assistance, London: Frank Cass. Burnell, Peter and Ware, Alan (eds) (1998) Funding Democratization. Manchester: Manchester University Press. 204 Peter Burnell

Carothers, Thomas (1999) Aiding Democracy Abroad. Washington, DC: Carnegie Endowment for International Peace. Elklit, Jørgen (1999) ‘Electoral Institutional Change and Democratization: You Can Lead a Horse to Water, But You Can’t Make It Drink’, Democratization, 6: 28–51. Friedman, Steven (1999) ‘South Africa: Entering the Post-Mandela Era’, Journal of Democracy, 10: 3–18. Gillespie, Richard, Waller, Michael and López Nieto, Lourdes (1995), Factional Politics and Democratization, London: Frank Cass. Katz, Richard S. and Mair, Peter (1995) ‘Changing Models of Party Organization and Party Democracy’, Party Politics, 1: 5–27. Lipset, Seymour M. (2000) ‘The Indispensability of Political Parties’, Journal of Democracy, 11: 48–55. McLean, Iain (1996) Oxford Concise Dictionary of Politics. Oxford: Oxford University Press. Mainwaring, Scott (1998) ‘Party Systems in the Third Wave’, Journal of Democracy, 9: 67–81. Mair, Stefan (2000) ‘Germany’s Stiftungen and Democracy Assistance: Comparative Advantages, New Challenges’, in Peter Burnell (ed.), Democracy Assistance. London: Frank Cass. Mattes, Rob and Thiel, Hermann (1998) ‘Consolidation and Public Opinion in South Africa’, Journal of Democracy, 9: 95–110. Phillips, Anne L. (1999) ‘Exporting Democracy: German Political Foundations in Central-East Europe’, Democratization, 6: 70–98. Pridham, Geoffrey (1999) ‘The European Union, Democratic Conditionality and Transnational Party Linkages. The Case of Eastern Europe’, in Jean Grugel (ed.), Democracy without Borders. London: Routledge. Randall, Vicky and Svåsand, Lars (2001) ‘Party Institutionalisation and the New Democracies’, in Jeff Haynes (ed.), Democratic Consolidation in the Third World: Problems and Prospects. London: Routledge. Taagepera, Rein (1998) ‘How Electoral Systems Matter to Democratization’, Democratization, 5: 68–91. van Cranenburgh, Oda (1999) ‘International Policies to Promote African Democratization’, in Jean Grugel (ed.), Democracy Without Borders. London: Routledge. Ware, Alan (1996) Political Parties and Party Systems. Oxford: Oxford University Press. White, Gordon (1994) ‘Civil Society, Democratization and Development (I): Clearing the Analytical Ground’, Democratization, 1: 375–90. Widner, Jennifer (1997) ‘Political Parties and Civil Societies in sub-Saharan Africa’, in Marina Ottaway (ed.), Democracy in Africa. The Hard Road Ahead. Boulder, CO: Lynne Rienner. 13 Regulatory Accountability: Towards a Single Citizen-Consumer Model? Martin Lodge

Policies of regulatory reform and liberalization are said to have challenged the nature of the public sector in the past two decades. This policy shift from the so-called positive to the regulatory state (Majone, 1997) has been characterized by a move from public to private ownership, the creation of quasi- autonomous supervisory agencies and an increased application of formal rules instead of informal ‘high trust’ relationships (Loughlin and Scott, 1997, pp. 205–7). Regulatory reform in the area of utilities has been particularly prominent with reform initiatives spreading from telecommunications to energy, water and transport throughout the 1980s and 1990s. At the same time, there has been an increasing concern over ‘public services’, reflecting the traditional obligation of the state to provide ‘Daseinsvorsorge’, the German administrative principle that obliges the state to secure an effective infrastructure for the pursuit of an individual’s economic activities. This chapter deals with three democratic concerns relating to the transparency of regulatory regimes: the impact upon citizenship; the potential lack of accountability; and the maintenance of national diversity with the possible predominance of a unified model. Regulatory transparency defines the changing relationship between state and citizen in the age of the regulatory state which arguably places less stress on traditional means of legitimization through input- oriented means of majoritarian institutions (Majone, 1999). The first concern relates to the impact of regulatory reform on the nature of citizenship. In the West European welfare state, the ‘normal’ way to deliver utility services has been through public bureaucracies, structured as politically directed hierarchies of permanent, full-time and specialized government officials. Accountability to the wider electorate was to be ensured by the principles of ministerial accountability and responsibility. Part of the intellectual attack on this ‘traditional’ model was targeted at the ‘high trust’ assumption that civil servants acted in the public interest, given incentives to ‘budget maximize’ or ‘bureau shape’. Instead, the importance of ‘low trust’ means of

205

K. Dowding et al. (eds.), Challenges to Democracy © Political Studies Association 2001 206 Martin Lodge control, including benchmarking, performance-related pay and privatization, was stressed. Such moves towards ‘low trust’ instruments were perceived to strike at the ‘conventional’ civil servant ethos and attracted considerable criticism. For example, the claim that privatized public services undermine bureaucratic ethics and democratic accountability relied on the ‘paternalistic’ argument that the nature of utility services prohibits informed choice (Haque, 1996, 1998). Similarly, privatized services were said to be on an inevitable downward trend given the replacement of ‘public interested’ ofﬁcials with ‘proﬁt-seeking’ businesses, leading to increased inequality, unemployment and social marginalization (for New Zealand, see Kelsey, 1995). Furthermore, the perceived lack of popular participation was criticized. By drawing on Hirschman’s categories of exit and voice, it was argued that the introduction of exit (or choice) undermines the exercise of voice (Hirschman, 1970; Falconer and Ross, 1999, pp. 341–2). The second concern relates to the perceived lack of accountability of existing regimes (Baldwin and Cave, 1999). The discussion has centred on methods of enhancing regulatory accountability by input-oriented means (ranging from procedural to monitoring devices, see Graham, 1997) or on the complexity of accountability (see Scott, 2000). Few studies have approached the issue of transparency by focusing on the notion of consumer sovereignty (Hood, 1986, pp. 171–93). Consumer sovereignty highlights as an ‘ideal type’ that a consumer should individually be able to choose the supplier, the nature and the amount of the provided good. By using this perspective, this chapter is therefore less concerned with ownership questions or constitutional law debates on parliamentary accountability, than with regulatory tools which facilitate the sovereign consumer. Finally, the literature on regulatory regimes still lacks comparative perspectives, despite numerous explanations of regulatory change (Hood, 1994, pp. 19–36). Studies on the welfare state and ‘big government’ have pointed to the persistence of ‘families of nations’ (see Castles, 1998) in the face of common, international challenges. Given the increasing amount of European legislation governing national regulatory regimes, this chapter questions whether there has been a move towards a common regulatory approach or a persistence of ‘regulatory diversity’ (Wilks, 1996). The following compares regulatory regimes within and across ‘families of nations’ (United Kingdom, New Zealand, Ireland, Sweden and Germany) in order to enquire whether there has been an emergence of a regulatory model with regard to ‘consumer transparency’. A typology of transparency mechanisms is developed before their existence is assessed in a comparative analysis of regulatory regimes. It concludes by assessing issues of convergence and the complexity of transparency mechanisms. Regulatory Accountability 207

Regulatory design and transparency mechanisms

The literature on regulatory design has focused on the advantages of ex post and ex ante controls on the behaviour of agents, both in terms of regulators and regulatees (see McCubbins et al., 1987; Macey, 1992). These controls are supposed to reduce the risks of agency and coalitional drift. Agency drift is used to describe attempts by agents to exploit asymmetric information advantages in order to pursue their own preferences rather than those of the principal. Coalitional drift is defined by changes in the preferences of the principal, such as a change of government, which alters regulatory priorities and therefore threatens constituencies benefiting from the regime established under the original ‘enacting coalition’. Agency drift is said to be prevalent given the cost of control, in terms of monitoring, providing rewards or imposing sanctions. The exercise of sanctions might be politically costly or might undermine the original regulatory objectives. Thus, besides ‘low cost monitoring systems’ such as ‘fire alarms’ by affected constituencies, self-revelatory mechanisms, such as the publication of information, and the establishment of political and administrative watchdogs (for example, committees and regulatory agencies) are promoted. Procedural controls are said to offer ‘deck-stacking’ means to steer agents towards desired outcomes without prescribing any definite targets. Such procedural instruments include regulatory decision-making procedures and the enfranchisement of specific constituencies in the decision-making process. Controls to enhance transparency operate at the ex post and the ex ante stages and represent attempts to minimize advantages stemming from asymmetric information. Ex post mechanisms rely on fire alarms and other monitoring devices to call a deviating agent to account, whereas ex ante mechanisms prescribe certain outputs and outcomes which agents have to fulfil. Without aiming to be fully exhaustive or mutually exclusive, four transparency ‘mechanisms’ can be distinguished: information, choice, representation and voice (see Figure 13.1). These four mechanisms can be differentiated in the way they emphasize ‘citizen’ or ‘customer’-oriented instruments, high- lighting the importance either of ‘participation’ or ‘market’ instruments. At the same time, instruments can be distinguished in the way they are either collectively or individually provided and/or exercised.

Choice The notion of consumer sovereignty stresses the importance of choice, the ability of complete exit and the ability to choose the amount of the good desired. Advocates of privatization and liberalization argue that competition provides the best incentives for enhancing consumer beneﬁts, pointing to 208 Martin Lodge

Figure 13.1 Mechanisms for transparency. lower prices and wider availability of services and goods. The extent of choice, however, cannot merely be measured in terms of market shares or the availability of choice options, but requires the ability to choose easily, requiring contracts, legal inter-operator provisions and speciﬁc technical installations. For example, competition in British telecommunications developed at far lower speed than in Germany in the absence of call-by-call and uniﬁed billing arrangements.

Voice The ‘voice’ mechanism mostly operates at the ex post level. Users are provided with procedural complaint and grievance handling mechanisms. Complaint- handling arrangements and the collection of data concerning complaints have increasingly become part of the ‘established’ menu of regulatory instruments. At the ex ante level, voice options exist in terms of openness of regulatory decision-making but, in general, the role of the individual utility-user is marginal.

Information Information asymmetries are among the key ‘benevolent’ reasons for regulation. Information deficits are likely to lead consumers to ‘lemon choices’, resulting in dissatisfaction but also in low trust, which is likely to disadvantage new market entry and thereby safeguards the incumbent. Informed customers and ‘naming and shaming’ practices force firms to offer ‘decent’ services and motivate service competition ‘to the top’. For example, an old Chinese rule to advance transparency forced physicians to display the number of patients who had died in their care (Hood, 1986, pp. 178–9). Information is provided by the regulated industries themselves as part of the licence conditions or as part of their competitive advertising, but also by consumer groups and the media. Information includes disconnection rates, repair responses, agreed codes of practices and financial information in order to investigate potential abuses of market positions. Indirect tools of information gathering and dissemination have also increasingly become part of the arsenal of European regulators, such as the European Environmental Protection Agency (Dehousse, 1997). Regulatory Accountability 209

Representation The empowerment of ‘public interest groups’ (so-called PIGs) has been advocated as a solution to regulatory problems of capture and corruption (Ayres and Braithwaite, 1992). In this model, PIGs not only participate in the direct punishment of non-compliant ﬁrms, but also compete regularly for their privileged status. Tripartite arrangements are widespread in regulatory regimes (Lodge, 1999, p. 47). Consumer organizations are consulted and advise regulators or ministers on consumer issues. For example, in Portugal, rate-setting was conducted by the regulatory agency, Erse, after consultation with a so- called tariff council. This committee was composed of an equal number of representatives from consumer bodies and Portugal’s six regulated electricity companies where the committee’s chairperson, the president of Portugal’s ‘Consumer Institute’, had the power of the casting vote. The protection of consumer interests has also become part of the ‘usual’ statutory objectives of regulatory agencies. The OECD advocates the combination of economic and consumer-oriented regulation in one regulatory ofﬁce, claiming that consumer groups have often been liable to adopting ‘anti- liberalization’ rhetoric (OECD, 1995). Table 13.1 summarizes the various transparency mechanisms. The next section explores these mechanisms in comparative perspective. The limitations of this undertaking should, nevertheless, be noted. The list of transparency mechanisms does not aim to be exhaustive or mutually exclusive. Furthermore, the discussion of the country experiences is mainly illustra- tive and represents an examination of continuously evolving regulatory regimes. Finally, the following assesses the existence rather than the effectiveness of transparency mechanisms.

English-speaking family of nations

New Zealand New Zealand’s public sector reform arguably represents one of the most comprehensive reforms among OECD countries. While most studies of New

Table 13.1 Transparency mechanisms

Ex ante instruments Ex post instruments Choice Competition and choice Ease of switching providers Voice Individual access to Complaint-handling decision-making procedures Information Service quality standards Benchmarks and performance leagues Representation Consumer councils involved Consumer advocacy in decision-making 210 Martin Lodge

Zealand have focused on administrative reforms and their basis in institutional economics (see Pollitt and Bouckaert, 2000), less attention has been paid to the ‘light handed’ approach towards utility regulation. This approach combined a strong emphasis on structural liberalization prior to ‘privatization’ and a reliance on competition law. The 1986 State-Owned Enterprise Act led to the corporatization of the former trading units of government departments. Outright privatization was initially not pursued in order not to alienate Labour’s electoral support. Reforms were aimed at overcoming the perceived failure of traditional ministerial and bureaucratic oversight mechanisms and poor reporting and auditing practices. Great stress was placed on financial transparency and accountability with social objectives only playing a secondary objective (Boston, 1998, p. 34). Once the National Party had resumed power in 1990, reforms were advanced with the sale of the telecommunications operator in 1990 (with the retention of a ‘Kiwi’ share to safeguard particular ‘universal’ services). In the gas sector, a policy of privatization led to the lifting of the statutory monopoly franchises in 1992. In electricity, distribution was originally provided by a mixture of departments within local government and local electricity power boards. These purchased electricity from the country-wide grid of the Ministry of Energy, which was also responsible for electricity generation. Following the corporatization of generation and transmission operations (into Electricity Corporation New Zealand) in 1986, structural separation of transmission from generation was initiated in 1988. Franchise restrictions on distribution were abolished in 1993 and extensive information disclosure regimes established, including the possibility for government imposed price controls. State-owned generation operations were separated in 1995 and 1999 and partly privatized. An enhanced regime for consumer switching was established in 1999. The emphasis across transparency mechanisms rested on the notion of consumer sovereignty and on contractual arrangements. The absence of sector- specific regulatory regimes also meant that issues such as interconnection, line charges, metering, direct access and competition were not regulated (Bergara and Spiller, 1997). Choice was facilitated across all domains (despite problems in interconnection negotiations in telecommunications and electricity), including, for example, an industry-led numbering transfer system in telecommunications and a gradual extension of choice in electricity supply. The absence of licences meant that operators were not required to establish specific voice instruments, such as complaint handling procedures or compen- zation mechanisms. Electricity companies did not establish wide-ranging complaint handling procedures (Ministry of Consumer Affairs, 2000). With regard to information, regulation was to enhance the extent of available information for financial-regulatory and consumer-oriented purposes. Disclosure rules were gradually increased in their detail, although not in their scope. The impor- Regulatory Accountability 211 tance attached to information disclosure reflected not only a response to past failings, but also aimed to minimize information asymmetries in the absence of sector-specific regulators. Information disclosure rules in telecommunications (1990) and electricity (1994) demanded separately audited financial statements for monopolistic and competitive businesses and on contracts, financial, efficiency and reliability performance measures and charges. In 1999, the electricity regulations were strengthened following dissatisfaction with the reporting practices of the transmission companies and public outcry about the prolonged effects of the 1998 Auckland power failure. While the impact of these information disclosure provisions on consumer transparency was doubtful, the so-called ‘Consumer Power Switch’ offered an online and interactive comparison between electricity providers. The Ministry of Consumer Affairs, an independent but subordinate ministry to the Ministry of Commerce, played the main representative role. Its task was to ‘promote a fair and informed market place’ by facilitating ‘self-help’ consumer choice and by encouraging voluntary business self-regulation. From 1990, the Ministry monitored the performance of the private telecommunication operators. In 1997, following customer complaints about the electricity companies’ contracts, the Ministry criticized the failure to offer ‘good contracts’ in terms of metering arrangements, consultation, liability questions and disconnection procedures (Ministry of Consumer Affairs, 1997a). The implementation of the industry’s voluntary code of practice was criticized in later reviews (Ministry of Consumer Affairs, 1998, 2000). Consumer interests were represented by the self-funded Consumers’ Institute which operated (with government support) the ‘Consumer Power Switch’ mechanism. The Labour-Alliance minority government launched major reviews of telecommunications and electricity regulation in early 2000. Inquiries recommended that regulation should become more sector-specific and mandatory, which also included the increased spread of price controls. The position of consumers was to be enhanced by the creation of sectoral ombudsmen. Provisions in the Commerce Act were to be strengthened to prevent discrimination by market-dominant firms.

United Kingdom The privatization process in the United Kingdom has been characterized by an incremental adaptation of regulatory models. The process of privatizing utilities, starting with telecommunications (1984), gas (1986), water (1989), electricity (1989/90) and railways (1993), can be deﬁned by four key themes. First, the usage of the price control system (RPI-X) aimed to avoid the juridiﬁed processes of US-style rate of return regulation. Second, quasi-autonomous sectoral regulatory bodies were established, following the initial rejection of the 212 Martin Lodge

UK’s competition authority, the Office of Fair Trading (OFT), to become responsible for telecommunications regulation. The third theme which emerged by the late 1980s emphasized the importance of structural change. Initially, the importance of universal service and the financability of these services were prioritized (copied from the 1969 Post Office Act and the 1973 Fair Trading Act), while competition and social objectives were secondary objectives (Hall, Scott and Hood, 2000, pp. 24–5). Drawing on negative experiences with vertically integrated private monopolies (British Telecommunications (BT) and British Gas), structural change was implemented in electricity and railways (vertical separation), water (yard- stick competition) and, finally, also in gas. The fourth theme was a growing interest in quality and service standards. Following complaints about BT in 1987, the Office of Telecommunications (Oftel) forced BT to resume the publication of performance achievements and established the threat to impose financial penalties. The 1992 Competition and Services (Utilities) Act, itself an offspring of the Citizen’s Charter, extended consumer regulation across all utility domains, including the promotion, prescription, publication and enforcement of performance standards as well as the provision of complaint and dispute handling procedures. These themes were developed in the 2000 Utilities Bill which called for a strengthening of the role of consumer councils. However, instead of the envisaged uniformity of consumer regulation across all utility sectors, the final Act effected mainly energy-related regulation after lobbying by telecommunications companies and departmental turf fights on water regulation. The extent of choice increased over time. The regulatory authorities took a strong interest in facilitating network access for new entrants, but, over time, moved away from being proactive liberalizing authorities towards becoming competition watchdogs. The 1992 Act advanced individual voice by becoming part of the licence requirements. Furthermore, information was enhanced by the required publication of disconnection rates, late and deferred payment procedures as well as service standards and targets. The status of voice and representation were advanced by the Labour government’s plans to give consumer councils the explicit task to deal with and monitor customer complaints. Criticisms about the regulatory agencies’ non-transparent decision-making patterns led to an increased procedural openness, especially in the cases of Oftel and Ofwat (Prosser, 1997). In terms of information, the regulatory office launched major changes as a response to Labour’s initiatives in utility regulation. Direct price information in telecommunications was only established (following ‘encouragement’ by Oftel) in November 1999 by a few operators. Similarly, in gas and electricity, the merged regulator (Ofgem) offered Internet-based price and complaint record comparison on a non-interactive basis. While Oftel had a history of publicizing consumer-relevant material, in Regulatory Accountability 213 the case of energy regulators, these efforts represented an attempt to pre-empt moves by the government. The UK’s representation mechanisms built on previous arrangements in the era of public ownership and varied considerably between utility sectors (Prosser, 1997). They ranged from local and regional committees in telecommunications and energy to the centralized Gas Consumers’ Council which investigated complaints and acted as consumer advocate. The Labour government’s Utilities Bill aimed to harmonize these structures across sectors along the lines of the Gas Consumers’ Council. Despite the rhetoric of enhancing the role of consumer councils in concentrating policy advice, complaint handling and representation, the past experience did not suggest that these were likely to play a crucial role in the policy process. The regulatory agencies’ consumer role was ambivalent. Despite much praise for their procedural openness, their emphasis on quality standards and their enhanced powers under the 1998 Competition Act, in terms of representing consumers the offices had a less praiseworthy record. For example, the customer complaint unit in Oftel played only a minor role (Hall, Scott and Hood, 2000, pp. 188–92). Nevertheless, the Utilities Act established the recognition of consumer interests as the primary objective for all regulators.

Ireland In terms of regulatory reform, the Republic of Ireland has arguably been a ‘laggard’ until the late 1990s. Developments in telecommunications were mainly driven by the evolving EU agenda. In the key area of telecommunications, fixed-line voice telephony, competition was introduced on 1 December 1998, although Ireland had secured derogation from EU-required liberalization until 1 January 2000. This acceleration occurred after strong lobbying of telecommunications companies and a governmental decision to utilize regulatory policy to attract foreign investment. A regulatory office, the Office of the Director for Telecommunications Regulation, was established along the lines of the UK office, Oftel. In electricity, developments were similarly shaped by EU legislation (Directive 96/92/EC). Despite initial government proposals to protect the Electricity Supply Board (ESB) by establishing a ‘single buyer system’ with a vertically integrated operator, a ‘third party access’ system was adopted with the separation of a state-owned company for transmission. A regulator, the Commission for Electricity Regulation, was established to promote competition. Given the EC Directive’s requirement to gradually increase the scope of liberalized markets, the market only opened to customers whose total consumption represented 28 per cent of the market (increasing to 32 per cent in 2003). In gas, liberalization was introduced in 1995 which allowed for competition for large users of natural gas and offered third-party access to the trans- 214 Martin Lodge mission network (thus prior to the EC Directive 98/30). However, due to the strict definition of ‘large users’, the main ‘eligible customers’ were the ESB and the Irish Fertilizer Industries. Access and interconnection charges were set by the Minister for Public Enterprise. In terms of transparency mechanisms, choice became increasingly widespread in telecommunications. In terms of voice, a code of practice for operators established procedures for consumer complaints, disconnection and non-payment of bills. A so-called Service Level Agreement guaranteed service levels. Although the Consumers’ Association of Ireland was consulted and institutions, such as the Small Claims Court, the Office of the Director of Consumer Affairs and the Regulator of Premium Rate Telephone Services, were involved in consumer-related regulation, no specific telecom-related consumer representative existed. However, the establishment of an independent consumer ombudsman and a reform of the telecommunication regulatory authority were being considered at the time of writing.

The scandinavian family of nations: Sweden

In contrast to the UK, where New Public Management reforms were said to aim to create the ‘minimalist state’, in Sweden, reforms ‘aimed to protect the state’ (Rhodes, 1998, p. 19). In 1982, the Social Democratic government initiated a policy of public sector reform to tackle the rising costs of the welfare state. Initiatives were, however, frustrated by conﬂicts between so-called ‘decentralizers’, traditionalists’ and ‘economizers’ (Premfors, 1998, p. 150). Although initially aiming to enhance ‘democratic participation’ and ‘decentralization’, increasingly the emphasis shifted to cost reduction. By 1989, the main initiatives were driven by the ‘economizers’ in the Ministry of Financial Affairs. Privatization and private sector competition were recognized as ‘legitimate’ policy options in the decentralized delivery of social services. The increased autonomy of administrative agencies and local governments was paralleled by new forms of ﬁnancial accountability to central government (Clayton and Pontusson, 1998, p. 92). Initial reforms in the utility domains focused on corporatization, including the Swedish National Defence Factories, the Swedish State Power Board, Swedish Telecom and Sweden Post. Corporatization was accompanied by a large extension of formal liberalization. In railways, the 1988 Transport Policy Act vertically separated the infrastructure as public administration from railway operator, which was established as a state-owned enterprise. Regional and local services were opened to competitive tender. Further measures to increase the extent of private and competitive services were taken during the 1990s. The postal and telecommunications monopolies were abandoned in 1993 and a post and telecommunications regulator, the Post & Telestryelsen Regulatory Accountability 215

(PTS) was established. Its competencies were strengthened in 1998 when it obtained explicit competition authority in telecommunications. The electricity market was fully liberalized in 1996. Choice was introduced across utility domains throughout the 1990s, most extensively in telecommunications, but gradually also in energy, railways and postal services. Choice in electricity was facilitated in 1999 by the introduction of a standard profile which ended the requirement to install costly electricity meters for measuring hourly consumption. In telecommunications, so-called ‘pre-selection’ was introduced in 1999. No specific licence conditions existed to require voice options, although the licences established standard conditions of services. The PTS was obliged to report on the complaints received by itself and the operators. It also published voice-option ‘recommendations’ for the operators (PTS, 1999, p. 10). However, given the lack of enforcement powers, standardized procedures remained elusive and the operators supplied incompatible information (PTS, 2000). Surveys were used to investigate the impact of liberalization on different sections of the population. While no specific consumer-related institutions existed beyond the overall consumer advocacy groups such as the National Consumer Board and the Consumer Agency, the PTS increasingly took on a more consumer-related role, given rising consumer complaints and EU legislative requirements.

The continental family of nations: Germany

Efforts in (West) Germany to ‘roll back the state’ are said to have been relatively modest during the 1980s (see Benz and Goetz, 1996, pp. 5–14). However, the ‘privatization’ of the former public monopolies in railways, telecommunications and postal services, the large-scale sale of municipal water providers and the liberalization of energy markets suggested that Germany was ‘catching up’ in terms of the public sector reform ‘bandwagon’. In telecommunications, the liberalization and privatization of ‘Deutsche Telekom’ evolved over three stages, starting in 1989 with the corporatization and separation of telecommunications, postal and banking activities. In 1995, these activities were formed into individual joint stock companies due to the costs of unification. Full liberalization of telecommunications markets and a regulatory agency for post and telecommunications were established in 1998. The railway operator, ‘Deutsche Bahn’, was established in 1994 as a publicly- owned joint stock company with internally divided business sectors for freight, regional and long-distance passenger services and infrastructure. As in Sweden, regional services were transferred to subnational governments, the Länder, which used competitive tendering procedures. No economic regulator was established; the ‘Eisenbahn Bundesamt’ supervised technical and safety aspects. As also in electricity, the Federal Cartel Office was responsible for 216 Martin Lodge disputes concerning network access. In the case of electricity, infrastructure charges and interconnection were regulated by a corporatist arrangement of business associations, ending the system of private regional distribution and (often municipal) supply monopolies. Germany witnessed a rapid development of choice in electricity and telecommunications with convenient switching mechanisms being introduced at the earliest stages of liberalization. In terms of voice licences did not prescribe complaint handling procedures, although the telecommunications regulator was active in consumer advice and reported, superficially, on the number of submitted complaints in the annual report. Comparative information on price, unlike information on service standards, was widely available from private sources. In terms of representation, consumer councils were not established, but consultation procedures were to ensure wide-ranging societal participation.

Concluding reﬂections

The shift from the positive to the regulatory state is said to have challenged underlying assumptions of representative democracy, in particular by removing economic policy domains from formal political involvement and by shifting the emphasis from the ‘citizen’ to the ‘consumer’. Traditional accounts of regulatory accountability have continued to be attached to the notion that public administration (the ‘producer’) represents the key to improved service delivery. Furthermore, these arguments generally assume public interested officials and effective citizen participation. This chapter has attempted to move beyond normative analysis by suggesting four transparency mechanisms to facilitate ‘consumer sovereignty’. This final section compares the different regulatory regimes, then assesses whether there has been a shift towards a unified consumer model and concludes that the variety and complexity of the ‘sovereign consumer’ has led to differentiated choices across national regulatory regimes. Despite the increasing significance of European legislation which directed national regulation in areas of licensing, interconnection, accounting and universal standards (extending in the case of voice telephony (98/10/EC) to quality standards, repair times, disconnection and billing arrangements), the analysis of the European states suggested diversity between and within families of nations. While across all states a notable shift towards choice occurred, particularly in telecommunications, the extent of choice in other domains varied considerably. More importantly, the significance of other transparency mechanisms varied – ranging from a strong emphasis on financial disclosure in New Zealand and wide-ranging price comparisons in Germany to the focus on representation and service standards in the UK. Despite the emphasis on standards, comparative information in the UK only emerged in late 1999. In Regulatory Accountability 217 terms of the two ‘citizen-oriented’ mechanisms, representation and voice, the British system allowed for extensive ex post options in terms of complaint handling. In Germany and New Zealand, these options were not stressed, while in Sweden, they became increasingly prominent. The British approach of institutionalizing specific consumer councils was also unique, despite doubts concerning their effectiveness (Prosser, 1986). While in Ireland and New Zealand, the creation of consumer ‘ombudsmen’ was under consideration at the time of writing, in Germany, consumer organization had to rely on lobbying and judicial action. National differences were enhanced by wider institutional variations between the national regulatory regimes. For example, the spread of quasi- independent sectoral regulatory agencies was limited to telecommunications. In energy, for example, Germany’s approach resembled New Zealand’s reliance on general competition law. Even in the case of telecommunications, a comparative study revealed a lack of systematic monitoring and information gathering across EU member states (European Commission, 1999, p. 12). These differences were underlined by varying conceptions of the notion of regulation. Furthermore, the post-reform regimes reflected a large degree of persistence; for example, the UK Consumer Councils represented a legacy of the ‘positive state’ era. Regulatory change throughout the 1990s also reflected institutional adaptation rather than substantial shifts, with, for example, the UK gradually increasing its regime of standards and voice, but also choice, and New Zealand, relying increasingly on more intensive information disclosure requirements while continuing to stress the importance of industry-organized self-regulation. This brief study did not reveal a converging shift towards a ‘consumer’ model. While the increased reliance on choice suggested that states moved towards emphasizing the ‘sovereign consumer’ rather than the ‘represented citizen-voter’, in all states the use of other transparency mechanisms suggested a growing awareness of the differentiated and complex nature of individual ‘consumer sovereignty’. As a consequence, an increasing use of effecting instruments beyond choice, such as information, voice and representation, was accompanied by a growing emphasis on utilizing detectors to investigate the impact of regulatory reform. For example, studies of the Maori and Pacific Island communities showed that direct communication through grass-root organizations rather than top-down information dissemination were effective tools for detecting concerns and effecting information transfer and, therefore, choice (Ministry of Consumer Affairs, 1997b). In the UK and Sweden, regulators conducted studies into the impact of liberalization among socio- economic groups (for example, the ‘fuel poor’ – Ofgem, 1999). The analysis of transparency mechanisms reveals that the proclaimed shift from the ‘positive’ to the ‘regulatory state’ did not constitute a simple shift 218 Martin Lodge between a ‘citizen’ and a ‘consumer’ model of regulatory transparency. Besides the shift from input-to output-oriented legitimacy and the broad increase of choice across domains and states, the increased emphasis on consumer sovereignty has been accompanied by the parallel development of transparency mechanisms to effect and detect the complex and differentiated distribution of ‘consumer sovereignty’ across populations. Furthermore, given national dis- tinctiveness in regulatory regimes, it is difficult to sustain the claim that the regulatory state has heralded a shift from ‘a’ citizen towards ‘a’ consumer model; despite the international liberalization and privatization rhetoric, the underlying regulatory instruments and concepts as well as the varying instruments used to enhance (and constrain) ‘consumer sovereignty’ continued to remain nationally distinct.

Bibliography

Ayres, Ian and Braithwaite, John (1992) Responsive Regulation. New York: Oxford University Press. Baldwin, Robert and Cave, Martin (1999) Understanding Regulation. Oxford, Oxford University Press. Benz, Arthur and Goetz, Klaus H. (1996) ‘The German Public Sector: National Priorities and the International Reform Agenda’, in Arthur. Benz and Klaus H. Goetz (eds), A New German Public Sector? Reform, Adaptation and Stability. Aldershot: Dartmouth. Bergara, Mario, E. and Spiller, Pablo T. (1997) ‘The Introduction of Direct Access in New Zealand’s Electricity Market’, Utilities Review, 6(2): 97–105. Boston, Jonathan (1998) ‘Public Sector Management, Electoral Reform and the Future of the Contract State in New Zealand’, Australian Journal of Public Administration, 57(4): 32–43. Castles, Francis G. (1998) Comparative Public Policy, Cheltenham: Edward Elgar. Clayton, Richard and Pontusson, Jonas (1998) ‘Welfare-State Retrenchment Revisited: Entitlement Cuts, Public Sector Restructuring, and Inegalitarian Trends in Advanced Capitalist Societies’, World Politics, 51(1): 67–98. Dehousse, Reneaud (1997) ‘Regulation by Networks in the European Community’, Journal of European Public Policy, 4(2): 246–61. European Commission (1999) ‘Fifth Report on the Implementation of the Telecommunications Regulatory Package’, COM, 537, 10 November. Falconer, Peter K. and Ross, Kathleen (1999) ‘Citizen’s Charters and Public Service Provision: Lessons from the UK Experience’, International Review of Administrative Sciences, 65(3): 339–51. Graham, Cosmo (1997) Is There a Crisis in Regulatory Accountability? London: CIPFA. Hall, Claire, Scott, Colin and Hood, Christopher (2000) Telecommunications Regulation. London: Routledge. Haque, Shamsul (1996) ‘Public Service under Challenge in the Age of Privatization’, Governance, 9(2): 186–216. Haque, Shamsul (1998) ‘The Paradox of Bureaucratic Accountability in Developing Nations under a Pro-Market State’, International Political Science Review, 19(4): 357–73. Hirschman, Albert O. (1970) Exit, Voice and Loyalty, Cambridge, MA: Harvard University Press. Regulatory Accountability 219

Hood, Christopher (1986) Administrative Analysis. Hemel Hempstead: Harvester Wheatsheaf. Hood, Christopher (1994) Explaining Economic Policy Reversals. Buckingham: Open University Press. Kelsey, Jane (1995) The New Zealand Experience: A World Model for Structural Adjustment?, Auckland: Auckland University Press. Lodge, Martin (1999) ‘Competing Approaches to Regulation’, in Kjell Eliassen and Marit Sjovaag (eds), European Telecommunications Liberalization. London: Routledge. Loughlin, Martin and Scott, Colin (1997) ‘The Regulatory State’, in Patrick Dunleavy, Andrew Gamble, Ian Holliday and Gillian Peele (eds), Developments in British Politics 5. London: Macmillan – now Palgrave. McCubbins, Matthew D., Noll, Roger G. and Weingast, Barry R. (1987) ‘Administrative Procedures as Instruments of Political Control’, Journal of Law, Economics and Organization, 3(2): 243–77. Macey, Jonathan R. (1992) ‘Organizational Design and Political Control of Administrative Agencies’, Journal of Law, Economics and Organization, 8(1): 91–110. Majone, Giandomenico (1997) ‘From the Positive to the Regulatory State – Causes and Consequences from Changes in the Modes of Governance’, Journal of Public Policy, 17(2): 139–67. Majone, Giandomenico (1999) ‘The Regulatory State and its Legitimacy Problems’, West European Politics, 22(1): 1–24. Ministry of Consumer Affairs (1997a) An Analysis of Contracts, Metering and Disputes Procedures for Domestic Electricity Customers, January. Ministry of Consumer Affairs (1997b) Final Report on the Provision of Consumer Information, Education and Advice, 22 July. Ministry of Consumer Affairs (1998) A Comparison of Domestic Electricity Contracts against the Electricity Supply Association of New Zealand Domestic Contracts Code of Practice, October. Ministry of Consumer Affairs (2000) Submission to the Ministerial Enquiry into the Electricity Industry, March. OECD (1995) OECD Reference Checklist for Regulatory Decision-making, 9 March. Ofgem (1999) Review of the Development of Competition in the Domestic Gas Market, November. Pollitt, Christopher and Bouckaert, Geert (2000) Public Management Reform – A Comparative Analysis. Oxford: Oxford University Press. Premfors, Rune (1998) ‘Reshaping the Democratic State: Swedish Experiences in a Comparative Perspective’, Public Administration, 76(1): 141–59. Prosser, Tony (1986) Nationalised Industries and Public Control. Oxford: Basil Blackwell. Prosser, Tony (1997) Law and the Regulators. Oxford: Clarendon Press. PTS (1999) Telecom Supervision 1998. Stockholm: PTS. PTS (2000) Telecommunications Supervision, Annual Report. Stockholm: PTS. Rhodes, Rod (1998) ‘Different Roads to Unfamiliar Places: UK Experience in Comparative Perspective’, Australian Journal of Public Administration, 57(4): 19–31. Scott, Colin (2000) ‘Accountability in the Regulatory State’, Journal of Law and Society, 27(1): 38–60. Wilks, Stephen (1996) ‘Regulatory Compliance and Capitalist Diversity in Europe’, Journal of European Public Policy, 3(4): 536–59. 14 E-governance: Weber’s Revenge? Perri 6

Across the world, politicians make much of their strategies for modernizing government, using new technologies (Heeks, 1999). Under that rubric are three different clusters of activities:

• Electronic service delivery. Most of the effort, money and political attention available for electronic government is devoted to the provision of services on-line to citizens and businesses through the telephone, the personal computer or the digital television (6 et al., 2000). • Electronic democracy. New legislatures, such as those in Scotland and Wales, are using electronic voting systems in their chambers, and there is some interest in on-line consultations with citizens (6, 2000b). • E-governance. Less attention has been devoted to digital support for policy- making, decision-making, group work between ministers and their juniors, senior civil servants working on policy formulation, development and management, and with policy advisers who are contracted to provide conﬁdential policy support

This third cluster is the present focus. It has been too little studied (Kraemer and Dedrick, 1997), and yet may well be of fundamental importance to the nature of democratic life. One way to classify e-governance systems is roughly according to the main purpose for which the tools are used; of course, there are overlaps because some tools are put to more than one use. There are tools for the following:

1. Generating understandings – simple data systems providing dictionaries of key terms in the dialects of different policy-makers to enable those from different professional or organizational cultural backgrounds who are collaborat- ing to understand one another’s vocabulary; idea generation tools; graphical problem-structuring tools (modelling in software procedures such as soft system methodology, robustness analysis, strategic options development

220

K. Dowding et al. (eds.), Challenges to Democracy © Political Studies Association 2001 E-governance: Weber’s Revenge? 221

and analysis: Rosenhead, 1989); mental mapping and mental representation tools that enable users to develop graphical representations of their own or others’ basic conceptual approach to problems; argumentation support tools to help multiple groups of decision-makers working simultaneously to generate options, identify pros and cons and track multiple flows of argument and debate (Buckingham Shum, 1998; Dennis et al., 1997); electronic whiteboards for graphical representation of connections between ideas (Massetti, 1998); and scenario building tools. 2. Collecting data or observations – search agents; digital agents based on neural nets for context sensitive searching, editing, précis or translation (6, 1999a); sensors; and communication recording and storage systems. 3. Organizing and analysing data on events, conditions, problems, processes that have been observed – spreadsheets and budget systems – one of the earliest areas of large-scale e-governance (e.g. the French 1980s ‘SIAD Mairie’ system: Klein, Roux and Villedieu, 1991); organizational memory tools (Buckingham Shum, 1998); document profiling systems in shared work spaces that enable users, on saving documents, to provide hypermedia linkages to related documents, and to identify key relationships with key organization documents; some electronic document management systems use bar codes on paper documents to enable their linkage with electronic versions (Prinz and Syri, 1997); hypermedia geographic information systems (GIS); training simulation systems for decision-makers, for example for crisis management; and formal models. 4. Supporting communication and transaction – e-mail, electronic conferencing; video-conferencing systems (Mosier and Tammaro, 1994); meeting management tools (Niederman et al., 1996); tools to model and manage conflict, come to consensus or operate a decision procedure such as weighted voting or arbitration (Watson et al., 1994); argumentation support systems; and electronic document interchange. 5. Modelling decisions and advising on possible consequences – spreadsheets; expert systems, for example to test consistency and precision in draft legislation in social security (Portman, 1988, pp. 77–9) or immigration (Frøkjaer, 1989); neural nets (Berry et al., 1998); modelling systems for problems such as criminal activity, integrated with resource allocation tools (Borins, 1998, pp. 132–3). And there are 6. Environments that provide integration and storage for the other categories – intranets and the world wide web are the most important examples; however, in many cases only the simplest functionality of intranets is actually used (NAO, 1999)

What are arguably the most promising e-governance technologies are still not yet in widespread use, although there have been pilots and experiments, 222 Perri 6 and Britain may be somewhat behind other developed countries in the use of many of the most advanced systems. The reasons why Britain lags behind include: design weaknesses due to assumptions by designers of excessive ratio- nalism in policy-making (Kling, 1997); institutional forces (Bellamy and Taylor, 1998), policy network structures (Pratchett, 1999), perceived costs of setting up and limited gains from use (Bugler and Bretschneider, 1993; Kraemer and Dedrick, 1997); and difﬁculties in persuading publics to approve of additional expenditure on decision support when service provision spending is constrained (Pratchett, 1999; Ranerup, 1999). However, the global trend toward more ‘joined-up’ government (6, 1997, 6 et al., 1999), more syndicated cross-national coordination in governance (6, 2000a), generational change among policy-makers and design improvements may well increase the demand for and use of e-governance tools.

Theories of e-governance

What might be the consequences of signiﬁcantly increased take-up of more sophisticated and recently developed e-governance systems across government? Will these tools sustain ‘better’ governance, by the standards and interests of any of the principal interests that shape policy processes? Is it meaningful even to ask whether they could ‘better’ support anything recognizable as ‘the public interest’? Broadly, there are four rival types of theory on offer. Each can be read as one possible future scenario for understanding the potential impact of e-governance.

Rationalization First, there are those who argue uncompromisingly that the use of these technologies represents a major once-for-all improvement in the capabilities of governance and in at least the possibility of rationality in decision-making (Stevens and McGowan, 1985, pp. 177–83; Tapscott, 1997). The only price is the cost of investment and some running costs. Indeed, on this view, these systems will steadily reduce the costs of acquiring, ordering, coding, organizing, selecting, managing and using information. Therefore, the systems will more than repay their initial costs over their life (argued by, for example, Reschenthaler and Thompson, 1996). This optimistic view is based on classical cybernetic theory (Wiener, 1948): it holds that (using Overman and Loraine’s 1994 summary) information decreases uncertainty, slows entropy and increases control (decreases variance) by feedback and deviation correction, and in general that more information enables more control (or at least up to a point that we have not yet reached, if the relationship between information and control is conceived as curvilinear). It is relatively straightforward to obtain evidence for the control theory if one limits oneself to asking no-longer-new, now familiarized decision- E-governance: Weber’s Revenge? 223 making users about their satisfaction with a particular technology, as studies of e-governance in Norwegian (Ytterstad and Watson, 1996) and Australian local government have done (Hasan and Hasan, 1997). However, people tend to engage in rationalization and cognitive dissonance reduction, and so they tend to report that the technology provides them with valuable information and with greater control. Van de Donk (1998) presents some qualitative evidence for reduced incrementalism in budget-setting arising from the introduction of spreadsheets and more advanced ﬁnancial planning tools, but the thesis still lacks the comparative quantitative support it would need. Such subjective evidence needs to be treated with great caution, for all the reasons now well-established in the studies on limited rationality in the policy process.

The price of reason A second group of theories accepts at least the possibility of greater control, quality and rationality in decision-making, but insists that this comes at a price. These theories argue that safeguards are needed, lest that price be too great in terms of:

• citizens’ individual liberty and privacy; • citizens’ collective (democratic) influence over governmental decisions (Raab, 1997); • politicians’ control over decision-making agenda in favour of civil servants or officers’ ‘infocracy’ (Zuurmond, 1998) or an oligopoly of private contrac- tors (Margetts and Dunleavy, 1995); and/or • civil servants’ capability to exercise constraint upon the populism of politicians; or alternatively • bureaucratization, as decisions that can be made using algorithms are pushed down the hierarchy from politicians to salaried officials, because they cease to be regarded as policy matters at all (Bugler and Bretschneider, 1997); • more personal benefits for policy-makers (Berry, Berry and Foster, 1998); • erosion of meaningful relationships among decision-makers (Wilson, 1999); • even in the short term, loss of quality in decision-making because the focus on quantitatively measurable dimensions leads to neglect of less easily measured aspects of public programmes (Power, 1997); • loss of commitment to decisions made on the advice of computer-based neural net models and expert systems, due to the reduced user control and understanding of the model, and the larger numbers of options generated (Landsbergen et al., 1997); or • in the long term, humane values – Weber’s ‘iron cage of rationality’ (Weber, 1958, 1976; Winner, 1977; van de Donk, 1998). 224 Perri 6

The most extreme version of this view is perhaps that of the ‘governmentality’ theorists who argue that government technologies and rationalities, including e-governance tools, have become steadily more invasive through their extension of knowledge and information about populations, and their exercise has supported socialization steadily to shape citizens into more docile subjects of authority (Burchell, Gordon and Miller, 1991; Barry, Osborne and Rose, 1996; Dean, 1999; Rose, 1999). None of these claims is wholly compelling, and often depends on highly selective use of anecdote. The bureaucratization thesis, for example, is vulnerable to the powerful counter-example of monetary policy: the automatic correction mechanisms popular in the 1980s have been largely abandoned and the importance of judgement recognized. Again, evidence for loss of meaning and of commitment is rather anecdotal or confined to experiments, and does not necessarily support generalization. Sometimes, norms, customs, rules, roles, practices that have worked well in face-to-face or telephone settings will be carried over successfully to the group- ware context. This was the finding of a study of a network of British general practitioners in the National Health Service who began to exchange information about a range of practice management, prescribing, purchasing and health policy issues in which they were involved (Fox and Roberts, 1999). This finding seems to be explained by the particular institutional setting within which the on-line network was embedded. But it cannot be assumed in advance that this transfer will occur automatically. One study attempted to test for such changes by predicting what roles a group of people in an experiment would play in such a setting, based on how they behaved in face-to-face groupwork, and what roles they would want to delegate to the software, and then compared these predictions with roles observed. The main finding was that people played fewer roles than expected, and that at least some – but, to the surprise of the experiment designers, relatively few – participants described the software as playing the roles of recording the group memory and monitoring procedure, and a few saw the software as group facilitator and motivator (Zigurs and Kozar, 1994). It may be that, unlike the medical case, the institutional setting within which the experiment was embedded was simply too thin to provide existing norms that could be carried over. It is very difficult to avoid Hawthorne effects in designing experiments of this kind, and even in designing studies in working organizations, and so it is hard to know what weight to put on such results. So far, the move to reduce arbitrary administrative discretion by street-level bureaucrats in public service decision-making about individual entitlements such as cash benefits (stressed by Snellen, 1998) has not been the thin end of the wedge for the substitution of automated decision-making in policy judgement, or the final coming of Weber’s (1958, 1976) ‘iron cage’ of bureaucratic E-governance: Weber’s Revenge? 225 rationality. Despite the burgeoning investment by governments in the technologies listed above, there is little evidence of policy-makers in finance min- istries or land use planning departments following wholesale and mechanically the recommendations that are cranked out casually for them by junior staff from often still crude software models.

Noise, fragmentation and the erosion of reason The third view is the most pessimistic. The claim is that e-governance will actually erode rationality generally. It points to the pathologies of excessive demand for policy analysis thus delaying action – ‘paralysis by analysis’ – the bloating out of the policy advice industries among think-tanks and consul- tancy firms, the problems of sheer information overload, the allegedly lesser ability of the public sector to manage information well than of private citizens or businesses, and the obsession with the already measured that distracts policy-makers’ attention away from tacit, implicit, qualitative, unstructured factors and toward formal, explicit, quantitatively measured, structured factors and information. Finally, and perhaps most crucially, this theory fears that, due to mechanical rule following as suggested by overly simple data interpretations, overly simple modelling and overly simple expert system flows from analysis to recommendation, the cultivation and the exercise of judgement in decision-making will be crowded out. This view wholly rejects the cybernetic faith that information is control, and prefers the trope of information as noise. Much of the evidence for the noise theory comes not so much from studies of the use of computers by policy-makers as from studies of accountability in public management. Successive studies of the actual use made of performance indicators – which are, of course, now typically collected, coded, submitted, analysed, collated and now published electronically – have found that much of the information is not read, not used at all or provides more complex pictures than those who hold public services to account really want or think that they need (Carter, Klein and Day, 1993; Heinrich, 1999). Again, if one asks a different set of questions about coping with workload from those asked by many studies that find satisfaction with growing control, then by the power of suggestion in the phrasing of a question (for example, using leading phrases that hint at ‘information overload’), one typically gets data that can be used to support the noise theory of information.

Technology as totem, fetish arena and foil in ritualized social conflict The theories in the fourth and final cluster argue that there will be no very fundamental and independent impact of technology itself on the technical or political rationality of decision-making. Rather, both continuities and changes in governance are driven socially and politically, and technologies are means, 226 Perri 6 occasions and totems for conflict over preservation (‘conservative’ social and political shaping) or change (‘radical’ social and political shaping) of styles of governance (Mackenzie and Wajcman, 1985; Bijker and Law, 1992). Indeed, e-governance technologies are not ‘neutral’. Rather, design responds to social pressures in order to make social organization embodied and implicit in the technologies (Akrich, 1992). For example, it could be argued that the Dutch research on e-governance which shows the promotion of horizontal coordination styles of policy- making and public management (Frissen, 1999), and modest experimentation in public consultation, shows more conservative than radical social and political shaping at work. While the rigidities of the traditional pillars may be almost dismantled by now, the Dutch tradition is perpetuated, of relying upon relatively autonomous networks of organizations in both implementation and rule-making which are not dominated by politicians. By contrast, the British case shows conservative social shaping of the opposite kind: technologies of e-governance are selected and used to perpetuate the British commitment to the dominance of the political executive, and secondly to continue the role of elected politicians as storytellers and leaders in public. Therefore, we find British politicians committed to seeing the development of digital television and to chat rooms – mass media for which politicians already possess the requisite communication skills – and to more leadership-centred technologies of e-governance for decision-making. However, the prospect that most citizens might access digital services using mobile phones, which provide far fewer opportunities for the kinds of political communication that politicians need to engage in, rather than digital television is one that many British politicians privately find disconcerting. All this reflects the prior political culture, at least at the national level, that characterizes the Westminster system. Crucial to the shaping of technology in the exercise of governance on this view are the rituals that policy-makers perform that legitimate their decisions (Kertzer, 1988; Edelman, 1988; 6, 2000c). New intellectual technologies can provide an array of ritual resources for demonstrating to wider publics the grounding and authority of their decision-making practices, but in fact the underlying style or rationale of decision-making remains fundamentally politically driven. Information, on this view, is neither control nor noise. Rather, it is a totem – not in any derogatory sense, but rather in the vitally important political sense that all societies depend for their cohesion on symbols and rituals of their collective decision-making and the judgement of their policy- makers, and the real importance of information is to be found here. Historical and ethnographic work on long-run usage by policy-makers of information technologies tends to provide more support for the totem theory (e.g. Nidumolu et al., 1996), as does some survey work with a less conven- E-governance: Weber’s Revenge? 227 tional design (e.g. Overman and Loraine, 1994). Ethnographic research on the use of e-mail provides further good evidence that all projects for social shaping provoke counter projects. Romm (1999) shows that even a supposedly ‘simple’ and ‘low bandwidth’ technology like e-mail does not greatly constrain its users; Lee (1994) shows that ‘bandwidth’ in use is as much a function of interpretation, implicit and tacit knowledge as of technical characteristics of the medium. For it is used by rival groups, subaltern and superior, and hori- zontally opposed factions, to their own ends: Romm shows that each exploits certain features of the technology – multiple addressing, recordability, processing and routing – to pursue their own goals. In this sense, e-mail merely reminds us of the longstanding finding in organization studies that the flows of power, information and knowledge in organizations almost never follow the official organization chart (Scott, 1992). E-mail does not deepen or exacerbate these informal networks of power and information, but merely provides them with new means to pursue their particular goals. This suggests that the hypothesis that, because of e-mail, certain groups of policy-makers will gain power at the expense of others because they are gaining control (rationalization, perhaps at a price – perhaps local chairs at the expense of leaders, or ministers at the expense of premiers, or civil servants at the expense of politicians generally) should be rejected. The technology is almost certainly not responsible. Rather, people use technologies as occasions, foils, symbols, tools, ritual objects, to develop and institutionalize the forms of solidarity they are already committed to. If there are changes in norms, customs and practices associated with moving to computer-supported decision-making among policy-makers, that may at first sight provide evidence for the rationalization at a price theory, we should be alive to the possibility that what is really taking place is radical social shaping. That is to say, the introduction of new technologies into groups of decision-makers is also being used as the occasion and the opportunity for making other changes that superior decision-makers want to make. Indeed, the conventional management literature argues that this is exactly what a good manager should do with technology in order to avoid wasting investment on automating bad old ways of doing things (e.g. Hammer and Champy, 1995). There is an extensive body of work that shows that where superior decision-makers use the opportunity to introduce more control, surveillance and accountability upon the work of lower decision-makers, the technology will be experienced as introducing rationalization at the price of discretion, autonomy and freedom of action (Overman and Loraine, 1994; Shapiro et al., 1991). Academic researchers, being sensitive to threats to their own intellectual freedom, tend to be more sympathetic to claims to autonomy in decision-making by other groups than, for example, politicians or civil servants, and this can lead them to represent social shaping – whether conserva- 228 Perri 6 tive or radical – as rationalization at a price. However, many of the findings support the view that projects for radical social shaping of decision systems typically provoke counter-shaping projects of a more conservative or differently radical nature (and vice versa), because users are ingenious in using such systems for their own private benefits, subverting the intentions of superior decision-makers who may be seen likewise as shaping the technology as much for their private ends as for organizational goals (Shapiro et al., 1991). The noise theory’s reading of the findings about the non-use of indicators is not uncontested. Many recent commentators looking at audit and accountability mechanisms in the public sector see these as evidence for the totem theory – namely, that rituals of accountability are being performed and symbols displayed without necessarily achieving the substance of control (Hogwood, Judge and McVicar, 1998; Power, 1997). One important substantive effect, these theories would suggest, of the development of e-governance tools principally to handle quantitative information has been a reallocation of that scarcest of resources in government, namely policy-makers’ attention to problems (March and Olsen, 1976), to those problems that lend themselves to highly formal treatment. Had there been, for example, greater development and use in government of charting, creativity and soft systems methodology tools (Checkland and Scholes, 1990; Ballantine and Cunningham, 1999), which have developed some of the ideas of the late great Sir Geoffrey Vickers (Checkland, 1994; Vickers, 1995), attention to problems might be allocated differently, but that selection of technologies would only have taken place had the institutional pressures upon policy-makers been rather different. Precisely because the noise and totem theories differ in their basic conception of information from the qualified cybernetic consensus of the first two, it is possible for them to come to opposite conclusions about the impacts of new intellectual technologies on leadership and collective action from their conclusions about the impacts on decision-making. For the very characteristics of being either noise and totem that can undermine the quality of decision- making may in fact increase the capabilities of politicians to engage in leadership and collective action, for the cover of fog and the power of symbolism are powerful tools in the pursuit of these activities of governance.

Comparing the theories

Table 14.1 summarizes some of the key differences between the theories. Is a straightforward choice to be made between these theories? Or are they recon- cilable – perhaps by allocating them to different domains or else to different levels of explanation? As we have seen, each theory has some empirical support, although most empirical studies have been of rather limited scope and not in general designed to test, let alone falsify, these rival theories. E-governance: Weber’s Revenge? 229

Table 14.1 Theories of e-governance at a glance

Theory Rationalization Rationalization Loss of Social shaping speciﬁc view at a price rationality Information Control Control (negative) Noise Totem as (positive) Public Integration Integration Fragmentation Depends on management balance of forces Planning Anticipation Anticipation Resilience Depends on style balance of forces Power shift To policy Captured by Undermining Depends on makers special interests power balance of forces

There are differences between the theories in the explanatory level at which they work. Social science distinguishes between proximate variables – close to the phenomenon they purport to explain – and distal variables – at some explanatory distance, working at deeper levels, not necessarily directly observable. For example, the control theory works with proximate variables – individual decision-makers’ satisfaction, their sense of efﬁcacy and control over their work, the collective action capabilities of groups of decision- makers. The noise theory is basically the converse of this – focusing on decision-makers’ sense of being overwhelmed. By contrast, social shaping or totem theory works with distal variables – institutions, symbols, meanings, ritual formations, group identities and the ways in which they shape group interests. Indeed, social shaping theory can offer an institutional explanation of the phenomena to which the control, rationalization at a price and noise theories point, as if they were data rather than explanations. Indeed, it suggests that the fears expressed in the ‘price of rationalization’ and ‘noise’ theories are essentially resentments of particular forms of social organization for which technologies are mobilized (Douglas, 1994). In this sense, e-governance technologies are no different from the development of governmental statistics in the eighteenth and nineteenth centuries as decision support tools (Hacking, 1990): there is nothing politically new in the e-governance debates. Although multi-theoretic approaches have their uses, especially for policy- makers, it should be clear from the foregoing discussion that I lean toward the totem or social shaping theory, partly on wider theoretical grounds such as a preference for distal variables (6, 1999b), and partly for reasons set out above about the interpretation of the empirical research data offered by the other theories. In the next section and the conclusion, I develop some of the main implications of the totem or social shaping theory for the uses of e-governance tools. 230 Perri 6

Technologies to support judgement and appreciation in e-governance

Judgement and appreciation (Vickers, 1995) are the craft skills of governance and decision-making (for an account of political judgement, see 6, 2000a). If tools of e-governance are used to economize upon or substitute for judgement, the consequences could be very damaging. One danger of rationalization theories is that they can give legitimacy to uses of e-governance tools as ways of streamlining and homogenizing decision-making when the political strength of many decision-making systems is precisely their diversity and responsiveness. In the name of removing arbitrariness, information technologies can too easily substitute explicit and formal information for the tacit and informal, and in consequence lose proper discrimination, destroy knowledge and undermine judgement. The importance of the social shaping theory is its recognition that, more signiﬁcantly than the design features of particular e- governance tools, ritual in judgement is often the way in which this tacit rich- ness is encoded and applied, and through which technologies are ‘domesticated’. The challenge for policy-makers is to develop norms, customs and rituals for the appropriate use of these technologies which will help to sustain and cultivate judgement. It is common in policy studies to be rather dismissive of ritual, on the grounds that ritual behaviours are supposedly suboptimal, less than rational if not outright irrational: in mainstream policy sciences, it is often a debunking strategy to suggest that something is ‘mere’ ritual or is ‘symbolic’ (Edelman, 1985 [1964]). But this is inadequate (6, 2000c). Ritual is neither essentially irrational (as suggested by the Carnegie school, e.g. March and Olsen, 1989, 1995) nor insincerely and cynically conducted public ceremonial, as the Puritan tradition has misled us into thinking (and as many studies of political ritual suggest, for example Edelman, 1988; Kertzer, 1988): rather, it is the repeated enactment of social organization and of settlement between rival and conﬂicting forms (Douglas, 1970, 1986). Anthropologists have long known that ritual is never empty, never formalis- tic, but endlessly inventive, and essential to the ways in which cultures, habits, expectations and solidarities are developed and judgement is practised, and to the social shaping and impact of technologies (Bell, 1992; Turner, 1974, 1982). Rituals, institutions, customs and norms in the use of technologies govern expectations of speed and time (Adam, 1990; 6, 2000c). Many policy-makers report that the impact of global media scrutiny and interviewing after events, of the culture of e-mail that has grown up on the Internet, and the customs of responsiveness associated with on-line voting press them to make decisions and respond at greater speed than permits the exercise of judgement. The effect can be to reinforce the sensibility that treats information as noise. E-governance: Weber’s Revenge? 231

Developing alternative customs requires more than simply assertion and effort: it requires work on institutionalized expectations and commitments among wider publics. One of the customs and norms that we ought to expect in democratic societies is that the electronic tools of e-governance will increasingly be put into the public domain and be tested by experts and by ordinary citizens against a variety of assumptions and against rival systems. Freedom of information should extend to public availability for all the models, simulations, problem structuring tools, geographical information systems and content analytic agents that policy-makers themselves use. (A recent report from the Performance and Innovation Unit (2000) in the British Cabinet Office has recommended this.) This may not be appropriate for all models, such as those generated as normative heuristics in workshop settings, but is generally appropriate for most formal positive quantitative and qualitative causal models (except, of course, for those used in military contexts). For the aim, in a democratic society, of the cultivation of judgement in governance should not be the empowering of a political elite, but the strengthening of the competence, maturity and self-governing capabilities of the citizenry (Elkin and Soltan, 1999). Indeed, every democrat must hold that in the medium to long run, only the robustness of the judgement capabilities of citizens can guarantee that those of policy-makers will be similarly stout. Similarly, the challenge to make e-governance tools serve the cultivation of policy judgement requires not only better-designed software systems, but the institutional buttressing of the practice of judgement as the political settlement through ritual processes between rival forms of social organization (6, 2000a, c). To this end, it would, at least in my view, be helpful to develop e-governance systems that explicitly model conflicts over policies as conflicts over forms of social organization, and that related rival claims explicitly to the particular styles of organization that they typically reflect. Secondly, it should be a key element in the professional ethics of solution designers and system developers that their systems should make palpable to policy-makers who use them just what the political presuppositions of the systems under development are. Thirdly, as politicians explain to their electorates why these systems justify the expense of investment, they need to make clear just how the central political virtues of judgement as settlement between rival forms of social organization are being cultivated through the use of such tools. Only in that context can we hope that e-governance tools will support democratic renewal.

Acknowledgements

I am grateful to Cisco Systems Ltd and to Nick Penston in particular for ﬁnancial support for the programme of research on which this paper is based. 232 Perri 6

I am grateful for their comments on earlier drafts of a longer paper from which the present one is taken, to Christine Bellamy, Paul Frissen, Brian Hogwood, Helen Margetts, Lawrence Pratchett, Fred Thompson and Steve Woolgar. None of them should be assumed to agree with my arguments, still less be held responsible for my errors.

References

Adam, B. (1990) Time and Social Theory. Cambridge: Polity Press. Akrich, M. (1992) ‘The De-scription of Technical Objects’, in Bijker, W. and Law, J. (eds), Shaping Technology/Building Society. Cambridge, MA: Massachusetts Institute of Technology Press, pp. 205–24. Ballantine, J. and Cunningham, N. (1999) ‘Strategic Information Systems Planning: Applying Private Sector Frameworks in UK Public Healthcare’ in Heeks, R. (ed.), Reinventing Government in the Information Age: International Practice in IT-enabled Public Sector Reform. London: Routledge, pp. 293–311. Barry, A., Osborne, T. and Rose, N. (eds) (1996) Foucault and Political Reason: Liberalism, Neo-liberalism and Rationalities of Government. London: UCL Press. Bell, C. (1992) Ritual Theory, Ritual Practice. Oxford: Oxford University Press. Bellamy, C. and Taylor, J. (1998) Governing in the Information Age. Buckingham: Open University Press. Berry, F., Berry, J. and Foster, S. (1998) ‘The Determinants of Success in Implementing an Expert System in State Government’, Public Administration Review, 58, 4: 293–305. Bijker, W. and Law, J. (eds) (1992) Shaping Technology/Building Society. Cambridge, MA: Massachusetts Institute of Technology Press. Borins, S. (1998) Innovating with Integrity: How Local Heroes are Transforming American Government. Washington, DC: Georgetown University Press. Buckingham Shum, S. (1998) ‘Negotiating the Construction of Organisational Memories’, in Borghoff, U. M. and Pareschi, R. (eds), Information Technology for Knowledge Management. Berlin: Springer-Verlag. Bugler, D. and Bretschneider, S. (1993) ‘Technology Push or Program Pull? Interest in New Technologies within Public Organisations’, in Bozeman, B. (ed.), Public Management: the State of the Art. San Francisco, CA: Jossey-Bass, pp. 275–93. Burchell, G., Gordon, C. and Miller, P. (eds) (1991) The Foucault Effect: Studies in Governmentality. Hemel Hempstead: Harvester Wheatsheaf. Carter, N., Klein, R. and Day, P. (1993) How Organisations Measure Success: the Use of Performance Indicators in Government. London: Routledge. Checkland, P. B. (1994) ‘Systems Theory and Management Thinking’, in Blunden, M. and Dando, M. (eds), Rethinking Public Policy-making: Questioning Assumptions, Challenging Beliefs – Essays in Honour of Sir Geoffrey Vickers on his Centenary. London: Sage, pp. 75–91. Checkland, P. B. and Scholes, J. (1990) Soft Systems Methodology in Action. Chichester: John Wiley & Sons. Dean, M. (1999) Governmentality: Power and Rule in Modern Society. London: Sage. Dennis, A., Valacich, J., Carte, T., Garﬁeld, M., Haley, B. and Aronson, J. (1997) ‘Research Report: the Effectiveness of Multiple Dialogues in Electronic Brainstorming’, Information Systems Research, 8, 2: 203–11. Douglas, M. (1970) Natural Symbols: Explorations in Cosmology. London: Routledge. E-governance: Weber’s Revenge? 233

Douglas, M. (1986) How Institutions Think. London: Routledge & Kegan Paul. Douglas, M. (1994) Risk and Blame: Essays in Cultural Theory. London: Routledge. Edelman, M. (1985) [1964] The Symbolic Uses of Politics. Urbara, IL: University of Illinois Press. Edelman, M. (1988) Constructing the Political Spectacle. Chicago: University of Chicago Press. Elkin, S. L. and Soltan, K. E. (eds) (1999) Citizen Competence and Democratic Institutions. University Park, PA: Pennsylvania University Press. Fox, N. and Roberts, C. (1999) ‘GPs in Cyberspace: the Sociology of a “Virtual Community” ’, Sociological Review, 4: 643–71. Frissen, P. (1999) Politics, Governance and Technology: a Postmodern Narrative on the Virtual State. Aldershot: Edward Elgar. Frøkjaer, E. (1989) ‘Controversial Issues of Expert Systems in Public Administration’, in Snellen, I., van de Donk, W. and Baquiast, J.-P. (eds) (1989) Expert Systems in Public Administration. Amsterdam: Elsevier, pp. 35–54. Hacking, I. (1990) The Taming of Chance. Cambridge: Cambridge University Press. Hammer, M. and Champy, J. (1995) Re-engineering the Corporation: a Manifesto for Business Revolution. London: Nicholas Brearley Publishing. Hasan, H. and Hasan, S. (1997) ‘Computer-based Performance Information for Executives in Local Government’, Australian Journal of Public Administration, 56, 3: 24–9. Heeks, R. (ed.) (1999) Reinventing Government in the Information Age: International Practice in IT-enabled Public Sector Reform. London: Routledge. Heinrich, C. (1999) ‘Do Government Bureaucrats Make Effective Use of Performance Management Information?’, Journal of Public Administration Research and Theory, 9, 3: 363–94. Hogwood, B., Judge, D. and McVicar, M. (1998) Agencies and Accountability, paper to the Whitehall Programme conference, University of Birmingham, 17–19 December. Kertzer, D. I. (1988) Ritual, Politics and Power. New Haven, CT: Yale University Press. Klein, M., Roux, D. and Villedieu, T. (1991) ‘Decision Support for Municipality Financial Planning in France: Recent Progress with “SIAD Mairie”, a Knowledge-based DSS’, in Traunmüller, R. (ed.), Governmental and Municipal Information Systems II: Proceedings of the 2nd IFIP TC8/WG8.5 Working Conference on Governmental and Municipal Information Systems, Balatonfüred, Hungary. Amsterdam: North-Holland, pp. 117–30. Kling, R. (1997) ‘Multivalent Social Relationships in Computer Supported Workplaces’, in Kiesler, S. (ed.), Research Milestones on the Information Highway. New Jersey: Lawrence Erlbaum. Kraemer, K. and Dedrick, J. (1997) ‘Computing and Public Organisations’, Journal of Public Administration Research and Theory, 7, 1: 89–112. Landsbergen, D., Coursey, D., Loveless, S. and Shangraw, R. (1997) ‘Decision Quality, Conﬁdence and Commitment with Expert Systems: an Experimental Study’, Journal of Public Administration Research and Theory, 7, 1: 131–57. Lee, A. (1994) ‘Electronic mail as a Medium for Rich Communication: an Empirical Investigation Using Hermeneutic Interpretation’, Management Information Systems Quarterly, 18, 2: 143–58. Mackenzie, D. and Wajcman, J. (eds) (1985) The Social Shaping of Technology: How the Refrigerator Got its Hum. Buckingham: Open University Press. March, J. and Olsen, J.-P. (1976) Ambiguity and Choice in Organisations. Bergen: Universitetsforlaget. March, J. and Olsen, J.-P. (1989) Rediscovering Institutions: the Organisational Basis of Politics. New York: Free Press. 234 Perri 6

March, J. and Olsen, J.-P. (1995) Democratic Governance. New York: Free Press. Margetts, H. and Dunleavy, P. (1995) ‘Public Services on the World Markets’, in Missionary Government, Demos Collection, 7: 30–2. Massetti, B. (1998) ‘An Empirical Examination of the Value of Creativity Support Systems on Idea Generation’, Management Information Systems Quarterly, 20, 1: 83–98. Mosier, J. and Tammaro, S. (1994) ‘Video Teleconference Use among Geographically Dispersed Workgroups: a Field Investigation of Usage Patterns and User Preferences’, Journal of Organisational Computing, 4, 4: 343–65. National Audit Office (NAO) (1999) Government on the Web, HC 87. London: Stationery Office. Nidumolu, S., Goodman, S., Vogel, D., Dankowitz, A. (1996) ‘Information Technology for Local Administration Support: the Governorates Project in Egypt’, Management Information Systems, 20, 2: 197–225. Niederman, F., Biese, C. and Beranek, P. (1996) ‘Issues and Concerns about Computer Supported Meetings’, Management Information Systems Quarterly, 20, 1: 1–22. Overman, E. and Loraine, D. (1994) ‘Information for Control: Another Management Proverb?’, Public Administration Review: 55, 2, 193–196. Performance and Innovation Unit (2000) Adding It Up: Improving Analysis and Modelling in Central Government. London: Cabinet Office. Portman, E. (1988) ‘The Alvey DHSS Large Demonstrator Project’, in Duffin P. (ed.), Knowledge-based Systems: Applications in Administrative Government. Chichester: Ellis Horwood, and London: CCTA, pp. 76–88. Power, M. (1997) The Audit Society: Rituals of Verification. Oxford: Oxford University Press. Pratchett, L. (1999) ‘New Technologies and the Modernisation of Local Government’, Public Administration, 77, 4: 731–50. Prinz, W. and Syri, A. (1997) ‘Two Complementary Tools for Co-operation in a Ministerial Environment’, Journal of Universal Computer Science, 3, 8: 843–64. Raab, C. (1997) ‘Privacy, Information and Democracy’, in Loader B. D. (ed.), The Governance of Cyberspace: Politics, Technology and Global Restructuring, London: Routledge, pp. 155–74. Ranerup, A. (1999) ‘Internet-enabled Applications for Local Government Democratisation: Contradictions of the Swedish experience’, in Heeks, R. (ed.), Reinventing Government in the Information Age: International Practice in IT-enabled Public Sector Reform. London: Routledge, pp. 177–93. Reschenthaler, G. and Thompson, F. (1996) ‘The Information Revolution and the New Public Management’, Journal of Public Administration Research and Theory, 6, 1: 125–43. Romm (Livermore), C. T. (1999) Virtual Politicking: Playing Politics in Electronically Linked Organisations. Cresskill, NJ: Hampton Press. Rose, N. (1999) Powers of Freedom: Reframing Political Thought. Cambridge: Cambridge University Press. Rosenhead, J. (ed.) (1989) Rational Analysis for a Problematic World: Problem Structuring Methods for Complexity, Uncertainty and Conflict. Chichester: John Wiley & Sons. Scott, W. (1992) Organisations: Rational, Natural and Open Systems, 3rd edn. Englewood Cliffs, NJ: Prentice-Hall. Shapiro, D., Hughes, J., Harper, R., Ackroyd, S. and Soothill, K. (1991) ‘Policing Information Systems: the Social Context of Success and Failure in Introducing Information Systems in the Police Service’, in Traunmüller, R. (ed.) Governmental and E-governance: Weber’s Revenge? 235

Municipal Information Systems II: Proceedings of the 2nd IFIP TC8/WG8.5 Working Conference on Governmental and Municipal Information Systems, Balatonfüred, Hungary. Amsterdam: North-Holland, pp. 183–97. 6, P. (1997) Holistic Government. London: Demos. 6, P. (1999a) Morals for Robots and Cyborgs: Ethics, Society and Public Policy in the Age of Autonomous Intelligent Machines. Brentford: Bull Information Systems. 6, P., (1999b) Neo-Durkheimian Institutional Theory, paper to the University of Strathclyde conference, ‘Institutional Theory in Political Science’, Ross Priory, 18–19 October. 6, P. (2000a) Governing by Technique: Judgment and the Prospects for Governance of and with Technology, paper to the OECD ‘Forum for the Future’ conference, ‘21st Century Social Governance: Power in the Global Knowledge Economy and Society’, 24–26 March, Hanover, Germany. 6, P. (2000b) London_Mayor@Your_Service: the Prospects for Digital Democracy in the Capital. London: A New Voice for London. 6, P. (2000c) From Affect to Citizenship: the Social Organisation of Emotion and Ritual Politics, paper to the University of East London seminar series ‘Affect, Ethics and Citizenship’, 1 June, Birkbeck College, London. 6, P., Leat, D., Seltzer, K. and Stoker, G. (1999) Governing in the Round: Strategies for Holistic Government. London: Demos. 6, P., Stoddart, P., Penston, N., Shire, C. and Laird, A. (2000) If You’re Serious About Modernising Government . . . Cambridge: Smart Card Club Snellen, I. (1998) ‘Street Level Bureaucracy in an Information Age’, in Snellen I. and van de Donk, W. (eds), Public Administration in an Information Age: A Handbook. Amsterdam: IOS Press: 497–508. Stevens, J. and McGowan, R. (1985) Information Systems for Public Management. New York: Praeger. Tapscott, D. (1997) ‘The Digital Media and the Reinvention of Government’, Canadian Public Administration, 40: 328–45. Turner, V. (1974) Dramas, Fields and Metaphors: Symbolic Action in Human Society. Ithaca, NY: Cornell University Press. Turner, V. (1982) From Ritual to Theatre: the Human Seriousness of Play. New York: PAJ Publications. van de Donk, W. (1998) ‘Beyond Incrementalism? Redistributive Policy-making, Information Systems and the Revival of Synopticism’, in Snellen, I. and van de Donk, W. (eds), Public Administration in an Information Age: A Handbook. Amsterdam: IOS Press, pp. 381–404. Vickers, G. (1995) [1963] The Art of Judgment: a Study in Policy Making, centenary edition. London: Sage. Watson, R., Alexander, M., Pollard, C. and Bostrom, R. (1994) ‘Perceptions of Facilitators of a Keypad-based Group Support System’, Journal of Organisational Computing, 4, 2: 103–5. Weber, M. (1958) ‘Bureaucracy’, from Economy and Society, in Gerth, H. and Mills, C. (eds) [1946], From Max Weber: Essays in Sociology. New York: Galaxy/Oxford University Press, pp. 196–245. Weber, M. (1976) The Protestant Ethic and the Spirit of Capitalism. London: Allen & Unwin. Wiener, N. (1948) Cybernetics: the Emerging Science at the Edge of Order and Chaos. New York: Simon & Schuster. Wilson, F. (1999) ‘Cultural Control Within the Virtual Organisation’, Sociological Review, 4: 672–94. 236 Perri 6

Winner, L. (1977) Autonomous Technology: Technics-out-of-control as a Theme in Political Thought. Cambridge, MA: Massachusetts Institute of Technology Press. Ytterstad, P. and Watson, R. (1996) ‘Teledemocracy: Using Information Technology to Enhance Political Work’, Management Information Systems Quarterly, 20, 3: 347. Zigurs, I. and Kozar, K. (1994) ‘An Exploratory Study of Roles in Computer Supported Groups’, Management Information Systems Quarterly, 18, 3: 277–98. Zuurmond, A. (1998) ‘From Bureaucracy to Infocracy: are Democratic Institutions Lagging Behind?’, in Snellen, I. and van de Donk, W. (eds), Public Administration in an Information Age: A Handbook, Amsterdam: IOS Press, pp. 259–72. 15 Decentralization and Development: The New Panacea Elinor Ostrom1

The current recommended panacea for the best governance of natural resources in developing countries is a form of radical devolution involving a massive ‘turning over’ of resources to local users. While I will argue in this chapter that the users of a local resource are frequently capable of developing well-crafted rule systems over time that manage local resources for which they have secure property rights, and of doing so better than many government- owned resources, I wish to raise a strong cautionary note about the current brand of a decentralization craze that has swept through the development projects funded by more than one donor. I will focus in this chapter on a particular aspect of the governance of human affairs related to the sustainable use of natural resources rather than on all forms of collective action.

Common-pool resources and the conventional theory

Most natural resource systems used by multiple individuals can be classified as common-pool resources. Common-pool resources generate finite quantities of resource units and one person’s use subtracts from the quantity of resource units available to others (Ostrom et al., 1994). Most common-pool resources are sufficiently large that multiple actors can simultaneously use the resource system and efforts to exclude potential beneficiaries are costly. Most theoretical studies of common-pool resources by political economists have analysed simple common-pool resource systems using relatively similar assumptions (Feeny, Hanna and McEvoy, 1996). In such systems, it is assumed that the resource generates a highly predictable, finite supply of one type of resource unit (one species, for example) in each relevant time period. Since appropriators are viewed as being trapped in these dilemmas, repeated recommendations were made that external authorities must impose a different set of institutions on such settings. Some recommend private property and others recommend government control as the most efficient form of ownership

237

(Simmons et al., 1996; Ophuls, 1973). Implicitly, theorists assume that regulators will act in the public interest and understand how ecological systems work and how to change institutions so as to induce socially optimal behaviour.

Self-organized resource governance systems in the ﬁeld

Most common-pool resources are more complex than the base theory of homogeneous appropriators taking one type of resource unit from a resource system that generates a predictable flow of units. The rich case study literature illustrates a wide diversity of settings in which appropriators dependent upon common-pool resources have organized themselves to achieve much higher outcomes than is predicted by the conventional theory (Wade, 1994; Sengupta, 1991).2 Small- to medium-sized irrigation systems come closer than many biological resources to approximating these conditions and are, thus, an appropriate setting in which to examine these patterns of relationships quantitatively. One resource unit – water – is the focus of efforts to organize and coordinate activities. Recent research on small- to medium-sized irrigation systems in Nepal has found a very substantial difference in performance between those systems owned and governed by the farmers themselves as contrasted to those systems owned and operated (but in some cases, not governed) by a national governmental agency. While most farmers own land in Nepal, most own very small parcels of less than 1 hectare. They are relatively homogeneous with similar preferences in regard to obtaining water for rice production during the monsoon and winter seasons and various crops during the spring. Farmer-managed irrigation systems (FMIS) have operated until recently as entities minimally recognized by the government. In a detailed analysis of data from 150 farmer-governed and national government irrigation systems in Nepal, Lam (1998) develops three performance measures: (1) the physical condition of irrigation systems; (2) the quantity of water available to farmers at different seasons of the year; and (3) the agricultural productivity of the systems. Controlling for environmental differences among systems, Lam finds several variables strongly related to these dependent variables. One is the form of governance of the system. Holding other variables constant, irrigation systems governed by the farmers themselves perform significantly better on all three performance measures. This variable has the largest explanatory power of any variable in Lam’s analysis, including the physical size of the system, terrain characteristics and the number of farmers. Thus, farmers with long-term ownership claims, who can communicate, develop their own agreements, establish the positions of monitors and sanc- tion those who do not conform to their own rules, are more likely to grow Decentralization and Development: The New Panacea 239 more rice, distribute water more equitably and keep their systems in better repair than is done in government systems. While there is variance in the performance of these Nepali systems, few perform as poorly as government systems holding other relevant variables constant. Similar evidence of farmer capabilities is derived from the study of 47 farmer-governed systems in the Philippines described by de los Reyes (1980). In a study of 48 irrigation systems in India (having diverse governance regimes), Bardhan (1999) finds higher performance in those where the farmers themselves have devised rules to regulate the use and maintenance of the systems as contrasted to those systems where the government exercises control. Since many of the government systems in Nepal, as well as in India, rely on high-tech engineering, the capability of farmers to increase agricultural production on their ‘primitive systems’, while they also provide the labour to maintain and operate the system, is particularly noteworthy.

On the origin of self-governed common-pool resources

Evidence from ﬁeld research thus challenges the generalizability of the conventional theory.3 While the theory is successful in predicting outcomes in settings where appropriators are alienated from one another or cannot communicate effectively (see Ostrom, 1999), it does not provide an explanation for settings where appropriators are able to create and sustain agreements to avoid serious problems of overappropriation. Nor does it predict well when government ownership will perform appropriately or how privatization will improve outcomes. A fully articulated, reformulated theory encompassing the conventional theory as a special case does not yet exist. On the other hand, scholars familiar with the results of ﬁeld research substantially agree on a set of variables that enhances the likelihood of appropriators organizing themselves to avoid the social losses associated with open-access, common-pool resources (McKean, 1992, 1996; Wade, 1994; Tang, 1992; Ostrom, 1990; Baland and Platteau, 1996; Ostrom et al., 1994). Considerable consensus exists that the following attributes of resources and of appropriators are conducive to an increased likelihood that self-governing associations will form.

Attributes of the resource:

R1. Feasible improvement. Resource conditions are not at a point of deteriora- tion such that it is useless to organize or so underutilized that little advantage results from organizing. R2. Indicators. Reliable and valid indicators of the condition of the resource system are frequently available at a relatively low cost. R3. Predictability. The ﬂow of resource units is relatively predictable. 240 Elinor Ostrom

R4. Spatial extent. The resource system is sufﬁciently small, given the trans- portation and communication technology in use, that appropriators can develop accurate knowledge of external boundaries and internal microenvironments.

Attributes of the appropriators:

A1. Salience. Appropriators are dependent on the resource system for a major portion of their livelihood or other important benefits. A2. Common understanding. Appropriators have a shared image of how the resource system operates (attributes R1, 2, 3, and 4 above) and how their actions affect each other and the resource system. A3. Low discount rate. Appropriators use a sufficiently low discount rate in relation to future benefits to be achieved from the resource. A4. Trust and reciprocity. Appropriators trust one another to keep promises and relate to one another with reciprocity. A5. Autonomy. Appropriators are able to determine access and harvesting rules without external authorities countermanding them. A6. Prior organizational experience and local leadership. Appropriators have learned at least minimal skills of organization and leadership through participation in other local associations or learning about ways that neighbouring groups have organized.

It is very important to stress that many of these variables are in turn affected by the type of larger regime in which users are embedded. Larger regimes can facilitate local self-organization by providing accurate information about natural resource systems, providing arenas in which participants can engage in discovery and conflict-resolution processes, and providing mechanisms to back up local monitoring and sanctioning efforts. The probability of participants adapting more effective rules in macro-regimes that facilitate their efforts over time is higher than in regimes that ignore resource problems entirely or, at the other extreme, presume that all decisions about governance and management need to be made by central authorities. In addition to the above variables for which there is considerable consensus, the variables of size and heterogeneity are also thought to be very important by many scholars, but the direction of their impact and the very definition of the variables themselves are subject to considerable controversy. There is not sufficient room here to discuss these strongly contested variables. The key to further theoretical integration is to understand how these attributes interact in complex ways to affect a basic benefit–cost calculation of a set of appropriators (A) using a resource (Ostrom, 1990, ch. 6). Each appropriator Decentralization and Development: The New Panacea 241 i (i ∈ A) has to compare the expected net benefits of harvesting continuing to use the old rules (BO) to the benefits he or she expects to achieve with a new set of rules (BN). Each appropriator i must ask whether his or her incentive to change (Di) is positive or negative.

Di = BNi – BOi

If Di is negative for all appropriators, no one has an incentive to change. If Di is positive for some appropriators, they then need to estimate three types of costs:

C1 – the up-front costs of time and effort spent devising and agreeing upon new rules; C2 – the short-term costs of adopting new appropriation strategies; and C3 – the long-term costs of monitoring and maintaining a self-governed system over time.

If the sum of these expected costs for each appropriator exceeds the incentive to change, no appropriator will invest the time and resources needed to create new institutions. Thus, if

Di < (C1i + C2i + C3i) for all i ∈ A, no change occurs. In field settings, everyone is not likely to expect the same costs and benefits from a proposed change. Some may perceive positive benefits after all costs have been taken into account, while others perceive net losses. And, as I discuss below, the number of rules that could be changed is so large that one should not presume that users evaluate some optimal combination of rules as against an existing set. Rather, one should presume that users do have a relatively good idea of the likely consequences of keeping a particular set of rules in place. They can use this base expectation and compare it to proposals made to change one or several rules to something different. Even with this assumption, we can expect that users will sometimes be dead wrong in their expectations. Thus, the choice of rules results from a rough estimate of the expected benefits and costs of continuing with existing rules or changing to proposed replacements. The collective-choice rules used to change the day-to-day operational rules related to appropriation affect whether a proposed institutional change favoured by some and opposed by others will occur. For any collective-choice rule, such as unanimity, majority, ruling elite or one-person rule, there is a 242 Elinor Ostrom minimum coalition of appropriators, K ⊂ A, that must agree prior to the adoption of new rules. If for any individual k, a member of K,

≤ Dk (C1k + C2k + C3k) no new rules will be adopted. And if for at least one coalition K ⊂ A, it is such that

Dk > (C1k + C2k + C3k) for all members of K, it is feasible for a new set of rules to be adopted. If there are several such coalitions, the question of which coalition will form, and thus which rules will result, is a theoretical issue beyond the scope of this chapter.4 For now, all I need to assume is that some set of new rules will be adopted (perhaps temporarily) when at least a minimal winning coalition is in favour of the proposed set. The rule used to change institutional arrangements in field settings varies from reliance on the decisions made by one or a few leaders, to a formal reliance on majority or super-majority vote, to reliance on consensus or close to unanimity. If there are substantial differences in the perceived benefits and costs of appropriators, it is possible that K appropriators will impose a new set of rules on the A-K other appropriators that strongly favours those in the winning coalition and imposes losses or lower benefits on those in the losing coalition (Thompson et al., 1988). If expected benefits from a change in institutional arrangements are not greater than expected costs for many appropriators, however, the costs of enforcing a change in institutions will be much higher than when most participants expect to benefit from a change in rules over time. Where the enforcement costs are fully borne by the members of K, operational rules that benefit the A-K other appropriators lower the long-term costs of monitoring and sanctioning for a governing coalition. Where external authorities enforce the rules agreed upon by K appropriators, the distribution of costs and benefits is more likely to benefit K and may impose costs on the A-K other appropriators (see Walker et al., 2000). The attributes of a resource (listed above) affect both the benefits and costs of institutional change. If resource units are relatively abundant (R1), there are few reasons for appropriators to invest costly time and effort in organizing. If the resource is already substantially destroyed, the high costs of organizing may not generate substantial benefits. Thus, self-organization is likely to occur only after appropriators observe substantial scarcity. The danger here, however, is that exogenous shocks leading to a change in relative abundance of the resource units occur rapidly and appropriators may not adapt quickly enough to the new circumstances (Libecap and Wiggins, 1985). Decentralization and Development: The New Panacea 243

The presence of frequently available, reliable indicators about the conditions of a resource (R2) affects the capacity of appropriators to adapt relatively soon to changes that could adversely affect their long-term benefit stream (Moxnes, 1996). A resource flow that is highly predictable (R3) is much easier to understand and manage than one that is erratic. In the latter case, it is always difficult for appropriators (or, for that matter, for scientists and government officials) to judge whether changes in the resource stock or flow are due to overharvesting or to random exogenous variables. Unpredictability of resource units in microsettings, such as private pastures, may lead appropriators to create a larger common-property unit to increase the predictability of resource availability somewhere in the larger unit (Netting, 1972; Wilson and Thompson, 1993). The spatial extent of a resource (R4) affects the costs of defining reasonable boundaries and then of monitoring them over time. The attributes of the appropriators themselves also affect their expected benefits and costs. If appropriators do not obtain a major part of their income from a resource (A1), the high costs of organizing and maintaining a self- governing system may not be worth their effort. If appropriators do not share a common understanding of how complex resource systems operate (A2), they will find it extremely difficult to agree on future joint strategies. Given the complexity of many common-pool resources – especially multispecies or mul- tiproduct resources – understanding how these systems work may be counter- intuitive even for those who make daily contacts with the resource. Appropriators who trust one another (A4) to keep agreements and use reciprocity in their relationships with one another face lower expected costs involved in monitoring and sanctioning one another over time (Ostrom, 1998a). Appropriators who lack trust at the beginning of a process of organizing may be able to build this form of social capital (Coleman, 1988) if they initially adopt small changes that most appropriators follow before trying to make major institutional changes. Autonomy (A5) tends to lower the costs of organizing. A group that has little autonomy may find that those who disagree with locally developed rules seek contacts with higher-level officials to undo the efforts of appropriators to achieve regulation. Prior experience with other forms of local organization (A6) greatly enhances the repertoire of rules and strategies known by local participants as potentially useful to achieve various forms of regulation. Appropriators in the field rarely face a setting that generates clear-cut benefit-cost ratios, and the collective-choice rules in some settings give a small elite substantial power to block suggested changes that may generate overall positive gains but some losses for those in power. Consequently, the growing theoretical consensus does not lead to a conclusion that most appropriators using common-pool resources will undertake self-governed regulation. Many settings exist where the theoretical expectation should be the opposite: 244 Elinor Ostrom appropriators will overuse the resource unless efforts are made to change one or more of the variables affecting perceived costs or benefits. Many aspects of the macroinstitutional structure surrounding a particular setting affect the perceived costs and benefits (Agrawal and Yadama, 1997). Thus, external authorities (including international donors) can do a lot to enhance the likelihood and performance of self-governing institutions. Their actions can also seriously impede these developments as well. Further, when the activities of one set of appropriators, A, have ‘spillover effects’ on others beyond A, external authorities can either facilitate processes that allow multiple groups to solve conflicts arising from negative spillovers or take a more active role in governing particular resources themselves. When the benefits of organizing are commonly understood by participants to be very high, appropriators lacking many of the attributes conducive to the development of self-governing institutions may be able to overcome their lia- bilities and still develop effective agreements. The crucial factor is not whether all attributes are favourable but the relative size of the expected benefits and costs they generate as perceived by participants. All of these variables affect the expected benefits and costs of appropriators. It is difficult, however, particularly for outsiders, to estimate their impact on expected benefits and costs given the difficulty of making precise measures of these variables and weigh- ing them on a cumulative scale. Further empirical analysis of these theoretical propositions is dependent on the conduct of careful comparative studies over time of a sufficiently large number of field settings using a common set of measurement protocols (see Ostrom, 1998b).

Can external authorities design optimal rules?

Underlying much policy analysis (including the advice given to international donors) of the last 50 years has been a presumption that the responsibility for designing institutional rules to overcome various kinds of collective-action problems – including common-pool resource problems – should be assumed by national governments. Policy analysts have had great faith in their own capacity to analyse theoretical models of typical problems and come forward with a set of institutional fixes that will lead the individuals involved to take actions leading to higher rather than lower outcomes. In doing so, policy analysts tend to think of citizens or users as motivated by narrow, self-interested preferences and of themselves as motivated by a general public interest. Both assumptions are probably exaggerated. The evidence presented above is consistent with a broader theory of a boundedly rational, learning and potentially norm-using individual (see Ostrom, 1998a). The evidence gathered from studies of the behaviour of policy analysts and government officials is consistent with a presumption that officials are similar to citizens in regard to their Decentralization and Development: The New Panacea 245 internal motivation. Officials may be placed in institutional settings, however, that bring substantial temptations for personal gain without much chance for others to learn about behaviour. Officials may also be isolated from the people and locations they are supposed to represent and not have much knowledge about time and place information. A better foundation for public policy is to assume that individuals may not be able to analyse all situations fully, but given a conducive, macro-political regime, they may make efforts to solve complex problems through trial-and- error testing out of different rules for solving various collective-action problems (see Ostrom, 1998a; Bendor and Mookherjee, 1987). Colleagues at the Workshop in Political Theory and Policy Analysis have been studying the kinds of rules that users of common-pool resources in many parts of the world have adopted (the methodology is described in Ostrom, 1999). Studying these rule systems provides essential information about the complexity of the situations that individuals find themselves facing and the consequent complexity of the rule systems they devise. In this chapter, I can only provide a short glimpse of that complexity, but I hope I can convey its extent. Let us turn first to the type of boundary rules – the rules that define who is eligible to use a common-pool resource – that users in the field have crafted. A key problem of all common-pool problems, such as managing a forest, an irrigation system, an inshore fishery or the Internet, is that of excluding free riders. A commitment that one set of users is authorized to use a resource and others are not is necessary if one is to exclude free riders. Empirical studies of boundary rules used in field settings have identified 27 different types of boundary rules used alone or in combination (Ostrom, 1999; Ostrom et al., 1994). A second essential set of rules relates to the harvesting and other activities in which individuals participate. Empirical studies have identified over 100 types of rules that specify when, where, how and how much of the products of a commons may be appropriated by someone who is authorized to do so. The number of rules related to incentives and sanctions, to information conditions and to procedural requirements that could be used to regulate activities related to a common-pool resource is also very large. Since rules are used in combination with one another, the potential configuration of rules that could be used to commit participants to follow strategies that improve the efficiency, equity and sustainability of common- pool resources approaches infinity. Consequently, instead of assuming that the choice of institutional rules to improve the performance of human systems that utilize common-pool resources – or any other complex task for that matter – is a process of designing optimal rules, we need to understand the policy design process as involving an effort to tinker with a large number of component parts (see Jacob, 1977). Those who tinker with any tools – including rules – try to find 246 Elinor Ostrom combinations that work together more effectively than other combinations. Policy changes are experiments based on more or less informed expectations about potential outcomes and the distribution of these outcomes for participants across time and space (Campbell, 1969, 1975). Whenever individuals agree to add a rule, change a rule or adopt someone else’s proposed rule set, they are conducting a policy experiment. Further, the complexity of the ever- changing biophysical world combined with the complexity of rule systems means that any proposed rule change faces a non-trivial probability of error. When only a single governing authority makes decisions about rules for an entire region, policy-makers have to experiment simultaneously with all of the common-pool resources within a jurisdiction with each policy change. And, once a change has been made and implemented, further changes will not be made rapidly. The process of experimentation will usually be slow. Information about results may be contradictory and difficult to interpret. Thus, an experiment that is based on erroneous data about one key structural variable or one false assumption about how actors will react can lead to a very large disaster (see Wilson et al., 2000). In any design process where there is substantial probability of error, having redundant teams of designers has repeatedly been shown to have considerable advantage (see Landau, 1969, 1973; Bendor, 1985). The important point is: if the systems are relatively separable, allocating responsibility for experimenting with rules will not avoid failure, but will drastically reduce the probability of immense failures for an entire region. Due to the disillusion of many analysts with the performance of highly centralized systems in the management of resources, frequent proposals are made (and many have been implemented) to use development assistance to support radical decentralization of centralized regimes (see Agrawal et al., 1999, and Agrawal and Ribot, 1999, for an overview of the relevant literature). One form of decentralization that is conceptualized is a complete layer of government that is composed entirely of self-organized, local bodies each of which govern a smaller-scale resource system. Such a form has advantages and disadvantages.

Advantages and disadvantages of fully decentralized systems

Among the advantages of assigning the authority to regulate smaller-scale and separable common-pool resources to the users are the following:

• Local knowledge. Appropriators who have lived and appropriated from a resource system over a long period of time have developed relatively accurate mental models of how the biophysical system itself operates, since the very success of their appropriation efforts depends on such knowledge. Decentralization and Development: The New Panacea 247

They also know others living in the area well and what norms of behaviour are considered appropriate by the community. • Inclusion of trustworthy participants. Appropriators can devise rules that increase the probability that others are trustworthy and will use reciprocity. This lowers the cost of relying entirely on formal sanctions and paying for extensive guarding. • Reliance on disaggregated knowledge. Feedback about how the resource system responds to changes in actions of appropriators is generated in a disaggregated way. Fishers are quite aware, for example, when the size and species distribution of their catch changes over time. Irrigators learn whether a particular allocation system is efﬁcient by comparing the net yield they obtain under one set of rules versus others. • Better adapted rules. Given the above, appropriators are more likely to craft rules that are better adapted to each common-pool resource than any general system of rules. • Lower enforcement costs. Since local appropriators have to bear the cost of monitoring, they are apt to craft rules that make infractions highly obvious so that monitoring costs are lower. Further, by creating rules that are seen as legitimate, rule conformance will tend to be higher. • Redundancy. The probability of failure for an entire region is greatly reduced by the establishment of parallel systems of rule making, interpretation and enforcement.

There are, of course, limits to all ways of organizing the governance of common-pool resources. Among the limits of a fully decentralized system are the following:

• Some appropriators will not organize. While the evidence from the ﬁeld is that many local appropriators do invest considerable time and energy into their own regulatory efforts, other groups of appropriators do not do so. There appear to be many reasons why some groups do not organize, including the presence of low-cost alternative sources of income and thus a reduced dependency on the resource, conﬂict among appropriators along multiple dimensions, lack of leadership and fear of having their efforts overturned by outside authorities. Tragically, some of the most needy groups may be those least likely to self-organize (Brett, private correspondence). • Some self-organized efforts will fail. Given the complexity of the task involved in designing rules, some groups will select combinations of rules that generate failure instead of success. They may be unable to adapt rapidly enough to avoid the collapse of a resource system. • Local tyrannies. Not all self-organized resource governance systems will be organized democratically or rely on the input of most appropriators. Some 248 Elinor Ostrom

will be dominated by a local leader or a power elite who only change rules that will be of advantage to them. This problem is accentuated in locations where the cost of exit is particularly high and reduced where appropriators can leave. • Stagnation. Where local ecological systems are characterized by considerable variance, experimentation can produce severe and unexpected results, leading appropriators to cling to systems that have worked relatively well in the past, and stop innovating long before they have developed rules likely to lead to better outcomes. • Inappropriate discrimination. It is always necessary to exclude some individuals from using a resource who do not have a legal right to use it and are not contributing to the sustainability of that resource. However, exclusion can be based on inappropriate grounds or on ascribed characteristics that have nothing to do with legal rights or the trustworthiness of individuals to follow a set of agreed-upon rules. • Limited access to scientific information. While time and place information may be extensively developed and used, local groups may not have access to scientific knowledge concerning the type of resource system involved. • Conflict among appropriators. Without access to an external set of conflict- resolution mechanisms, conflict within and across common-pool resource systems can escalate and provoke physical violence. Two or more groups may claim the same territory and may continue to make raids on one another over a very long period of time. • Inability to cope with larger-scale common-pool resources. Without access to some larger-scale jurisdiction, local appropriators may have substantial difficulties regulating only a part of a larger-scale common-pool resource. They may not be able to exclude others who refuse to abide by the rules that a local group would prefer to use. Given this, local appropriators have no incentives to restrict their own use.

Thus, while there are advantages of such a form of decentralization, it is based on a theory of governance that thinks there is a discrete choice between organizing resource regimes strictly at a local level or organizing them from the centre. It is also based on a naive theory of community (Agrawal and Gibson, 1999). There is a better way.

The advantages of polycentricity

Many of the capabilities of a parallel adaptive system can be retained in a polycentric governance system. By polycentric, I mean a system where citizens are able to organize not just one but multiple governing authorities at differing scales (see Ostrom et al., 1961; McGinnis, 1999a, 1999c). Each unit may Decentralization and Development: The New Panacea 249 exercise considerable independence to make and enforce rules within a circumscribed scope of authority for a specified geographical area. In a polycentric system, some units are general-purpose governments while others may be highly specialized. Self-organized resource governance systems, in such a system, may be special districts, private associations or parts of a local government. These are nested in several levels of general-purpose governments that also provide civil equity as well as criminal courts. In a polycentric system, the users of each common-pool resource would have some authority to make at least some of the rules related to how that particular resource will be utilized. Thus, they would achieve many of the advantages of utilizing local knowledge as well as the redundancy and rapidity of a trial-and-error learning process. On the other hand, problems associated with local tyrannies and inappropriate discrimination can be addressed in larger, general-purpose governmental units that are responsible for protecting the rights of all citizens and for the oversight of appropriate exercises of authority within smaller units of government. It is also possible to make a more effective blend of scientific information with local knowledge where major universities and research stations are located in larger units, but have a responsibility to relate recent scientific findings to multiple smaller units within their region. Because polycentric systems have overlapping units, information about what has worked well in one setting can be transmitted to others who may try it in their settings. Associations of local resource governance units can be encouraged to speed up the exchange of information about relevant local conditions and about policy experiments that have proved particularly successful. And, when small systems fail, there are larger systems to call upon – and vice versa. Polycentric systems are themselves complex, adaptive systems without one central authority dominating all of the others. Thus, there is no guarantee that such systems will find the combination of rules at diverse levels that is optimal for any particular environment. In fact, one should expect that all governance systems will operate at less than optimal levels, given the immense difficulty of fine-tuning any very complex, multi-tiered system. In the United States, many examples of dynamic, polycentric resource governance systems exist where there is strong evidence of high performance. One example is the Maine lobster fishery, which is noteworthy because of the long-term, complementary roles adopted by both local and state governance systems. Maine is organized into riparian territories along most of the coast. Boundary rules and many of the day-to-day fishing regulations are organized by harbour gangs (Acheson, 1988). At the same time, the state of Maine has long-established formal laws that protect the breeding stock and increase the likelihood that regeneration rates will be high. ‘At present, the most important conservation laws are minimum and maximum size measures, a prohibi- 250 Elinor Ostrom tion against catching lobsters with eggs, and a law to prohibit the taking of lobsters that once had eggs and were marked – i.e., the “V-notch” law’ (Acheson et al. 1998, 400). Neither the state nor any of the harbour gangs have tried to limit the quantity of lobster captured. The state does not make any effort to limit the number of fishers since this is already done at a local level. However, the state has been willing to intercede when issues exceed the scope of control of local gangs. In the late 1920s, for example, when lobster stocks were at very low levels and many local areas appear to have had substantial compliance problems, the state took a number of steps – including threats to close the fishery – that supported informal local enforcement efforts. By the late 1930s, compliance problems were largely resolved and stocks had rebounded. Recently, in response to changes that were breaking down the harbour gang system, the state formalized the system by dividing the state into zones with democratically elected councils. Each council has been given authority over rules that have principally local impacts – trap limits, days and times fished, etc. Interestingly, the formalization of local zones was followed, almost immediately, by the creation of an informal council of councils to address problems at a greater-than-local scale (Wilson, 1997). Today, the state needs only about six patrol officers on the water to police the activities of 6800 lobstermen, all the other fisheries and coastal environmental laws. During the 1990s, the fishery has been growing substantially with increased yields (Acheson, 1993). At the same time, there is strong evidence that the number of reproductive-age females in the Maine waters is very large and that the recruitment levels will continue at a high level.

Some recommended forms of decentralization: a cautionary note

The research briefly reviewed above has been cited from time to time as supportive for a naive policy of ‘handing over’ resources – particularly irrigation systems and forests – to the users. In regard to irrigation, the stimulus has frequently been the advice of donors to reduce the size of the public sector drastically. In regard to forests, the stimulus has also been the problem that some government forests are entirely degraded and a hope that local users will invest the time and energy needed to encourage substantial regrowth.5 Unfortunately, the image behind these policies is that of a radical decentralization whereby government totally pulls out and ‘turns over’ the responsibility to local groups. Further, the process of ‘turning over’ has frequently involved a quickly called meeting between a district irrigation or forestry officer with those who show up at a meeting (usually the better connected in a community). Arrangements differ by country and sector, but many reports from the Philippines and from Nepal indicate a relatively casual withdrawal Decentralization and Development: The New Panacea 251 by government and a uniform agreement that creates the same rules for a user group throughout the country. While the responsibility for managing the resource is suddenly turned over to the users, the users are not perceived as actively engaged in the design of their own institutions. Fujita et al. (1999), for example, describe the processes of turning over government irrigation systems to farmers in the Philippines – a country where farmer-managed systems have proved themselves to be comparatively effective (Siy, 1982). The systems had been constructed, managed and operated by the National Irrigation Association (NIA). Until the end of the 1970s, most of the operation and maintenance of these systems was undertaken by the NIA based on subsidies from the national treasury augmented by foreign aid inflows. In the early 1980s the international rice market crashed and the receipt of foreign aid was reduced. A policy of handing over NIA systems to the farmers was initiated.

Results of the handing-over policy, however, have not been very encouraging. There have been some success cases but more failure cases. … The failure has been, to a large extent, based on hasty handing-over by state agencies under the rather suddenly emerged ﬁnancial crisis in the 1980s. Typically absent in designing the policy has been due knowledge on the side of bureaucracy on the capacity and mechanism of local communities to organize collective action for managing local commons. Appropriate policies to enhance the communities’ organizational capacity have rarely been instituted. The difﬁculty here is that major constraints in organizing collective action for irrigation management are different across different communities with different environmental as well as economic and social conditions. (Fujita et al., 1999, pp. 3–4)

Fujita and his co-authors conducted a quantitative study of 46 irrigation systems that had been government-owned and were handed over to farmers. They obtained information for each system as to whether particular activities were jointly undertaken (such as cleaning the canals) as well as information about contextual variables that would affect the difﬁculty of organizing collective action. They found that many of the variables identiﬁed above as affecting the likelihood of organizing did indeed have an independent effect in the multiple regressions they conducted. The variables that consistently affected performance were:

• ample water supply (R1) (negative); • large differences in water supply between headenders and tailenders (heterogeneity) (negative); • large service area (R4) (negative); 252 Elinor Ostrom

• little interaction among the farmers (A2 and A4) (negative); • easy exit to non-farm economic activities (A3) (negative); and • prior experience as a communal irrigation system prior to the NIA (A6) (positive).

This recent research thus strongly supports the theoretical arguments developed above. Further, it provides evidence that simply turning over government property to users in a casual process is not a policy that has a firm theoretical or empirical foundation. Citations to the research showing the capabilities of farmer groups to self-organize and cope with many collective- action problems misrepresent the research. It is one thing to self-organize to create your own property and slowly develop the rules of association that enable a group to benefit from the long-term management of that resource. It is quite something else to have a government tell you that now you have to manage something that the government can no longer handle itself! Especially after you have been told that it is the government’s responsibility to do this for you. Similar outcomes occurred in the early stages of handing over government forests to users in Nepal. Most of these efforts involved allocating degraded or completed deforested land to users without much discussion about who would be involved and how they would manage. A study conducted in one district in 1985 of 419 ‘chairpersons’ of a forest committee uncovered that ‘most of them did not know if they were members of a community forest committee, or what they were expected to do’ (Britt, 2000, p. 22). Despite intensive inputs from the FAO, the World Bank and bilateral donors, only a small proportion of the designated forests was actually turned over, and very few management plans were prepared by the users. Further, whether the users actually owned the forest thereafter was never made clear and most users expected the government to take it back from them if they were to make substantial improvements. More recent developments have involved a higher level of community involvement and the creation of a federation of forest user groups that has provided an important overlapping level to individual forest user groups (see Britt, 2000). If decentralization is operationalized as these policies were, they are bound to fail and lead to a confirmation of the arguments made by public officials that users can’t manage these resources (see also Bienen et al., 1990). If one is to build a vibrant democratic and accountable system of governance from a system that is highly centralized, it is obvious that some forms of ‘decentralization’ will be needed. The casual turning over of resources to users (or of major government assets to ‘private’ owners), however, can turn out to be counterproductive. Decentralization and Development: The New Panacea 253

Conclusion

The major conclusion of this chapter is that the development of effective institutions is a process that takes time and is strongly affected by a variety of factors related to the resource (or problem at hand) and the individuals involved. Trying to jump-start the process of institutional development through substantial development assistance without a solid theoretical understanding of institutional development can retard rather than advance human well-being. One of the important lessons of extensive research is that there is no single blueprint for an effective organization to solve similar problems – let alone substantially different ones. Enabling citizens and their officials to create collective-action organizations in the public and/or private sphere at multiple levels takes considerable time and effort and will rarely approach either a fully centralized or a fully decentralized system. Polycentric political systems are frequently rejected by policy analysts on the basis of their surface appearance. They look too messy and chaotic. In studies of local public economies, however, messy polycentric systems significantly out- performed metropolitan areas served by a limited number of large-scale, unified governments (see literature synthesized in McGinnis, 1999b; Bish, 1999). As discussed above, evidence that complex, multi-tiered systems for governing common-pool resources are comparatively successful is now accumulating. The good society may not be achievable by systems that look neat, but do not have the information and adaptive capabilities needed to achieve efficiency, sustainability and equity. Those of us in academic life need to learn how better to analyse complex systems in order that we can distinguish between those complex systems that are truly counterproductive, and those that are well matched to the ecological systems they are intended to govern. The choice is not simply between pure centralization and pure decentralization.

Notes

1. This chapter forms the basis of a plenary address given to the Political Studies Association annual conference, London School of Economics, 11 April 2000. Support of the National Science Foundation (SBR 9521918) and the Food and Agriculture Organization of the United Nations is gratefully acknowledged. The comments of Teddy Brett on an earlier version and the editorial assistance of Anne Leinenbach and Patty Zielinski are deeply appreciated. 2. There is also a very rich experimental literature demonstrating that when subjects face analytical problems with the mathematical structure of the conventional theory, but are able to communicate, they are able to reach agreements, and stay with them, that come much closer to the optimal use of the resource. Experimental conditions can create predictable variance in the level of efﬁciency achieved. See Ostrom et al. (1994) for an overview of this literature. 254 Elinor Ostrom

3. Segments of this chapter draw on Ostrom (2000) and Ostrom (1999). 4. Which decision rules are used at a collective-choice level do indeed make a substantial difference in the distribution of beneﬁts (see Walker et al., 2000). 5. In other cases, apparent decentralization may be an effort to quiet opposition and delay a real form of democratization from occurring (Mandondo, 2000).

References

Acheson, James M. (1988) The Lobster Gangs of Maine. Hanover, NH: New England University Press. Acheson, James M. (1993) ‘Capturing the Commons: Legal and Illegal Strategies’, in T. L. Anderson and R. T. Simmons (eds), The Political Economy of Customs and Culture: Informal Solutions to the Commons Problem. Lanham, MD: Rowman & Littlefield. Acheson, James M., Wilson, J. A. and Steneck, R. S. (1998) ‘Managing Chaotic Fisheries’, in Fikret Berkes and Carl Folke (eds), Linking Social and Ecological Systems. Management Practices and Social Mechanisms for Building Resilience. Cambridge, MA: Cambridge University Press. Agrawal, Arun, with Britt, Charla and Kanel, Keshav (1999) Decentralization in Nepal: A Comparative Analysis, A Report on the Participatory District Development Program. Oakland, CA: ICS Press. Agrawal, Arun and Gibson, Clark (1999) ‘Enchantment and Disenchantment: The Role of Community in Natural Resource Conservation’, World Development, 27: 629–49. Agrawal, Arun and Ribot, Jesse (1999) ‘Accountability in Decentralization: A Framework with South Asian and West African Cases’, Journal of Developing Areas 33(4):473–502. Agrawal, Arun and Yadama, Gautam N. (1997) ‘How Do Local Institutions Mediate Market and Population Pressures on Resources? Forest Panchayats in Kumaon, India’ Development and Change, 28: 435–65. Baland, J. M. and Platteau, J. P. (1996) Halting Degradation of Natural Resources. Is There a Role for Rural Communities? Oxford: Clarendon Press. Bardhan, Pranab (1999) Water Community: An Empirical Analysis of Cooperation on Irrigation in South India. Berkeley: University of California, Working Paper. Bendor, Jonathan B. (1985) Parallel Systems. Redundancy in Government. Berkeley: University of California Press. Bendor, Jonathan B. and Mookherjee, Dilip (1987) ‘Institutional Structure and the Logic of Ongoing Collective Action’, American Political Science Review, 81: 129–54. Bienen, Henry, Kapur, Devesh, Parks, James and Riedlinger, Jeffrey (1990) ‘Decentralization in Nepal’, World Development, 18: 61–75. Bish, Robert L. (1999) ‘Federalist Theory and Polycentricity: Learning from Local Governments’, in Donald P. Racheter and Richard E. Wagner (eds), Limiting Leviathan. Cheltenham: Edward Elgar: 203–20. Britt, Charla (2000) Forestry and Forest Policies. Bloomington: Indiana University, Workshop in Political Theory and Policy Analysis, Working Paper. Campbell, Donald T.(1969) ‘Reforms as Experiments’, American Psychologist, 24: 409–29 Campbell, Donald T. (1975) ‘On the Conflicts between Social Evolution and between Psychology and Moral Tradition’, American Psychologist, 30: 1103–26. Coleman, J. (1988) ‘Social Capital in the Creation of Human Capital’, American Journal of Sociology, 91: 309–35. de los Reyes, R. P. (1980) 47 Communal Gravity Systems: Organization Profiles. Quezon City, Philippines: Ateneo de Manila University, Institute of Philippine Culture. Decentralization and Development: The New Panacea 255

Feeny, D., Hanna, S. and McEvoy A. F. (1996) ‘Questioning the Assumptions of the “Tragedy of the Commons” Model of Fisheries’, Land Economics, 72: 187–205. Fujita, Masako, Hayami, Yujiro and Kikuchi, Masao (1999) The Condition of Collective Action for Local Commons Management: The Case of Irrigation in the Philippines, IRRI- Japan Shuttle Project. Social Science Division Working Paper. Jacob, F. (1977) ‘Evolution and Tinkering’, Science, 196: 1161–6. Lam, Wai Fung (1998) Governing Irrigation Systems in Nepal: Institutions, Infrastructure, and Collective Action. Oakland, CA: ICS Press. Landau, M. (1969) ‘Redundancy, Rationality, and the Problem of Duplication and Overlap’, Public Administration Review, 29: 346–58. Landau, M. (1973) ‘Federalism, Redundancy, and System Reliability’, Publius 3: 173–96. Libecap, G. and Wiggins, S. N. (1985) ‘The Influence of Private Contractual Failure on Regulation: The Case of Oil Field Unitization’, Journal of Political Economy, 93: 690–714. McGinnis, Michael (1999a) Conflict Dynamics in a Three-Level Game: Local, National, and International Conflict in the Horn of Africa. Paper presented at the 33rd North American meeting of Peace Science Society (International), University of Michigan, Ann Arbor, 8–10 October. McGinnis, Michael (ed.) (1999b) Polycentric Governance and Development: Readings from the Workshop in Political Theory and Policy Analysis. Ann Arbor: University of Michigan Press. McGinnis, Michael (ed.) (1999c) Polycentricity and Local Public Economies: Readings from the Workshop in Political Theory and Policy Analysis. Ann Arbor: University of Michigan Press. McKean, Margaret A. (1992) ‘Management of Traditional Common Lands (Iriaichi) in Japan’, in Daniel W. Bromley et al. (eds), Making the Commons Work: Theory, Practice, and Policy. San Francisco, CA: ICS Press: 63–98. McKean, Margaret A. (1996) ‘Common Property Regimes as a Solution to Problems of Scale and Linkage’, in Susan Hanna, Carl Folke, and Karl-Göran Mäler (eds), Rights to Nature. Washington, DC: Island Press: 223–43. Mandondo, Alois (2000) Forging (Un)democratic Resource Governance Systems from the Relic of Zimbabwe’s Colonial Past. Harare, Zimbabwe: University of Zimbabwe, Institute of Environmental Studies, Working Paper. Moxnes, E. (1996) Not Only the Tragedy of the Commons: Misperceptions of Bioeconomics. Bergen, Norway: Foundation for Research in Economics and Business Administration, SNF, Working Paper. Netting, Robert McC. (1972) ‘Of Men and Meadows: Strategies of Alpine Land Use’, Anthropological Quarterly, 45: 132–44. Ophuls, W. (1973) ‘Leviathan or Oblivion’, in H. E. Daly (ed.), Toward a Steady State Economy. San Francisco, CA: Freeman: 215–30. Ostrom, Elinor (1990) Governing the Commons: The Evolution of Institutions for Collective Action. New York: Cambridge University Press. Ostrom, Elinor (1998a) ‘A Behavioral Approach to the Rational Choice Theory of Collective Action’, American Political Science Review, 92: 1–22. Ostrom, Elinor (1998b) ‘The International Forestry Resources and Institutions Research Program: A Methodology for Relating Human Incentives and Actions on Forest Cover and Biodiversity’, in F. Dallmeier and J. A. Comiskey (eds), Forest Biodiversity in North, Central and South America, and the Caribbean: Research and Monitoring, Man and the Biosphere Series, vol. 21. Paris: UNESCO; New York: Parthenon: 1–28. 256 Elinor Ostrom

Ostrom, Elinor (1999) ‘Coping with Tragedies of the Commons’, Annual Review of Political Science, 2: 493–535. Ostrom, Elinor (2000) ‘Collective Action and the Evolution of Social Norms’, Journal of Economic Perspectives 14(3):137–58. Ostrom, Elinor, Gardner, Roy and Walker, James M. (1994) Rules, Games, and Common- Pool Resources. Ann Arbor: University of Michigan Press. Ostrom, Vincent, Tiebout, Charles M. and Warren, Robert (1961) ‘The Organization of Government in Metropolitan Areas: A Theoretical Inquiry’, American Political Science Review, 55: 831–42. Sengupta, Nirmal (1991) Managing Common Property: Irrigation in India and the Philippines. New Delhi: Sage. Simmons, R. T., Smith, Jr. F. L. and Georgia, P. (1996) ‘The Tragedy of the Commons Revisited: Politics versus Private Property’, Washington, DC: The Center for Private Conservation. Siy, R. Y., Jr (1982) Community Resource Management: Lessons from the Zanjera. Quezon City: University of the Philippines Press. Tang, Shui Yan (1992) Institutions and Collective Action: Self-Governance in Irrigation. San Francisco, CA: ICS Press. Thompson, L. L., Mannix, E. A. and Bazerman, M. H. (1988) ‘Negotiation in Small Groups: Effects of Decision Rule, Agendas, and Aspirations’, Journal of Personality and Social Psychology, 54: 86–95. Wade, Robert (1994) Village Republics: Economic Conditions for Collective Action in South India. San Francisco, CA: ICS Press. Walker, James M., Gardner, Roy, Herr, Andrew and Ostrom, Elinor (2000) ‘Collective Choice in the Commons: Experimental Results on Proposed Allocation Rules and Votes’, Economic Journal, 110: 212–34. Wilson, James A. (1997) ‘Maine Fisheries Management Initiative’, in G. Palsson (ed.), The Social Impacts of Individual Transferable Quotas. Copenhagen: TemaNord. Wilson, James A., Low, Bobbi, Costanza, Robert and Ostrom, Elinor (2000) ‘Scale Misperceptions and the Spatial Dynamics of a Social-Ecological System’, Ecological Economics, 31: 243–57. Wilson, P. N. and Thompson, G. D. (1993) ‘Common Property and Uncertainty: Compensating Coalitions by Mexico’s Pastoral Ejidatarios’, Economic Development and Cultural Change, 41: 299–318. Index

accountability Brazil 56 and regulatory regimes 205–6, 216 Bretton Woods system 190, 203 in European Union 115 Britain of international organizations 15 age of democracy xi Africa xi, 188–9 and European integration 171 African Americans 4, 5, 6 anti immigration 141 Albania 141, 183 citizenship 138, 150 American Declaration of Independence civic engagement 104 3–4 class 3, 5 Angola 140 deliberative polls 16n animal rights movement xv, 118 democracy assistance 191 anti-capitalist movement xv, 118, 125, democratic socialism 12 127, 129 DiY culture 126–8 Argentina 56 e-governance 222, 224, 226, 231 Armenia 56 freedom and happiness 57 Australia hackers 120 e-governance 223 inequality xi, 3, 12 freedom and happiness 56 local government system xi local elections 89 movement towards equality 4 news media 110 new labour 43n rights and liberties of citizens 9 newspapers 102, 104, 106–10, 112, voter turnout 89–90, 95n 135 Austria participation in politics xiv, 112 and European Union 171 privatization xvii, 208, 211–13, 214, freedom and happiness 56 216–17 news media 106–10 Reform Act (1832) 16n voter turnout 89–90 rights and liberties of citizens 9 authoritarian regimes xi, 9, 101, 172 status 3 Azerbaijan 56 TV news 106–10, 112 unelected bodies xi Back Oriﬁce hacking program 121–2, unelected upper chamber xi, 3, 5 124, 127, 129n university admissions 3 Baltic States 182 voter turnout xi, xiv, 61–77: Bangladesh 56 Conservative voters 65, 74–7, Barcelona, participatory experiment in 79n; constituency scale 65–72, 77; advisory councils 153, 156–8, European Parliamentary elections 161–4, 165n 61; individual scale 65, 72–6; women in 158, 162, 164 labour voters xiv, 65, 71–2, 75–7, Bavaria 179 79n; liberal democrat voters 74–7, Belarus 56 79n; local government elections Belgium 61, 64–5, 89; national scale 65–7; freedom and happiness 56 1997 general election xiv, 61, 64, news media 106–10 65, 66, 67–77; rational voter models Berisha, President 183 of 62–77; Scottish Parliamentary Blocher, Christoph 142, 143 elections 61; Welsh Assembly Brandt, Willy 172 elections 61; women’s suffrage 5

257 258 Index

British Commonwealth 191 decentralized system xvii, 237, 246–8, Bulgaria 250–3, 254n freedom and happiness 57 development assistance xvii, 244, political parties 180, 181, 183 246, 253 polycentric governance system Canada 248–50, 253 freedom and happiness 57 self-organized governance systems in local elections 89, 96n the field 238–44 satisfaction in 39 fisheries xvii, 245, 247–8, 249–50 voter turnout 89–90, 96n forests xvii, 245, 250–2 Central Europe, democratization of irrigation systems xvii, 238–9, 245, conditionality 170, 173–5 250–2 convergence 170, 173–7, 185 sustainable use of 237 European Union and 169–85 communism membership of European Union 169, European i, xi, 169, 171–2, 175–6 173, 174, 180, 182, 183, 185, 190 fall of xi, 169, 172 party systems xii, xvi, 169–85 reform communists 175–6 transnational party cooperation xvi, computer security 170–85, 186n, 190 hackers and 119–22, 123 see also specific countries conditionality 170, 173–5 centralization consumer sovereignty 206, 207–8, of natural resources xvii, 246, 253 210–11, 216–18 Chile consumer transparency 206, 207–8, democratic breakdown 9, 11, 198 218 freedom and happiness 57 convergence 170, 173–7, 185 China crime 23, 24 and regulatory design 208 Croatia 57, 179, 183 computer hackers in 125–6 Cult of the Dead Cow (CDC) 121–2, economy of 12 123, 124, 125, 127 freedom and happiness 57 cynicism xi, 100–1, 104 citizens see happiness; participation Czech Republic citizenship freedom and happiness 57 transparency of regulatory regimes and political parties 176–7, 179, 182 205–9, 216–18 TV news 110 civic disengagement Czechoslovakia 140, 175, 176, 180 and political communications xv, 100–1, 104, 113, 114 decentralization civic engagement of natural resources xvii, 237, 246–8, and political communications 100–1, 250–3, 254n 104, 111–15 decision-making and voter turnout 61, 81–2, 86–9, among leaders 11–12, 14 91, 92–4 by and among leaders and consumers civic literacy xiv-xv, 91–4 12 civil rights 4, 5, 6 by leaders 11–12, 14 civil society 154–5, 157, 160, 169, 177, computer assisted 221–3, 224–6, 200, 202–3 227–8 class 3, 5 democratic xii, xiv, xv Cohen, Joshua 131–2, 137, 143, 150n, participation in xii, xiv, xv 155, 163 deliberative democracy, ideal of 131–2 common pool of natural resources participation in xii, xv-xvi, 155 centralization xvii, 246, 253 theory of 153–64 conventional theory 237–8, 239 see also referendums Index 259 democracy assistance xvi, 190–1, 203 economic enterprises Democracy in America (Tocqueville) democratic governance of 15, 16n 4–5 economic equality 13, 31 democratic competition economic freedom and satisfaction xiii, 40–2, 51 and happiness 46, 47–50, 51, 52–3, democratic politics 54–7 and quality of life 33–4 and satisfaction xi, xiii and satisfaction 33–42 in poor nations 48–50, 52–3 Denmark egalitarianism and European integration 171 and levelling down xiii, 18, 23–4, 32n freedom and happiness 57 and Pareto improvements xiii, 18 news media 106–10 Socially-Located Egalitarianism (SE) developing countries 23–5 centralization xvii, 246, 253 e-governance common pool of natural resources Britain 222, 224, 226 xvii, 237–53 collecting data 221 decentralization xvii, 237, 246–8, decision-making 221–3, 224–6, 227–8 250–3, 254n France 221 development assistance xvii, 244, generating understanding 220–1 246, 253 information and 222, 224, 225, 226, development projects 237 229, 231 economic and political development integration and storage 221 xvii judgement and appreciation 230–1 fisheries xvii, 245, 247, 249–50 modelling decisions 221 forests xvii, 245, 250–2 organizing and analysing data 221 irrigation systems xvii, 238–9, 245, supporting communication 221 250–2 theories 222–9 sustainable use of natural resources elections 237 and quality of life 33 development assistance xvii, 244, 246, campaigns for 103, 111 253 electoral system xv dictatorial regimes xi, 9 local government 61, 64–5, 82, 87–94, DiY culture 118, 119, 126–8 96n Downs, Anthony 62–4, 69–70, 72–4, 77, international monitoring 201 83–4 turnout xi, xii, xiv-xv, 61–77, 81–94 electronic civil disobedience 119, 123–6 Eastern Europe, democratization of electronic direct movements xv, 118–29 conditionality 170, 173–5 environmental movements xv, 118, convergence 170, 173–5, 185 126–7, 129 European Union and 169–85 environmental pollution 24 membership of European Union 169, envy 173, 174, 180, 182, 183, 185 and well-being 28–9 party systems xii, xvi, 169–85 politics of 18, 31 transnational party cooperation equality 170–85, 186n, 190 and justice 19–21 see also specific countries and liberty xi, xii, 7–11 ecological movements xv, 118 and rights xii economic development and satisfaction xiii and inequality 23–4 and welfare democracies xii and satisfaction 34, 37–9 economic 13, 31 developing countries xvii egalitarians and xiii freedom and 51 human 3–4, 5–6 260 Index equality continued fairness in franchise xi and levelling down 27–31 political 3–15 and well being 27–30 social 23–5 fascism xi, 185 spread of xii female suffrage xi, 4, 5 see also levelling down feminists 155, 197 Estonia 57 Finland Eurobarometer xiii, 37–9, 107 freedom and happiness 57 Europe news media 106–10 broadcasting 102, 105–6 satisfaction in 39 communism xi, 169, 171–2, 175–6 fisheries xvii, 245, 247, 249–50 democratic socialism 12 forests xvii, 245, 250–2 fascism xi, 185 France integration 169–71, 173–5 e-governance 221 news media 102, 105–9, 113 freedom and happiness 57 transnational party cooperation xvi, local elections 89 170–5, 186n movement towards equality 4 see also European Union; specific news media 106–10 countries rights and liberties of citizens 9 European Community socialists 43n news media 108–9 Frasier Institute 49, 50, 51, 52, 54–5, satisfaction in 37–9, 42n 56–7 see also European Union; specific freedom countries and happiness 46–57 European Court of Justice 174 and satisfaction xi, xiii European Empire xi definition of 51–2 European Environmental Protection economic xi, xiii, 46, 47–50, 51, 52–3, Agency 208 54–7 European integration 169–71, 173–5 for whom? 8 European Parliament government and 48–50, 51–3, 54–5 elections to 61 negative 51, 52 political parties in 170, 173, 174, personal xi, 47, 51 178–9, 183, 185 political xiii, 46, 47–8, 52, 53, 56–7 European Union press xiii, 46, 47–8, 51, 52, 53, 56–7 Central and Eastern Europe and 169–85 Georgia 57 conditionality 173–5 Germany convergence 173–7, 175 democracy assistance 190, 197 Council if Europe 178 freedom and happiness 57 Council of Ministers 174 local elections 89, 96n democratic deficit xi news media 102, 104, 106–10 federal state xi privatization xvii, 205, 208, 215–17 lack of accountability 115 SPD 43n membership of xvi, 169, 173, 174, voter turnout 89–90, 96n 180, 182, 183, 185, 190 government news media 106–7 and satisfaction xiii-xiv political parties in 169–85 and size of democratic unit 13–14 privatization 213, 217 and society 8–9 public disengagement 100 and use of natural resources 244–6 transnational party cooperation xvi, authoritarian xi, 9, 101, 172 170–85, 186n freedom and 48–50, 51–3, 54–5 see also specific countries of a democratic state 10–11 exclusion 7, 25–31, 132–3 responsiveness to citizens 14 Index 261

Greece 106–10, 173 Iceland 9, 57 Greek city-states 13 immigration and cultural conflicts 15 Habermas, J. 154–9, 164 anti immigration referendums in hackers and hacktivism Switzerland 137–50 and electronic civil disobedience 119, in Britain 141 123–6 India and electronic direct action movements freedom and happiness 57 xv, 118–29 hackers 120 and exploration of computer security scheduled castes 6, 15–16n 119–22, 123 voting 6 anti capitalist principles xv, 118, 125, industrialization 23–4 127, 129 inequality Back orifice hacking program 121–2, and economic development 23–4 124, 127, 129n and market distribution 35–7 communities of 119–23 and market economies 36 connection to DiY culture 118, 119, increasing levels in UK and USA xi 126–8 political 103 Cult of the Dead Cow (CDC) 121–2, state institutions and xi 123, 124, 125, 127 information defence of virtual cultures 119, 122, and privatization 208, 209–11, 123 212–13, 216–18 environmental principles 126–7, and referendums 132, 134–5 129 and voter turnout 62–4, 71, 77, hacking in USA 124–6 82–4,92 politically motivated hackers 119, computer hackers and 119, 122–3, 123–6 126, 127–8 Hambuger, Joseph 16n e-governance and 222, 224, 225, 226, happiness 229, 231 and quality of life 46 information and communication cross-national differences 34–5 technologies democratic politics and 33–42 e-governance xvii, 220–31 freedom and 46–57 see also hackers; internet human needs and 35 informational politics in rich and poor countries 34, 47–50, hackers and 119, 122–3, 126, 127–8 52–3 Institute for Democracy and Electoral party control and xiii, 38–41 Assistance (IDEA) 85–7, 189 providing for xi international assistance stated preferences xiii definition of international community unemployment and 47 190–1 see also satisfaction democracy assistance xvi, 190–1, 203 Hemings, Sally 15n development assistance xvii, 244, Heritage Foundation 47, 50, 52, 55–7 246, 253 Human Development Index 46 for political parties in third wave human equality 3–4, 5–6 democracies 188–203 human rights 15, 179, 191, 200 International Monetary Fund 56 Human Rights Foundation 174 international organizations 14, 15, 16n, Hungary 191 freedom and happiness 57 see also specific organizations political parties 176–7, 179, 180–2, internet 184, 186n and democratic values 15, 16n refugees from 140 and electronic civil disobedience 119, TV news 110 123–6 262 Index internet continued liberty and electronic direct action movements and equality xi, xii, 7–11 xv, 118–29 and satisfaction xiii and participation of citizens xii, xiv individual xiii and political inequality 103 literacy xv, 51, 106 and political knowledge xv Lithuania 57 as a news source 105, 107, 110, 111–13 livability of nations 46 hackers and hacktivism xv, 118–29 local government elections 61, 64–5, in USA 105, 110, 111, 112–13 82, 87–94, 96n Ireland Luxembourg 106–10 freedom and happiness 57 news media 106–10 Macedonia 183 privatization xvii, 213–14, 217 market, the irrigation systems xvii, 238–9, 245, and equality 12–13 250–2 and inequalities 35–7 Italy and unhappiness 35–7 freedom and happiness 57 capitalism xii, 12–13, 35–7 newspapers 102, 106–10 distribution 35–7 political parties 171, 176 state intervention in xiii-xiv social capital 81 Meciar, Vladimir 177, 180, 184 TV news 106–10 media malaise 100–4, 105–13, 114–15 Mexico 57 Japan 39, 57, 102 Milosevic, Slobodan 183 Jefferson, Thomas 4, 12, 15n Moldavia 56 justice Montenegro 183 and equality 19–21 moral virtue and solidarity 26–7 and levelling down 18–23, 32n multinationals 51 knowledge Native Americans 4, 5 and news media xv, 101 NATO 171 and political participation xv Nazis 9, 140 and voter turnout 83–4, 91–4 Nepal 238–9, 250–1 in USA 84–5, 95n, 101 Netherlands internet and xv e-governance 226 Kohl, Helmut 185n freedom and happiness 57 Kosovo 141 news media 106–10 voting 95n Latin America xi New Zealand Latvia 57 local elections 89 levelling down privatization of public services xvii, and justice 18–23, 32n 206, 209–11, 216–17 egalitarianism and xiii, 18, 23–4, referendums 132 32n rights and liberties of citizens 9 fairness and 27–31 voter turnout 89–90 moral virtue and solidarity 26–7 womens’ suffrage 5 preference satisfaction of the worst off news media 18–19 and civic engagement 100–1, 104, racial segration and 25–31 111–15 real-Pareto maxim 18–23, 25, 31 and public disengagement xv, 100–1, redistribution and 18–19, 29 104, 113, 114 social equality 23–5 Europe 102, 105–6, 113 well-being 22–3, 25, 27–31 internet 105, 107, 110, 111–13 Index 263

mass propaganda 101 knowledge and xv media malaise 100–4, 105–13, 114–15 levels of xi, xii newspaper circulation xv, 93, 105 new forms of xii newspaper reading xv, 92–3, 104, new technologies xii 105–8, 112–13 see also voter turnout radio 106–10, 112 personal freedom trends in 104–11 and happiness 47, 51 TV news xv, 101–5, 106–13 and state institutions xi USA 100–13 Philippines 57, 250–2 videomalaise 101 Phillips, Ann 31 virtuous circle in xv, 113–15 Poland newspapers freedom and happiness 57 and knowledge xv, 101 political parties 175, 176, 179, 180–1 and moral decline 100–1 TV news 110 circulation xv, 93, 105 political communications concentration of ownership 104–5 and civic disengagement xv, 100–1, press freedom xiii, 46, 47–8, 51, 52, 104, 113, 114 53, 56–7 and civic engagement 100–1, 104, reading xv, 92–3, 104, 105–8, 112–13 111–15 tabloidization 102 in USA 100–13 ‘yellow press’ 101, 102 internet 105, 107, 110, 111–13 Nigeria 57 media malaise 101–4, 105–13, Norman, Richard 23–5 114–15 Norway newspapers xv, 92–3, 100–1, 104–8, and European Union membership 112–13 171 professional marketing 103–4, 111 e-governance 223 radio 106–10, 112 freedom and happiness 57 trends in the news industry 104–11 rights and liberties of citizens 9 TV news xv, 101–4, 106–13 TV news 110 TV viewing 105 virtuous circle in 113–15 OECD countries see also internet newspapers 105, 106–7 political equality privatization 209 assumption of intrinsic equality 7 satisfaction in 39–40 barriers to 3–4, 12–13, 31 TV 105, 108, 110 changes toward 6 voter turnout xiv conflict with liberty xi, xii, 7–11 see also specific countries desirability of 6–14 exclusion from political participation Pareto improvements 7 and levelling down 27 feasibility of 3–6, 11–14 egalitarians and xiii, 18 market capitalism 12–13 ‘real-Pareto’ maxim 18–23, 25, 31 referendums and xv, 131 participation of citizens rights and xii, 9–11 and system effectiveness xii, 13–14 solutions 14–15 changing patterns of xiv system effectiveness versus citizen deliberative forms of democracy xii, participation 13–14 xv–xvi, 55 unequal distribution of resources 13 democratic decision-making xii, xiv, political freedom xv and happiness 46, 47–8, 52, 53, 56–7 in referendums xii, xv, 131–50 and satisfaction xiii internet xii, xv in poor nations 48, 52–3 264 Index political parties evaluation of 34 and happiness xiii, 38–41 happiness and 46 and policy regimes 38 indicators of 46 categorization of 43 objective 35 conditionality 170, 173–5 convergence 170, 173–7, 185 radical social movements development in Central and Eastern and hackers xv, 118–29 Europe xii, xvi, 169–85 animal rights xv, 118 European Union and 169–85 anti-capitalist xv, 118, 125, 127, 129 in ‘third wave’ democracies democracy ecological xv, 118 assistance xvi, 190–1, 203 electronic civil disobedience 119, election monitoring 201 123–6 international assistance 188–203 electronic direct action movements political intervention 190–1 xv, 118–29 transnational party cooperation xvi, environments 126–7, 129 170–85, 186n, 190 radio 106–10, 112 politics of envy 18, 31 Rae, Douglas 8 polycentric political systems xvii, racial segration 248–50, 253 and levelling down 25–31 poor nations redistribution and economic and political freedom and envy 29 48–50, 52–3 and levelling down 18–19, 29 and satisfaction xiii-xiv and satisfaction xi, xiii happiness in 34, 48, 49–50, 52–3 referendums Portugal and political equality xv, 131 freedom and happiness 57 and popular control xv, 131 news media 106–10 anti immigration referendums in political parties 172–3 Switzerland 137–50 privatization 209 consensus 132, 136–7, 150n press freedom equality of resources 132, 133–4 and happiness 47–8, 51, 52, 53, 56–7 exclusions 132–3 and satisfaction xiii, 46 information and 132, 134–5 privatization participation in Swiss referendums information and 208, 209–11, xii, 131–7 212–13, 216–18 regulatory regimes public sector 205–18 accountability 205–6, 216 public sector, privatization citizenship and 205–9, 216–18 accountability 205–6, 216 consumer sovereignty 206, citizens and 205–9, 216–18 207–8,210–11, 216–18 consumer sovereignty 206, 207–8, consumer transparency 206, 207–8, 210–11, 216–18 216–18 consumer transparency 206, 207–8, national diversity 205–6 218 public interest groups (PIG) 209 public interest group (PIG) 209 regulatory design 207–9 regulatory design 207–9 transparecny mechanisms 206, 207–8, traditional model 205 209–18 transparency mechanisms 206, 207–8, unified model 205, 216–18 209–18 religion 47, 183, 197 unified model 205, 216–18 rich nations and satisfaction xiii-xiv quality of life economic and political freedom in democratic politics and 33–4 48–50, 52–3, 54 elections and 33 happiness in 34, 47, 49–50, 52–3, 54 Index 265 rights xii, 8–11 Soares, Mario 172 Romania 5, 7, 179, 183 social capital Rousseau, Jean-Jacques 88 and voter turnout xiv-xv, 81–5, 91–4, 96n satisfaction definition of xiv and economic freedom xi, xiii social equality 23–5 and equality xiii South Africa 57, 197 and government intervention South-East Asia xi, 46 xiii-xiv South Korea 57, 110 and liberty xiii Soviet Union 12, 140, 171 and political freedom xiii Spain and press freedom xiii, 46 democratization of Spanish polity and redistribution xi, xiii 153 and welfare systems xi, xiii-xiv, freedom and happiness 57 37–41, 42–3n membership of NATO 171 consumption norm 34–5 news media 106–10 cross-national differences in 33–5, political parties 172–3 42n see also Barcelona cultural norm 34–5, 37–9, 41 Sri Lanka 140 democratic competition and xiii, suffrage 40–2, 51 expansion 13 democratic politics and 33–42 female xi, 4, 5 economic development and 34, universal xi, xiv, 5 37–9 Sweden human needs and 35 and Central and Eastern Europe 182 in European Community 37–9, 42n freedom and happiness 57 in OECD countries 39–40 news media 106–10 in poor nations xiii-xiv privatization xvii, 214–15, 217 in rich nations xiii-xiv satisfaction in 39 liberty and xiii voter turnout 96n nature of democracy and xiii Switzerland of the worst off 18–19 freedom and happiness 57 political intervention and 34–5 referendums political parties and 38–41 anti-immigration 137–50 study of life satisfaction 34–5 participation in xv, 131–7 unemployment and 37–9 rights and liberties of citizens 9 see also happiness satisfaction in 39 Scandinavia 4, 105, 107, 182 self-image 137–8, 140 see also specific countries voter turnout 89–91 Schwarzenbach, James 143 system effectiveness Scott, James 6, 15–16n and citizen participation xii, 13–14 Scottish Parliament 61, 220 segregation technologies, new and levelling down 25–31 see e-governance; hackers; internet Sen, Amartya 8, 46 ‘third wave’ democracies Serbia 183 democracy assistance xvi, 190–1, Sinn Fein 197 203 slavery 4, 15n electoral competition xvi Slovakia party systems 188–203 freedom and happiness 57 political reform xvi, 190 political parties 176–7, 179, 180–1, Tocqueville, Alexis de xii, 4–5, 7–9, 11, 184 83, 88 Slovenia 57, 179, 183 totalitarian regimes 9 266 Index trade unions 51 universal suffrage xi, xiv, 5 transnational organizations 14–15 Uruguay 9, 11, 57 see also specific organizations utilities, privatization of xvii, transnational party cooperation xvi, 205–16 170–85, 186n, 190 transparency mechanisms 206, 207–8, Vietnam 140 209–18 virtual cultures, defence of 119, 122, Tudjman, Franjo 179, 183 123 Turkey 57, 110 virtuous circle TV news xv, 101–4, 106–13 in the news media and political activism xv, 113–15 Uganda 192–3 voter turnout Ukraine 57 and social capital xiv-xv, 81–5, unelected chambers xi, 3, 5 91–4,96n unemployment 9, 37–9, 47 ‘civic community index’ 81 United Nations High Commission for civic engagement 61, 81–2, 86–9, 91, Refugees 190 92–4 United States of America civic literacy xiv-xv, 91–4 African Americans 4, 5, 6 compulsory voting 89–90, 95n agrarian republicanism 12 costs of non-voting 82–5 campaign activism 112–13, 115 in Britain xi, xiv, 61–77 capitalism 12 in mature democracies xii, 87 civic disengagement 100–1, 113 in OECD countries xiv civic engagement 104 in USA 81, 86–91, 95n, 96n civil rights 4, 5, 6 information 62–4, 71, 77, 82–4, 92 constitutional assembly xi interpersonal trust 82, 85–7, 91, 94 cynicism xi, 100–1, 104 knowledge and 83–4, 91–4 Declaration of Independence 3–4 local elections 82, 87–94, 96n deliberative polls 16n democracy assistance 191, 196, 199 wars 9, 172 equality 3–4 Weberian rationality xvii, 223 freedom and happiness 57 Weimar Republic 9, 11 hacking in 124–6 Weiss, Peter 176 inequality xi welfare systems influence in Italy 171 and satisfaction xi, xiii-xiv, 37–41, internet 105, 110, 111, 112–13 42–3n Maine Lobster Fishery 249–50 well-being Native Americans 4, 5 and levelling down 22–3, 25, 27–31 news media 100–4, 106–13 envy and 28–9 political ignorance 84–5, 95n, 101 fairness and 27–30 propositions on the ballot in markets and 35–7 California133–4, 135, 136–7 politics and 35–7, 40, 42 satisfaction in 39 social 33 slaves 4, 15n subjective 34, 35–7, 40, 42, 44n social capital 81, 83–4 Welsh Assembly 220 theories of media malaise 101–4 West Germany 170–1, 172–3 Trilateral Commission 101–2 women TV news 101–4, 106–13 feminists 155, 197 videomalaise 101 in participatory experiment in voter turnout 81, 86–91, 95n, 96n Barcelona 158, 162, 164 womens’ suffrage 4, 5 voting rights xi, 4, 5 Index 267

World Bank 56 Yugoslavia, former 140, 182–3 World Trade Organization (WTO) 118, 125, 127 Zambia 199 World Value Study xiii, 39–40 Zimbabwe 194