Social Capital and the Transition to Democracy

Social Capital and the Transition to Democracy

The concept of social capital has been used by political scientists to explain both the transition to democracy in Eastern Europe and the decline of social cohesion in Western societies. This edited volume presents the latest quantitative research on how post-communist countries are adapting to Western models of society. It discusses important questions such as: • • •

To what extent does social capital influence democratic transition in the former communist countries? How do levels of social capital in the region compare with those in Western Europe and the United States? What policies can governments pursue to promote the creation and maintenance of social capital?

Social Capital and the Transition to Democracy combines theoretical and institutional analysis with detailed case studies looking at Russia, Poland, Bulgaria, Lithuania, Romania and the former East Germany. It will interest students and researchers of social capital, democracy and postcommunist politics. Gabriel Badescu is Associate Professor of Political Science at the BabesBolyai University, Cluj-Napoca, Romania. Eric M. Uslaner is Professor of Government and Politics at the University of Maryland, College Park.

Routledge Studies of Societies in Transition 1 The Economics of Soviet Break-up Bert van Selm 2 Institutional Barriers to Economic Development Poland’s incomplete transition Edited by Jan Winiecki 3 The Polish Solidarity Movement Revolution, democracy and natural rights Arista Maria Cirtautas 4 Surviving Post-Socialism Local strategies and regional response in Eastern Europe and the Former Soviet Union Edited by Sue Bridger and Frances Pine

13 Experimenting with Democracy Regime change in the Balkans Geoffrey Pridham and Tom Gallagher 14 Poverty in Transition Economies Edited by Sandra Hutton and Gerry Redmond 15 Work, Employment and Transition Restructuring livelihoods in postcommunism Edited by Al Rainnie, Adrian Smith and Adam Swain 16 Environmental Problems of East Central Europe: 2nd Edition Edited by F. W. Carter and David Turnock

5 Land Reform in the Former Soviet Union and Eastern Europe Edited by Stephen Wegren

17 Transition Economies and Foreign Trade Jan Winiecki

6 Financial Reforms in Eastern Europe A policy model for Poland Kanhaya L. Gupta and Robert Lensink

18 Identity and Freedom Mapping nationalism and social criticism in twentieth century Lithuania Leonidas Donskis

7 The Political Economy of Transition Opportunities and limits of transformation Jozef van Brabant

19 Eastern Europe at the Turn of the Twenty-First Century A guide to the economies in transition Ian Jeffries

8 Privatizing the Land Rural political economy in postcommunist socialist societies Edited by Ivan Szelenyi

20 Social Capital and the Transition to Democracy Edited by Gabriel Badescu and Eric M. Uslaner

9 Ukraine State and nation building Taras Kuzio

21 The Former Yugoslavia at the Turn of the Twenty-First Century A guide to the economies in transition Ian Jeffries

10 Green Post-Communism? Environmental aid, innovation and evolutionary political economics Mikael Sandberg

22 The Former Soviet Union at the Turn of the Twenty-First Century A guide to the economies in transition Ian Jeffries

11 Organisational Change in PostCommunist Europe Management and transformation in the Czech Republic Ed Clark and Anna Soulsby

23 Federalism and the Dictatorship of Power in Russia Mikhail Stoliarov

12 Politics and Society in Poland Frances Millard

24 Elites and Democratic Development in Russia Edited by Anton Steen and Vladimir Gel’man

Social Capital and the Transition to Democracy

Edited by Gabriel Badescu and Eric M. Uslaner

First published 2003 by Routledge 11 New Fetter Lane, London EC4P 4EE Simultaneously published in the USA and Canada by Routledge 29 West 35th Street, New York, NY 10001 Routledge is an imprint of the Taylor & Francis Group

This edition published in the Taylor & Francis e-Library, 2004. © 2003 editorial matter and selection, Gabriel Badescu and Eric M. Uslaner; individual chapters, the contributors All rights reserved. No part of this book may be reprinted or reproduced or utilized in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging in Publication Data Social capital and the transition to democracy / edited by Gabriel Badescu and Eric M. Uslaner. p. cm. Includes bibliographical references and index. 1. Democratization—Europe, Eastern. 2. Democratization— Europe, Central. 3. Democratization—Former Soviet republics. 4. Social capital (Sociology)—Europe, Eastern. 5. Social capital (Sociology)—Europe, Central. 6. Social capital (Sociology)—Former Soviet republics. I. Badescu, Gabriel, 1969– II. Uslaner, Eric M. JN96 .A58 S63 2003 320 .947—dc21 2002153873

ISBN 0-203-42809-9 Master e-book ISBN

ISBN 0-203-43891-4 (Adobe eReader Format) ISBN 0–415–25814–6 (Print Edition)

Contents

List of illustrations Notes on contributors Acknowledgements 1 Introduction

vii xi xiv 1

GABRIEL BADESCU AND ERIC M. USLANER

2 Trust networks and democratic transition: Yugoslavia in the mid-1980s

10

HAJDEJA IGLICˇ

3 Communist societies, the velvet revolution, and weak ties: the case of East Germany

28

HENK FLAP AND BEATE VÖLKER

4 Social networks in Russian politics

46

NICOLAS HAYOZ AND VICTOR SERGEYEV

5 Social networks, civil society, and the prospects for consolidating Russia’s democratic transition

61

JAMES L. GIBSON

6 Trust and civic engagement in East and West

81

ERIC M. USLANER

7 Social capital, ethnicity and support for democracy in the post-communist states

95

KATHLEEN M. DOWLEY AND BRIAN D. SILVER

8 Social trust and democratization in the post-communist societies GABRIEL BADESCU

120

vi

Contents

9 Civic engagement in a post-communist state

140

JEFFERY J. MONDAK AND ADAM F. GEARING

10 Why post-communist citizens do not join voluntary organizations

165

MARC MORJÉ HOWARD

11 Social capital in Poland

184

JERZY BARTKOWSKI

12 What does social capital add to democratic values?

200

RICHARD ROSE AND CRAIG WELLER

13 Legacies and conflicts: the challenges to social capital in the democratic transition

219

ERIC M. USLANER AND GABRIEL BADESCU

Index

233

Illustrations

Figures 3.1 3.2

7.1 7.2

7.3

7.4

7.5 7.6

8.1

8.2 10.1 12.1 12.2

Typical pattern of personal networks in a communist society such as the former GDR Institutional embeddedness of relational investment and resulting consequences for personal networks in a communist society Differences in pride in country among forty-nine ethnic groups in twenty post-communist countries Differences in confidence in the legal system among forty-nine ethnic groups in twenty post-communist countries Differences in confidence in the Parliament among forty-nine ethnic groups in twenty post-communist countries Differences in satisfaction with government performance among forty-four ethnic groups in seventeen postcommunist countries Differences in support for democracy among forty-four ethnic groups in seventeen post-communist countries Differences in percentages of those who belong to any non-religious voluntary organisation among forty-nine ethnic groups in twenty post-communist countries Relationship between the proportion of people who ‘trust other people’ and the proportion of volunteer members in associations Relationship between the proportion of ethnic majority and odds ratio between trust and membership Average number of organizational memberships per person, by country Endorsement of democracy as an ideal Attitudes toward undemocratic alternatives

32

33 104

106

107

108 109

111

126 134 168 203 203

viii

Illustrations

Tables 2.1 2.2 3.1 3.2

3.3 3.4 3.5 5.1 5.2 5.3 5.4 5.5 5.6 6.1 6.2 6.3 7.1 7.2 7.3 7.4

7.5 8.1 8.2

Discussion networks in Serbia and Slovenia Strength of ties in Serbia and Slovenia Design of the study on changing personal networks in the former GDR before and after the political turn, 1989–94 Logistic random coefficient model for association of relational functions for the discussion of political matters and for provision ties in East Germany, 1989 Logistic random coefficient model on similarity of niche and provision relationships in East Germany, 1989 Patterns of triads within niches, between niches and weak ties, and outside niches Logistic random coefficient model for the discussion of political matters for 1994 data in East Germany Attributes of social networks Cross-national differences in interpersonal trust Different facets of interpersonal trust, Russia, 1998 The relationship between trust in strangers and trust in acquaintances, Russia, 1998 The impact of trust on support for democratic institutions and processes, Russia, 1998 The impact of network political capacity on support for democratic institutions and processes, Russia, 1996 Summary of trust and group membership in the West and former communist states Models for trust from two-stage least-squares estimations Models for group membership from two-stage least-squares estimations Freedom House rating of post-communist countries in the study Correlations between social capital indicators and Freedom House indicators of democratisation Correlations between social capital indicators and support for democratic institutions Product–moment correlations between social capital indicators and support for institutions and democracy in post-communist states Effects of social capital and ethnicity on support for the democratic system in post-communist countries Correlations between aggregate measures of generalized trust in three cross-national surveys Correlations between proportion of members in volunteer associations in three cross-national surveys

21 22 32

37 38 40 41 66–67 71 72 73 75 75 89 90 92 98 100 101

112 115 127 127

Illustrations ix 8.3

Proportions of volunteers in associations, as estimated by several surveys, representative for the adult population of Romania 8.4 Structural model with two latent variables for trust 8.5 Relationships between social trust and volunteer membership of associations 8.6 Correlations between social trust and trust in strangers across several categories of population, defined by the proportion of ethnic minorities and type of locality 9.1 Levels of interpersonal discussion 9.2 Political participation and media exposure 9.3 Retrospective economic perceptions 9.4 Accuracy and independence of neighborhood-level economic perceptions 9.5 Economic perceptions and presidential approval 10.1 Changes in levels of membership, 1995–97 to 1999: Russia, East Germany, and West Germany 10.2 OLS regression analysis of individual-level membership in voluntary organizations: Russia, East Germany, and West Germany, combined 10.3 OLS regression analysis of individual-level membership in voluntary organizations: Russia, East Germany, and West Germany, separately 10.4 Mistrust of communist organizations and organizational membership today 10.5 Changes in reliance on friends from the communist period to today, and organizational membership 10.6 Disappointment and organizational membership 10.7 OLS regression analysis of individual-level membership in voluntary organizations: Russia and East Germany, combined and separately 11.1 Trust in institutions 11.2 Corruption as social problem 11.3 Political alienation 11.4 Trade union membership 11.5 Perceived sources of wealth 11.6 Activity in associations according to public opinion surveys (1972–98) 11.7 Activity in associations (1975–99) 11.8 Activity at community level (1999) 11.9 Participation in community action 11.10 Frequency of personal contacts (relatives) 11.11 Frequency of personal contacts (non-relatives) 11.12 Trust in people

128 131 132

135 147 149 152 153 155 169

171

172 174 175 176

178 186 186 188 189 190 191 192 194 195 196 196 197

x

Illustrations

11.13 12.1 12.2 12.3 12.4 13.1

Trust in different social groups Impact of human capital, economic and civic resources Impact of social capital on democratic values Integrating influences on democratic values Impact of significant influences on democratic values Generational differences in social capital in Romania and Moldova from CID surveys of Romania and Moldova

197 208 209 211 213 229

Contributors

Gabriel Badescu is Associate Professor of Political Science at the BabesBolyai University, Cluj-Napoca, Romania, where he has taught since 1996. His publications include Social Capital and Political Participation in Romania and Roma Population in the 2000 Romanian Local Elections. Jerzy Bartkowski is a political sociologist and teaches at the University of Warsaw. He has published two books on local elites in Poland (in Polish) and has just finished a book on the impact of tradition on political and social behavior in Poland. Kathleen M. Dowley is Assistant Professor of Political Science at SUNY New Paltz. She has published articles on the relationship between subnational and national loyalty, ethnic mobilization in democratic transitions and the measurement of political culture in multiethnic states. As a 2002 recipient of an American Political Science Association’s Small Research Grant, she is currently working on local government reform in East-central Europe. Henk Flap is Professor of Sociology in the Department of Sociology and the ICS Research School at Utrecht University. His research focuses on the development and testing of social capital theory in different organizational and institutional contexts. Professor Flap’s recent publications include ‘No man is an island’, in Conventions and Structures, edited by Favereau and Lazega, ‘Assortative meeting and mating: unintended consequences of organized settings for partner choices’, coauthored with Matthijs Kalmijn, in Social Forces, and the edited volume Creation and Returns of Social Capital, together with Beate Völker. He is currently engaged in a historical study of the organization of the persecution of the Dutch Jews during the Second World War. Adam F. Gearing teaches history in the City Honors School, Buffalo, New York. He holds a graduate degree in international affairs from Georgetown University.

xii

Contributors

James L. Gibson joined the Political Science Department, Washington University in St Louis, as Sidney W. Souers Professor of Government in 1999. He studies mass behavior and democratization in the United States, Europe and Africa, and is interested in understanding why people think the way they do about political issues (especially political tolerance), and how such thinking translates into public policy and democratic reform. His numerous articles have appeared in a wide range of national and international social-scientific journals, including all the leading political science journals, and he has also written five books, including the recently published Overcoming Intolerance in South Africa: Experiments in Democratic Persuasion (with Amanda Gouws). Nicolas Hayoz is Associate Professor of Political Science and Director of the Interfaculty Institute of East and Central European Studies at the University of Fribourg (Switzerland). He has published articles on politics and state reform in the Former Soviet Union and Russia. His research interests include state reforms, good governance and legal and judicial reforms in Eastern and Central Europe. Marc Morjé Howard is an Assistant Professor in the Department of Government and Politics at the University of Maryland. His research focuses on a variety of topics related to democracy and democratization, including civil society, citizenship, right-wing extremism, ethnicity, democratic performance and public opinion. His chapter is drawn from his recent book, The Weakness of Civil Society in Post-Communist Europe. Hajdeja Iglicˇ is Associate Professor of Sociology at the University of Ljubljana, and at the Institute of Social Sciences in Ljubljana. Her Ph.D. is from Columbia University (2000). She researches social networks, social capital and political participation, and has also published on elite networks and democratic transition. Jeffery J. Mondak is LeRoy Collins Professor of Political Science, Florida State University. He is the author of Nothing to Read: Newspapers and Elections in a Social Experiment, and articles in the American Journal of Political Science, the British Journal of Political Science, the Journal of Politics, and several other journals. Richard Rose created the New Democracies Barometer, monitoring mass response to post-communist transformation (see www.cspp.strath.ac.uk). His latest book in the field is Elections without Order: Russia’s Challenge to Vladimir Putin (with Neil Munro). He is Professor of Public Policy at the University of Strathclyde, Glasgow, and has authored hundreds of articles and books translated into eighteen languages. Victor M. Sergeyev is Professor of Comparative Politics and Director of the Centre for International Studies, MGIMO-University of the Min-

Contributors xiii istry for Foreign Affairs of the Russian Federation (Moscow). He is the author of many books and articles, including Russia’s Road to Democracy (with N. Biryukov), Russian Politics in Transition (with N. Biryukov) and The Wild East. His current interests include comparative studies of political culture, globalization processes in international relations and the analysis of decision-making. Brian D. Silver is Professor of Political Science at Michigan State University (East Lansing, MI), where he is also Director of the State of the State Survey. He specializes in the comparative politics of the postcommunist states, with particular interest in ethnic differences in political attitudes and political mobilization as well as in demographic change. He has published extensively in the leading journals in political science, demography and Eurasian studies. He is currently engaged in a national survey assessing American public opinion on the trade-offs between security and civil liberties in the wake of the terrorist attack on America on 11 September 2001. Eric M. Uslaner is Professor of Government and Politics at the University of Maryland-College Park, where he has taught since 1975. He has written seven books including The Moral Foundations of Trust, The Decline of Comity in Congress, and The Movers and the Shirkers: Representatives and Ideologues in the Senate, and over fifty articles. Uslaner’s edited books include Social Capital and Participation in Everyday Life (with Paul Dekker). He is currently conducting research on the linkage between trust, inequality and civic engagement in the United States under a grant from the Russell Sage Foundation in addition to his work on the grant from the Starr Foundation to IREX on ‘Civil Society and Development on the Black Sea: Social Involvement in Romania and the Republic of Moldova’ (with Gabriel Badescu and Paul Sum). Beate Völker is Associate Professor in the Department of Sociology, Utrecht University, the Netherlands. Her research focuses on the impact of social institutions and settings on people’s personal networks and relationships and is currently directed at the influence of social contexts and places such as the neighborhood and the place where one works for the composition of people’s networks and relationships. In addition, she is interested in the consequences of these settings and networks for individual behavior (e.g. for solidarity behavior). Together with Henk Flap she conducted a national representative survey on the personal networks of the Dutch. She has published her work in Social Networks, Rationality and Society, Acta Sociologica and the European Journal of Sociology, among others. Craig Weller, who has studied Russian politics at Cornell, the University of London and Oxford, was formerly a research fellow at the Centre for the Study of Public Policy, University of Strathclyde, Glasgow.

Acknowledgements

We are grateful to the Starr Foundation, through the Institutional Research and Exchanges Board of the United States Department of State, for a grant to us, in conjunction with Paul Sum of the University of North Dakota, and Mihai Pisica and Cosmin Marian of Babes-Bolyai University under the IREX Caspian and Black Sea Collaborative Program (2001). Eric Uslaner is also grateful to the Russell Sage Foundation and the Carnegie Foundation for a grant under the Russell Sage Program on The Social Dimensions of Inequality (see http://www.russellsage.org/ programs/proj_reviews/social-inequality.htm) and to the General Research Board of the Graduate School of the University of Maryland-College Park. Gabriel Badescu’s research for this volume was supported in part through the Blue Bird Project, which is financed by a consortium of international donors and administered by the Central European University. Specifically, contributions from the international donor, Bank of Sweden Tercentenary Foundation, were used to fund his work. We are particularly grateful for the helpful comments of Marc Morjé Howard, Paul Sum and Rita Moore. The editors and publishers would also like to thank the following publishers for their permission to reprint text: Blackwell Publishers for permission to reprint ‘Civic Engagement in a Post-communist State’, originally published in Political Psychology, 19 (3) (September 1998): 615–37. Sage Publications for permission to reprint a shortened and revised version of ‘Communist Societies, the Velvet Revolution, and Weak Ties’, originally published in Rationality and Society, 13 (4) (November 2001): 397–429. Taylor & Francis (www.tandf.co.uk) for permission to reprint ‘Social Capital, Ethnicity and Support for Democracy in the Post-Communist States’, originally published in Europe-Asia Studies 54 (June 2002): 505–27. ‘Social Networks, Civil Society, and the Prospects for Consolidating Russia’s Democratic Transition’ was originally published in the American Journal of Political Science 45 (2001): 51–69.

1

Introduction Gabriel Badescu and Eric M. Uslaner

The transition to democracy among the formerly communist nations of Central and Eastern Europe has been both slow and uneven. All of these states now have democratic constitutions and institutions but some have made ‘democracy work’ (Putnam 1993) better than others. Slovenia has become ‘Westernized’ far more quickly than Russia, Romania, and even Poland and Hungary. Even in the most ‘democratic’ of these countries – the former East Germany, which has been reunified with the Federal Republic of Germany – citizens are slow to become participants. The Western model of democracy, which posits a trusting and active citizenry, seems far away from the countries making the transition from communism. In this volume we seek to explain why democratic attitudes and participation lag in Central and eastern Europe – and why some countries fare better than others. We thus seek to explain why there is a lower level of social capital in the formerly communist countries and what the prospects are for a civic culture in the future. The fall of the communist regimes in Eastern Europe and the Soviet Union from 1989 to 1990, along with the change of the political landscape in other regions of the world in the 1970s and 1980s, were considered by many observers to be part of a global democratic trend. This movement has become widely known as a ‘third wave of democracy’ (Huntington 1991). Despite many variations, these political changes began simultaneously in several countries in each region. All shared a turn from dictatorial rule toward more democratic governance. This trend led to the assumption that the underlying conditions, such as the economic level, institutional legacies and socio-cultural traditions are not major factors in either the onset or the outcome of the transition process. The expectation that political changes will follow a steady and convergent path towards democratization, albeit perhaps at a different pace, was even more apparent in the case of the former communist countries. In the more than twenty years that have passed since the beginning of the third wave, it is obvious that only a few of the transitional countries have achieved a relatively well-functioning democracy, and that most of the others do not seem to be deepening their level of democratization

2

Gabriel Badescu and Eric M. Uslaner

(Carothers 2002). The former communist countries show a slightly better record, but also a widening gap between the Central and European countries on one side and the former Soviet republics, excluding the Baltics, on the other (Karatnycky 2001). Some countries, such as Hungary and Poland, are considered to be significantly more democratic than others. Even those scholars who consider that the differences among postcommunist countries are of secondary importance in comparison with commonality and who are also very critical of the course of the reform progress recognize the differences (Poznanski 2002: 215). In analyses of why differences exist among the former communist countries with respect to their level of democratization, as well as why initial expectations about region as a whole have not been fulfilled, several categories of potentially influent conditions have been identified. These include the institutional design and type of leadership characteristic of the ancien régime, the identity and degree of control exercised by those who initiate the transition, the foreign policies of neighboring polities, the type and outcome of economic processes and cultural variables. Little agreement exists on how large the impact of each of these factors is and which are relevant for democratization. Perhaps the most disputed issue is the existence of relationships between cultural variables and democratization, and the identification of the mechanisms through which the two categories influence each other. The general aim of this volume is to clarify if democratic transitions in the former communist countries have been significantly affected by certain cultural attributes. The idea that culture is linked to democracy had great impact during the 1960s and 1970s, following the publication of Almond and Verba’s The Civic Culture (1963). Later, the interest in a cultural approach suffered a serious loss of adherents,1 and had reached an all-time low by the beginning of the 1990s, but the fall of the European communist regimes brought strong arguments in favor of revitalizing the field. First, the concept of civil society, which is associated closely with that of civic culture, has become a major objective of study, largely as a result of the developments surrounding the collapse of communism. Although employed in studies on Latin America and Southern Europe in the 1980s, the idea of a civil society achieved recognition largely as a result of being used as an interpretation of the decline and breakdown of the former soviet block (Di Palma 1991; Tismaˇneanu 2001). Second, it has been easier to include cultural factors in empirical analyses of democracy as cross-national databases become available. In addition to the three waves of the World Values Survey, which now cover sixty-five societies containing 75 percent of the world’s population (Inglehart 2000: 82), there are several other comparative surveys assessing various cultural factors, and conducted in both developing and consolidated democracies. Third, the notion of social capital, also a concept referring to cultural phenomena and linked closely to that of civil society (e.g. van Deth 2001),

Introduction 3 has attracted enormous attention during the past ten years. Generally understood as referring to values, norms and social networks that influence the capacity of people to cooperate, social capital has been used in several disciplines for many years (Feldman and Assaf 1999). However, it became one of the most frequently used concepts in the social sciences only after the publication of Making Democracy Work (Putnam 1993). Putnam’s book succeeded in renewing the research program of political culture studies, due in part to the fact that its demonstration that democratic institutions rest on foundations of civic community is a serious alternative to theories based on social alliances, economic development, and political strategies (Laitin 1995: 172). The contributions in this volume seek to identify the extent to which social capital influences democratic transition in the former communist countries. Robert Putnam suggests that these countries enjoy little social capital, with dire consequences for the quality of governance and democratization. To date, however, few studies have examined the levels of social capital in the region and their impact. How do levels of social capital in the region compare with those in Western Europe and the United States? Are there differences among post-communist countries? How can we explain any differences and what is their impact on transition? Has social capital increased or decreased since the fall of communism? Does marketization create tensions that might hinder the development of social capital? What policies can governments pursue to promote the creation and maintenance of social capital? The chapters in this volume will address these questions and others. Some will be broadly theoretical, attempting to make the necessary distinctions among different dimensions of social capital and their corresponding measures. Some will make comparisons between former communist countries and Western democracies. Others will be comparative within the former communist countries.

Social capital and democratization While discourse centered on the term social capital has recently become more prevalent among academics, policy-makers, journalists and even the general public, the imprecision and lack of convergence in definitions of the main concepts structuring this discourse have increased: Social capital is in danger of going the way of political culture – a potentially powerful concept that is given many different meanings by many different people for many different purposes. (Newton 1997: 575) For some authors, social capital is ‘nothing more than an elegant term to call attention to the possible individual and family benefits of sociability’

4

Gabriel Badescu and Eric M. Uslaner

that is also compatible with ‘a nuanced understanding of the pros and cons of groups and communities’ (Portes and Landolt 1996: 21). Others argue that studies which use the concept of social capital should try to avoid any debate over definitions and focus more on achieving progress in terms of empirical inquiry. In their view, what should unite the work on social capital is not ‘their commitment to advocacy of a common theoretical missive or theoretical agenda, but rather their common emphasis on objective, cautious analysis’ (Mondak 1998: 433). Most of the works on social capital follow conceptualizations of James Coleman and Robert Putnam by considering that social capital refers to aspects of social structure that facilitate certain actions of actors. Putnam’s distinction among three types of entities that comprise social capital is especially influential: networks, norms and trust, which help to solve collective action problems (1993: 36). Many recent definitions and applications of the concept, however, depart from Putnam’s original assumption that the various types of social capital are closely interrelated, being part of a unidimensional syndrome.2 The difference is that the existence of a strong relationship between social networks, norms and trust is not taken for granted but, instead, is subject to empirical investigation. Moreover, dimensionality of each of the three sub-concepts of social capital is examined, as well as their links to other notions. A consequence of this perspective is that the assessment of the relationship between social capital and democratization involves distinct analyses for each separate type of social capital. The contributions in this volume follow this approach, each author making a clear specification of the type of social capital on which he or she has focused. Together, they present a comprehensive list of forms of social capital that are analyzed in their relationship to aspects of democratic processes. There are different kinds of trust, as well as different types of social capital norms and networks that are considered from both an individual and aggregate level perspective.

The chapters in this book This volume relies on the basic assumption that any democracy needs a minimum level of civic engagement in order to function. Its chapters provide a fresh and detailed account of the level of political participation, membership in volunteer associations and political support within the former communist countries. Not only does it challenge the common view that East European citizens are characterized by apathy towards political processes and that this trend is increasing, but some contributions also provide a detailed account of various forms of democratic engagement, their prerequisites and their effects on democratic processes. Special attention has been given to the case of citizen activism that has the potential to produce undemocratic outcomes, such as that of increasing intolerance

Introduction 5 among ethnic groups. The causal effects between social capital factors and forms of civic engagement are explored throughout this volume, within a variety of contexts, and by using different types of data and methodological tools. Chapters 2 to 5 concentrate on the relationships between certain types of social networks as one main component of social capital, and democratic transition. Hajdeja Iglicˇ (Chapter 2) analyzes how personal networks affect democratization by influencing the types of political mobilization and political networks. She does so by comparing survey data on personal networks of people in different regions of the former Yugoslavia during the mid-1980s. Her study concentrates on trust networks, understood as patterns of social integration which are established through routine participation in various social settings, and which are followed when people become members of political networks. She hypothesizes that trust networks explain not only differences in quantity of contentious action but also in quality, by distinguishing among community, social, and elite-sponsored actions. Chapter 3, written by Henk Flap and Beate Völker, contributes further to the idea that personal networks represent a significant form of social capital for democratic transition. Their study focuses on East Germany, a society where democratic transition has been one of the most rapid and successful among the former communist countries. The research design consists of a three-wave panel survey that allows a longitudinal perspective on social ties of individuals, covering the period between 1989 and 1994. The main hypothesis is that some of the attributes of the personal networks which the social capital theory contends are relevant for social, economic and political processes have been significantly influenced by the institutional change that has followed the collapse of communist regimes. Nicolas Hayoz and Victor Sergeyev’s contribution (Chapter 4) is based on a similar theoretical approach, but is applied to the case of Russia, generally considered as a country with one of the least successful transitions. The two authors distinguish among several types of networks and relations of trust between and inside state and society. This conceptual framework is applied to a number of concrete examples, and aims to establish how particularistic reciprocity within personal networks could be transformed in networks of generalized reciprocity and civic engagement. The relationships between social networks and democratization in Russia are analyzed further in James Gibson’s contribution (Chapter 5). First, Gibson evaluates the claim that Russian social networks lack most of the characteristics that are required for the development of a vibrant civil society by making use of a seven-country comparative survey. He examines the structure of social networks across mass publics in Bulgaria, Hungary, Poland, France, Spain, the United States and Russia. Most crucially, he examines the types of social connections in Russia, using a panel survey of the public taken in 1996 and 1998. Russians, he argues, have

6

Gabriel Badescu and Eric M. Uslaner

strong rather than weak ties, but these bonds are often politically relevant and thus increase support for processes of democracy and nourish democratic attitudes. The main focus within the next chapters is on the causal mechanisms between trust, civic engagement, and democratic processes. Eric M. Uslaner (Chapter 6) asks why both generalized trust and civic engagement are lower in the former communist countries than in the West. By using cross-national data of the 1990 to 1995 World Values Surveys, Uslaner examines possible similarities between the roots of trust and civic activism, or if there is a different individual level psychology in the West and in the former communist countries. His main hypotheses are that generalized trust depends upon a sense of optimism about the future and that this causal link is applicable to former communist countries as it is to the West. In addition, Uslaner discusses the possible links between the increase of economic inequality and the low levels of generalized trust in the transitional countries. Using Western data, he found this link to be significant. Is the relationship between trust and equality the same in post-communist regimes or is it different because past equality was enforced by the state and was not the product of normal social interactions and market forces? In Kathleen Dowley and Brian Silver’s contribution (Chapter 7), trust and civic engagement are key concepts as well, but emphasis is placed on their explanatory potential for democratic support. By making use of the 1995 to 1997 World Values Surveys, the two authors investigate the link between several indicators of social capital, including generalized trust and the level of support for political institutions and, more generally, commitment to the principles of democracy. They not only examine if correlations exist, but also analyze the causal mechanisms that could produce associations at the individual level. An important question they ask is: Does social capital function in the same manner in post-communist Europe as in Western countries, or are its usual markers more likely to indicate the ethnic polarization of society? Gabriel Badescu (Chapter 8) explores further the issue of measurement and analyzes how some of the standard indicators of generalized trust and civic engagement function when applied to different national contexts. He asks if the type of volunteer associations that exist in a country and the distribution of ethnic minorities influence the validity of trust and civic engagement indicators. His main hypothesis, tested on the 1999 European Values Surveys and on several surveys representative for the Romanian population, is that when measurement problems are taken into account, a high level of generalized trust is linked to increased involvement in associational life, and that this result holds in post-communist countries characterized by different degrees of success in democratization, as well as in Western countries. Jeffery Mondak and Adam Gearing (Chapter 9) investigate both the determinants and effects of civic engagement by using a distinct methodo-

Introduction 7 logical strategy. They analyze data from surveys conducted in a city from a post-communist society, Romania, and in a city from a consolidated democracy, the United States. This research design enabled them to explore the links between individual responses, including indicators of civic activism, and community-level factors. Their comparative inquiry examines three interrelated questions. First, they ask whether community-level civic engagement in post-communist Europe is affected adversely by characteristics of the local political context. Second, they examine whether the quality of people’s social perceptions suffers when civic engagement is limited. Third, they explore the link between civic engagement and attitudes toward politics. Marc Howard (Chapter 10) extends the examination of what influences the level of civic engagement to the case of two other countries, Russia and Germany. In addition to the secondary analysis of the 1995 to 1998 World Values Surveys, Howard makes use of the Post-Communist Organizational Membership Study survey, a specially designed survey conducted in the two countries. In addition to the general socio-economic factors, he tests the effect of several factors on civic engagement that describes consequences of the prior communist experience. Howard hypothesizes that the legacy of mistrust of formal organizations, the persistence of private friendship networks and general assessments of the post-communist period are significant predictors of membership in volunteer organizations. More generally, he asserts that attitudinal norms and behavioral patterns can have a powerful effect that lasts beyond the institutional context within which they first arose, and traces some of the possible consequences for the post-communist transition. The remaining chapters offer broad views of social capital forms and discuss how some of these are linked to transition. Jerzy Bartkowski (Chapter 11) proposes a comprehensive view of social capital in Poland, an East European nation characterized by one of the highest levels of civic activism under communism and by a relatively successful democratic transition. He presents the dynamics of different types of social networks and trust, and discusses how they are related to the post-communist social and economic transformations. Bartkowski contends that the low level of social capital in Poland is largely a consequence of the communist legacy but also the result of the fast institutional transformations of a traditional society. Richard Rose and Craig Weller (Chapter 12) conduct a similarly encompassing analysis, applied to the case of Russia, of the relationship between social capital and democratization. They ask to what extent social capital influences democratic values of citizens. More specifically, the influence of trust, involvement in political networks and social integration on support for democratic values is assessed when human capital, economic resources and civic attitudes are taken into account. In answering these questions, the two authors use data of a specially designed survey,

8

Gabriel Badescu and Eric M. Uslaner

the 1998 New Russia Barometer, containing measures of several forms of social capital, including those that are specific to politics. Finally, Eric M. Uslaner and Gabriel Badescu’s contribution (Chapter 13) is a review of the major themes and debates in the literature of social capital, applied to the case of democratic transition of East European nations. They begin with a discussion of why evidence on post-communist transition challenges some of the main claims of social capital theory: (1) democratic regimes stimulate participation, (2) civic engagement increases generalized trust, and (3) democratization breeds trust. The authors argue, using results presented in the previous chapters in this volume, as well as their own analyses on cross-national and national survey data, that the legacy of communism, with regard to widespread distrust and civic disengagement, is still present, and is slow to overcome. On the other hand, they show that what separates East and West is largely the experience people had under communism, not psychology; therefore policies which aim to shape individual experiences so as to increase trust and civic engagement are possible. Written around the common theme of the relationships between social capital factors and democratic processes, these chapters contribute collectively to our understanding of how democracy works in the former communist countries and what are the missing ingredients for their transition to democracy to be complete.

Notes 1 ‘In light of the career choices of the political scientific community, work on civic culture today has an aura of working out anomalies in Ptolemaic astronomy’ (Laitin 1995: 169). 2 ‘Our discussion of trends in social connectedness and civic engagement has tacitly assumed that all the forms of social capital that we have discussed are themselves coherently correlated across individuals. This is in fact true. Members of associations are much more likely than nonmembers to participate in politics, to spend time with neighbors, to express social trust and so on’ (Putnam 1995: 73).

Bibliography Almond, G. A. and Verba, S. (1963) The Civic Culture. Princeton, NJ: Princeton University Press. Carothers, T. (2002) ‘The End of the Transition’, Journal of Democracy 13, 1: 5–21. Di Palma, G. (1991) ‘Legitimation from the Top to Civil Society: Politico-Cultural Change in Eastern Europe’, World Politics 44, 1 (October): 49–80. Feldman, T. R. and Assaf, S. (1999) ‘Social Capital: Conceptual Frameworks and Empirical Evidence. An Annotated Bibliography’. The World Bank Social Capital Initiative Working Paper No. 5. Huntington, S. (1991) The Third Wave. Democratization in the Late Twentieth Century. Norman: University of Oklahoma Press.

Introduction 9 Inglehart, R. (2000) ‘Culture and Democracy’ in L. E. Harrison and S. P. Huntington (eds) Culture Matters. How Values Shape Human Progress. New York: Basic Books. Karatnycky, A. (2001) ‘Nations in Transit: Emerging Dynamics’, available at http://www.freedomhouse.org/research/nitransit/2001/cover-materials/essay1.htm. Laitin, D. (1995) ‘The Civic Culture at 30’, American Political Science Review 89, 1: 168–73. Mondak, J. (1998) ‘Editor’s Introduction to the Special Issue: Psychological Approaches to Social Capital’, Political Psychology 19, 3 (September): 433–41. Newton, K. (1997) ‘Social Capital and Democracy’, American Behavioral Scientist 40: 575–86. Portes, A. and Landolt, P. (1996) ‘The Downside of Social Capital’, The American Prospect 26 (May–June): 18–21. Poznanski, K. (2002) ‘Transition and Its Dissenters: An Introduction’, East European Politics and Societies 15, 2: 207–20. Putnam, R. (1993) Making Democracy Work: Civic Traditions in Modern Italy. Princeton, NJ: Princeton University Press. Putnam, R. (1995) ‘Bowling Alone: America’s Declining Social Capital’, Journal of Democracy 6 (January): 65–78. Tismaˇneanu, V. (2001) ‘Civil Society, Pluralism, and the Future of East and Central Europe’, Social Research 68, 4 (winter): 979–91. van Deth, J. (2001) ‘The Proof of the Pudding: Social Capital, Democracy and Citizenship’. Paper presented at the European Science Foundation EURESCO Conference on Social Capital: Interdisciplinary Perspectives at the University of Exeter, 15–20 September.

2

Trust networks and democratic transition Yugoslavia in the mid-1980s Hajdeja Iglicˇ

Political uncertainty and the repressive nature of the state apparatus under the conditions of communism, combined with economic insecurity and competition for limited material goods, made trust a very scarce resource in personal relations in Central and Eastern Europe.1 It is a widely accepted thesis today that those whose lives are more insecure can afford to trust less, since for them betrayed trust is relatively more consequential (Inglehart 1999; Offe 1999; Uslaner 1999). The lack of trust in East European societies was seen as the main obstacle to mass democratic resistance. But contrary to such expectations, people proved to be capable of putting considerable pressure on their states. The visible activism of East European populations has been explained in terms of strong personal ties. Opp and Gern (1993) argue that while distrust was pervasive in the communist regime at large, small islands of trustful ties among friends and colleagues could be found which served as mobilization vehicles. This chapter extends the argument about the crucial role of microintegration for the East European democratic resistance, suggesting that what matters for political mobilization is the type of personal networks in which people are embedded rather than the presence or absence of particular social ties. We thus change the focus from the existence of certain ties in one’s network that serve as a vehicle for mobilization, to whole personal networks and trust potential in these networks. We talk about trust networks rather than recruitment networks. Trust potential is a crucial network characteristic that affects the process of political mobilization by constraining the formation of new political ties with previously unknown people (‘strangers’) and the creation of broader political identities. I also propose an alternative view of the political mobilization of populations in different parts of Eastern and Central Europe, and argue that personal networks, with their trust potential and sociability style, affect not only the capacity of society for collective action, but also the form and content of political networks and political mobilization. This is what Calhoun (1994) calls the ‘social fabric’ of political action. The term refers to ties, networks, and collective identities which emerge around collective action and may consist either of pre-existing social groups which take on a

Yugoslavia in the mid-1980s

11

political role and identity, newly formed and purpose-specific political groups, or mass publics that are integrated by means of intense identity politics. It is crucial to study the character of political networks since the success of democratic transition is a result of both the intensity of popular mobilization and the way politics is done. As suggested by Calhoun (1993), in the democratic transitions in Eastern Europe, nationalist ideology was stronger and more exclusive, and consolidation of democracy less likely in those countries which did not experience decentralized and autonomous mobilization from below, and the articulation of multiple collective identities and diverse political networks. The explanation for differences in political networks is usually sought in the repressive capacity of the state. The argument runs as follows: the more repressive the state, the less likely the emergence of autonomous and multiple political networks within society, and the more likely an elite-led mobilization and creation of an overarching collective identity. In this chapter, I argue that in addition to the opportunities framed by the state, one has to look also at the capacity of society to create diverse political networks and multiple collective identities. I show that this capacity depends very much on social networks and the way people form and maintain social ties prior to political mobilization. In the empirical analysis I look at personal networks of the Yugoslavs in the mid-1980s. This was a high period of contentious action that took very different forms, from political protests and demonstrations, to strikes, new social movements, petitions and so on. Yet, despite the generally high level of political mobilization, contentious action was carried out in a form that was conducive to democracy in only some of the Yugoslav regions. As a result, democratic transition was not successful in all parts of Yugoslavia. When the country broke up in the early 1990s after a decade of intense mass politics, its constituent parts found themselves on very different paths of political development: from potentially stable democracies in the northern part, to the authoritarian regime in the central part, and unstable democracy in the south. The case of Yugoslavia thus allows us to look at both successful and unsuccessful cases of democratic transition, and the role of social and political networks in this process. I do this by comparing personal networks of people in different regions of the former Yugoslavia, and show how these networks allowed for different types of political mobilization and political networks, and consequently, for different outcomes – successful and unsuccessful democratization. I begin by describing political mobilization in different parts of the former Yugoslavia with a focus on the kind of political networks and collective identities that emerged in this process. I continue with a more general discussion of the relationship between social and political networks, and propose that different network configurations may be best for different kinds of mobilization (elite-sponsored, social and community mobilization). I end with the empirical analysis of personal networks in

12 Hajdeja Iglicˇ different regions of Yugoslavia in 1987, in order to see to what extent we can explain the variation in the types of mobilization and political networks with personal networks and trust potential embedded in these networks.

Political mobilization in Yugoslavia in the mid-1980s Pluralization of political life in the former Yugoslavia in the 1980s resulted in the formation of new political subjects outside the institutional channels (new social movements and committees), the formation of new political organizations, and the politicization of existing professional associations. Many new forms of political action emerged which had not been practised up until then: civic initiatives, petitions, rallies and marches. In addition, the number of strikes increased dramatically from the second half of 1985. The rise of informal political networks with new subjects, repertoires of action, organizational structures and political language was accompanied by decreasing membership in and identification with socialist political institutions. However, this pluralization of political life came about in very different ways and led to the diversification of the Yugoslav political space. Regions became characterized not only by the stronger or weaker presence of noninstitutionalized political networks, but also by their different political styles, languages and priorities. In the northern regions of the former Yugoslavia, the politicization of the population followed the trends started in Western Europe at the beginning of the 1980s which included the emergence of new social movements and alternative cultures, a decline in membership in political parties and trade unions, and the rise of new political subjects within civil society. In the central and southern regions, political mobilization took a more ‘traditional’ road. Contentious activity was carried out in the form of ‘en bloc’ protests. The social basis of such protests was region specific: in the central region, Serbia, these were deliberately called street rallies which attracted people from all over Serbia. In southern regions, in particular Macedonia and Kosovo, protests were more likely to be grounded in the collectivities of routine social life, such as work communities and villages. In this chapter I focus in particular on differences in the political activities in two republics of the former Yugoslavia: Slovenia and Serbia. The reason is that these two republics experienced very high levels of nationwide mobilization.2 Serbia: mass rallies and populist political leadership Political processes in Serbia in the second half of the 1980s, starting with the dismissal of Ivan Stambolic´ from the position of President of Serbia in December 1987 and the rise of Slobodan Milosˇevic´, are well described in numerous monographs which deal with Yugoslavia’s disintegration. Changes within the political elites were accompanied by very strong nation-

Yugoslavia in the mid-1980s

13

alist mobilization. The most characteristic form of this mobilization was massive street rallies. Participants in the rallies demanded the subordination of Kosovo – a ‘cradle of the Serbian nation’ – to the republic of Serbia.3 Serbia’s leadership succeeded in mobilizing the masses on nationalist grounds and threatened the leadership of other republics with the ‘power of the street’. The threat of street rallies was real. Throughout Yugoslavia’s biggest republic, Serbs turned out by the million for rallies. In the summer of 1988 rallies took place in almost every major city in Serbia. When the rallies spread to other parts of Yugoslavia, participants demanded the resignation of republican and provincial leaderships. Changes in the leadership of Montenegro and Vojvodina were direct consequences of such rallies. In both cases the old leaders resigned under pressure a day after the rallies took place, and Milosˇevic´ appointees replaced them. The leadership of mass rallies was assumed by a small group of ten to twelve individuals, mostly retired military and police officers, and local officials of the Communist Party. Although they coordinated their activities with the political elite, they retained a degree of autonomy. Members of the Committee for Protest Rallies from Kosovo Polje planned the place and time of the rallies and appeared as spokespersons at most of the events (Doderovic´ 1990). The actual organization of rallies was in the hands of local militants, in most cases local party officials. Mass meetings were well organized and financed.4 The organizers of rallies brought large numbers of participants from towns and cities all over Serbia, as well as from other regions of Yugoslavia that contained significant Serbian minorities. Many reporters describe how people were recruited from their workplaces and driven to the rallies by bus (Telegraph, November 1994). As a result of careful organization, the rallies never led to violence, although the atmosphere was very emotional and rife with fiery rhetoric. While street rallies brought together several hundreds of thousands of people and represented a very intense repertoire of action, they did not lead to the creation of distinct political networks among the broader population. The social structure of rallies consisted of a diverse and amorphous constituency mobilized from above. This was a populist mobilization that relied upon loyalty and support to a particular person and an overarching collective identity, rather than on multiple structural or political collectivities present within civil society. The dominant collective identity which combined ethnic and religious elements was so intense that it suppressed all other potential collective identities emerging within the public sphere. Different types of civic initiatives and new social movements were relatively weak and, for the most part, limited to the largest cities. Slovenia: new social movements and civic initiatives Contentious action in Slovenia consisted of the ongoing participation of people in a multitude of special-purpose groups and civic initiatives,

14 Hajdeja Iglicˇ movements (peace, ecological, feminist, gay, and petition movements), and politicized professional associations (Silber and Little 1996: 48–58). By February 1989 Radio Free Europe reported that about a hundred grassroots organizations and ten independent political groups were active in Slovenia (Stokes 1993: 237). The integration of various political groups at the national level took the form of an ‘associational network.’ While the opposition media, especially student radio and the political weekly Mladina, played a central role in providing an information infrastructure, there was never a single individual, group or organization that could lay claim to the top leadership role. The initiatives formulated by new political actors challenged the state on very concrete and specific issues. The corresponding repertoire of action included the organization of public discussions, the presentation of proposals and initiatives through the mass media and street-stands, and the establishment of services for subgroups of the population through selforganization. Its most important terrain was the mass media and the public sphere rather than the streets. The movement’s goal was to influence politics and political elites through the mobilization and expression of public opinion, thus opening the door for the institutionalization of political pluralism by contributing to the rise of social pluralism and autonomous collective identities within civil society. These activities were only loosely organized and cooperation among different initiatives was usually project-based rather than permanent. Even in the case of the largest mass demonstrations, organized to protest the unconstitutional decision of the Yugoslav People’s Army to put four prominent local journalists on trial in a military rather than a civilian court, coordination was managed by a loose network of groups joined specifically for this purpose into an umbrella organization (Committee for the Protection of Human Rights). Over 500 organizations (professional organizations, schools, and enterprises) and thousands of ordinary citizens joined them later. Due to the differences in their political ideologies, these groups defined their collective identity in the language of ‘interest,’ and their unity as ‘unity in action.’ This contributed to a de facto de-homogenization of political society, and made the development of an overarching nationalist ideology that would override other collective identities rather unlikely.

Social and political networks Every political action can be viewed as a result of the formation of new political ties and political identities among participants, existing alongside or replacing old political ties and networks. New political ties may either be forged among previously connected individuals by virtue of the addition of new relations to an already existing tie, or they may develop among previously unconnected or only weakly connected individuals and social groups by establishing new ties among them. Similarly, we can distinguish

Yugoslavia in the mid-1980s

15

between embedded and disjoined political identities (Tilly 1997). Embedded political identities are rooted in the networks and identities of routine social life: neighbor, co-worker, co-member of an association. Disjoined political identities appear chiefly in public life and are only weakly related to routine identities: movement leader, demonstrator, state official. The transformation of pre-existing social ties into political ties by means of the addition of new relational contents is a common topic in the literature on political action. Social networks have been widely linked with the processes of mobilization for contentious action. This is what scholars have called ‘micro mobilization contexts’ (McAdam 1988). These are social sites within people’s daily routines where informal ties among people serve as conduits for political solidarity, communication, and recruitment to various kinds of political organizations and groups. The well-known concept of ‘netness’ (Tilly 1978) was one of the first to express precisely this notion: that the high density of social relations among movement constituencies arises from daily associations and involvement in communal life. The ideal social structure for successful recruitment and mobilization seems to be one with ‘a lot of weak bridging ties which link together cohesive local groups bound together by means of strong interpersonal ties’ (McAdam and Paulsen 1993; Goldstone 1994). The theory predicts that members of such social groups are very likely to become political allies, which gives their routine identity a political anchorage. Empirical evidence suggests that this view is too simplistic. Pre-existing social ties and networks do not necessarily overlap with political networks. Pre-existing social networks sometimes split when faced with the challenge of conflicting political loyalties; alternatively, political networks can emerge around new political ties, and people end up in a familiar situation where there is little overlap between their social and political networks. It is also possible that only some social ties are converted into political ties without repercussions for the social ties that remain. Finally, whole social networks can be converted into political networks. I propose that there are three patterns in the relationship between personal and political networks. First, high overlap due to the ‘en bloc’ conversion of social into political networks that is characteristic of community mobilization. Second, little overlap between political and personal networks due to the strict distinction people make between their private and public identity. This is a situation of elite-sponsored mobilization, where participants do not seek to establish more durable and trusting (in a personal sense) ties with each other. Third, some overlap between the political and personal networks that are idiosyncratic rather than ‘en bloc’ as in community mobilization. In this case we talk about social mobilization. In social mobilization, personal relations of various strengths that are established among the participants of newly created and purpose-specific political groups and organizations contribute to the generation of personal trust within these political entities.

16 Hajdeja Iglicˇ In community mobilization, political networks and identities are embedded firmly in the identities and networks of routine social life. The recruitment into political organizations and protest groups occurs along the pre-existing community and neighborhood ties, ties established in the associational life of the public sphere, and at work. Mobilization along the lines of pre-existing communities usually poses a problem of coordination on a wider scale. While stronger identification with a social group contributes to the mobilization of that group, it limits the capacity of the group to generalize its identification across the groups (e.g. Gould 1993). In community mobilization, across-group coordination and cooperation is usually established with the help of ‘natural’ brokers and leaders, and political organizations that link various communities together within a large-scale movement. When present, political organizations mobilize contentious action in interaction with the pre-existing communities. They use the given forms of sociability for the penetration of their ideas, and they draw support from horizontal networks of social solidarity. But the relationship between the informal networks and political organization can also be riven with tensions since the political identities involved in informal networks and political organization may start to contradict each other over time.5 In the second type of mobilization, political and personal networks are disjoined. Such uncoupling of personal and political networks presupposes a sharp distinction between the public and private spheres: political networks belong to the public domain, while personal networks belong to the private domain. As a result, social ties established in the public sphere remain weak and impersonal, and make no contribution to the individual’s social integration within the political settings in which he or she participates. Halebsky (1976) presents a version of mass society in which an individual’s identification with a group remains weak and distrust pervasive, even if social ties to the group are abundant. Such mass society is prone to totalitarianism and various large-scale ideological (religious, ethnic, social) movements, since common participation in organizations and other exchange contexts, in the absence of plurality of intermediary collective identities, provides the communication structure suitable for elitesponsored mobilization on the basis of an undifferentiated and overarching identity. In addition, a strict distinction between the private and public spheres results in the contraction of the private sphere and personal networks into small, dense, and homogeneous groups (with close kin at the core of these groups), that act as strong reinforcement agents of externally provided ideology and collective identities. Mass demonstrations, rallies and protests are the most common forms of this kind of contentious action because they do not require participants to form more durable political ties of medium strength that would span the difference between the public and private, the strong and weak areas of the networks. The role of leadership, which provides political unity and identity to the otherwise unconnected mass of protestors, is usually played by an existing organizational

Yugoslavia in the mid-1980s

17

hierarchy, or by an external group of actors. This gives the mobilization a top-down character, especially if external actors are from the elite, or from groups sponsored by the elite. In social mobilization, the distinction between the public and private spheres is well established, but the political and personal networks nevertheless overlap to a certain degree. It depends on personal affinities and decisions by the partners as to whether or not they want to turn their otherwise impersonal relationship into a personal one, and what degree of intimacy they allow in the relationship. The relationships can be disembedded from the organizational and other kinds of impersonal settings, and brought into the private sphere either as part of a wider group of friends, or individually as very close friends. The coexistence of personal and organizational ties under the conditions of a clear a priori separation between them comes in the form of personal relations providing ‘a social infrastructure to the corporate ones’ (Paine 1972: 1). Because individuals regard impersonal and personal ties as belonging to two distinct normative systems, the former governed by self-interest and instrumental concerns and the latter by affection, trust, intimacy, and generosity, the influence of personal ties over one’s political affiliation is limited. The issue of political loyalty is not a constituent part of personal ties (McAdam and Paulsen 1993: 659). The mixing of personal and impersonal elements in social ties provides personal anchoring to the kind of universalistic sociability described by Adam Smith (1976) in his theory of personal relations in commercial society. More recently, authors have used the terms ‘universal otherhood’ (Nelson 1969) and ‘generalized trust’ (Uslaner 1999) – as opposed to ‘clan brotherhood’ and ‘particular trust’ – in order to describe a society in which the relations between people who do not belong to the same social groups and circles are driven by positive motivations and trust. But both ‘friendship’ and ‘strangership’ are constitutive elements of universal sociability (Silver 1994). While most of the time people are involved in relations with neutral or indifferent ‘strangers’ who occupy different social roles, some of these strangers become their personal friends. Personal trust that emerges between friends in a society characterized by universal sociability is a dyadic rather than group relationship. As a result, a social group has much less control over the political loyalties of its members than communities in which trust is a group relationship. As noted by Nelson, the price paid for the emergence of universal society is an attenuation of the love and loyalty that held clans together. Still, the personal element which results from the above described process of ‘turning strangers into friends’ is strong enough to ‘furnish one with the experience of trust and loyalty somewhere within real networks of solidarity’ (Chazel 1990) even under the conditions of universal sociability. This experience is especially important for the investment of trust when people establish new, purpose-specific ties within the field of voluntary

18 Hajdeja Iglicˇ associations and non-institutionalized political participation.6 Thus dyadic personal trust enables people to cooperate outside pre-existing social networks by providing them with the significant experience of trustful relations, without posing any boundaries to their political and social involvement.7 This is why in social mobilization the contentious action represents a space where new ties and new identities are formed along preexisting social ties. Political networks are not bounded by the pre-existing social structure as in community mobilization. Rather, people establish political ties according to their political affinities and gradually transform these originally impersonal ties into personal ties, which increase the solidarity of the political group. Since personal trust is a dyadic rather than a group relationship, it is not equally realized in all social ties which people establish within a given social setting. These ties may vary from strong to weak, from sociability ties to intimate ties. The model of new social movements best describes this kind of collective action. Although the ties that connect activists do not originate in communal or kin contexts, they are not necessarily weak. Participants usually strengthen the ties among themselves through participation. This is how social movements and other kinds of civic initiatives, as well as voluntary associations, contribute to the creation of social networks and thereby to the strengthening of the civil society, as has already been argued by the proponents of the new social movements perspective on political action (Touraine 1985; Melucci 1989) and theorists of social capital (Putnam 1993, 1995). We may add that in a society where generalized trust is widespread, more formalized social settings, such as the workplace, school, and political organizations, also contribute to the creation of social networks. These social contexts have been disregarded so far as a possible source of social capital, and treated as relatively unimportant compared to new social movements and voluntary associations. To sum up: while trust is consequential for the political life of societies in general, and for political mobilization in particular, it originates in social contexts and social networks in which people are embedded when they go about their routine daily activities: family, neighborhood, workplace, voluntary associations, and friendship groups. Social networks should be conceived of as those social practices in which patterns of social integration are established and followed when people enter the political sphere and become members of political networks. It is within social networks that people tacitly develop the rules that determine what are the appropriate and permissible ways of forming, maintaining, and breaking trustful relations. Thus social networks tell us whether people are in principle open to strangers whom they meet in various secondary social contexts, and in which way and under what conditions they are ready to turn these formal and semi-formal ties into personal ties. It is these aspects of social networks that affect the formation of political ties and the creation of political networks.

Yugoslavia in the mid-1980s

19

Social networks in the former Yugoslavia Most of the recent empirical studies on social capital have taken the normative approach and measured social capital in terms of generalized versus particular trust. Generalized trust has been measured by a question posed in a large number of surveys throughout the world asking whether most people can be trusted. Similarly, particular trust has been measured with a question that asked whether people trusted members of their own group. This chapter takes a structural approach, measuring networks instead of beliefs and drawing conclusions about trust on the basis of how people construct their personal networks. Following the discussion in the previous section about the importance of forming trustful relations in impersonal social contexts, it is particularly important to examine to what extent people choose personal contacts in these settings. This tells us whether or not they are in principle open to establishing personal ties with strangers. We can hypothesize that generalized trust is widespread when people are ready to include social ties established at work and in voluntary associations into their personal networks, especially when they bring these social ties into the private domain and consider them as friends. On the other hand, we conclude that particular trust is prevalent when networks remain limited to close kin, despite the fact that people participate in secondary social settings and thus have the opportunity for establishing social ties in these settings.

The survey This chapter analyzes survey network data collected in Yugoslavia in 1987.8 The survey used a structured questionnaire in which respondents were asked about their discussion networks. Information on discussion networks was elicited with a single name generator. Respondents were asked to name those persons with whom they had discussed important matters during the past six months. It is important to note that the question did not ask specifically about which personal matters, but left it to the respondent to decide on what are important matters in his or her case. The analysis of the overlap of different exchange relations (Burt 1983) suggests that ‘discussing important matters’ is a moderately intense relational content which captures significant kin as well as non-kin social ties. This is its advantage compared to other name generators which are more likely to capture only kin or non-kin ties. From the discussion networks we can thus expect information about the important social ties beyond the first circle of close ties. Respondents could name as many persons (‘alters’) as they wished in response to this name generator question, but the additional network data on (1) relationships between ego and alters, (2) relationships among alters, and (3) personal characteristics of alters were collected for the first five

20 Hajdeja Iglicˇ alters only. The relationships between ego and alters were described in terms of closeness, contact frequency, length of acquaintance, and role relationship with the respondent. Relationships among alters were described in terms of closeness for every pair of alters. We also obtained data on the gender, age, nationality, religion, education, and occupation of alters. The results Let us now turn to the analysis of discussion networks in the two republics, in Serbia9 and Slovenia. The comparison of personal networks in these two republics provides the central evidence for our hypothesis about the relationship between social networks and political mobilization. The study conducted in East Germany by Völker (1995) suggests that small, kin centered, and ‘private niche’ networks were characteristic of totalitarian political regimes. This notion of atomization and fragmentation of a society due to its political regime was not unfamiliar in Yugoslav sociology. According to Milic´ (1995) the trend toward social fragmentation in Yugoslavia started in the late 1960s. At the center of privatized social life was the nuclear family embedded in a highly supportive environment consisting of close kin (mother, father, siblings, children). At first, the trend toward privatization brought liberation from more rigid corporate structures from the early period of real socialism and more traditional, neighborhood, and village-based communities. However, this same trend was later regarded with a great deal of skepticism, since there were no indications that the social space beyond the first circle of close kin which surrounded the privatized family would become inhabited with the wider range of non-kin ties. The analysis of the relative presence of different network types in the former Yugoslavia shows that Yugoslavs were indeed very likely embedded in privatized kinds of personal communities (Iglicˇ 2001). Privatized or kin-centered networks represented 41 percent of all discussion networks. The community networks were the second most common network type. About 30 percent of the population were embedded in this network type. Only 18 percent of people were involved in differentiated personal communities, similar to those found in Western democracies (Fisher 1982; Wellman 1979). However, while a relatively large part of the Yugoslav population maintained a privatized social life, there was a huge variation in discussion networks throughout the country. It would appear that people from different regions lived in different social worlds. The comparison of social networks in Serbia and Slovenia illustrates this point. First, discussion networks in Serbia were much smaller than in Slovenia (3.12 compared to 3.66: see Table 2.1). Network size in Serbia was somewhat smaller, and in Slovenia it was well above the Yugoslav average (3.26). Networks in Serbia were not only smaller, but also denser than in Slovenia (0.77 compared to 0.73), although the difference is not

Yugoslavia in the mid-1980s

21

Table 2.1 Discussion networks in Serbia and Slovenia Variable

Network size Network density: • overall • kin density • non-kin density • kin-non-kin density Number of kin: • overall • spouse • close kin • extended kin Number of non-kin: • overall • co-workers • co-members • neighbors • ‘just friends’ Proportion of kin Proportion of non-kin N

Serbia

Slovenia

Mean/(st. dev)

Mean/(st. dev)

3.12

(1.58)

3.66

(1.61)

3.93

0.77 0.87 0.73 0.68

(0.26) (0.23) (0.26) (0.27)

0.73 0.85 0.71 0.63

(0.25) (0.20) (0.27) (0.27)


1.34
0.89
1.05
1.45

1.36 0.33 0.81 0.22

(0.12) (0.49) (1.06) (0.60)

1.19 0.30 0.68 0.21

(0.19) (0.48) (1.03) (0.52)


1.63
0.48
2.15
0.33

1.67 0.30 0.08 0.55 0.74 0.50 0.50

(1.05) (0.23) (0.49) (0.98) (1.21) (0.40) (0.40)

2.10 0.49 0.22 0.55 0.98 0.39 0.61

(1.15) (0.29) (0.52) (1.03) (1.18) (0.36) (0.36)

3.05 2.11 2.15 0.09 1.93
3.36 3.36

358

t-value

205

Notes Network density was operationalized as a ‘mean tie strength linking pairs of alters in respondent’s egocentric network’ (Marsden 1993). Strength of ties between alters was coded in the following way: 1 – if alters knew each other well, 0.5 – if they knew each other but were not close, 0 – if they did not know each other. Number of different kinds of social ties (kin, spouse, neighbor, co-worker, co-member, and just friend) was calculated for the average egocentric network. Figures in parenthesis are SDs.

significant at 0.05 levels. The difference in the density of networks is especially high when we observe the strength of ties between the kin and nonkin parts of networks. In Serbia, people tended to embed their various social ties in a group structure, rather than maintaining them in separate parts of networks. This practice was very different from what has been described in Western democracies where networks are highly integrated at the core (kin ties) and loosely integrated at the edges.10 Although networks in Slovenia were also very dense (for example, comparable data for USA shows a density of only 0.61), they were characterized by a higher separation between kin and non-kin social ties than in Serbia. Smaller and denser networks in Serbia also included more of kin ties, in the absolute and relative sense. The proportion of kin in Serbian networks was 50 percent, compared to 39 percent in Slovenian networks. Serbian networks in particular included more ties with close kin. Slovenian networks were larger and sparser than Serbian, and included

22 Hajdeja Iglicˇ fewer close kin ties and more ties with non-kin. Among non-kin ties, the most important were ties with ‘just friends’.11 But in addition to ‘just friends’, Slovenian networks also included larger number of ties with coworkers and co-members. The proportion of co-workers in Slovenian networks was similar to American networks (14 percent reported by Marsden (1987) and Burt (1990)). The proportion of co-members in Slovenian networks was half of the American proportion, due largely to the low number of voluntary associations in Slovenia at that time. A comparison of the strength of ties in Serbia and Slovenia shows that social ties in Serbia were in general stronger than in Slovenia (Table 2.2). Slovenian networks were larger since they included more weaker and nonkin ties. But ties with kin were also somewhat weaker in Slovenia than in Serbia, indicating that social ties within the kin system were more likely to be perceived according to the personal affinity of partners rather than their social role.12 Thus, the composition of Slovenian networks and the nature of social ties were more similar to the networks found in Western democracies, while Serbian networks resembled the privatized types of networks predicted by social theory. These were small, kin-centered networks that did not penetrate the wider social structure. They lacked ties with ‘just friends’ that would connect them within the community, as well as ties with coworkers and co-members that would help them to integrate within organizational settings. As discussed earlier, privatized networks do not free up people for selfTable 2.2 Strength of ties in Serbia and Slovenia Variable value Strength of ties: • overall • especially close • equally close • less close Strength of ties: • spouse • kin • ‘just friends’ • neighbors • co-workers • co-members N

Serbia

Slovenia

% Mean/(st. dev)

% Mean/(st. dev)

0.94 (1) (1) (0.5)

0.87 11 80 9

(0.22)


7.09 15 64 21

0.96 0.89 0.85 0.91 0.83 0.82

(0.13) (0.20) (0.22) (0.19) (0.24) (0.24)


2.06
3.81
2.86
1.03
2.05
1.79

0.99 0.96 0.92 0.92 0.89 0.90

(0.16)

(0.06) (0.14) (0.18) (0.18) (0.21) (0.20) 1,348

t-value

828

Notes Strength of ties was calculated for the average social tie, on the scale 1 – especially close or equally close, and 0.5 – less close. Figures in parenthesis are SDs.

Yugoslavia in the mid-1980s

23

organization, since their trustful relations do not reach beyond the circle of close kin. People in Serbia were involved in a very intense form of political mobilization, but, as we pointed out, this was a top-down mobilization. The intense identity politics helped establish trust among ‘familiar strangers’ (Patterson 1999) who otherwise lacked meaningful social ties with one another. In contrast, personal networks in Slovenia included weaker social ties established within formal and semi-formal social contexts. This indicates that Slovenians had the capacity for political selforganization and autonomous action, which was expressed in the large number of new social movements and civic initiatives active throughout Slovenia in the late 1980s. The pluralism of overlapping political groups and circles contributed to the much less intense identity politics in Slovenia compared to Serbia.

Conclusion In this chapter we perceived networks as trust networks. Trust networks, as opposed to recruitment networks, are those patterns of social integration established through routine participation in various social settings, and which are followed when people enter the political sphere and become members of political networks. They set the conditions under which one establishes trustful relations with different acquaintances and strangers within the political realm, and impose boundaries on political networks. So far, network analysis of collective action perceived social networks as inhibitors or promoters of collective action, and looked for the ideal social structure that would encourage it. This study suggests instead that different network configurations in Serbia and Slovenia allowed for different kinds of contentious action (community, social and elitesponsored). It is also argued that the form of political mobilization is affected by the societal capacity for establishing a certain kind of political network. This is different, on the one side, from structural arguments which trace differences in political action to the political opportunity structures, and on the other side, from the rational choice model that relates the form of collective action to different types of goals and levels at which political claims are made (local, national, global). By comparing personal networks in Serbia and Slovenia we showed that it was the personal networks in which people were embedded in their everyday lives that affected the way they formed political ties when it came to the rather extraordinary situation of political mobilization. In the discussion about the relationship between social and political networks we indicated some mechanisms that generated different kinds of trust through personal interactions at the micro level that were then translated to the macro level. For example, in the case of Serbia we showed how restricted trust in personal interactions resulted in small and dense personal networks, and how this restricted trust at the micro level combined with very

24 Hajdeja Iglicˇ strong ‘familiar trust’ (Patterson 1999); i.e. trust in others who were of the same ethnic and religious affiliation when it came to political mobilization at the macro level. By comparison, in Slovenia we found that the greater presence of generalized trust led to differentiate personal networks which included personal ties of various strengths that connected people to different social contexts. The capacity for forming trustful personal ties – weaker and stronger – within various semi-formal and formal contexts allowed for the self-organization of the Slovenian population and the creation of multiple political networks and identities. When we observe a country characterized by a very high level of social animosity, we usually look at the extent of organizational participation among people, and ask whether people who belong to different categories and social circles are meeting each other with sufficient frequency to establish trustful ties. But structural factors or opportunities for contact are not enough. In order to develop personal ties with strangers and engage with them in a cooperative relationship, one has to be embedded in a social network that carries trust potential. Such potential for social integration within a group of strangers is a source of tolerance for social minorities and competing political groups, as well as a capacity that affects one’s participation in various forms of political and civic self-organization. Among other things, it is crucial for the kind of political mobilization that is carried out by purpose-specific political groups requiring a relatively high involvement and identification on the part of their members, but which do not have any prior anchoring in the social structure. Members of such political groups usually do not share strong collective identities, memories and traditions, but base their participation on common political interests and concerns, and gradually develop trust by forming personal relations of different strengths with one another.

Notes 1 Rose (1994: 29) writes: ‘East Europeans know those whom they trust, and trust those whom they know.’ In Eastern Europe, the recipients of trust are likely to be an extension of friendship or collegial networks or of local and primordial communities, rather than internally diverse constituted political communities. 2 Kosovo also experienced a very high level of national mobilization on the part of the Albanian and Slavic (mainly Serbian) population throughout the 1980s. Contentious activity declined suddenly in the winter and spring of 1990, when the Albanian political leadership opted for a strategy of non-violent resistance, and reappeared again in 1996 with the emergence of the Kosovo Liberation Army. However, since our data on personal networks are highly unrepresentative for Kosovo, we decided to omit this region from the analysis. 3 The nationalist appeals were combined with the requests for the ‘debureaucratization’ of the Serbian state. 4 ‘The participants were given free transport, technical services and even money. Those who appeared as direct organizers were all party members’ (Imsimrovic´, in Magasˇ 1993: 221).

Yugoslavia in the mid-1980s

25

5 Historical literature offers numerous cases of community mobilization: from peasants revolts in ‘early modern’ Europe (Sabean 1976), to contentious action among artisan workers in the France of the old Regime (Sewell 1990), mass mobilization of peasantry and small-town workers in the resistance to the coup d’état in 1851 (Margadent 1979), and neighborhood-based mobilization in the Paris Commune in 1871 (Gould 1993). 6 Max Kaase shows that while interpersonal trust does not relate to political trust, it is an important pre-condition of non-institutionalized political involvement (Kaase 1999). 7 This argument builds on the so-called ‘indirect effect’, which says that a person learns concrete behavior in a specific aspect of life and transfers this behavior to other fields (Elster 1997; Jordana 1999). 8 A national stratified and multi-stage sample was used in order to select a representative cross-section of the Yugoslav population between 15 and 75 years old. The sample size was 2,241. 9 The republic of Serbia also included two autonomous provinces, Kosovo and Vojvodina, which were given a large degree of political and administrative autonomy. In this analysis I focus on the data from inner Serbia, thereby excluding respondents from the autonomous provinces. 10 Wellman (1979: 1211) describes the density of social ties in the personal network of East Yorkers in the following way: ‘Kin and non-kin intimates tend to be in different clusters of their intimate networks and not to have intimate ties with each other.’ 11 ‘Just friends’ are social ties that do not belong to any recent communal or organizational context. Generally, people who are ‘just friends’ have met each other within past social settings and maintain their relationship, although they no longer participate in these settings, or through common friends. 12 This finding is similar to what Fisher called the ‘self-selection by familialism’ (1982: 82–83). He suggested that in urban contexts non-kin ties replace kin ties, and kinship involvement becomes selective concerning with whom and when people will be involved.

Bibliography Birnbaum, P. (1992) ‘Nationalism: a comparison between France and Germany’, International Social Science Journal 133: 375–82. Burt, R. S. (1983) ‘Distinguishing relational contents’, in R. S. Burt and M. J. Minor (eds) Applied Network Analysis. Beverly Hills, CA: Sage. —— (1990) ‘Kinds of relations in American discussion networks’, in W. M. Marshall and C. J. Calhoun (eds) Structures of Power and Constraints. Cambridge: Cambridge University Press. Calhoun, C. J. (1993) ‘Nationalism and civil society: democracy, diversity and selfdetermination’, International Sociology 4: 387–411. —— (1994) ‘Social theory and the politics of identity’, in C. J. Calhoun (ed.) Social Theory and the Politics of Identity. Oxford: Blackwell. Chazel, F. (1990) ‘Individualism, mobilization, and collective action’, in P. Birnbaum and J. Leca (eds) Individualism. Oxford: Oxford University Press. Doderovic´, M. (1990) Kako se dogodio Sˇolevic´?. Zagreb: Gradina. Elster, J. (1997) Egonomics. Barcelona: Gedisa. Fisher, C. S. (1982) To Dwell among Friends: Personal Networks in Town and City. Chicago, IL: University of Chicago Press.

26 Hajdeja Iglicˇ Goldstone, J. A. (1994) ‘Is revolution individually rational? Groups and individuals in revolutionary collective action’, Rationality and Society 1: 139–66. Gould, R. (1991) ‘Multiple networks and mobilization in the Paris Commune 1871’, American Sociological Review 56: 716–29. —— (1993) ‘Trade cohesion, class unity, and urban insurrection: artisanal activism in the Paris Commune’, American Journal of Sociology 58: 721–54. Halebsky, S. (1976) Mass Society and Political Conflict: Towards a Reconstruction of Theory. Cambridge: Cambridge University Press. Iglicˇ, H. (2001) ‘Formation of political ties: personal networks and political action in the former Yugoslavia’, unpublished Ph.D. thesis, New York: Columbia University. Inglehart, R. (1999) ‘Trust, well-being and democracy’, in M. E. Warren (ed.) Democracy and Trust. Cambridge: Cambridge University Press. Jordana, J. (1999) ‘Collective action theory and the analysis of social capital’, in J. W. van Deth et al. (eds) Social Capital and European Democracy. London: Routledge. Kaase, M. (1999) ‘Interpersonal trust, political trust and non-institutionalized political participation in Western Europe’, West European Politics 3: 1–21. McAdam, D. (1988) ‘Micro-mobilization context and recruitment to activism’, International Social Movement Research 1: 125–54. McAdam, D. and Paulsen, R. (1993) ‘Specifying the relationship between social ties and activism’, American Journal of Sociology 99: 640–67. Magasˇ, B. (1993) The Destruction of Yugoslavia. London: Verso. Margadant, T. W. (1979) French Peasants in Revolt: The Insurrection of 1851. Princeton, NJ: Princeton University Press. Marsden, P. V. (1987) ‘Core discussion networks of Americans’, in American Sociological Review 52: 122–31. —— (1993) ‘The reliability of network density and composition measures’, Social Networks 15: 399–422. Melucci, A. (1989) The Nomads of the Present: Social Movements and Individual Needs in Contemporary Society. Philadelphia, PA: Temple University Press. Milic´, A. (1995) ‘Svakodnevni zˇivot porodica u vrtlogu drusˇtvenog rasula: Srbija, 1991–1995 godine’, in S. Bolcic´ (ed.) Drusˇtvene Promene i Svakodnevni Zˇ ivot: Srbija Pocˇetkom Devedesetih. Beograd: Insˇtitut za sociolosˇka istrazˇivanja. Nelson, B. (1969) The Idea of Usury: From Tribal Brotherhood to Universal Otherhood. Chicago, IL: University of Chicago Press. Offe, C. (1999) ‘Can we trust our fellow citizens?’, in M. E. Warren (ed.) Democracy and Trust. Cambridge: Cambridge University Press. Opp, K. D. and Gern, C. (1993) ‘Dissident groups, personal networks, and spontaneous cooperation: the East German revolution of 1989’, American Sociological Review 58: 659–80. Paine, R. (1972) ‘Scale and personal dimension in society’, paper presented at the Burg Wartenstein Symposium, No. 55, Wener-Gren Foundation. Patterson, O. (1999) ‘Liberty against the democratic state’, in M. E. Warren (ed.) Democracy and Trust. Cambridge: Cambridge University Press. Putnam, R. D. (1993) Making Democracy Work: Civic Traditions in Modern Italy. Princeton, NJ: Princeton University Press. —— (1995) ‘Bowling alone: America’s declining social capital’, Journal of Democracy 6: 65–78.

Yugoslavia in the mid-1980s

27

Rose, R. (1994) ‘Post communism and the problem of trust’, Journal of Democracy 5: 18–30. Sabean, D. W. (1976) ‘The communal basis of pre-1800 peasant uprisings in Western Europe’, Comparative Politics 8: 355–64 Sewell, W. H. Jr. (1990) ‘Collective violence and collective loyalties in France: why the French revolution made a difference?’, Politics & Society 4: 527–52. Silber, L. and Little, A. (1996). Yugoslavia: Death of a Nation. London: Penguin, TV Books. Silver, A. (1994). ‘“Two different sorts of commerce” – Friendship and strangership in civil society’, in J. Weintraub and K. Kumar (eds) Public and Private in Thought and Practice. Chicago, IL: Chicago University Press. Smith, A. ([1759,1791] 1976) ‘The theory of moral sentiments’, in D. D. Raphael and A. L. Macfie (eds) The Theory of Moral Sentiments. Oxford: Clarendon Press. Stokes, G. (1993) The Walls Came Tumbling Down: The Collapse of Communism in Eastern Europe. New York: Oxford University Press. Tilly, C. (1978) From Mobilization to Revolution. New York: Random House. —— (1997) ‘Political identities’, unpublished manuscript, Columbia University. Touraine, A. (1985) The Voice and the Eye: An Analysis of Social Movements. Cambridge: Cambridge University Press. Uslaner, E. M. (1999) ‘Democracy and social capital’, in M. E. Warren (ed.) Democracy and Trust. Cambridge: Cambridge University Press. Völker, B. (1995) ‘“Should Auld Acquaintance Be Forgot. . .?” Institutions of Communism, the Transition to Capitalism, and Personal Networks: The Case of East Germany’. Amsterdam: Thesis Publishers. Wellman, B. (1979) ‘The community question: the intimate networks of East Yorkers’, American Journal of Sociology 84: 1201–31.

3

Communist societies, the velvet revolution, and weak ties The case of East Germany Henk Flap and Beate Völker

Introduction East Germany’s institutional overhaul began with the fall of the Berlin Wall in the autumn of 1989. The transformation in the GDR from communism to market and democracy is among the best traceable cases of the velvet revolution in East European societies (Stark and Bruszt 1998). The institutional changes had extensive consequences for almost every aspect of life of former GDR citizens. For a sociologist these events constitute a unique opportunity to inquire into key elements of social research: the social conditioning of life and the limits to malleability through social institutions. We focus on the degree to which seemingly private decisions such as with whom to interact are influenced by institutional contexts. In particular we study why the promise of communism, the creation of social cohesion, did not come true. We argue that every institutional context leads to specific personal experiences and problems for which there are no standardized solutions. Personal networks are a means of solving such problems. Therefore, people invest in social relationships according to the social institutional environment in which they live.

Communist societies, communist networks? Social capital theory interprets personal networks and the resources to which they provide access as social resources. Personal networks are positioned alongside symbolic, political, and economic capital as another type of capital resource (Bourdieu 1981; Coleman 1990; Flap 1999, 2002; Lin 2001). The more capital individuals have, the more likely they are to achieve their goals. These goals come down to improving physical wellbeing, gaining social approval, and bettering living conditions in general. Social capital has at least three dimensions: the number of people in one’s network, their ability, and their willingness to lend support. The idea of social capital implies that people will invest in relationships according to the expected value of future support. This ‘shadow of tomorrow’ suggests

The case of East Germany

29

that social institutions will influence the returns of and thereby investments in social capital. Through time, institutions stabilize people’s social situations and their problems as well as the instrumentality of certain ways of solving these problems. Marxist society has two main organizing principles: (1) the political control of most spheres of life by a Leninist party and state, and (2) the organized dependency on the party for all goods and opportunities (Walder 1994: 299). Marxism is an activist ideology, propounding collective norms of comradeship according to which men can control their fate on Earth. Human nature is considered to be malleable: qualities such as morality and social consciousness are products of a specific social situation. Changes in the social institutions, from private to collective property rights and organizations that control these institutions, will create a new socialist individual for whom collective interests reign superior to self-interest. Practically all institutions in the GDR were devised according to norms of collectivism. People worked in brigades and had to join organizations for spending their leisure time. Neighborhood and house committees collectivized life around the home (see Völker and Flap (1997) on communist neighborhood relationships). This organizational embedding of citizens – officially meant to create social cohesion – enabled the one-party state to control nearly all aspects of life. Our interviewees assumed that in every house committee, cadre, or brigade at least one person worked as an unofficial informant of the security police. The encompassing political control of daily life presented people with the acute problem of whom to trust and how to decide whether intentions of others were honest. This problem became particularly prominent if one wanted to discuss politics: people had to assume that in every public circle informants were listening and the damage potential of misplaced trust was high. A note in one’s personnel record of being politically unreliable could be detrimental to one’s own or one’s family’s life chances. The absence of public political debate created the often-quoted ‘double reality’ in which many people lived. The GDR was sometimes called the society of ‘niches’ (Gaus 1983; cf. Ash 1981). Citizens tried to escape the ‘collectivism and communism by design’ and searched for more individuality, freedom, and areas of life that were not contaminated by communist ideology. People created niches in their personal networks as a refuge from the meddling of government in their private lives. We expect that East Germans discussed politics frankly only with people whom they trusted; that is, those with whom they also discussed their personal, private problems. The part of a network in which personal matters are discussed is usually referred to as the ‘core’ network (Marsden 1987). Accordingly, we call the part of the core network in which people also discuss politics the ‘niche’. We expect niches to consist of strong ties to trustworthy others, which allow an uncensored exchange of political opinions and which provide social approval.

30 Henk Flap and Beate Völker Repeated transactions embedded in a network of highly connected strong ties enable everybody in the circle to have the same information about all other members. A person’s reliability can be easily confirmed and a reputation for being untrustworthy spreads quickly. Strong ties give people some leverage in their relationships with each other, since they can threaten to withhold future support (Granovetter 1985). Therefore, such a network structure promotes mutual trust and inheres least risk of betrayal. Consequently, we expect the niches in personal networks to be highly dense and to consist of strong ties. We also expect the niches to be composed of relationships with similar others, since feelings of security and trust are more likely if interaction partners share the same background. Finally, since members of the niche are selected cautiously and the relationships cost time to maintain, we expect a limited number of niche relationships. Rather, these relationships have a high multiplexity. The second organizing principle of the communist system has been the Plan, designed by government, describing what goods and services should be produced, and how they should be distributed. This command economy led to shortages of various commodities, from electric equipment to fruit (Kornai 1980). People tried to compensate for these shortages and safeguard their physical well-being through the creation of an informal economy. This ‘shadow economy’ consisted of second jobs, a gray or black market, and informal exchange. Unlike people in other Eastern European countries only a few East Germans had a second job, and, because of the government’s control, black markets were inaccessible to many. Although there were undoubtedly differences among citizens, in general everybody was confronted with the problem of how to obtain goods in short supply. Because black markets and second jobs were not popular solutions, we believe people invested in their personal networks to get scarce commodities. ‘Provision networks’ as a source of material commodities are often mentioned in studies on the organization of daily life in the ex-GDR (Srubar 1991; Hölder 1992; Diewald 1995; cf. Sik and Wellman 1999). Since the variety of resources accessible through others increases with the network heterogeneity, we believe that the instrumentality of the provision network was higher the more different occupations were present. Moreover, ‘provision ties’ will have been weak rather than strong, because these kinds of transactions needed no special investment. This argument does not imply that provision ties were always weak, or that weak ties were always provision ties. In our view, provision ties are an important case of weak ties, whose emergence and value are institutionally conditioned. Of course, someone who was able to provide family and friends with scarce commodities will have done so. Niche relationships were embedded in a dense network of strong ties, and such a network structure warrants trustworthiness of the interaction partners. In weak relationships, such as provision ties, the problem of whom to trust will have been even more prominent. In empirical studies

The case of East Germany

31

(see e.g. Burt and Knez 1996) weak ties turn out to be least trustworthy in general. In a communist society trustworthiness of weak ties becomes a greater issue because of the increased damage potential. In a one-party state most social institutions were party controlled and opportunities for having a good life depended largely on a politically correct image. Further, people knew about the considerable control apparatus. Being distantly related to somebody implied uncertainty about that person’s intentions and the circles in which he or she might be involved. In addition, a weak tie provides little leverage for influencing decisions. Therefore, a weak relation would have been suspect. Hence, we argue that the communist ideology and its related policies had perverse effects because people know about the ‘strength of weak ties’ (Granovetter 1973). Since people in the former GDR were suspicious of weak ties they probably had fewer weak relationships than people in democratic countries. Earlier we found that ties between direct neighbors barely existed in the former GDR because they were mistrustful of each other (Völker and Flap 1997). Further, weak ties are even more suspicious if they are interconnected. Consequently, someone with unconnected weak ties will have been better off. In addition, weak ties will not only have been separated from each other but also from strong ties; that is, from the niches, in order to maintain their functions. Therefore, there will have been a ‘gap’ between the niche and the remainder of the personal network, inter alia, the provision ties. It may seem implausible that the focal actor influences the connections among network members. However, both partners who intended to discuss personal and political matters were similarly interested in a cautious selection of third parties. Friends of friends will have been friends because in the former GDR there was a strong selection on forbidden triads in the evolution of networks. If they emerge at all, they will not last long. The structural features that, according to our argument, were typical of personal networks in a communist society such as the GDR are presented in Figure 3.1. If our argument on the institutional conditioning of networks is correct, these networks will change after the institutional overhaul, as there will no longer be a need to concentrate on a few trustworthy others. The segregation in the network will disappear and the niches themselves will vanish. In the old days niches helped people to develop a sense of individuality and had many relational functions. Nowadays, an individual’s interests might be served better if different relational activities are undertaken with a greater number of different others; this way, one avoids being dependent upon just a few. We also expect a decline in the average strength of niche relationships. Further, after the fall of the Wall, there will no longer be a need to keep provisional ties or, more generally, weak ties apart from each other or apart from stronger ties. Political trustworthiness is no longer an issue in everyday life and politics will be discussed with strangers too (see Figure 3.2).

32 Henk Flap and Beate Völker

B

D 0 

A






0 


0

C

E

Figure 3.1 Typical pattern of personal networks in a communist society such as the former GDR Notes Dense niches (A, B, and C connected via thick lines), disconnected provision ties (D and E connected only with the focal actor A through thin lines), and a gap between niche and provision network; ,
, and 0 indicate the notation for the triad analysis.

Design, measurement, and methods The research design includes three points of measurement, each time collecting data on retrospective and actual networks both before and after the political turn (see Table 3.1). Respondents were sampled randomly in Leipzig and Dresden. Interviews began in 1992. In 1993, respondents were re-interviewed and respondents from another sample were interviewed for the first time. In 1994, both samples were interviewed for the third and second time, respectively. Respondents had to be between 30 and 55 years of age. Table 3.1 Design of the study on changing personal networks in the former GDR before and after the political turn, 1989–94 April 1991 Pilot study n  35

Sample 1

Sample 2

May 1992

April 1993

March 1994

1st measurement data on 1989 and 1992, n  189

2nd measurement data on 1993, control questions for 1989 and 1992, n  126 1st measurement data on 1989 and 1993, n  300

3rd measurement data on 1994,  control questions n = 83 2nd measurement data on 1994,  control questions n  221

The case of East Germany

33

Institutional framework

Communism/ totalitarianism

Command economy

Collective outcomes

Party and state control/collectivism

Economy of shortage

Individual problems

Trust and individual identity

Obtaining scarce goods

Individual investment

In niches, being aware of weak ties

Provision networks

Niches are: small, strong ties, multiplex, homogeneous, dense, separated from weak ties.

Provision networks are: small, weak ties, uniplex, heterogeneous, open, separated from niches.

Network consequences

Figure 3.2 Institutional embeddedness of relational investment and resulting consequences for personal networks in a communist society

At t1 we paid most attention to personal networks before the turn; at t2 we focused strongly on the actual situation and asked a subset of questions only about the time before the turn in order to control for memory biases (Bernard et al. 1984). Similar retrospective networks mentioned at each point of measurement indicate the absence of biases. We found a slight tendency to forget weaker ties: in later measurements less weak ties were mentioned. If we assume that weaker ties are forgotten, tests of our hypotheses will be conservative, apart from the one on the small number of weak ties during communism. Further, it has been controlled for critical life events between t1 and t2 (see Völker 1995). Panel attrition was unrelated to common socio-demographic characteristics. The retrospective questions had a time frame of six months. We concentrated on actual, not hypothetical interactions, which is an additional safeguard against retrospective biases. To avoid methodological artifacts, we changed the order of name-generating questions for the actual and retrospective networks at the different points of measurement: in the second

34 Henk Flap and Beate Völker and third interviews we asked first for the ‘actual’ network. Here, we use the data referring to the situation in 1989 and before (n  489) and to the situation in 1994 (n  304). The questions generating names of the network members were fashioned after Fischer (1982) and the General Social Survey in the USA (Burt 1984), and adapted to the situation in East Germany. The names extracted in response to the question ‘With whom did you discuss important personal matters?’ are considered as the core network (see Marsden (1987) on Americans’ core networks). In total, we used eighteen name-generating questions to describe how our respondents’ networks looked before and after 1989. We refer to name generators also as ‘relational functions’. We consider ‘niches’ to consist of ties that are important for the discussion of personal matters (i.e. core ties) as well as for the discussion of political opinions and events. Provision ties are generated straightforwardly by asking who helped to obtain scarce commodities. In 1994 we replaced this question with one on to whom our respondents had turned for advice on financial matters. Information about network members was collected by asking additional questions on the relationship between network member and respondent (role, contact frequency, duration, strength) and on personal characteristics of the alter (occupation, age, gender). Tie strength is best measured by ‘closeness’, since this indicator is not contaminated by other measures (Marsden and Campbell 1984). We used a five-point rating scale. Multiplexity was measured by the sum of different simultaneous functions of a relationship. Similarity of occupations was measured on a Guttman scale indicating whether the occupational domains are comparable, similar, or exactly the same. We calculated heterogeneity by an index for qualitative variation (IQV) which measures the likelihood that two network members, randomly selected, belong to different defined categories (Agresti and Agresti 1978). In order to establish network density, the number and strength of connections between ties were measured for five name-generating questions: those alters with whom the relationship implied discussion of personal matters, provision of goods, career advice, being an enemy at work, and small repairs in a household. One measure of density is the average tie strength among network members, and the other is the fraction of existing links and maximum possible links. In 1992, the interlinkages were measured for the ‘discussion of personal matters’ only, while in the later measurement in 1993 the other four questions were added. The ‘gap’ between the niche and provision network was established by counting triads of ties consisting of the focal actor, one niche network member and one member of the provision network. The absence of a tie between the two alters indicates a gap between niche and provision ties. To count the triads, we used a simple program called ‘SIGRPOS’ (Snijders 1999) that calculates positional statistics for the vertices of a signed digraph.

The case of East Germany

35

Background characteristics of ego and alter were measured using standard instruments: educational level was measured as the highest educational certificate obtained, and occupational class was determined using Goldthorpe’s class scheme (Erikson et al. 1979). An additional coding was used developed especially for the situation in the GDR (Mayer and Solga 1993) which involved three types of information: domain, vocational training needed, and the specific occupational code. Other features were ascertained in a straightforward manner (e.g. by asking people whether they felt controlled, whether they missed any commodities in their household, and how they coped with shortages). All relational data and characteristics of alters are reported by the respondents and these relationships are dependent upon each other. The mutual dependence of relationships in a network advises against the use of ordinary least-squares regression (see Van Duijn et al. 1999: 188). Aggregation of the data is statistically correct but implies a loss of information, and differences between relations cannot be considered. Multilevel modeling accounts for the nested structure of the data (see Snijders and Bosker 1999). In this chapter, the models were estimated with the program VARCL (Longford 1988).

Results The results of our analysis are presented in the order of the hypotheses. First, differences between niche and provision network in 1989 are analyzed; second, we looked at characteristics of ties within each network part. Third, we inquired into the network structures, the interconnections and the gap. We conclude our analysis by comparing the 1989 networks to the 1994 networks, thereby approximating a comparison between a communist and non-communist setting. The shape of personal networks The networks of East Germans were small in the old days, especially given the large number of name-generating questions. Their mean size was eleven individuals. Comparable studies in the United States and The Netherlands found larger personal networks while using similar but fewer namegenerating questions (e.g. Fischer (1982) found nineteen alters with eleven questions in California, and Busschbach et al. (1999) found twenty-two alters with nine name-generating questions in The Netherlands). Thus it is unlikely that our findings are completely due to the retrospective nature of the data. To estimate the importance of informal exchange in the former GDR, we asked what kind of good had been missed in particular and how people had coped with its shortage. We inquired into the importance of the various alternatives, i.e. black market, second job, or personal network. Most people mentioned the personal network as the only serious option.

36 Henk Flap and Beate Völker About 70 percent of respondents acquired goods via informal networks. Only 8 percent said they had contacts with the former Federal Republic of Germany (FRG) providing an additional means to obtain commodities. There were no reports of buying in one of the 300 Exquisite shops where one could buy western goods at incredibly high prices or having regular access to western marks (Völker 1995: 17). If it did exist, a ‘provision network’ consisted on average of three persons; the mean number of suppliers for all respondents is 1.9. Most East Germans (89 percent) in the sample also had a ‘niche’: a circle of close associates who were important interlocutors for discussing personal and political matters. If one had a niche, it averaged 2.6 persons. Strength and multiplexity Most provision ties were weak. We did not exclude family members, who may have been providers too, but who might also be closely related to the respondent. The average strength of provision relations per network is 2.7 (sd. 1.0); for the whole network it is 3.4 (sd. 1.1). Provision ties are the weakest of all ties in the personal networks (excluding ties to adversaries). Their average multiplexity is 1.1 (sd. 0.7), which is low compared to the multiplexity of other ties. The niches were formed by an individual’s most intensive relationships: the average closeness of these relationships is 4.4 (sd. 1.5). Niche relationships before the turn have an average multiplexity of 1.9, which is higher than the multiplexity of the whole network (1.3 on average). Differences in multiplexity were tested by multivariate analysis of variance for repeated measurements (not shown here). As expected, the multiplexity of niche ties differs significantly from that of provision ties. We conclude that niche and provision ties were different with regard to strength and multiplexity. Both parts of the network and also the whole network of East German citizens before the political upheaval were rather small. To inquire deeper into the coincidence of relational functions that define niche ties, that is, ‘discussing important personal matters’ and ‘discussing political opinions and events’, and to study the multiplexity of niche ties compared to provision relationships, we estimated a logistic random coefficient model for relations that were important for the discussion of political matters and for provision ties, respectively (Table 3.2). To control for conditions that might affect the likelihood of a tie being important for the discussion of personal or political matters or being a provision tie, we included close family, partner, sample, town, and network size in the analyses. Talking about politics requires a trustworthy relationship and we consider the discussion of personal matters as an indicator of trust. Hence, the relational function ‘discussion of political matters’ will have been well predicted by ‘discussion of personal matters’. In addition, other relational functions will have been good predictors, since niche

The case of East Germany

37

Table 3.2 Logistic random coefficient model for association of relational functions for the discussion of political matters and for provision ties in East Germany, 1989 Logistic regression parameter estimates

Political matters

Provision ties

Coefficient (SE)

Coefficient (SE)

Functions of relationships Getting a job Discussion of professional matters Discussion of problems related to work Advice in questions of career Getting a house Help with small repairs Provision of scarce items Leisure time Discussion of political matters Discussion of personal matters


0.18 (0.16) 0.66 (0.10)** 0.28 (0.12)** 0.08 (0.17)
0.40 (0.23) 0.45 (0.09)**
0.31 (0.10)* 0.81 (0.08)** – 1.10 (0.09)**


0.53 (0.21)**
0.50 (0.14)**
0.03 (0.17)
0.33 (0.24)
0.28 (0.25) 0.11 (0.11) –
0.33 (0.10)**
0.28 (0.10)**
0.73 (0.13)**

Control variables Partner Close family Sample Town Network size

0.10 (0.05)* 0.35 (0.08)**
0.09 (0.10) 0.02 (0.10)
0.10 (0.10)


0.16 (0.23)
0.17 (0.10)
0.80 (0.10)**
0.07 (0.11)
0.01 (0.01)

Intercept


1.36

Variance component estimates Random intercept variance Deviance (
2 log likelihood)

0.65 (0.06) 5,603.87

0.05 0.64 (0.10) 3,899.32

Notes * and ** represent p  0.05 and p  0.01 respectively. Data contain 4,524 relationships of 476 respondents; unstandardized coefficients. Emboldened words indicate variables on the level of the respondent.

relationships were multiplex. Our expectation is that provision ties did not coincide with other ties. The results show that the discussion of political matters indeed required a trustworthy relationship. If individuals discuss personal matters with one of the network members, they are likely to discuss political matters with that network member as well. These relations are also important for the discussion of professional matters and problems related to work, helping with small repairs, as well as for spending leisure time, although to a lesser degree. A relation that is important for provision ties is not at all important for discussing politics. Provision ties are not associated positively with any other relational function. All effects found are negative, which indicates that provision ties, as expected, are comparably uniplex. No significant random effects were found in the models, indicating that there were no major differences among individuals.

38 Henk Flap and Beate Völker This analysis confirms the previous results on the aggregated level, as well as our general hypothesis: providing scarce items is not co-mingled with discussing personal and political matters or with any other activity that requires trustworthiness. There was a sharp division of labor among different types of ties in personal networks in the former GDR.

Heterogeneity and density According to our hypotheses, the occupational and educational heterogeneity of niches should be low, whereas the heterogeneity of provision networks should be high. A comparison of the indices of qualitative variation (IQV) of occupational domains for niches and provision networks before 1989 corroborates this hypothesis. The niche and provision network differ significantly from each other in heterogeneity, while using a multivariate analysis of variance for repeated measurements and controlling for the two samples. At the aggregated level niches are homogeneous while provision networks are far more heterogeneous. Next, effects were tested for the particular relationship between ego and alter. For this purpose, logistic random coefficient models were again estimated with ‘niche relation’ and ‘provision relation’, respectively, as the dependent variable and similarity of occupation as the independent variable, using the coding described above. In order to control for strong relationships, we included roles such as partner and family in the model. Table 3.3 shows that, in contrast to niche relationships, provision ties under communist conditions were socially different from ego. Table 3.3 Logistic random coefficient model on similarity of niche and provision relationships in East Germany, 1989 Logistic regression parameter estimates

Similarity Control variables Partner Close family Sample Town Size of network Intercept Variance component estimates Random intercept variance Deviance (
2 log likelihood)

Niche

Provision

Coefficient (SE)

Coefficient (SE)

0.16 (0.04)**


0.18 (0.05)**

3.03 (0.15)** 1.40 (0.07)**
0.20 (0.12)
0.05 (0.09)
0.74 (0.88)
2.00


0.16 (0.12)
0.14 (0.10)
0.24 (0.10)**
0.07 (0.10) 0.01 (0.01)
0.95

0.31 (0.06) 3,929.6

0.59 (0.08) 4,788.9

Notes * and ** represent p  0.05 and p  0.01 respectively. 4,524 relationships of 476 respondents; unstandardized coefficients. Emboldened words indicate variables on the level of the respondent.

The case of East Germany

39

Our hypothesis on the homogeneity of niches and the heterogeneity of provision networks turned out to be correct. Niche relationships were relations to alters who were similar to ego as well as to each other with regard to education and occupation, while provision network members were different from ego and from each other. We also tested the hypothesis that the density of the niche will have been much greater than that of the provision network. Results of multivariate analyses of variance show that niches and the weak part of the network differ significantly in their density, regardless of which indicator for density is used. Density in niche networks turned out to be high, compared not only to the network parts outside of the niche, but also to what is known from density in (core) networks in other countries (Marsden 1987).

The division of labor and changes in this division We argued that niches were kept apart from provision ties or, more generally, from weak ties, and called this the ‘gap’ (see Figure 3.1). In order to test this argument we counted specific triads in the personal network that indicate the existence of a gap. We distinguished between weak ties (indicated by
), strong ties (indicated by ) and no ties (indicated by 0). Accordingly, as depicted in Figure 3.1, the expected typical pattern of triads in the niches is ‘  ’, implying a strong connection among firm ties. The expected pattern between niche and weak ties, the gap, is ‘
0’ or ‘
 0’, indicating that the focal actor is connected strongly to one network member, weakly to another, and that these two alters are not connected themselves. Finally, the expected pattern within weak ties is ‘

0’, indicating that the focal actor is related loosely to two network members who themselves are not connected. Table 3.4 shows the frequency of different types of triads in 1989 and 1994. For 1989 the expected pattern within the niches (  ) occurred most often, i.e. almost 70 percent. Further, between niches and weaker, provisional ties the triad pattern expected for the gap 
0, occurs quite frequently (30 percent). However, the pattern,   0, occurs even more frequently (38 percent), indicating a strong connection of ego with the two alters, who are not connected with each other. This finding is due to our coding of strong ties: ties with a medium strength (category 3 on a 5-point scale) are considered strong. Outside the niches, the same pattern occurs most frequently and the pattern

0 occurs relatively frequently (18 percent), indicating that the relationships between ego and alters are weak and that the alters themselves are not connected. In sum, only in the niches did we find many closed triads indicating high interconnection. In the other investigated parts of the networks we found the expected open triad structures.

40 Henk Flap and Beate Völker Table 3.4 Patterns of triads within niches, between niches and weak ties, and outside niches (triads in networks of 300 respondents, East Germany, 1989 and 1994, in percentages, n in parentheses) Within niches

Between niches and weak ties

Outside niches

1989 1994 (n  217) (n  153)

1989 1994 1989 1994 (n  666) (n  529) (n  186) (n  153)

Closed triads  69.1


0.0

 0.0 

/

0.9 
 1.4 
3.5

45.8 ** 2.0 2.0 2.6 0.0 34.0 **

10.4 0.2 1.1 5.6 6.0 6.8

19.4 * 0.0 0.0 26.9 ** 4.5 17.9 *

1.1 3.8 0.0 8.1 1.6 6.3

10.7 * 4.9 1.0 13.6 1.9 16.5 *

Open triads 
0/
0 0

0

0.7 12.4 * 0.7

30.3 38.6 1.7

22.4 * 9.0 ** 0.0

38.7 22.6 17.7

26.2 ** 22.3 2.9 **

1.4 21.7 0.0

Notes  indicates a strong relationship,
a weak relationship, and 0 indicates no relationship. See text for coding details. Notation is ij, ik, jk, whereas the focal actor (ego)  i. * indicates significant differences between 1989 and 1994 at p  0.05 and ** at p  0.01.

Networks after the velvet revolution When inquiring into the implications of our argument for the period after communism one should bear in mind that our data provide insufficient information on personal networks in a market-democratic or capitalist institutional system since the transition was still ongoing in 1994. Yet, compared to other Eastern European societies, the transition in the former GDR has probably been most rapid (Stark and Bruszt 1998). In 1994 we did not ask about ties which helped to attain items that have been scarce in supply, since the kinds of commodities that had been exchanged through these relationships were no longer scarce. Instead, we asked about another kind of provision tie; that is advice on all types of financial matters, which we regarded as a new system-induced problem for individuals. Personal networks in the former GDR changed mostly in ways that we predicted. The size of personal networks did not increase; neither did that of niches or of the part made up of weak ties. Many relations were broken, especially relations to enemies and untrustworthy others. People also engaged in new relationships, mostly in the context of work, which were important for all kinds of activities. The strength of niche relationships decreased significantly. Of all niche relationships about 34 percent weakened. Relationships that were import-

The case of East Germany

41

ant for information about all kinds of financial matters, the ‘new’ kind of provision relationships, were somewhat stronger than provision relations had been. Many of the former provision relationships became weaker (about 58 percent) or were terminated (22 percent). Average multiplexity of niche relations decreased, whereas it increased slightly for provision relations. Average homogeneity of the whole network increased, while that of the niches decreased and the density of all the network parts considered decreased. There was a lessening of differences between niche and weaker ties. In order to understand how the coincidence of relational functions changed, we estimated the same random coefficient model as in Table 3.2 Table 3.5 Logistic random coefficient model for the discussion of political matters for 1994 data in East Germany Logistic regression parameter estimates

Discussion of political matters Coefficient (SE)

Intercept Functions of relationships Getting a job Discussion of professional matters Discussion of problems related to work Advice in questions of career Getting a house Help with small repairs Leisure time Discussion of personal matters Control variables Partner Close family Sample Town Network size Critical life event Variance component estimates Random intercept variance Discussion of personal matters Covariance (intercept, slope) Deviance (
2 log likelihood)


0.82
0.98 (0.44)** 0.28 (0.14)* 0.72 (0.29) ** 0.70 (0.16)**
0.27 (0.31) 0.75 (0.13)** 0.47 (0.10)** 0.70 (0.12)**

Difference in coefficients 1989–94 (absolute t-valuea) – 1.8 2.9** (1.3) 3.5** (1.0) 1.8 6.2** 2.7**


0.08 (0.18)
0.56 (0.11)**
0.11 (0.12) 0.05 (0.11)
0.18 (0.12) 0.09 (0.06) 0.45 (0.08) 1.02 (0.28)** 0.25 (0.11) 3,063.3

Notes *, ** and  represent p  0.05, p  0.01 and p  0.10 respectively. For 1989, data contain 4,524 relationships of 476 respondents, in 1993, 3,246 relationships of 417 respondents and in 1994, 3,156 relationships of 305 respondents; unstandardized coefficients. Emboldened words indicate variables on the level of ego. a Values within brackets indicate absolute t-values for non-significant coefficients.

42 Henk Flap and Beate Völker for the data collected in 1994. The results are given in Table 3.5. The right-hand column shows the absolute t-values for the difference between the 1989 and the 1994 model. Discussion of personal matters is still strongly associated with talking about politics but the coincidence is lower than before. The variable ‘discussion of personal matters’ has a significant random slope, indicating that the relationship between this variable and the dependent one differs among individuals. In 1989 no such differences were found. In 1994, talking about politics did not require talking about personal matters as well, or at least not in the same manner as before. Political issues can also be discussed with other persons, since confining these discussions to a circle of trustworthy people is now unnecessary. The changes in triad structures are documented in Table 3.4. Relationships within the niches are still highly connected (46 percent of all triads are closed and show the pattern   ). Further, there is a remarkable incline in the pattern  
, from 4 percent to 34 percent of all triads in the niches. This latter finding indicates the decline of strength of ties in niches which fall apart. In the relationships between niche and weak ties the number of open triads has become considerably smaller. The pattern   0 occurs in only 9 percent of the triads and the pattern 
0 and
 0 decreased from 30 percent to 22 percent. The patterns 

and

occurred more frequently in 1994, a finding that is also due to the decreasing strength of the niche relationships. While in 1989 the pattern   0 occurred most frequently, several different patterns became prominent in 1994. It appears that the relationships between the alters became more diverse, while the gap between niches and weak ties disappeared. The same applies for the triads outside of the niches. Although open triads still occur quite frequently among these ties, they occur less often than before and more alternative patterns are found instead.

Conclusions Our hypotheses on the effects of communism on social networks are well confirmed by our data. East Germans created niches that compensated for the collective organization of daily life, the lack of privacy, and the absence of political debate. Niche relationships were remarkably close, multiplex, characterized by repeated transactions and high occupational similarity. The niches were also particularly dense. People indeed seem to have closed ranks in their efforts to avoid adverse effects of institutional conditions. Further, East Germans created provision networks that compensated for the economy of shortages. Provision relationships were weak rather than strong, with low multiplexity. People avoided occasional acquaintances and kept those whom they did not know well apart from each other. The number of linkages among members of the niche and the weak relationships such as the provision network, as well as their strengths, was rather

The case of East Germany

43

low. There was a clear gap between the niche and the weaker part of the network. In general, East Germans’ networks were relatively small. Our analyses confirm the social capital argument that weak ties have perverse effects in communist societies and that there is therefore a specific division of labor in personal networks. Individuals select different partners for different kinds of transactions and the institutional context determines which transactions are made and with whom by presenting individuals with particular problems for achieving their general goals of well-being and approved identity (Lindenberg 1992). At the level of the entire network of all citizens of the GDR our argument implies that there were small islands of intimacy in a sea of instrumental sociability (see Misztal 2000). A promise of the Marxist utopia was to create social cohesion as well as social equality. Yet Marxist practice in the former GDR did not result in a society of comrades, but in a society with a sharp division between private and public parts of personal networks; that is, a society that lacked overall social integration. After the velvet revolution it appears that personal networks became disconnected from institutional conditions. The division of labor between the niche and the rest of the network consisting of weaker ties began to disappear. During the transition the niches became vague and people included more weak ties in their networks. Contrary to what we expected networks did not grow, although there where important changes: people got rid of untrustworthy others and established new, diverse contacts. Strength and multiplexity of niche relationships decreased. Average homogeneity of the whole network increased, while that of the niches decreased. Further, the density of all the network parts considered decreased. The coincidence of the relational functions ‘talking about politics’ and ‘talking about personal matters’ became weaker than before. As we expected, there is no need anymore to confine discussions of politics to a small circle of trustworthy others, to the niche. In 1994 the former GDR was still very much a society in transition and not yet a full-fledged capitalist, democratically well-established society. Many people did not know what their own interests were in the new situation or who might be helpful in promoting them. This probably explains why people’s networks did not grow. However, one can also argue that people did not greatly change the number of activities undertaken with members of a niche, since it saves time and costs if a few others are relevant for many relational functions; that is, if one’s existing network ties fulfill many of one’s needs, there is no great urge to enlarge one’s network (Lindenberg 1997). Perhaps most fundamental is that communism taught people not to trust strangers (for a more general discussion, see Misztal 2000: 208–28). Moreover, former GDR citizens nowadays suspect and eschew all organized life. Since all state-sponsored associations have collapsed and there were no others apart from the church, hardly any voluntary organizations exist under the new regime. Thus, apart from work, there are few places

44 Henk Flap and Beate Völker where people are forced to interact with relative strangers. Moreover, the absence of voluntary organizations, according to the classic Toquevillean argument (Putnam 1993), bodes ill for the chances of democracy and good government in ‘the new countries’ of Germany. In more general terms, our analysis showed that because there is an institutional conditioning of individuals’ returns of social capital, there is also such conditioning of their investments in social networks. This implies for recent discussions among political scientists that social capital at the aggregated level of a community or a society is also institutionally conditioned although decisions to invest in social capital are made by individuals (Glaser 2001).

Bibliography Agresti, A. and Agresti, B. F. (1978) ‘Statistical Analysis of Qualitative Variation’, in K. F. Schuessler (ed.) Sociological Methodology. San Francisco, CA: JosseyBass, pp. 204–37. Ash, T. G. (1981) ‘Und willst Du nicht mein Bruder sein. . .’, Die DDR heute. Spiegel Buch 15, Hamburg: Rowohlt. Bernard, R., Killworth, P. D., Kronenfeld, D. and Sailer, L. (1984) ‘The Problem of Informant Accuracy: The Validity of Retrospective Data’ Annual Review of Anthropology 13: 495–517. Bourdieu, P. (1981) ‘Le capital social. Notes provisoires’, Actes de la Recherche en Sciences Sociales 31: 2–3. Burt, R. S. (1984) ‘Network Items and the General Social Survey’, Social Networks 6: 293–339. Burt, R. S. and Knez, M. (1996) ‘Trust and Third-Party Gossip’, in R. M. Kramer and T. R. Tyler (eds) Trust in Organizations. Frontiers of Theory and Research. London: Sage Publications, pp. 68–89. Busschbach, J. van, Flap, H. and Stokman, F. (1999) ‘Stability and Change in Personal Networks. A Test of the Investment Theory’. Paper presented at the National Dutch Sociological Conference, Utrecht, 29 May. Coleman, J. S. (1990) Foundations of Social Theory. Cambridge: Belknapp Press. Diewald, M. (1995) ‘Informelle Beziehungen und Hilfeleistungen in der DDR: Persönliche Bindungen und instrumentelle Nützlichkeit’, in B. Nauck, N. F. Schneider and A. Tölke (eds) Familie und Lebensverlauf im gesellschaftlichen Umbruch. Stuttgart: Ferdinand Enke Verlag, pp. 56–75. Duijn, M. van, Busschbach, J. van and Snijders, T. (1999) ‘Multilevel Analysis of Personal Networks as Dependent Variables’, Social Networks 21: 187–209. Erikson, R., Goldthorpe, J. and Portocarero, L. (1979) ‘International Class Mobility in Three Western European Societies: England, France and Sweden’, British Journal of Sociology 30: 415–41. Fischer, C. S. (1982) To Dwell Among Friends. Personal Networks in Town and City. Chicago, IL: University of Chicago Press. Flap, H. (1999) ‘Creation and Returns of Social Capital’, La Revue Tocqueville 20: 1–21. —— (2002) ‘No Man is an Island: The research programme of a social capital theory’, in E. Lazega and O. Favereau (eds) Conventions and Structures in Economic Organization: Markets, Networks and Hierarchies. Cheltenham: Edward Elgar, pp. 29–59.

The case of East Germany

45

Gaus, G. (1983) Wo Deutschland liegt. Hamburg: Hoffman & Campe. Glaser, E. L. (2001) ‘The Formation of Social Capital’, Isuma. Canadian Journal of Policy Research 2: 35–40. Granovetter, M. (1973) ‘The Strength of Weak Ties’, American Journal of Sociology 78: 1360–80. —— (1985) ‘Economic Action and Social Structure: The Problem of Embeddedness’, American Journal of Sociology 91: 481–510. Hölder, E. (ed.) (1992) Im Trabi durch die Zeit – 40 Jahre Leben in der DDR. Stuttgart: Metzler Poeschel. Kornai, J. (1980) The Economics of Shortage. Amsterdam: North Holland. Lin, N. (2001) Social Capital. A Theory of Social Structure and Action. Cambridge: Cambridge University Press. Lindenberg, S. (1992) ‘An Extended Theory of Institutions and Contractual Discipline’, Journal of Institutions and Theoretical Economics 148: 123–54. —— (1997) ‘Grounding Groups in Theory: Functional, Cognitive, and Structural Interdependencies’, Advances in Group Processes 14: 281–331. Longford, N. T. (1988) ‘VARCL: Software for Variance Component Analysis of Data with Hierarchical Nested Random Effects (Maximum Likelihood)’. Manual and Software. Marsden, P. V. (1987) ‘Core Discussion Networks of Americans’, American Sociological Review 52: 122–31. —— (1990) ‘Network Data and Measurement’, Annual Review of Sociology 16: 435–63. Marsden, P. and Campbell, K. E. (1984) ‘Measuring Tie Strength’, Social Forces 63: 480–501. Mayer, K. U. and Solga, H. (1993) DDR-adaequate Schemata für die Abbildung der beruflichen Stellung, der beruflichen Tätigkeiten und Ausbildungen. Forschungsbericht, Berlin: Max-Plank-Institut für Bildungsforschung. Misztal, B. A. (2000) Informality. Social Theory and Contemporary Practice. London: Routledge Putnam, R. (1993) Making Democracy Work. Civic Traditions in Modern Italy. Princeton, NJ: Princeton University Press. Sik, E. and Wellman, B. (1999) ‘Network Capital in Capitalist, Communist and Postcommunist Countries’, in B. Wellman (ed.) Networks in the Global Village. Life in Contemporary Communities. Oxford: Westview Press, pp. 225–53 Snijders, T. A. B. (1999) SIGRPOS. Groningen: Groningen University. Snijders, T. A. B. and Bosker, R. (1999) Introduction to Multilevel Analysis. London: Sage. Srubar, I. (1991) ‘War der reale Sozialismus modern? Versuch einer strukturellen Bestimmung’, Kölner Zeitschrift für Soziologie und Sozialpsychologie 43: 415–32. Stark, D. and Bruszt, L. (eds) (1998) Postsocialist Pathways. Transforming Politics and Property in East Central Europe. Cambridge: Cambridge University Press. Völker, B. (1995) ‘Should Auld Acquaintance Be Forgot. . .?’ Institutions of Communism, the Transition to Capitalism, and Personal Networks: The Case of East Germany’. Amsterdam: Thesis Publishers. Völker, B. and Flap, H. (1997) ‘The Comrade’s Belief – Some Implications and Consequences: Neighborhood Relations in the Former GDR’, European Sociological Review 3: 241–65. Walder, A. G. (1994) ‘The Decline of Communist Power: Elements of a Theory of Institutional Change’, Theory and Society 23: 297–323.

4

Social networks in Russian politics Nicolas Hayoz and Victor Sergeyev

This chapter explores power relations in Russia and the corresponding structures of personalized trust on the basis of an analysis of three different types of distinctions and relations: state (capacity) versus society, institutionalized power versus personalized power networks and institutional structures of trust versus personalized networks of trust. As the case of Russia shows, personal networks of power can be presented as a compensation for the absence of strong institutions or institutionalized structures of trust. Moreover, in ‘low trust societies’ such as Russia, highly personalized power systems are not necessarily expressing only illegal or semi-legal structures of trust; in an environment without institutionalized social capital forms, they can also be the engine for reform. It was foreseeable that the social capital discussion would move east, as Martin Aberg (2000) already observed, to the countries of East-central Europe and, of course, to Russia, the country of the ‘hundred friends’ and connections. That networks of all types could be considered as resources for cooperation, whatever their form, has been noticed already by many observers of such societies in transition such as Russia, the Ukraine or Bulgaria. But so far the search for social capital and structures of trust in these countries has given only mixed results. Therefore, as everywhere in scientific descriptions, it may be useful to stress the obvious fact that we can see only what our concepts allow us to see. Speaking of positive or negative effects of social capital is not only a matter of the indicators used and their normative implications; it also points to the problem of defining and limiting the notion of social capital (Fukuyama 1997; Rose 1999). An observer with a ‘bright’ or optimist view of social capital will probably tend to find in every associational activity the seeds for creating generalized trust. By adopting a more realistic approach one may discover that such an image of society and state–society relations is too simple and that the structures of trust constitute something more than what can be found with a pattern, reducing the problems of societal development and its features (including structures of power) to a question of cooperation. In fact, even if basic mechanisms motivating cooperation such as coercion, interests, values and personal relations can be fixed in a theoretical manner, we

Social networks in Russian politics 47 are facing considerable difficulties in finding in practice causal and teleological explanations concerning the aggregated collective effects of different cooperative behavior. It is easy to notice that armed with a ‘civic community index’ (Marsh 2000) one may find social capital in the darkest ‘black holes’ of Russian provinces. Cooperation itself is not enough to create modernity. With regard to the above-mentioned lack of conceptual differentiation the purpose of this chapter is to investigate on the basis of a number of concrete examples the relation between different types of networks in Russian politics. It is also important to show the significance of a differentiation of networks of trust and networks of power within the power structures on the one hand and the general relations of trust in the state on the other. This points to the question about how Russia will be able to pass from the negative to the positive side of social capital, from the pattern of ‘particularistic’ reciprocity within personal networks to the norms of generalized reciprocity and networks of civic engagement (Dinello 1999: 32). First, we will present a conceptual framework of different types of networks and relations of trust between and inside networks, state and society. We offer a general view about studies in the field of social capital and its effects. Most social capital studies are concerned with more or less positive effects of certain social interactions and social institutions – a given amount of social capital – on civil society and the social competitiveness of a given society in general, thus creating the conditions for economic growth by reducing transaction costs of every kind and the development of a strong, trustworthy state. What has been said about positive social capital since the already classical studies of Coleman (1988) and Putnam (1993)1 is that there is a causal nexus between the density and quality of social networks and the efficiency of institutions of all kinds, whether political, economic, pedagogic or cultural. Typically the indicators measuring stocks of social capital in a society are found in the forms of civic engagement or the degree of trust (Levi 1998; Warren 1999). The crucial importance of resources symbolized by social capital has also been stressed by international organizations concerned with sustainable development such as the World Bank (2000: 18) or OECD. One may call this aspect the ‘good news’ or the optimist version of what social capital does or should do. However, it should be obvious, as many cases in Eastern Europe and other regions show, that social capital also has its ‘dark’ or perverse side (Petro 1999a); its negative implications: for example, the Sicilian mafia. Is it not a part of civil society? Do the corrupt power networks in Russia represent an engine for economic and political reform? As far as social networks are concerned, particularly networks of power, several authors are stressing those negative aspects of social capital (Levi 1996; Portes and Landolt 1996).2 But things are not so clear here since in the ‘gray zones’ of

48 Nicolas Hayoz and Victor Sergeyev particularistic networks there is room for many different patterns of personal networks which can contribute to the formation of certain forms of trust important for complex modern institutions. Therefore it may be important to know something more about the forms of this so-called ‘negative social capital’. How can we be sure about the negative or positive effects of social capital? Is it not rather so that the world is neither ‘black’ nor ‘white’ but rather ‘gray’, implying ambiguity (Sergeyev 1998)? Authors such as Bruszt and Stark have good reasons to consider the two faces of networks: they can be agencies of development or ‘rent-seekers depleting the public treasury and inhibiting economic growth’ (Bruszt and Stark 2000: 115). And is it really so clear that only horizontal networks should be associated with positive social effects due to norms of cooperation and trust? Norms of cooperation in any particular network are likely to change from one situation to another. They can represent social capital in one context and unsocial capital in another (Krishna and Shrader 1999).3 Of course in Russia they may be associated with corruption, nomenclature, post-communist clientelism and so on, but they can also play an important coordination role in and between the economy and the political system. Instead it would be more appropriate to consider a personal network in terms of ‘clique’; that is, a non-kinship informal association, within which exists group feeling and intimacy, as a group norm of behavior (e.g. Easter 1996: 558). What kinds of network should we consider particularly in Russia? Can we make a clear distinction between networks of civil society and criminal networks? It is also obvious that networks of trust with access to power systems have to be distinguished from networks without such links to the systems of political power. Looking closer, we may discover at least three problematic distinctions in the discussion about positive and negative effects of social capital which are either not reflected or are reflected only in a rather general way. First, there is the distinction between generalized interpersonal and institutional or systemic trust.4 It is a distinction between horizontal and vertical trust. Many consider institutional trust only as a problem of attitudes towards institutions, and neglect the difference between institutions and the holders of public functions. A difference which is also based on an epistemological gap is faced by individuals when they have to trust anonymous mechanisms of modern institutions rather than personally wellknown public actors (Offe 1999). Thus the modern state has to deal with the paradox (Migdal 1997) of an impartial ruler standing ‘above’ a society, which has nevertheless to be, and to be seen as, a trustworthy component of that society. Second, the clarification of the general problem is related to the distinction between networks of trust and networks of power, which should also include the distinction between old or pre-existing forms of social capital (for example, under communist rule) and present forms of networks. In

Social networks in Russian politics 49 addition, we have to be aware of the functional similarity of trust and power with regard to the problem of organizing collective cooperation. If cooperation cannot be based on trust it may be enforced by state power – a ‘suboptimal’ solution, as Putnam would call it. One should distinguish here the ‘Leviathan-like’ political power under conditions of absence of social capital and political power in the more modern sense of a legitimate, trustworthy and responsive state able to promote generalized trust in society. But trustworthiness is not enough to get things done and to solve problems of collective action. A legitimate power has to mobilize resources such as law, force and money in order to attain its goals and to protect civil society or markets. The problem is not in the means but in the way they are used: a repressive power does not need to rely on trust – at least not in the short term. On the other hand, incoherent or inefficient laws erode every power, legitimate or not, and affect the chance of generalized trust to become a factor of cooperation based more on trust in the rationality of general behavior than on personal bonds. We could say, then, that the more political power fails in preserving generalized trust, in being responsive, the more it has to ‘inflate’ its other means. In this regard one should not forget that the observance of laws by citizens is, as Margaret Levi has shown, conditioned by the public perception of the state, its institutions, and office holders as actors acting on behalf of the people (Levi 1998: 88). This brings us to the state/society distinction which points to the question of the role of the state in framing social interactions. The question of the negative and positive role of the state and the state-related networks is more crucial than the general discourse about positive and negative aspects of social capital. In this regard, Nicolai Petro makes an interesting distinction between a ‘non-statist’ view of social capital and a ‘statist’ view, which stresses the crucial role of the state as an agent promoting and creating social capital (Petro 1999b). For this purpose and in relation to state characteristics and activity, analysts of social capital forms in Eastern Europe are qualifying social networks mainly by using distinctions such as non-communitarian versus communitarian social capital or vertical versus horizontal networks (Putnam 1993; Aberg 2000; World Bank 2000). Richard Rose, for example, uses the distinction of modern, premodern and anti-modern networks in analyzing resources the Russians are mobilizing in their contacts with ‘the Byzantine’ Russian bureaucracy (Rose 2000). Noncommunitarian or ‘premodern’ social capital and particularistic networks seem to fulfill functions in transitional societies with ‘weak’ institutions or state–society relations dominated by generalized distrust.5 These particularistic networks are used by a significant part of the population to compensate for the absence of trustworthy and responsive state structures for attaining individual or group goals (see Aberg (2000: 313), who reached similar conclusions to those made by Rose (2000)). The compensatory

50 Nicolas Hayoz and Victor Sergeyev function of networks, especially clientelist networks, is also stressed by Andras Sajo who describes them as ‘a viable form of social organization where other networks and forms of social organization are nonexistent or at least underdeveloped’ (Sajo 1998: 42). At this point we have to differentiate. Looking at these types of network implies a bottom-up perspective, a view through the eyes of citizens and voters concerned by institutions and administrations which are not working the way they should. In this perspective – the daily life experience – non-communitarian social capital and particularistic networks seem to have a social function in those countries in the sense that they compensate for the lack of efficient formal institutions (Aberg 2000: 313; Rose 2000: 53). Personal networks are a problem-solving strategy (‘to get things done’). In some post-socialist countries they go hand-in-hand with the lack of generalized trust and trust in formal institutions and civic networks. Why should one trust the world beyond one’s family and the wider ‘family’ of one’s friends if this world is, as the case of Russia shows, perceived in ‘hobbesian’ terms, full of discriminations and exclusions, inequalities, greed, crime, and corruption?6 On the other hand, in a ‘topdown’ perspective things are different, which is why we have to remind ourselves that we stress particularly political elite networks, where personal trust and trustworthiness are means for achieving and maintaining power. Networks of clientelism or networks of power are being described as vertical quasi-institutional structures, for example, of political groups or business representatives around the heads of the regional executive (Lapina 1997; Ledeneva 1998). Here we face ‘amicable’ and informal structures of negotiation which deploy stabilizing political and social effects (Lapina 1997, 1999). Studies of clientelist networks in Eastern and East-central Europe refer to the survival of old network structures among the new political structures, which should be distinguished from civic networks on the one hand, and new power networks on the other (Holmes 1998; Dinello 1999; Hardin 1999). On the other hand, ‘traditional’ networks of power organize – by means of corruption, insider information and other strategies of exclusion – the uncertainty of ‘untrustworthy’ competitive market mechanisms, and thus contribute to vertical structures of exchange associated with ‘oligarchic capitalism’.7 The question of whether Palermo represents the future of Moscow may make sense here, but shows also some determinism (Hedlund and Sundström 1996: 32). Or should this mean that democratization has no chance in countries such as Russia and the Ukraine? Of course ‘Palermo’ also has the meaning of the survival of particularistic networks of trust based on the protection of corruption. Rose-Ackerman shows precisely how such a type of personal networks could maintain a system of generalized and normalized corruption (Rose-Ackerman 1999: 97).8 Obviously the kind of trust necessary for keeping illegal networks has to be distinguished from

Social networks in Russian politics 51 trust or distrust of political institutions. Under this aspect generalized distrust of institutions may very well go hand-in-hand with illegal personal systems of trust. Corruption has to do with the confusion of the border between public and private interests, a confusion that a ‘weak’ state can only reinforce (Levin and Satarov 2000). Furthermore, in Russia, corruption, together with patron–client relations, show a systemic character insofar as the dominant symbiotic pattern of political and economic ‘nomenclature elites’ on all levels continue the Soviet ‘tradition’ of personal networks, including the normalization of corruption. Corruption may indeed be considered ‘antisocial’ or ‘unsocial’ structures of trust in regard to their negative collective output (World Bank 2000). In such networks trust is of course limited and conditioned by the power resources of the participants in the system of contacts. For example, specific information about other ‘friends’ in the network may create a kind of strategic ‘community’ with some ‘mutual deterrence potential’.9 If social capital may be considered as a catalyst in improving the performance of economic and political institutions, it is at the same time likely that certain forms of social capital are obstacles to the establishment of those very same institutions.10 If we consider not only the ‘quasitraditional’ networks of everyday life and the corruption at all levels of the political system but look also at the degree of penetration of political and economic elites by the networks of organized crime we can hardly afford to be optimistic.11

Networks of trust as a negative social capital It is obvious that there are different types of networks in politics: to the networks of power and the networks of trust mentioned above we could add the related networks of influence and information. It is not our task here to present a typology of social networks. Instead we concentrate our attention on one specific type, i.e. networks of trust. On the one hand, we try to demonstrate how these networks contribute to the positive and negative social capital on different societal levels. On the other hand, we intend to show how crucial the networks of trust are to the survival of political groups in Russia. Let us look first at networks of trust conceived as negative social capital. The best examples of a negative role of social capital are the complex and un-balanced relations between the bureaucratic structures of the state and economic pressure groups of different types. In Russia, the most important and widely discussed subject is the so-called ‘oligarchic power’ of a small number of businessmen with ‘special relations’ within the government and presidential administration. For example, the role of media companies during the presidential elections of 1996 is widely acknowledged (Gel’man and Elizarov 2000: 33). Of course it is possible to say that these ‘corrupted relations’ also played a positive role in Russian

52 Nicolas Hayoz and Victor Sergeyev politics, preventing the victory of the communists. Nevertheless, such a position leads inevitably to a kind of double standard in the evaluation of ‘how democracy works’ in different countries. The widely known corruption scandals, such as those regarding privatization of the biggest Russian oil companies for very low prices and falsified tenders (Sergeyev 1998: 149),12 are obviously examples of how the networks of trust between government representatives and business destroy public trust in democratic institutions and the market economy. Another striking example of depleting public trust is the so-called ‘Harvard case’, in which a group of scientific advisers from Harvard University and a group of Russian specialists in privatization were accused of illegal use of confidential economic information about the state of some Russian enterprises and bureaucratic decisions, in favor of their own enrichment. In this case, the close relations of collaboration at international level may be considered as a kind of positive social capital. But obviously this capital has been instrumentalized in a negative sense. It is reasonable to ask about the institutional conditions which allowed so many respectable people to make these kinds of deal, and it is clear that such a case was possible only through the existence of a configuration of complicated networks of trust in both Russia and the United States (Stringer 2000).

Networks of trust as a tool for the survival of political groups In order to investigate these problems, the focus should be placed on concrete situations. For this purpose we present three cases, which illustrate some of the questions mentioned above and give us the opportunity to show how different types of social network interact. These cases are (1) the process of formation of the political movement ‘Democratic Russia’ during the final years of the USSR, (2) the formation of Gaidar’s government in October 1991, and (3) the appearance of the ‘Nijny Novgorod group’ during the spring of 1997. Other cases could have been selected, but, in order to cover not only different configurations of political power but also different periods including the first democratization process in the USSR after 1988, we choose to focus on these three cases, stressing the dynamic and interdependent relationships between networks of trust and networks of power. Multiparty politics was not institutionalized in Russia until 1991. This means that the most important role in political life has had to be played by informal networks. In a situation where politics was not a secure field of activity the problem of interpersonal trust was crucial. That is why not only dissidents but also the first legitimate political groups, which emerged at the beginning of the so-called ‘perestroika’ period, were based entirely on personal networks of trust. People organized within these groups had

Social networks in Russian politics 53 usually known each other for a long time, since they had been studying or working together for years. The most politically active part of the population comprised intellectuals from the Academy of Science and from different universities. Surprisingly enough, student activity was relatively low. The core of the new politics in 1988 to 1989 was represented by academics from the institutes of the Academy of Science of the USSR, mainly from Moscow, Leningrad and Novosibirsk. At the end of 1988, these people started to organize so-called clubs of voters, intending to give support for the March 1989 elections to the pro-reform candidates (Biryukov and Sergeyev 1997: 71–72). These clubs quickly assumed not only the task to support, but also to select some of the candidates. All this activity was basically of a grass-roots kind. Nobody paid for it and a lot of the work was done through private contacts. The most important role in this period was played by Sacharov, with his enormous intellectual and moral influence and through a very wide personal network. A successful campaign in favor of Boris Yeltsin has been launched through this network. These networks of academics later became the main information and public relations network for the pro-reform deputies of the Congress of People Deputies of USSR and for the interregional group created by them. The same networks of people became the core of the new political movement Democratic Russia, which, after the 1990 election for the Russian Congress of People Deputies, started to dominate the Soviet political landscape. It is important to note that the composition of this network of political activists was very uneven. It consisted of a number of prominent academics, but the vast majority of network members were junior researchers with rather unsuccessful scientific careers. Most worked in the field of physics, chemistry and computer science, and had only a scattered knowledge of politics and economics. They tended to believe in the wonders of market economy and democracy which should transform Soviet society immediately for the better, and they seemed to have a very limited understanding of the role of institutional design in these reforms. It was no surprise that many of them changed their political orientation immediately after the beginning of the reform process initiated by Yeltsin, and joined the communist and nationalist camps. Others abandoned any ideas of democracy in favor of pro-market and authoritarian anti-communist policy. The decay of Soviet science dealt another blow to this network. As a result it became weak, disabled and practically disappeared between 1992 and 1993 (Urban et al. 1997). The events of October 1993 and the dissolution of the Congress of People Deputies of Russia destroyed it completely and, at the same time, Democratic Russia disappeared. Some conclusions may be drawn from this experience. First of all, this network was a network of trust created within a social group that enjoyed high esteem in Soviet society, but that could never have gained direct access to political power. This fact has been well known by each member

54 Nicolas Hayoz and Victor Sergeyev of Soviet society. In the eyes of the public they could not represent the network of power, and could not be trusted as power holders. This is why they needed Yeltsin, who became famous simultaneously as a member of the old power network and as a democrat. The members of Democratic Russia failed to maintain influence over Yeltsin. They were not trusted by him, apart from a few individuals who soon lost their connections with the initial network. After the death of Sacharov, they also lost trust within the intellectual community, because none of them could present such an impressive record of dissidence to the Soviet regime and intellectual influence at the same time. Their support for the reduction of resources for Russian science and their effort to describe Soviet science as corrupt deprived them of the last traces of public trust in their own environment. They have since failed to transform their network of trust into solid political power. The existence of social capital in the sense of networks of trust as such is not enough to create new institutions or even to influence politics. This case illustrates how difference between the internal network of trust and the network of power reflected by the public can destroy political efforts at modernization. The second case that argues the importance of social networks in Russian politics regards the process of formation of Gaidar’s government. Strictly speaking, this government was created by Burbulis during September and October 1991 and was led by Yeltsin himself, yet Gaidar was its most important public figure (Biryukov and Sergeyev 1997: ch. 5). The government was formed largely on the basis of the network of trust, including mainly people from the Institute of Economics of Scientific and Technological Progress. The members of this government, including Gaidar himself and Chubais, succeeded in earning the reputation of ‘Western-educated economists’. This image was rather far from reality, since Gaidar had been a journalist of the communist party newspaper Pravda, Chubais had been a research associate in an obscure institute in Leningrad, and other members of the team could hardly be called Western-educated economists. The attitude towards members of the Gaidar government were very negative among the professional soviet bureaucracy. They were called ‘zavlaby’ (heads of laboratories) which signified that they were not considered as real administrators by the professionals (Gaidar 1996). In fact they were outside the real power network of the Soviet bureaucracy and could never cross this threshold to become ‘the same’ as traditional Soviet bureaucrats. This case illustrates a clear separation of the network of trust from the network of power. In only a few new governmental agencies, such as the State Committee for Property led at the time by Chubais, this network of trust corresponded to a certain extent to the network of power. It is worth noting that Gaidar’s network of trust without power could not endure for long in Russian politics.13 Most of these people were dismissed at the end of 1992. The example of Chubais is especially interesting, since

Social networks in Russian politics 55 his longer political life was due largely to the fact that he succeeded in maintaining an effective network of power. This case also shows another significant configuration of networks. Gaidar’s government has been considered by the members of Democratic Russia to be trustworthy specialists. Moreover, although Yeltsin trusted them, they have never been trusted by the administrative network of power, and they also failed to create relations of trust with the public, despite many efforts of the TV channels controlled by Yeltsin’s administration and the ‘oligarchs’. Finally, being connected by the network of trust only with the top leadership, they could not transform these relations into a real network of power (with the notable exception of Chubais). The third case refers to the promotion to power of a group of young politicians from Nijny Novgorod during 1997 to 1998. The leader of the group, Boris Nemtsov, established relations with Yeltsin after becoming a member of the Congress of Russian Deputies in 1990. From the beginning of 1988, several pro-reformist groups formed in Nijny. Among them was the Voter’s Club, which appeared at the same time in Moscow, and was later transformed into the ‘Candidates for Democracy’ Club. The beginning of Nemtsov’s political career was related to his active role in the struggle against the project to construct an atomic power-station near Nijny (then Gor’ky). Then, at the beginning of the 1990s, he positioned himself as one of the radical democrats championing economic reform. After being appointed by Yeltsin as a governor of Nijny, he nevertheless followed a rather authoritarian political course in relations with his political opponents, reproducing in Nijny the style of Yeltsin regime in Moscow. For this purpose he created a strong personal network of power which was, to a certain extent, of a clientelist type. When Nemtsov moved to Moscow as a first vice-prime minister, having this position guaranteed by Yeltsin at least until the following presidential elections, in 2000 he took with him to Moscow a group of people, including Kirienko14 and Brevnov. The first was appointed head of the Ministry of Fuel and Energy, the second head of RAO EES, the Russian electric power supply monopoly. Soon after, Brevnov was dismissed amidst accusations of mismanagement, but Kirienko managed to enter the circle of the close Yeltsin team and, in March 1998, after the resignation of Chernomyrdin, he was appointed prime minister. This was a bad time for the Russian economy. In December 1997, the Russian stock exchange collapsed and the state bonds (GKO) were issued in large quantities in an effort to improve the situation. The fall of oil prices in the spring of 1998 produced a new blow to the Russian stock exchange and state finances. In such a situation the government of Kirienko–Nemtsov refused to devaluate the currency. As experts argued later, a 30 percent devaluation in April could have saved the situation. Instead, the government started an unpopular and dangerous campaign for better tax collection, which at that moment was totally unrealistic. As a

56 Nicolas Hayoz and Victor Sergeyev result, the trust in government among the business circles and the public in general decreased. The sudden decision to proclaim default in August also destroyed the relations of trust between the Nijny group and Yeltsin, and has entirely broken public trust in Nemtsov and his group (Fish 2000: 20). The political combination of networks of trust and of power for the Nemtsov group was quite different from Gaidar’s case. In the situation of Nemtsov’s group there was a real network of power and trust, but the collapse of the group was determined by the collapse of public trust, produced by an extremely incompetent economic policy.

Conclusions We can see that the role of different types of network during a period of transition is very far from being strictly determined. If the history of the transformation process in Russia under the aspect of trust and particularly the dynamics of networks of trust were written, it would record a huge waste of popular goodwill, failure of office holders to build up trustworthy institutions, negative effects of the symbiosis between business and political networks of trust on political power and public trust, the power of ‘anti-modern’ networks and so on. Our aim has been more modest: to stress the two-faced character of networks, particularly in underinstitutionalized Russia, where nothing works without networks of trust and power, but where networks of trust are also useless if they are not connected or combined with the ‘right’ networks of power. In Russia, the problem of unbalanced informal connections with power is particularly virulent. The chances of achieving a positive collective output are diminished because these networks risk adopting strategies of exclusion in order to maintain the existing power configuration. These strategies of exclusion, enforced by a ‘weak’ state, promote personal trust instead of institutional trust. The absence of legitimate and efficient state structures offers opportunities for compensation structures represented not only by non-communitarian networks between the political system and society, but also by criminal networks such as Mafia groups or rent-seeking ‘oligarchic capitalists’ avoiding public control. The relationship of trust between public and political systems has to be distinguished from the networks of trust inside the political system and the relations of trust between them. Obviously the importance of public trust is very high, as the above mentioned examples have shown. Public trust is influenced positively by certain types of networks, such as information networks, by assuring the transparency of collective decisions and by diffusing balanced critical views towards the political institutions. Nevertheless, these networks are necessary but not sufficient for effective policy. An important question is: Which are the institutions allowing interpersonal networks of trust supported by the majority of citizens to transform and stabilize themselves into democratic networks of power necessary for

Social networks in Russian politics 57 the implementation of larger reforms? The cases analyzed in this chapter showed how networks formed outside of power structures, as elements of civil society, can be transferred to the state structures, where they become corrupted and lose touch with the networks outside of the state. Therefore we have no reason to connect directly the richness of societal networks with a certain type of social dynamics. It is important to distinguish between these networks in order to understand the specific interaction between them and their aggregation, and to be able to say what kinds of dynamic in society it is possible to expect.

Notes 1 For studies using (critically) the theories of Putnam and Coleman, see e.g. Braithwaite and Levi 1998; Levi 1996, 1998; Warren 1999; Pharr and Putnam 2000. 2 See also Putnam (2000) who speaks of ‘the dark side of Social capital’. 3 Portes and Landolt (1996) are concerned by the same problem when observing that ‘the same kinds of ties that sometimes yield public goods also produce “public bads”: mafia families, prostitution rings and youth gangs.’ 4 For generalized trust, whether in the sense of strategic or moralistic trust, see Levi 1998; Uslaner 2002. For institutional trust see Hardin 1998. For more general systemic trust or confidence – trust in social systems – see Luhmann 1988. 5 The notion of a ‘weak’ state is in fact misleading, since the state which is distrusted may be a very strong one and even cruel towards its citizens. It may be strong as an organization but absent as an institutional structure producing expectations and symbols of authority and legitimacy. 6 In this regard Russia is no different from the cases presented by Eric Uslaner (2002), referring among others to Edward Banfield’s study about the Italian village of Montegrano. In a society such as Russia with huge inequalities it should be no surprise that interpersonal trust is low and people do perceive strangers as threatening. Here also the truth is that ‘where you live – your context – shapes what you think’ (Uslaner 2002). 7 For the negative effects of such structures on reforms see Kukolev 1996; Afanasiev 1999; Dinello 1999; Diskin 1999; Hardin 1999. 8 See also Della Porta and Mény 1996; Gambetta 1996. With regard to Russia see Handelmann 1997; Sergeyev 1998; Lambroschini 1999. 9 This may include the publication of kompromat and the use of scandals in the struggle among and between different networks and clans. For the conception of social capital based on the cohesion of networks secured by fear see Jarmila Premusova, ‘I’ll Scratch Your Back, You Scratch Mine – Today’s “Social Capital”’, The New Presence (Prague), December 2000: 27–28. Gambetta (1988) would speak here, as in the case of the Mafia, of systems of distrust rather than of systems of trust. 10 See Hardin (1999) for a reflection on the mutually reinforcing relations between interpersonal capital and institutional capital. However, the author also observes that ‘interpersonal and institutional capital can similarly get in our way. Widespread customs and even very local practices of personal networks can impose destructive norms on people, norms that have all of the structural qualities of interpersonal capital. Institutional capital can also wreak its destructive hold, as in the Eastern nations that are now trying to build new

58 Nicolas Hayoz and Victor Sergeyev

11 12 13 14

economies and political systems while partially still in the grip of old institutions’ (Hardin 1999: 187). On the subject of organized crime, especially mafia groups, see Handelmann 1997; Sergejev 1998; Knabe 2000. The facts were revealed, for instance, by the Accountancy Chamber of the Federal Assembly (Boldyrev 1996). This point should also be considered in the context of reforms without institution building (See McFaul 1995). See Novoe Vremya. Interview with A. Chubais, N17, 3 May 1998.

Bibliography Aberg, M. (2000) ‘Putnam’s Social Capital Theory Goes East: A Case Study of Western Ukraine and L’viv’, Europe-Asia Studies 52: 295–317. Afanasiev, M. (1999) ‘Ispytyvaia politicheskie instituty’, Pro et Contra 4: 8–25. Biryukov, N. I. and Sergeyev, V. (1997) Russian Politics in Transition. Aldershot: Ashgate. Bova, R. (1999) ‘Democratization and the Crisis of the Russian State’, in G. B. Smith (ed.) State Building in Russia. London: Sharpe Publishing. Braithwaite, V. and Levi, M. (eds) (1998) Trust and Governance. New York: Russell Sage Foundation. Bruszt, L. and Stark, D. (2000) ‘Postcommunist Networking: Secret Agents, Mafiosi and Sociologists’, East European Constitutional Review, winter/spring: 115–20. Coleman, J. S. (1988) ‘Social Capital in the Creation of Human Capital’, American Journal of Sociology 94: 95–120. Coleman, J. S. (1990) Foundations of Social Theory. Cambridge: Cambridge University Press. Della Porta, D. and Mény, Y. (eds) (1996) Démocratie et corruption en Europe. Paris: La Découverte. Dinello, N. E. (1999) ‘The Russian F-Connection, Finance, Firms, Friends, Families, and Favorites.’ Problems of Post-Communism. January/February: 23–33. Diskin, I. (1999) ‘ “Rossiïskaia model” social’noï transformacii’, Pro et Contra 3: 5–40. Easter, G. M. (1996) ‘Personal Networks and Postrevolutionary State Building. Soviet Russia Re-examined’, World Politics 48: 551–78. Fish, M. S. (2000) ‘Superexecutive Power and Political Underdevelopment’, in V. E. Bonnell and G. W. Breslauer (eds) Russia in the New Century. Stability or Disorder. Boulder, CO: Westview Press. Fukuyama, F. (1997) ‘Social Capital: The Problem of Measurement’ (www.cipe.org/mdf1997/socicap.htm). Gaidar, E. (1996) Gosudarstvo i Evilutsiya. Moscow. Gambetta, D. (1988) ‘Mafia – The Price of Distrust’, in D. Gambetta (ed.) Trust: Making and Breaking Cooperative Relations. Oxford: Blackwell. Gambetta, D. (1996) ‘Civisme et corruption’, in D. Porta and M. Y. Donatella (eds) Démocratie et corruption en Europe. Paris: La Découverte. Ganev, V. I. (2000) ‘Notes on Networking in Postcommunist Societies’, East European Constitutional Review, winter/spring (http://www.law.nyu.edu.eecr/ vol9numonehalfreviews/notes.html).

Social networks in Russian politics 59 Gel’man, V. (1998) ‘The Iceberg of Russian Political Finance’, in A. W. Burnell (ed.) Funding Democratization. Manchester: Manchester University Press. Gel’man, V. and Elizarov, V. (2000) ‘Uchreditel’nye vybory v kontekste rossiyskoi transformazii’, Pervyi Electoral’nyi zikl v Rossii. Moscow: ves’ Mir. Gibson, J. L. (2001) ‘Social Networks, Civil Society, and the Prospects for Consolidating Russia’s Democratic Transition’, American Journal of Political Science 45: 51–69. Handelmann, S. (1997) ‘Stealing the Dream: “Bandit Capitalism” in the PostCommunist States’, in A. Karatnycky, A. Motyl and B. Shor (eds) Nations in Transit. New Brunswick, NC: Transaction Publishers. Hardin, R. (1998) ‘Trust in Government’, in V. Braithwaite and M. Levi (eds) Trust and Governance. New York: Russell Sage Foundation, pp. 9–27. Hardin, R. (1999) ‘Social Capital’, in J. E. Alt, M. Levi and E. Ostrom (eds) Competition and Cooperation. New York: Russell Sage Foundation, pp. 170–89. Hedlund, S. and Sundström, N. (1996) ‘Does Palermo Represent the Future for Moscow?’, Journal of Public Policy 19: 113–56. Holmes, S. (1998) ‘What Russia Teaches Us Now’, The American Prospect 33 (http://www.prospect.org/print/v8/33/holmes-s.html). Holmes, S. (1999) ‘Plaidoyer pour un Etat libéral et fort en Russie’, Esprit: 97–111. Knabe, B. (2000) ’Die Rolle der Systemmafia bei der Transformation Russlands’, in Bundesinstitut für ostwissenschaftliche und internationale Studien (ed.) Russland in Europa? Innere Entwicklungen und internationale Beziehungen – heute. Cologne: Böhlau. Krishna, A. and Shrader, E. (1999) ‘Social Capital Assessment Tool’, Prepared for the Conference on Social Capital and Poverty Reduction. The World Bank, Washington, DC, June 22–24. Kukolev, I. (1996) ‘Regional’nye elity: bor’ba za vedutchie roli prodolzhaetsia’, Vlast’ 1: 46–52. Lambroschini, S. (1999) ‘Russie: corruption et délitement de l’Etat’, Politique Internationale 83: 313–30. Lapina, N. (1997) Regional’nye elity Rossii. Moscow: INION. Lapina, N. (1999) ‘Milieux d’affaires et pouvoir politique dans les régions russes’, Le courrier des pays de l’Est 436: 27–33. Ledeneva, A. V. (1998) Russia’s Economy of Favours, Blat, Networking and Informal Exchange. Cambridge: Cambridge University Press. Levi, M. (1996) ‘Social and Unsocial Capital: A Review Essay of Robert Putnam’s Making Democracy Work’, Politics and Society 24: 45–55. Levi, M. (1998) ‘A State of Trust’, in V. Braithwaite and M. Levi (eds) Trust and Governance. New York: Russell Sage Foundation. Levin, M. and Satarov, G. (2000) ‘Corruption and Institutions in Russia’, European Journal of Political Economy 16: 111–32. Luhmann, N. (1988) ‘Familiarity, Confidence, Trust: Problems and Alternatives’, in D. Gambetta (ed.) Trust. Making and Breaking Cooperative Relations. Oxford: Blackwell. McFaul, M. (1995) ‘State Power, Institutional Change, and the Politics of Privatisation in Russia’, World Politics 47: 210–43. Marsh, C. (2000) ‘Social Capital and Democracy in Russia’, Communist and PostCommunist Studies 33: 183–99.

60 Nicolas Hayoz and Victor Sergeyev Migdal, J. S. (1997) ‘Studying the State’, in M. I. Lichbach and A. S. Zuckermann (eds) Comparative Politics. Cambridge: Cambridge University Press. Offe, C. (1999) ‘How Can we Trust our Fellow Citizens?’, in M. E. Warren (ed.) Democracy and Trust. Cambridge: Cambridge University Press. Petro, N. (1999a) ‘The Novgorod Region: A Russian Success Story’ in Post-Soviet Affairs 15: 235–61. Petro, N. (1999b) ‘Creating Social Capital in Russia: The Novgorod Model’, The Poland Library. (http://www.masterpage.com.pl/outlook/socialcapital.html). Pharr, S. J. and Putnam, R. D. (eds) (2000) Disaffected Democracies. Princeton, NJ: Princeton University Press. Portes, A. and Landolt, P. (1996) ‘The Downside of Social Capital’, The American Prospect 26 (http://www.prospect.org/cgi-bin/printable.cgi). Putnam, R. (1993) Making Democracy Work. Civic Traditions in Modern Italy. Princeton, NJ: Princeton University Press. Putnam, R. (2000) Bowling Alone. The Collapse and Revival of American Community. New York: Simon & Schuster. Rose, R. (1999) Modern and Pre-modern and Anti-modern Social Capital in Russia. Glasgow: CSPP Publications. Rose, R. (2000) ‘Uses of Social Capital in Russia: Modern, Pre-modern, and Antimodern’, Post-Soviet Affairs 16: 33–57. Rose-Ackerman, S. (1999) Corruption and Government. Causes, Consequences and Reform. Cambridge: Cambridge University Press. Sajo, A. (1998) ‘Corruption, Clientelism, and the Future of the Constitutional State in Eastern Europe’, East European Constitutional Review, spring (http://www.law.nyu.edu/eecr/vol7num2/special/special.html). Schulze, P. W. and Spanger, H. J. (eds) (2000) Die Zukunft Russlands. Staat und Gesellschaft nach der Transformationskrise. Frankfurt: Campus. Sergeyev, V. (1998) The Wild East. Crime and Lawlessness in Russia. New York: Sharpe Publications. Sergeyev, V. and Biryukov, N. (1993) Russia’s Road to Democracy: Parliament, Communism and Traditional Culture. Aldershot: Brookfield. Stark, D. and Bruszt, L. (1998) Postsocialist Pathways: Transforming Politics and Property in East Central Europe. Cambridge: Cambridge University Press. Urban, M., Igrunov, V. and Mitrokhin, S. (1997) The Rebirth of Politics in Russia. Cambridge: Cambridge University Press. Uslaner, E. (2002) The Moral Foundations of Trust. Cambridge: Cambridge University Press. Warren, M. E. (ed.) (1999) Democracy and Trust. Cambridge: Cambridge University Press. World Bank, The (2000) Entering the 21st Century. World Development Report 1999/2000. Oxford: Oxford University Press.

5

Social networks, civil society, and the prospects for consolidating Russia’s democratic transition James L. Gibson

Few concepts have captured the imagination of those studying democratization more than ‘civil society’. Dating back to ancient studies of political systems, theories of civil society claim that democracies require a particular set of autonomous institutions in order to check the power of the state (Putnam 1993). Furthermore, civil society is enhanced by efficacious attitudes and participatory behaviors on the part of the citizenry (Brehm and Rahn 1997). Civil society seems to be an essential condition for successful democratization. The importance of civil society has been reinforced recently by the momentous changes in the former Soviet Union and the states of Central and Eastern Europe. Many argue that nascent civil societies were instrumental to the democratic revolutions of the late 1980s, although to varying degrees across the countries experiencing democratic change (Stokes 1993; Tismaneanu 1995; Bernhard 1996). The democratization of Poland in particular is attributed to the ascendancy of Solidarity, a quintessential civil society organization (Bernhard 1996). Others, however, argue that the weakness of civil society in these countries today poses a threat to their ability to consolidate democratic change (Ekiert 1991; Arato and Cohen 1992). Although conceding that the emergence of a civil society was one of the reasons for the decline of Soviet communism in Central and Eastern Europe, they are nonetheless pessimistic that such a society can be expanded and consolidated through the subsequent stages of democratization. Many analysts argue that civil society in the mid- to late-1990s is being undermined by the radical individualism, social anomie and distrust, and plain, simple greed that characterize politics in these polities. Understanding civil society is therefore of substantial importance both for theories of democratization and the success of democratic transformations in Russia and the region. ‘Civil society,’ however, is a concept presenting formidable difficulties for empirically minded social scientists, especially since it has taken on myriad meanings. At the level of the political system, it refers most often to the existence of autonomous organizations (Putnam 1993). At the level of the individual citizen, civil society requires a specific set of attitudes and

62 James L. Gibson behavioral orientations toward politics, including a certain style of interpersonal interaction and collaboration. (Brehm and Rahn 1997; Lake and Huckfeldt 1998; Mondak and Gearing 1998). Inglehart (1997), for instance, argues that interpersonal trust, even without an explicitly political content, is the basic building block of civil society. When it comes to rigorous empirical work, not much agreement on the indicators of civil society characterizes the field (see the debate in Foley and Edwards 1998). I take a somewhat unconventional approach to analyzing civil society. Instead of examining the existence and functioning of formal organizations, I focus on the prevalence of social networks as a key attribute of civil society, especially in polities long dominated by totalitarianism. The antithesis of civil society is atomization – a condition in which each citizen is dissociated from every other citizen. In an effective civil society, I contend that citizens must not be atomized and that social networks must be well developed. Moreover, these social networks should be politically relevant – they must encourage and support discussion of politics among citizens. Most importantly, the social networks must be composed of ‘weak ties’; that is, they must span relatively heterogeneous segments of society, rather than being clan or kinship based. Thus, the central postulate of this research is that a vibrant civil society profits from well-developed social networks. Indeed, the existence of expansive social networks may well be the most important precursor to the development of the effective and autonomous political organizations believed to be necessary to the successful consolidation of a democratic transition. With the micro-level data at hand, I cannot test directly the hypothesis that such social networks contribute to systemic democratization. I can, however, determine whether those embedded in strong social networks tend to support the central institutions and processes of democracy and whether networks contribute to political discussion and deliberation. Relying on micro-level theories of social learning and political practice, I hypothesize that social networks are a means of transmitting innovative information and values in a society, and consequently that, in democratizing polities, those with more developed networks are more likely to adopt democratic values. Furthermore, to the extent that network interactions are politically relevant, the building blocks for the development of organizational activity are in place. Thus, this chapter not only describes the nature of social networks in Russia, but also tests hypotheses about the democratic consequences of social atomization versus integration. As I have noted above, interpersonal trust is often believed to be a crucial element in civil society. Trust is said to make cooperation (political, economic, social) possible, mitigating collective action dilemmas. While reconsidering the role of social networks in Russia, I also want to rethink the meaning of the concept ‘trust’ and re-examine how it too relates to attitudes toward democratic institutions and processes, and the willingness

Russia’s democratic transition 63 of citizens to engage in political discussions. Thus, trust and social integration are the two elements of civil society investigated in this chapter.

Russia’s uncivil society Few who study Russian political culture are optimistic about the development of a strong civil society in that country (Roeder 1999). In addition to the debilitating burden of hundreds of years of authoritarianism, contemporary Russia is said to lack two crucial elements of a civil society: interpersonal trust and a broad array of non-state voluntary organizations. For instance, Rose (1998: 11) found that 80 to 90 percent of Russians do not belong to any voluntary associations.1 Mishler and Rose (2001) report that Russians are the least trustful of any Central and Eastern Europeans (although this conclusion is undermined slightly by the fact that the metric for the responses to the question differed in Russia). Thus, many take contemporary data such as these as confirmation that Russian political culture is inhospitable to democratization. Too distrustful to form groups to promote their self and collective interests, Russians today continue their ancient tradition of extreme atomization, selfishness, and anomie.

Atomization, social networks, and democracy One important social impediment to democratic consolidation in Russia is that the legacy of totalitarianism may reassert itself. For instance, Bernhard (1996) argues that one of the primary objectives of Stalinism was precisely the destruction of civil society so that potential threats to monocratic rule could be exterminated. According to some, this legacy persists in contemporary political cultures and is manifested in learned helplessness, receptivity to paternalism, and a confrontational attitude towards conflict (Bernhard 1996: 323). Thus one form of ‘un-civil’ society is that which is characterized by social atomization: the absence altogether of social networks. In such a society, every individual is disconnected from every other. A second type of ‘un-civil’ society has entirely different characteristics – it is a society made up of strong but closed social networks. Granovetter refers to these as ‘strong ties,’ defining the ‘strength’ of a tie as a ‘combination of the amount of time, the emotional intensity, the intimacy (mutual confiding), and the reciprocal services which characterize the tie’ (1973: 1361). The family or clan is an exemplar of a network characterized by strong ties. Such networks tend to be internally homogeneous and cohesive, thereby inhibiting interactions with those outside the network and resulting in the atomization of small groups, if not individuals. Neither social atomization nor ‘strong ties’ is conducive to the development of civil society. ‘Weak ties,’ by contrast, facilitate civil society. ‘Weak ties’ characterize social networks that are not based primarily on family

64 James L. Gibson relationships and ‘are more likely to link members of different small groups than are strong ones, which tend to be concentrated within particular groups’ (Granovetter 1973: 1376; emphasis in original). Social interaction, especially outside the narrow confines of one’s family, contributes to the development of broader, less selfish, and more socially engaged attitudes. As Putnam put it: ‘Dense but segregated horizontal networks sustain cooperation within each group, but networks of civic engagement that cut across social cleavages nourish wider cooperation’ (1993: 175), which of course facilitates democratic governance. Thus, civil society profits from social networks characterized by relatively weak and hence permeable boundaries; such networks facilitate cooperation among citizens. In polities attempting democratic transitions, weak ties are especially effective at transmitting novel information about unfamiliar political institutions.2 In such societies, citizens are actively trying to learn the new ‘rules of the political game,’ and learning from fellow citizens is a crucial element of this process. New ideas spread with difficulty in a perfectly atomized society and find impediments to crossing strong ties.3 Thus a reasonable hypothesis is that those embedded in social networks characterized by weak ties are most likely to adopt the emergent values of a new regime, and, in this instance, to endorse democratic institutions and processes. Of course, weak ties are capable of diffusing any type of information, not just information favoring democracy. Novel information, of every sort, is the type of information transmitted most effectively across weak ties. It is possible, therefore, that weak ties might also facilitate the dissemination of information undermining democratic governance. Indeed, in stable democracies, weak ties may well contribute nothing to social learning about democratic institutions and processes, since the information is readily available to all. The effect of weak ties in promoting democratic values is most likely greatest in transitional regimes, where citizens have little experience of democratic governance, and where democratic institutions and processes are neither well understood nor discussed. Social networks, to the extent that they are heterogeneous and politically relevant, also contribute to democratic values through the simple process of political discourse. Mutz (1999) concludes that ‘cross-cutting exposure’ (by which she means political discussions with non-like-minded network members) contributes substantially to the development of democratic values in the American mass public. Thus ‘weak’ social networks that are politically relevant may contribute to the development of democratic values through processes of diffusion and through practice in democratic discussion. Soviet scholars have long recognized the importance of networks such as the ones I consider here. So-called ‘kitchen circles’ – ‘groups of friends who met in the kitchens of their apartments and led endless conversations about the meaning of creation, art, and politics’ (Greenfeld 1992: 23) – have long been an integral part of Russian political culture and politics.4

Russia’s democratic transition 65 Thus, in this chapter, I am especially interested in the following attributes of social networks: • • •

The size of social networks: How extensive they are. Network politicization: The degree to which network members engage in political discussions. The ‘strength’ of the ties among network members: The degree to which ties cross social groups.

Although I cannot claim that large, politicized, and ‘weak’ social networks are a sufficient condition for an effective civil society, they are most likely necessary for the consolidation of democratic transitions.

Research design This analysis is based primarily upon a panel study of the Russian mass public, initiated in 1996, and continued in 1998. The overall focus of the survey was on attitudes toward democratic institutions and processes. In the first wave, face-to-face interviews were completed in 1996 (in the period of the run-up to the first round of voting in the Russian presidential election). Non-institutionalized residents of Russia who were 16 years old and over were eligible to be interviewed. The sample is representative of the entire territory of Russia. Interviews were attempted with 2,442 respondents, with a resulting response rate of 84 percent. Individual respondents were selected using the Kish selection method (Kish 1965: 398–401) and consequently the data are weighted to reflect the size of the household. Local interviewers were used, but they were trained by project personnel traveling from Moscow. The average length of the interview was 87 minutes (standard deviation  31 minutes). As is common in surveys in Russia and elsewhere, women were slightly overrepresented in the sample (58 percent female in the sample versus 55 percent in the population). The second wave of the panel was fielded in 1998. The response rate was 82.7 percent. Generally, this second-wave response rate is quite high by comparative standards.5 For very limited purposes, I take advantage of a survey we conducted in the entire Former Soviet Union in 1992 (see e.g. Gibson 1997). This survey included a representative subsample of Russia. The methodology of that survey in virtually every respect duplicates the methodology of my 1996 to 1998 panel, even to the point of using the same survey team in Russia. In addition, I make use of a seven-country survey I conducted in 1995. Representative samples of the mass public were interviewed in Bulgaria, Hungary, Poland, Russia, France, Spain and the United States. Noninstitutionalized residents of each country aged 18 years and over were eligible to be interviewed.

66 James L. Gibson The structure of social networks Borrowing from the (U.S.) General Social Survey, we began our discussion of social networks with the following question: From time to time, most people discuss important matters with other people. Who are the people with whom you discuss such matters? Just tell me only their first names or initials. IF LESS THAN 3 NAMES MENTIONED, PROBE: Anyone else? ONLY RECORD FIRST 3 NAMES. Table 5.1 reports data on the number of network members identified. The Russian respondents in 1996 claimed fairly extensive social networks. The average number of network members identified was two (out of a possible maximum of three), and 43.4 percent were capable of naming Table 5.1 Attributes of social networks The size of social networks Country

Russia, 1995 United States Russia, 1996 France Poland Spain Bulgaria Hungary Russia, 1992

Number of members identifieda

Percentage No network members

Three network members

Mean

Std. dev.

19.8 10.4 12.9 19.9 16.1 13.2 13.1 9.8 7.6

39.1 31.6 43.4 55.0 54.9 55.0 56.1 61.4 60.8

1.79 1.80 1.99 2.02 2.08 2.13 2.13 2.27 2.33

1.16 1.00 1.06 1.22 1.15 1.10 1.11 1.04 0.97

N 772 810 2,059 762 824 775 1,180 777 2,536

The Politicization of Social Network – Talking Politics Country

Russia, 1995 United States Russia, 1996 France Poland Spain Bulgaria Hungary Russia, 1992

Percentage discussing politics (1st network alter)

Frequency of discussions (all network alters)b

Never

Daily

Mean

Std. dev.

34.5 5.7 6.0 12.0 5.5 8.2 11.1 9.0 3.1

8.4 18.1 38.1 14.3 27.3 27.4 33.8 33.0 48.1

2.36 4.34 4.84 3.97 4.50 4.38 4.41 4.45 5.04

1.32 1.20 1.16 1.36 1.19 1.35 1.49 1.39 0.95

N 611 723 1,783 610 686 662 1,013 701 2,334

Russia’s democratic transition 67 Table 5.1 Continued The prevalence of ‘weak’ network ties Country

Russia, 1995 United States Russia, 1996 France Poland Spain Bulgaria Hungary Russia, 1992

Number of weak tiesc

Percentage No weak ties

All weak ties

Mean

Std. dev.

44.7 49.6 39.5 42.5 51.3 38.8 46.7 52.5 24.5

8.8 4.7 9.6 11.8 10.1 12.4 10.7 6.5 17.7

0.89 0.73 1.01 1.03 0.86 1.08 0.92 0.77 1.39

0.98 0.87 1.00 1.06 1.04 1.05 1.03 0.95 1.04

Mean

Std. dev.

3.37 3.83 4.39 4.50 4.54 4.59 4.71 4.76 5.66

2.37 2.34 2.86 2.77 2.77 2.67 2.54 2.77 2.59

N 772 810 2,059 762 824 775 1,203 786 2,536

Network political capacityd

Russia, 1995 United States France Poland Bulgaria Russia, 1996 Hungary Spain Russia, 1992

N 772 810 762 824 1,203 2,059 786 775 2,536

Notes a   0.16, p  0.000. c   0.21, p  0.000. d   0.21, p  0.001. b   0.47, p  0.000. The capacity index ranges from 0 to 9, with high scores indicating greater network capacity.

an entire complement of network alters. Only a very small proportion of Russians was unable to name any network members. Compared to social networks in our seven-country survey, the number of network members mentioned by the Russians in 1996 is about average. The Russian mean of 1.99 is about the same as France (2.02), considerably higher than the United States (1.80), but lower than Hungary (2.27). The figure is also higher than the results from the comparable (but smaller) survey in Russia in 1995 (x¯ 1.79).6 These data yield little evidence of a distinctive legacy of social atomization in Russia, at least in terms of isolating Russians from each other. Table 5.1 also reports results from a survey we conducted in Russia in 1992. These data provide some purchase on the question of how social networks may have changed in Russia over the course of the 1990s. In 1992, most Russians (60.8 percent) named three network members, and the

68 James L. Gibson average number of alters named is the highest of any of the countries. The decline in the number of network alters identified in 1996 is substantial and no doubt reflects the diminishing necessity of social networks in the economic domain (e.g. the ready availability of goods), as well as the increasing strains on ordinary Russians to make a decent living (see Greenfeld 1992). No clear East–West differences exist in the size of the social networks. Indeed, Hungary has the highest percentage of people with extensive networks, while Russia (1995) and the U.S. have the lowest percentages. The figures in Bulgaria, France, Poland, and Spain are similar. Ironically, one of the most democratic countries in the group (the United States) has one of the lowest percentages of respondents with large networks, while perhaps the least democratic country (Bulgaria) has a considerably higher percentage. Thus, there seems to be little direct relationship between democracy and social networks. But are these networks politically relevant? We also asked the respondents to describe the extent of political discussions with each of their network members. Their responses are also reported in Table 5.1. Considerable variability exists in the extent of political discussions within these social networks.7 In Poland, Hungary, and Bulgaria, such discussions occur with remarkable frequency. The Americans are below average in the frequency of discussing politics, quite similar to the Spanish, and in every instance their mean score is higher than the French. In general, these social networks are important centers of political discussions in most of these countries. Cross-national differences in network politicization are greater than are such differences in network density (as indicated by the ETA coefficients in Table 5.1. Russia in 1995 differs considerably from the other countries, with between a quarter and one-third of the respondents claiming never to discuss politics with their network members. Russian social networks are by far the least politicized of the countries in the 1995 surveys. The differences between Russia and Bulgaria (to focus on another Slavic country undergoing difficult change) are stark indeed. By 1996, however, Russian social networks were extremely politicized, with the Russians scoring the largest mean level of political discussion of any of the countries (apart from the Russians in 1992). Consider first political discussions with the first-named network member. In Russia, political conversations are very common, with nearly 80 percent of respondents talking about politics at least weekly (data not shown). Political discussions with other network members are also quite common. In Russia in 1996, 38.1 percent of the respondents claimed to have political discussions nearly daily, as compared to 18.1 percent of respondents in our 1995 survey in the United States. Russian networks obviously carry considerable political content. It appears that the 1995 Russian data may have been affected heavily by political alienation and disgust during the particu-

Russia’s democratic transition 69 lar period of the survey (surrounding the parliamentary elections of 1995) and may well represent a short-term aberration. The final attribute of networks – whether they are based on ‘strong’ or ‘weak’ ties – is more difficult to measure. Following Granovetter (1973) and Huckfeldt et al. (1995), I assume that strong ties are present when networks are composed of members of the same family, and that weak ties characterize relationships across family boundaries. This assumption is based on the belief that, at least among those naming a family member as part of the social network, within-family relationships are stronger. Thus, the number of non-family network members may be used as an indicator of the prevalence of ‘weak’ ties. Table 5.1 also reports these data. A high percentage of people rely exclusively upon family members for their social networks (‘no weak ties’). The percentage naming no nonfamily network members ranges from 24.5 percent in Russia (1992) and 38.8 percent in Spain to 52.5 percent in Hungary.8 Much lower percentages draw all of the network members from non-family sources. According to the mean scores, ‘weak’ ties are most common in Russia (1992 and 1996), Spain, and France, and are least prevalent in Hungary. The American respondents score approximately at the mean of the seven countries. Yet cross-national differences are relatively small in comparison to withincountry differences. Once more, a simple ‘legacies of communism’ explanation is insufficient to account for these data. Weak ties are relatively common in France, but so too are they in Bulgaria. It appears from these data that social networks bear few scars from the era of communist domination in Russia and Central and Eastern Europe. A summary indicator of the potential for political effectiveness of each network tie would be useful for the hypothesis testing that follows. To construct such a measure, I scored each tie according to its strength and degree of politicization (and, of course, its existence). Summing across the three network ties yields an index of the political capacity of the respondent’s social network.9 Capacity is the ability of the social network to transmit political information and to provide experience of politics through political discussions. Networks with high capacity are broad (including many members), politicized (in the sense of talking about politics being common), and ‘weak’ (network ties transcend family boundaries). The index varies from 0 (social isolates, without any network partners) to 9 (all three network alters are weak ties, and the network is highly politicized). According to this measure (see Table 5.1), social networks should be most effective at facilitating social learning in Spain and Hungary, and least effective in Russia (1995) and the United States. The Russians in the 1992 survey stand out as distinctive. This capacity measure is moderately related to level of education, interest in politics, and (self-proclaimed) opinion leadership in each of the countries. It is not, however, associated with the respondent’s gender. This cross-national analysis reveals that Russians have extensive social

70 James L. Gibson networks that are highly politicized and that often transcend family units. Although such networks are also found in some parts of the West (e.g. Spain), Russian social networks (at least in 1996) are distinctly different from such networks in the United States. In terms of political capacity, it appears that Russian networks have considerable potential for shaping political values.

Summary Russian social networks may well have emerged primarily as a response to the repressive state. Unable to organize publicly, Russians may have substituted private social networks for formal organizations. But Russians are not atomized, and, as a consequence, Russian social networks have a variety of characteristics that may allow them to serve as important building blocks for the development of a vibrant civil society. In addition to carrying considerable political content, these networks are characterized by a relatively high degree of trust. Because the networks are not closed (strong), they link Russians together to an extent not often recognized by most analysts.

Interpersonal trust Many theories of civil society assume that interpersonal trust is essential to successful democratic governance. Mishler and Rose describe the conventional wisdom when they assert: ‘Trust is necessary [to the establishment of civil society] so that individuals may participate voluntarily in collective institutions, whether in political institutions, such as political parties, or in economic and social institutions, such as labor unions, business associations, and churches’ (1997: 419). As a consequence, there has been a tidal wave of research recently on interpersonal trust in a wide variety of contexts (Braithwaite and Levi 1998; Norris 1999). Several scholars argue that, at least at the macro level, interpersonal trust and effective democratic governance are intimately interconnected (Putnam 1993; Inglehart 1997), even if there is debate about the causal structure of trust (Brehm and Rahn 1997). How trustful are Russians? The conventional measure of interpersonal trust asks whether ‘most people can be trusted’ or whether one ‘can’t be too cautious when dealing with people.’ Table 5.2 reports the responses of the Russians in my surveys, as well as the responses from a variety of surveys compiled as part of the World Values Survey.10 The World Values Surveys were conducted in the early 1990s. Russians in 1998 expressed more distrust than trust – 59.0 percent said one cannot be too cautious, while only 30.5 percent claimed that other Russians can be trusted. At first glance these figures seem to indicate low levels of trust, but in comparison to the countries in the World Values

Russia’s democratic transition 71 Table 5.2 Cross-national differences in interpersonal trust Country

Sweden Switzerland Norway Finland China Denmark Netherlands Canada Russia, 1992 United States West Germany Ireland Japan Iceland Great Britain Northern Ireland Russia, 1996 Russia, 1998 India Italy Austria Mexico Belgium Spain Moscow East Germany Poland South Korea Bulgaria Lithuania South Africa Nigeria Czech-Slovak France Estonia Hungary Argentina Byelorus Chile Portugal Slovenia Latvia Romania Turkey Brazil

Percentagesa

Weighted N

Cautious

Uncertain

Trust

30.6 34.9 34.9 35.6 39.6 40.8 43.1 46.0 47.9 48.6 51.3 52.3 52.5 54.1 54.6 55.4 58.2 59.0 59.7 60.2 60.8 60.8 61.6 62.2 64.4 65.4 65.5 65.8 66.0 69.2 71.7 71.9 72.2 72.4 72.4 73.1 73.6 73.6 75.1 76.2 77.7 81.0 82.4 88.9 92.5

9.8 38.6 0.0 4.5 0.7 3.6 4.5 3.2 14.9 2.8 17.4 0.7 9.9 4.0 3.3 1.7 6.2 10.5 9.2 4.3 10.9 8.6 7.8 6.0 2.8 12.0 0.0 0.0 5.2 0.0 0.0 6.7 0.0 6.3 0.0 3.0 4.1 1.3 2.8 3.0 6.0 0.0 1.8 1.3 0.9

59.6 26.5 65.1 59.9 59.7 55.6 52.4 50.7 37.1 48.6 31.2 47.0 37.6 41.9 42.1 42.9 35.6 30.5 31.2 35.5 28.4 30.6 30.6 31.8 32.9 22.5 34.5 34.2 28.8 30.8 28.3 21.4 27.8 21.4 27.6 23.8 22.4 25.1 22.1 20.8 16.3 19.0 15.8 9.9 6.6

1,047 1,358 1,156 584 1,323 154 521 1,727 2,529 2,004 3,059 179 2,224 700 2,805 79 2,094 1,719 2,511 2,280 1,460 1,515 307 2,059 1,010 1,328 852 1,229 1,031 1,000 1,622 460 1,394 2,806 1,008 998 1,002 1,010 1,500 477 1,034 903 1,103 1,025 1,747

Source: World Values Survey, conducted in the early 1990s. The countries are ranked on the percentage asserting one cannot be too cautious in dealing with people. Note a Percentage total to 100% (except for rounding errors).

72 James L. Gibson Survey, Russians are not particularly distrustful. The percentage of Russians responding that one cannot be too cautious is lower, for instance, than the percentage in Spain, France, and Portugal (focusing on Western European countries) and is dramatically lower than most countries in Eastern and Central Europe. Indeed, if we take sampling error into account, the Russian data are not that different from the British data. This is a remarkable finding in light of the political, social, and economic turmoil in contemporary Russia. Unlike most research on interpersonal trust, we have measured trust along several different dimensions. First, we asked Russian respondents the traditional interpersonal trust question. Second, we modified that question by delimiting the object of trust to those whom the respondent knows. Third, we asked the respondents to estimate how much they are trusted by others. Finally, of course, we measured the level of trust between the respondent and each of the network alters. The replies to these queries are reported in Table 5.3. The frequencies of these different aspects of interpersonal trust differ markedly depending upon the specific object being addressed. Trust in people in general may be relatively low in Russia (less than one-third are Table 5.3 Different facets of interpersonal trust, Russia, 1998 Measure

Percentages

N

Distrustful Uncertain Trustful Total Generally speaking, do you agree that most people can be trusted, or that you can’t be too cautious when dealing with people?

59.0

10.5

30.5

100.0

1,719

And what about people you know – can they be trusted, or do you have to be very careful when dealing with people whom you know?

27.3

13.1

59.6

100.0

1,720

5.0

26.1a

69.0

100.0

1,773

Second network member

7.5

31.1

61.4

100.0

1,408

Third network member

7.9

30.6

61.6

100.0

1,881

33.1b

11.6

55.3

100.0

1,714

Trust in social network members First network member

Do you think that most people you know trust you?

Notes a This figure combines the center-point on a 5-point scale with those who are uncertain how much they trust the network member. b This figure includes 1.5% who say that ‘most people do not trust me’ and 31.6% who say that ‘some do trust me, others do not.’

Russia’s democratic transition 73 trustful), but trust in acquaintances is relatively high (nearly 60 percent). Members of social networks are trusted even more, and most people estimate that others trust them. But when it comes to ordinary Russians, levels of trust among people who know each other are high and approximately double the levels among strangers. These various measures of interpersonal trust are not strongly intercorrelated. Especially significant is the non-existent relationship between trust in strangers and trust within the social network (r  0.03). Those who trust people whom they know tend to trust strangers (although remember the differences in the frequency distributions: see Table 5.3), but clearly interpersonal trust has quite different manifestations depending upon the object of trust. Upon reflection, this finding is not surprising. It is instructive to consider further the relationship between trust in people in general and trust in acquaintances. Table 5.4 reports this relationship, using trichotomized versions of each variable. Almost all (93.7 percent) of those who trust strangers (people in general) tend to trust people whom they know. This finding is not surprising, but those who are uncertain about strangers also tend to trust people they know (62.2 percent). Most importantly, 41.4 percent of those who are cautious with strangers nonetheless tend to trust people whom they know (with another 15.7 percent claiming uncertainty). These figures provide a useful perspective on the meaning of trust in people in general. Of those distrustful of strangers, less than a majority (42.9 percent) are distrustful of people they know, and of those uncertain whether to trust strangers, only a tiny proportion (8.3 percent) is distrustful of acquaintances. These findings may well limit the debilitating effects of low levels of trust among Russians who do not know each other. Which type of trust is important to civil society? The answer to this question is not entirely unambiguous. To the extent that democracy requires political cooperation among strangers, these findings are discouraging for democrats. But to the extent that democratic cooperation is Table 5.4 The relationship between trust in strangers and trust in acquaintances, Russia, 1998 Trust in acquaintances

Cautious Uncertain Trust Total N

Trust in strangers Cautious

Uncertain

Trust

42.9 15.7 41.4 100.0 1,012

8.3 29.5 62.2 100.0 181

3.9 2.4 93.7 100.0 525

Note Entries are percentages (except for N).

74 James L. Gibson based on friendships and acquaintances (but not family members), these data are encouraging. Indeed, perhaps the significance of these data turns entirely on how easy it is for Russians to convert strangers into acquaintances – that is, the degree to which the line between people whom one knows and people whom one does not know is easily traversed. After all, political cooperation does not take place among perfect strangers. Weak, not strong, social networks facilitate converting strangers to friends.

The consequences of Russian social networks What consequences do these network attributes have for attitudes toward democratic institutions and processes? Following earlier research, I have measured support for democratic institutions and processes along five separate sub-dimensions: (1) the relative value of social order and individual liberty; (2) support for a free and pluralistic media; (3) support for competitive elections and a multi-party system; (4) support for dissent; and (5) rights consciousness. Through Common Factor Analysis, I have created a summary index of support for democratic institutions and processes.11 Given the unidimensional structures, I created a measure of support of democratic institutions and processes in 1996 and 1998 by averaging the responses to the five indices. The correlation between democratic values in 1996 and 1998 is 0.43, indicating considerable stability in these values. Following the logic outlined above, I hypothesize that those having social networks with strong political capacity are more likely to support democratic institutions and processes, as are those with higher levels of interpersonal and network trust. The equations relevant to these hypotheses are shown in Table 5.5. Table 5.6 reports two models. The first is based on an equation in which only the three measures of trust are included as independent variables. Model II adds four control variables often found to be useful predictors of support for democratic institutions and processes (see Bahry 1993). Consider first the degree to which the various measures of trust are related to these democratic values. None of the standardized regression coefficients for the trust measures exceeds 0.10 (which is a low threshold for survey data). Although trust generally has little predictive capacity, it is noteworthy that the strongest relationships are found with the measures of whether people trust the respondent and whether people whom the respondent knows are trustworthy, not the measure of trust in people in general. These relationships are certainly slight, but general trust is an especially weak (and statistically insignificant) predictor of democratic values. Particularly when it comes to support for democratic institutions and processes, trust seems to have only the most marginal influence (at best). Table 5.6 reports a similar analysis for the measure of network political capacity. Note first that the measure of network political capacity is

Russia’s democratic transition 75 Table 5.5 The impact of trust on support for democratic institutions and processes, Russia, 1998 Model I .r Trust in people in general Trust in people whom you know Trust in me Education Age Rural residence Gender N R Standard deviation – Dependent variable Standard error of estimate

.b

se

Model II .

.b

se

.

0.05
0.00 0.01
0.01

0.00 0.01

0.02

0.09

0.02 0.01

0.07*

0.02 0.01

0.06*

0.10 0.28
0.27
0.13
0.10

0.04 0.01

0.08**

0.04 0.03
0.00
0.05
0.08

0.01 0.07** 0.00 0.22*** 0.00
0.21*** 0.01
0.11*** 0.02
0.11*** 0.40***

0.38

1,702 0.11*** 0.38

0.38

0.35

1,708

Notes *** p  0.001; ** p  0.01; * p  0.05.

Table 5.6 The impact of network political capacity on support for democratic institutions and processes, Russia, 1996 Model I .r Network political capacity Education Age Rural residence Gender

.b

0.16

0.03 0.00

0.35
0.25
0.18
0.08

N R Standard deviation – Dependent variable Standard error of estimate Notes *** p  0.001; ** p  0.01; * p  0.05.

se

Model II . 0.16***

.b

se

0.02 0.00

. 0.10***

0.05
0.00
0.07
0.07

0.00 0.29*** 0.00
0.17*** 0.01
0.15*** 0.02
0.08*** 0.44***

0.43

2,038 0.16*** 0.42

0.42

0.38

2,053

76 James L. Gibson moderately related to support for democratic institutions and processes in the bivariate analysis. Those embedded in more politicized and ‘weak’ networks are more likely to support the main institutions and processes of democratic governance. The multivariate controls reduce the magnitude of this relationship somewhat, but even in Model II the standardized regression coefficient is highly statistically significant and exceeds 0.10. In general, social networks appear to be an important source of learning – from others and from experience – about the meaning of democratic institutions and processes. This finding is substantively significant since it withstands the control for powerful but somewhat amorphous predictors of democratic values (in the sense that demographic factors typically represent a multitude of specific micro-level processes). There are several reasons why these correlations are not stronger. First, Russia has been in the democratization process for several years now, making democratic ideas less novel than they once were. Second, being embedded in social networks with high political capacity is perhaps a necessary but not sufficient condition for access to democratic ideas. Without such networks, diffusion is difficult; but the existence of a network does not ensure that democratic ideas will flow if no network members are exposed to democratic values. Third, not all Russians require social networks to the same degree in order to learn about social norms.

Discussion and conclusions Several important conclusions emerge from this research. •

•

It is useful to think of ‘civil society’ as a concept including the social networks within which individual Russians are embedded. Ordinary Russians are far from being atomized, perhaps suggesting an unexpected legacy of the totalitarian period. Instead, Russian social networks are politically charged, dense, and characterized by high levels of trust, but only moderate levels of political agreement. Most importantly, Russian social networks are politically ‘strong’ because they are ‘weak.’ Contrariwise, interpersonal trust is a concept of limited utility for understanding processes of democratization. Trust generally seems to be fairly unimportant, but trust in strangers seems to be particularly insignificant. Earlier research suggesting (and, to a limited degree, finding) support for the trust–democratic values hypothesis has probably been confounded by the difference between trust in strangers and trust in acquaintances. For most Russians, interpersonal trust is certainly not a prerequisite to support for democratic institutions and processes.

Taking a step or two away from the data, I can construct a somewhat speculative picture of Russian civil society. Russia seems not to have the wide variety of autonomous, relatively formal organizations that are said to

Russia’s democratic transition 77 characterize vibrant democracies in the West. And social networks, even ‘weak ones,’ are not an alternative manifestation of a civil society. For democracy to be entirely consolidated in Russia, these social networks must at some point evolve into more formal organizations. Social networks develop in response to an organizational vacuum created by the old regime. For democracy to succeed, these networks must give rise to broader and more impersonal forms of social organization. Still, the crucial point is that the task of building social organizations is greatly facilitated by the ease with which Russians interact with each other. Weak social networks contribute to social cooperation. The task of building a civil society would be far more complicated were there in fact a legacy of totalitarian atomization. Weak social networks are not an alternative to civil society; but their existence makes the prospects for civil society much brighter. These networks continue to be a source of learning about the art of democratic governance. Weak social networks are probably more important during the early days of a major political transformation, but they remain significant in Russia today, since Russians have not learned all there is to know about democracy. For instance, they definitely have not learned to tolerate their political enemies. Learning how to apply democratic values is a demanding task. Social networks facilitate social learning, especially learning about how to govern democratically. I contend that interpersonal trust is not a necessary part of this equation. People do not base their decisions to become active in organizational life on the basis of whether they believe strangers can be trusted. These data have shown just how great the distinction is between trust in people in general and in people whom one knows. Organizations are not formed from strangers. People play football, discuss politics, and attend churches with people whom they know, and Russians generally trust people whom they know. In few countries does the existence of political parties depend upon the willingness of people to trust strangers.12 Careful attention must be given to the various forms of trust, and in particular to the relationship between trust in people and trust in institutions. In modern societies people do not need to trust one another, since they can rely upon institutions to rectify problems that arise. This is certainly true of economic relations (does one trust the dealer from whom one buys a car?), but probably pertains to political relations as well. When people in a community band together to try to solve a problem, they most likely do so out of common interests, without giving much consideration to whether their neighbors and comrades are trustworthy. Perhaps people must trust institutions in order to make use of them, but trust in people seems to be a much over-rated element of a democratic political culture. Why is organizational life so barren in Russia? First, as I noted above, it may not in fact be so barren. But, second, social networks may have enabled Russians to accomplish many of their goals without resorting to

78 James L. Gibson formal organizations. Third, economic necessity, caused by the greed of Russia’s ruling elites, has also limited the amount of time and energy left for recreational or social activities of any sort. But perhaps most important is the simple fact that such organizations take time to develop and evolve into formal organizations. Russians are surely distrustful of organizations imposed upon them from the top. Organizations that evolve from the grass roots take time to grow. Russian civil society may well be evolving just about as rapidly as can be expected. The existence of dense, weak, and politicized social networks should give encouragement to those favoring the consolidation of democracy in Russia.

Acknowledgments Funding for this project was provided by the National Science Foundation (SBR-9423614, SES-9023565, SBR-9710137, SES-9213237, and SBR9311403) and the National Council for Soviet and East European Research (#810-07).

Notes 1 This estimate discounts union membership, including only those union members ‘who trust local union leaders to represent their interests’ (Rose 1998: 11, footnote 4; see also page 27, which seems to provide some contradictory evidence). 2 Most analysts see novel political information and values as originating outside the borders of the formerly dictatorial systems (Gibson et al. 1992). Processes of diffusion appear to have been crucial in providing citizens with alternative models of political organization. 3 Of course, once information penetrates a group characterized by ‘strong’ ties, it diffuses throughout the network quite readily. But since ‘weak’ ties are more permeable, novel information is more readily accessible, and therefore Granovetter and I contend that ‘weak’ ties are more conducive to social learning in transitional regimes. 4 For an interesting anthropological study of the nature and content of the discussions held within Russian social networks, see Ries 1997. See also Ledeneva (1998) on blat networks. 5 See, for instance, Gibson 1996. 6 This difference is, of course, statistically significant, although note that the relationship between the year of the survey and network size is quite small (  0.08). 7 The frequency of political discussion with each network alter was measured with the following question: ‘How often do you discuss political questions with [each of the network members]?’ The possible responses were: (1) Practically every day, (2) At least once a week, (3) At least once a month, (4) Once a year, (5) Less than once a year, and (6) Never. 8 These figures include people in the denominator who were not able to name a full complement of network members. 9 The index was created from three smaller indices characterizing the capacity of each network relationship. Each dyad was scored first on the existence of a relationship (0 or 3), then for the degree of politicization of the relationship

Russia’s democratic transition 79 (normalized, 0 through 1.0), and finally for whether or not the relationship was ‘weak’ (0 through 3). Each of these indices ranges from 0 (no relationship) to 3 (weak relationship, very frequent political discussions). Thus, in effect, the political capacity of each existing relationship is a weighted function of its politicization and weakness. 10 Caution must be exercised in interpreting this table since the data were collected under a variety of national circumstances. Perhaps most important is how readily survey agencies accept ‘don’t know’ responses. In our surveys we are entirely willing to accept such a response. But given the infrequency of ‘don’t know’ responses in many of the World Value Surveys, it seems unlikely that ‘don’t know’ was as easily accepted in all the countries. 11 Factor analysis of the five subscales in the 1996 survey reveals a strongly unidimensional factor structure, with four of the five subscales loading on the factor approximately equally. The exception is the measure of rights consciousness, which has a factor loading of 0.44. The smaller relationship to the factor is most likely because virtually all Russians are highly conscious of claiming rights, and therefore the subscale suffers somewhat from degenerate variance. Similar results characterize the 1998 survey. A strongly unidimensional structure emerged; the loadings for four of the five subscales are approximately equal; and rights consciousness has the lowest loading on the factor. 12 Ironically, it may well be that dictatorial regimes depend more on trust than do democratic regimes, due in part to the lack of institutional checks on power.

Bibliography Arato, A. and Cohen, J. (1992) Civil Society and Democratic Theory. Cambridge, MA: MIT Press. Bahry, D. (1993) ‘Society Transformed? Rethinking the Social Roots of Perestroika’, Slavic Review 52 (Fall, #3): 512–54. Bernhard, M. (1996) ‘Civil Society after the First Transition: Dilemmas of Postcommunist Democratization in Poland and Beyond’, Communist and PostCommunist Studies 29 (September): 309–30. Braithwaite, V. and Levi, M. (eds) (1998) Trust and Governance. New York: Russell Sage Foundation. Brehm, J. and Rahn, W. (1997) ‘Individual-level Evidence for the Causes and Consequences of Social Capital’, American Journal of Political Science 41 (July): 999–1023. Ekiert, G. (1991) ‘Democratization Processes in East Central Europe: A Theoretical Reconsideration’, British Journal of Political Science 21: 285–313. Foley, M. W. and Edwards, B. (1998) ‘Beyond Tocqueville: Civil Society and Social Capital in Comparative Perspective’, American Behavioral Scientist 42 (September): 5–20. Gibson, J. L. (1996) ‘Political and Economic Markets: Changes in the Connections Between Attitudes Toward Political Democracy and a Market Economy Within the Mass Culture of Russia and Ukraine’, Journal of Politics 58 (November): 954–84. Gibson, J. L. (1997) ‘Mass Opposition to the Soviet Putsch of August 1991: Collective Action, Rational Choice, and Democratic Values in the Former Soviet Union’, American Political Science Review 91 (September): 671–84.

80 James L. Gibson Gibson, J. L., Duch, R. M. and Tedin, K. L. (1992) ‘Democratic Values and the Transformation of the Soviet Union, Journal of Politics 54: 329–71. Granovetter, M. S. (1973) ‘The Strength of Weak Ties’, American Journal of Sociology 78 (May): 1360–80. Greenfeld, L. (1992) ‘Kitchen Debate’, The New Republic 207 (#13, Issue 4,053, September 21, 1992): 22–25. Huckfeldt, R., Beck, P., Dalton, R. and Levin, J. (1995) ‘Political Environments, Cohesive Social Groups, and the Communication of Public Opinion, American Journal of Political Science 39 (4): 1025–54. Inglehart, R. (1997) Modernization and Postmodernization: Cultural, Economic, and Political Change in 43 Societies. Princeton, NJ: Princeton University Press. Kish, L. (1965) Survey Sampling. New York: John Wiley & Sons. Lake, R. La Due and Huckfeldt, R. (1998) ‘Social Capital, Social Networks, and Political Participation’, Political Psychology 19 (#3, September): 567–84. Ledeneva, A. V. (1998) Russia’s Economy of Favors: Blat, Networking and Informal Exchange. New York: Cambridge University Press. Mishler, W. and Rose, R. (1997) ‘Trust, Distrust and Skepticism: Popular Evaluations of Civil and Political Institutions in Post-Communist Societies’, Journal of Politics 59 (#2, May): 418–51. Mishler, W. and Rose, R. (2001) ‘What are the Origins of Political Trust? Testing Institutional and Cultural Theories in Post-Communist Societies’, Comparative Political Studies 34 (1): 30–62. Mondak, J. J. and Gearing, A. F. (1998) ‘Civic Engagement in a Post-Communist State’, Political Psychology 19 (September): 615–37. Mutz, D. (1999) ‘Cross-cutting Social Networks: Testing Democratic Theory in Practice.’ Paper delivered at the 1999 Annual Meeting of the Midwest Political Science Association, Chicago, Illinois. Norris, P. (1999) Critical Citizens: Global Support for Democratic Government. New York: Oxford University Press. Putnam, R. D. with Leonardi, R. and Nanetti, R. Y. (1993) Making Democracy Work: Civic Traditions in Modern Italy. Princeton, NJ: Princeton University Press. Ries, N. (1997) Russian Talk: Culture and Conversation during Perestroika. Ithaca, NY: Cornell University Press. Roeder, P. G. (1999) ‘The Revolution of 1989: Postcommunism and the Social Sciences’, Slavic Review 58 (#4, Winter): 743–55. Rose, R. (1998) ‘Getting Things Done in an Anti-Modern Society: Social Capital Networks in Russia’, Studies in Public Policy, #304, Glasgow, Scotland: Centre for the Study of Public Policy, University of Strathclyde. Stokes, G. (1993) The Walls Came Tumbling Down: The Collapse of Communism in Eastern Europe. New York: Oxford University Press. Tismaneanu, V. (ed.) (1995) Political Culture and Civil Society in Russia and the New States of Eurasia. Armonk, NY: M. E. Sharpe.

6

Trust and civic engagement in East and West Eric M. Uslaner

It is common knowledge that citizens in the former communist countries of Central and Eastern Europe have less trust in their fellow citizens than those in the West. It is also common knowledge that they participate less in voluntary organizations than their Western counterparts (see the chapters by Gibson and Howard in this volume, among others). It is not hard to see why both trust in fellow citizens and participation in civic groups is lower in formerly communist societies. Authoritarian political systems pit citizens against each other. When people feel compelled to turn on their friends lest the state turns on them, interpersonal trust may become too risky. In such a world, you really cannot be too careful in dealing with people, even if everyone would much prefer to treat others as if they were trustworthy. If people are wary of strangers they will limit their social activities to close friends whom they do see as trustworthy (see the chapters by Gibson, and Hayoz and Sergeyev in this volume). As the Russian pollster and sociologist Masha Volkenstein argues (quoted in Hoffman 1996: A40): It’s difficult to have a civil society when the country is corrupt and criminalized. . . . When society is under stress, it’s not a good time to talk about civil society. You need stability. . . . Now, we are just surviving. We don’t have enough energy, time, and money for this. It’s hard times, like during the war, and you have to survive on your own. When people are struggling, they have neither the time nor the energy for civic engagement. Nor is it clear what the pay-off is for this type of participation. The closed circle of personal friends has an immediate pay-off: They can help you get potatoes or bananas and other necessities of daily life. They are also a social safety net in a society where widespread trust may seem foolhardy. And, of course, civic participation as we know it in the West was simply not an option under communism. Many organizations were banned and others were severely restricted. There was little tolerance for any group that could challenge the status quo. Since people could disguise the real purpose of any organization, this meant that all groups were suspect. So at

82 Eric M. Uslaner least part of the explanation for low levels of civic engagement is simple: fewer groups, less participation. Add to this a widespread distaste among the people for the sorts of groups fostered by the regime, and the explanation becomes compelling. So it does not seem that we have to go very far to explain why the formerly communist states of Central and Eastern Europe have low levels of trust and civic engagement. Or does it? The argument may seem compelling, but it is too thin in a critical respect. It presumes that there is little in common between the experiences of people in the West and those in countries making the transition from communism to democracy. While communism was distinctive (though not unique) in how it destroyed trust and civic engagement, we do not know much about how authoritarian rule affected people’s levels of trust and civic engagement. There are a plethora of models seeking to explain why people develop trust in others and why they participate in civic groups. Almost all are based upon the experiences of the West. Are there any lessons in these models for the low levels of civic involvement in the formerly communist countries? I suggest that there are, and I shall focus on one model (rather immodestly, my own) that seems to explain trust rather well in the individual level in the United States and at the aggregate level in a wide range of societies. But it fails miserably in the aggregate as an explanation of trust in the formerly communist countries. The aggregate model is based upon the individual-level analysis in the United States. So the anomalous results for the formerly communist countries makes us step back and wonder whether civil society has different roots in these two types of regime. The answer, revealed in more detail below, is that citizens in the West and the formerly communist countries are not as different as some accounts would have us believe. Yes, the regimes are very different and this clearly affects both trust and civic engagement. But the differences in regimes work through the same underlying motivations for trusting others and taking part in civic groups.

The bases of trust and civic engagement Trust and civic engagement are the key elements of Putnam’s (1993) civic community. Trust and civic engagement give a community a cooperative spirit – and, in turn, better government and greater prosperity. Most treatments of trust and civic engagement either assume that they are part of the same general syndrome of social capital (Putnam) or part of a causal chain where one leads to the other. For Putnam (1993: 180) the causal arrow linking trust and civic engagement does not go one way: Social trust, norms of reciprocity, networks of civic engagement, and successful cooperation are mutually reinforcing. Effective collabora-

Civic engagement in East and West

83

tive institutions require interpersonal skills and trust, but those skills and that trust are also inculcated and reinforced by organized collaboration. In his more recent work, Bowling Alone, Putnam (2000: 137) goes further about the difficulty in establishing causality: [P]eople who trust others are all-around good citizens, and those more engaged in community life are both more trusting and more trustworthy . . . the critically disengaged believe themselves to be surrounded by miscreants and feel less constrained to be honest themselves. The causal arrows among civic involvement, reciprocity, honesty, and social trust are as tangled as well-tossed spaghetti. Others think not. Stolle (1998) presents evidence from surveys of members of voluntary associations in the United States, Sweden, and Germany, showing that length of membership has no impact on generalized trust. However, group members are more trusting than the mass public (as determined from public opinion surveys of the mass public). So Stolle concludes that there is a self-selection effect for voluntary associations. Civic groups do not make you more trusting, but more trusting people join voluntary associations. And Brehm and Rahn (1997) argue that the causal direction is precisely the opposite, from civic engagement to trust. Complicating the argument even further, I present evidence that there may be no link at all between trust and most forms of civic engagement. That said, we are way too early in the study of social capital in countries in transition from communism to argue that any of these perspectives is the correct one. I shall examine whether there is a linkage in both the West and the former communist countries below, but I focus primarily on the determinants of trust. In Uslaner (2002) I develop a framework for trust that is meant to be applicable across cultures. The basis of this model (that trust depends upon a sense of optimism about the future and one’s ability to make the world better) seems just as applicable to formerly communist countries as to the West. Indeed, the quote from Masha Volkenstein above seems to suggest that people in transitional societies are both pessimistic about the future and feeling helpless. This alone would make them mistrust others and thus withdraw from civic engagement. If you are uncomfortable with all but your closest friends, you will restrict your social circle accordingly. The key basis of generalized trust – and what separates it from both distrust and particularized trust – is a sense of optimism and control. For some people, a lifetime of disappointments and broken promises leads to distrust of others. A history of poverty with little likelihood of any improvement led to social distrust in the Italian village of Montegrano that Edward Banfield (1958: 110) described in the 1950s:

84 Eric M. Uslaner [A]ny advantage that may be given to another is necessarily at the expense of one’s own family. Therefore, one cannot afford the luxury of charity, which is giving others more than their due, or even justice, which is giving them their due. Generalized trust, at least in the West, is not based primarily on personal experiences. It rests on the premise that others will not let us down (Silver 1989: 276–77). People realize that it is not wise to extrapolate from individual cases to the general. Instead, we either seek some rationalization for our disappointing experience or simply wave it away as irrelevant (cf. Baker 1987: 5). Optimists are not worried that strangers will exploit them. If they take a chance and lose, their upbeat perspective leads them to try again. Setbacks are temporary; the next encounter will be more cooperative (Seligman 1991: 4–5). Optimism is a multifaceted phenomenon. An upbeat outlook has four components. The first two are central: a view that the future will be better than the past and the belief that we can control our environment to make it better. The other elements of optimism are a sense of personal wellbeing and a supportive community. Optimism leading to trust does not depend upon expectations in the short term. Optimists do not just expect tomorrow to be better than today. Each passing day should be better than the next. This may be a little unrealistic, since things do not always go our way. Yet bad days should be exceptions. Similarly, when we shift our view to how well a society is doing, expecting an upturn in next year’s economy is not sufficient to make people trusting. For there is an inevitable business cycle and what goes up will come down, and our trust may well prove to be unwarranted. If trust varies with perceptions of the near-term economy, then it will not be a stable value. Instead, it will fluctuate – perhaps sharply – over a short period of time. Optimism that undergirds generalized trust reflects a deeper sense that things are on the right track and will continue to get better (see Rahn and Transue 1998). As important as expectations for the future are, a sense of control is no less critical. Optimists believe that tomorrow will be better than today because they can make it better. Optimists are masters of their own fate (Seligman 1991: 5). Pessimists do not believe that they can control the world. They expect the worst and believe that it will last indefinitely. Just as optimists believe that they have the power to change the world, pessimists see a dark future as beyond their control. They may be tempted to blame sinister forces – the strangers in their midst – as the reason why their fate is so dire (Rosenberg 1956: 694; Banfield 1958: 111; Brehm and Rahn 1997: 1010). Optimists will thus be more likely to place their faith in strangers. Trusting people, according to at least some accounts, will also bond with strangers to join civic groups (Stolle 1998; Putnam 2000: 22). It is easy to

Civic engagement in East and West

85

see how communist regimes can destroy both optimism and a sense of control over one’s life. For most people living under communism, it is not at all clear that the future will be better than the past, and certainly they will not feel that they have control over their own destinies. Authoritarian governments that set people against each other, such as the former communist regimes in Eastern and Central Europe, can make trust hazardous. When people feel compelled to turn on their friends lest the state turns on them, interpersonal trust may become too risky. In such a world, one really cannot be too careful in dealing with people, even if everyone would strongly prefer to treat others as if they were trustworthy. Even with democratic institutions in place, people living in countries with legacies of oppression will neither trust their fellow citizens nor participate in civic life. The type of regime is obviously the key causal agent, but it should work through the same social psychology of optimism and trust that we find in the West. In the West, at least, the roots of trust and optimism do not lie in personal experiences, either in one’s financial status or in life more generally (Uslaner 2002: chs 2 and 4) but they do have a basis in experience, collective experiences. Wealth does not matter, but the distribution of resources does, for two reasons. First, optimism for the future makes less sense when there is more economic inequality. People at the bottom of the income distribution will be less sanguine that they too share in society’s bounty. How well the country is doing collectively, rather than how well any of us is doing individually, leads to changes in interpersonal trust (see Kinder and Kiewiet 1979). Second, the distribution of resources plays a key role in establishing the belief that people share a common destiny and have similar fundamental values. When resources are distributed more equally, people are more likely to perceive a common stake with others. If there is a strong skew in wealth, people at each end may feel that they have little in common with others. In highly unequal societies, people will stick with their own kind. Perceptions of injustice will reinforce negative stereotypes of other groups, making trust and accommodation more difficult (Boix and Posner 1998: 693). Putnam (1993: 88, 174) argues that trust will not develop in a highly stratified society. And Seligman (1997: 36–37, 41) goes further. Trust cannot take root in a hierarchical culture. Such societies have rigid social orders marked by strong class divisions that persist across generations. Feudal systems and societies based on castes dictate what people can and cannot do based upon the circumstances of their birth. Social relations are based on expectations of what people must do, not on their talents or personalities. Since inequality is an aggregate measure, I tested the relationship between trust and inequality for the United States over time and across a sample of sixty-three nations, using the Gini index of inequality from Deininger and Squire (1996) and measures of trust from the World Values

86 Eric M. Uslaner Study. For countries without a legacy of communism, the simple correlation of generalized trust and the Gini index is
0.684.1 Economic inequality is strongly related to trust, and this connection does not vanish in multivariate tests. It does go away in the formerly communist nations of Eastern and Central Europe (where the correlation falls to
0.239). The dynamic of economic inequality and trust clearly works differently in democracies and authoritarian societies. Why is the relationship between trust and inequality so different in a broad sample of countries and the formerly communist nations? There seem to be two reasonable explanations. The first, which I shall examine here, is that there is a different individual-level social psychology in the West and in the formerly communist countries. Communist regimes are so repressive that they not only lower levels of trust, but reshape it altogether. Second, the social psychology may be the same, but the aggregate data may tell a different story in Central and Eastern Europe. The relationship between trust and equality may not hold in communist regimes because equality was not the product of normal social interactions and market forces, but rather enforced by the state. Or it may not hold for the simple reason that economic statistics for communist states are not very accurate. I move now to an examination of the roots of trust and civic engagement at the individual level in the West and in the formerly communist states.

Modeling trust and civic engagement I shall present and test a model of trust and civic engagement using Wave 3 (1990–95) of the World Values Study (WVS), one of the few crossnational surveys to include questions about trust and membership in civic organizations. The ‘standard’ trust question asked in the WVS and many other surveys is: ‘Generally speaking, do you believe that most people can be trusted, or can’t you be too careful in dealing with people?’ The WVS asked whether people were members of religious, sports, arts, professional, charitable, or environmental groups, as well as political party organizations and labor unions. I estimate the model separately for ten Western nations and nineteen formerly communist countries.2 Because trust and group membership may be related to each other, I estimated a model using two-stage least squares, which permits me to examine reciprocal causation. The model for trust focuses on optimism and control, but the WVS is not strong on either of these measures. Thus we should not be surprised to find that some other variables that are not generally significant in more fully specified models – personal experiences and trust in government institutions – will be significant in these estimations. The measure of optimism is whether the future of humanity will be bright or bleak. And the measure of control is a scale asking how much control people have over decision-making in their daily lives. Since generalized trust is confidence in strangers, attitudes toward

Civic engagement in East and West

87

others will play a key role in shaping trust.3 People who say that it is important to understand others’ preferences will be more likely to have faith in others. And people who want their children to be tolerant and unselfish are displaying the sort of anti-authoritarian attitudes that underly generalized trust; similarly, people who want their children to be obedient will probably be less tolerant and trusting (Rosenberg 1956: 695; Adorno et al. 1964: 255). People who favor greater equality – namely efforts to reduce income gaps – should also be more trusting. Religious people may look inward to their own kind rather than trust more generally, so I posit that people who were raised as religious or who see themselves as religious may have less faith in other people than those who are secular. However, this may be a distinctively Western view. In the formerly communist states, the Church sometimes served as the bulwark against the totalitarian state, as in Poland. Thus we may find that religious people are more trusting in the formerly communist states. Postmaterialists – people who value beauty and a caring society – are more likely to trust others, while materialists are more concerned with their own welfare rather than with any public good (Rahn and Transue 1998; Inglehart 1999). Thus people who say that good pay is an important consideration in choosing a job will be less likely to have faith in others. Levi (1999) and Offe (1999), among others, argue that a state, particularly a democratic state, can generate trust in people. Levi (1999: 82) maintains that states build trust through ‘the use of coercion’ and that ‘democratic states may be even better at producing generalized trust than are nondemocratic institutions . . . because they are better at restricting the use of coercion to tasks that enhance rather than undermine trust.’ Rothstein (in press) elaborates the link between trust and coercion: If people believe that the institutions that are responsible for handling ‘treacherous’ behavior act in fair, just and effective manner, and if they also believe that other people think the same of these institutions, then they will also trust other people. Levi (1998: 87) holds that ‘[t]he trustworthiness of the state influences its capacity to generate interpersonal trust.’ Rothstein (in press) elaborates on this linkage: [I]f you think . . . that these . . . institutions [of law and order] do what they are supposed to do in a fair and effective manner, then you also have reason to believe that the chance of people getting away with such treacherous behavior is small. If so, you will believe that people will have very good reason to refrain from acting in a treacherous manner, and you will therefore believe that ‘most people can be trusted. A strong legal system will reduce transaction costs, making trust less risky. The more experience people have with compliance, the more likely they

88 Eric M. Uslaner are to have confidence in others’ goodwill (Brehm and Rahn 1997: 1008; Levi 1999; Offe 1999). I thus include measures of how beneficial democracy is, confidence in the national government and especially the legal system, and the extent of political corruption. The greater the support for democracy, the government, and especially the legal system, the more likely people should be to trust others. If people see their government as corrupt, they may lose faith in other people as well. In a wide range of studies, education is one of the strongest predictors of trust, so I include it in the model for trust (see Putnam 1995a). As a general measure of well-being, I also expect financial satisfaction to have a modest effect on trust. Trust may also be higher in smaller towns, where people are more likely to know each other, so I include the size of the community in the trust model. To test for reciprocal causality, I include group membership as a potential determinant of trust. The model for group membership is not as complex or as fully specified. The dependent variable is the number of groups people join. Many workers must join unions as a condition of their employment. Unions in communist states were often not voluntary organizations, and were generally controlled by the state. Thus I exclude union membership in the measure of voluntary organization participation. Clearly, we are interested in the effects of trust, but other factors matter as well. Political interest is a surrogate for a wider concern for public affairs. The WVS asks people to evaluate some public interest organizations. I include confidence in the women’s movement as another measure of how people view public participation more generally. Similarly, if people think that it is best for experts to run the government, they are expressing little confidence in civic engagement more generally, and in a corrupt system, people may see participation in civic groups as fruitless. On the other hand, people who attend church frequently are more likely to take part in civic groups. Church-goers learn important skills that carry over into political action. Clergy mobilize people into political and social action (Verba et al. 1993: 457). The otherwise passive flock would not participate if it were not involved in an organized religious community (Wuthnow 1991: 156). Finally, active membership in a church or a synagogue allows people to develop and practice skills (letter-writing, organizing) that translate easily into political action (Verba et al. 1995). I would also expect that materialists (good pay being the most important factor in a job) would be less likely to work with others in civic groups. Educated people should also be more likely to participate, as should older people, and people living in smaller towns (Putnam 2000). Putnam (1995a, 2000) has argued that much of the decline in civic engagement in the United States can be traced to the growth in television-watching: When people are at home watching television, they cannot be out and about in civic groups. The more time people spend watching television, the less likely they should be to join civic groups.

Civic engagement in East and West

89

Trust and membership Clearly trust and group membership are far lower in the former communist states than in the West (see Table 6.1). In neither group of states are most people trusting. About 44 percent of people in the West believe that most people can be trusted, compared to just 23 percent in the states in transition to democracy. Clearly the history of repressive regimes in these states has taken its toll on faith in strangers. Civic engagement is also far lower in the former communist states. Total memberships (excluding unions) are almost twice as high in the West, as is the percentage of people who belong to any group (again excluding unions). Yet the participation rate in the formerly communist states is not as low as we might have expected: 40 percent of people say that they belong to at least one organization. There is greater participation in the West across every type of group, differences that are significant at p  0.0001 or more. The only membership group where we find a small difference (2 percent) is for unions. This is hardly surprising, given the importance of unions in the communist state. Yet, even here, the difference is statistically significant. I present the equation for trust in Table 6.2. Of course, there are differences between the Western states and those in transition to democracy, but overall, the similarities are greater. The variables that are significant for the West are also generally significant for the formerly communist states. The most important variables, as measured by both the (unstandardized) regression coefficients and the t ratios, are optimism for the future, postmaterialism, the importance of understanding others’ preferences, support for equality, perceptions of political corruption, and education. In Table 6.1 Summary of trust and group membership in the West and former communist states Trust/group

Western states

Former communist states

Trust Total memberships* Any membership* Church group memberships Sports clubs Arts groups Political parties Environmental groups Professional groups Charitable organizations Other groups Unions

0.436 1.997 0.787 0.469 0.387 0.242 0.151 0.126 0.197 0.219 0.215 0.297

0.230 1.079 0.403 0.227 0.181 0.148 0.125 0.097 0.127 0.122 0.115 0.278

Notes All differences significant at p  0.0001 or less * Excludes unions.

90 Eric M. Uslaner Table 6.2 Models for trust from two-stage least-squares estimations Independent variable

Group membership Future bright or bleak Have control over life Understand others’ preferences Incomes made more equal Children should be tolerant Children should be unselfish Children should be obedient Raised as religious Is respondent religious Postmaterialism Good pay most important in job Financial satisfaction Education Size of community Need more respect for authority Democratic system beneficial Confidence in legal system Confidence in national government Extent of political corruption Constant

Western states

Formerly communist

Coefficient

t ratio

Coefficient

t ratio

0.008 0.115**** 0.005* 0.098**** 0.006** 0.024* 0.013
0.063****
0.026**
0.008 0.028****
0.022** 0.078** 0.014***
0.022**
0.053**** 0.054**** 0.047**** 0.036***
0.074**** 0.633****

0.642 8.535 1.511 6.868 2.235 1.471 0.894 0.015
1.738
0.620 5.040
1.664 2.317 3.546
2.750 6.684 5.909 5.218 3.867
8.473 9.608


0.013 0.090***
0.001 0.037*** 0.005**


2.019 10.054
0.472 4.203 3.032 N.A. 2.471
2.115 2.325
2.115 4.932
4.251 0.740 3.931 4.902 0.888 0.629 2.760 2.842
5.399 7.552

0.025*
0.021** 0.022**
0.017** 0.189****
0.058**** 0.001 0.008*** 0.008**** 0.005 0.004 0.016** 0.016**
0.032**** 0.305

Notes R2 0.158 0.032 SEE 0.459 0.412 N 5,497 9,708 **** p  0.0001; *** p  0.01; ** p  0.05; * p  0.10.

both the West and the formerly communist countries, these are generally the most powerful determinants of trust. Trust is lower in the transitional states because there is less tolerance for others, a greater perception of corruption, and fewer postmaterialists. (Ironically, the level of optimism is slightly higher in the formerly communist states.) Greater support for greater equality of income in the formerly communist states helps to push the level of trust a little higher. There are, of course, differences in the two models. Tolerance is more important in the West (which is more tolerant). Postmaterialist values have by the far the greatest impact in the former communist countries. Citizens in transitional countries are far more materialist than those in the West. This is hardly surprising, given the difficulties of daily life. Thus the smaller number of postmaterialists are very strongly over-represented among the people who trust others. Education matters more in the West (the impact is twice as great), and we see different impacts for religion. In the West, people who were raised as religious are less trusting, as expected. In the formerly communist countries, being raised as religious makes you more trusting. But

Civic engagement in East and West

91

being religious now makes you less trusting. This may well reflect the greater role of religion in building bonds during the communist era. Religion may not be quite so important with the overthrow of the authoritarian regimes. There is a modest impact for size of community in the West: people in smaller towns are more trusting. But in Central and Eastern Europe there is more trust in the big cities. The rural areas of transitional societies are poorer and less developed than those in the West. Rural life may foster trust only in people like oneself, rather than faith in strangers. The big cities are more diverse, and this may foster a greater sense of generalized trust. There are strong effects for support for democracy and confidence in the legal system and the national government in the West. Belief in democracy’s benefits is insignificant in the transitional states, and confidence in government matters less. Much of this reflects the lower level of support for democracy and confidence in the legal system in these countries. This is not surprising: Howard (Chapter 10, this volume) finds that many Russians and East Germans have little faith in the reformed polity, and thus do not take part in civic life. This is rather ironic. Authoritarian regimes depressed trust, yet democratization does not seem to build trust (cf. Mueller 1999; Uslaner 2002). There is relatively little pay-off from support for the government or democracy (Gibson, Chapter 5, this volume; Mishler and Rose 2001). Perhaps the greatest, and least comprehensible, difference lies in the effect of civic engagement on trust. In the West, joining lots of groups does not produce more trust (cf. Stolle 1998; Uslaner 2002: ch. 5). In the transitional states, civic engagement seems to lead to less trust. This is counterintuitive, to be sure. Perhaps the negative sign reflects an underspecified model, with trust acting as a surrogate for variables omitted in the model. The model for the West fares much better, with a higher R2 (0.158 compared to 0.032). The standard error of the estimate (SEE) is higher in the West. The SEE for the West is almost equal to the mean value (0.430), while the SEE for the formerly communist countries is much higher than the mean (by a factor of 1.8). This suggests that the basis for trust is shakier in the countries in transition. Yet the results for the West and the formerly communist countries are similar. The variables that perform best in one equation also fare best in the other. There is little evidence that the social psychological processes that give rise to trust or distrust are fundamentally different in the two societies. The equations for group membership are, with a few exceptions, also similar across the two groups of countries (see Table 6.3). In both sets of countries, but especially in the formerly communist nations, we find a selfselection effect: trusters are more likely to join voluntary organizations. Mistrusters simply opt out of civic engagement (Stolle 1998). The impact of trust in these states overwhelms almost everything else. More trusting people will join an additional 2.2 organizations. The average person in transitional countries belongs to only one organization (see Table 6.1), so the impact of trust is huge. The only variable that is more important than

92 Eric M. Uslaner Table 6.3 Models for group membership from two-stage least-squares estimations Independent variable

Western states Coefficient

Trust Political interest Church attendance Confidence in women’s movement Best for experts to run government Extent of political corruption Good pay most important in job Hours watch television Age Education Size of community Constant

Formerly communist t ratio

0.422** 2.400 0.270**** 9.444 0.267**** 21.264 0.128**** 4.126 0.031 1.165 0.113** 2.906
0.038
0.734
0.032
1.056 0.002 1.040 0.180**** 14.588
0.009
0.793 0.633**** 9.608

Coefficient

t ratio

2.193**** 0.217**** 0.211**** 0.311****
0.041* 0.091*** 0.186** 0.522****
0.006*** 0.060**** 0.029** 0.305****

5.755 7.639 14.811 11.226
1.490 2.482 2.300 21.614
3.814 4.891 2.958 7.552

Notes R2 0.173
0.039 SEE 1.779 2.355 N 5,497 9,708 **** p  0.0001; *** p  0.01; ** p  0.05; * p  0.10.

trust is television-viewing. Putnam (1995b) posited that spending lots of time watching television would lead people to foresake civic engagement. There is little evidence for this in the West. In both the East and the West, political interest, confidence in the women’s movement, perceptions of political corruption, education, and especially church attendance, all shape the level of civic engagement. There is weak support for the political efficacy surrogate (best for experts to run the government) and somewhat stronger evidence that materialism leads people away from civic engagement, but only in the transitional states. Younger people participate more in civic affairs in the former communist societies (see Howard, Chapter 10, this volume), as do people in the bigger cities, where there may well be more opportunities for civic engagement. Once more, the predictive power for the West is much greater, whether measured by R2 or the standard error of the estimate.4 Overall, the models for the two groups of states are more similar than different.

Reprise The differences in models between the West and the countries in transition pale by comparison to the similarities. People trust others and join civic groups generally for the same reasons in different political regimes and cultures. The levels of trust and civic engagement reflect the disparate experiences in authoritarian and democratic regimes, but the regime is not simply a ‘black box’ that begins with structure and ends with behavior.

Civic engagement in East and West

93

There is an underlying social psychology that I have detailed in this chapter, and it appears to operate in similar ways in different contexts. If the underlying social psychology is the same in the two groups of countries, then the anomalous weak relationship between trust and inequality must be spurious. The message from countries without a legacy of communism is straightforward: if we want a trusting society, then we must pay heed to the question of economic inequality. Optimism is the strongest determinant of trust in both the East and the West, but optimism seems to be in shorter supply in the formerly communist countries. Pursuing policies that foster equality may slow economic growth. Formerly communist countries rightly put a high premium on growth, yet they will have to face the issue of trade-offs. High levels of growth may leave some people further behind than they were under the old regime, and this may hinder development of the trust that is essential to making democracy work. Developing trust is thus a balancing act – and hardly an easy one.

Notes 1 The Gini index comes from Deininger and Squire (1996) and the database is available at http://www.worldbank.org. I used the Gini index with the best available data in the Deininger–Squire database and matched the Gini index to the year in which trust was measured for each case. I excluded China from the analysis because its score on trust seemed to be an outlier, and possibly the result of sampling error. 2 The Western countries are Great Britain, West Germany, Spain, the United States, Japan, Australia, Norway, Sweden, Finland, and Switzerland. The formerly communist countries are Poland, Belarus, East Germany, Slovenia, Lithuania, Latvia, Estonia, the Ukraine, Russia, Moldova, Armenia, Georgia, Azerbaijan, Montenegro, Macedonia, Croatia, and Bosnia Herzegovina. Other countries were part of the sample, but many did not ask the full complement of questions or were not available in the Wave 3 dataset available to me. 3 The variables in the next several paragraphs follow the general lines of thought in Uslaner (2002: chs 2 and 4). There I am less supportive of the notion that trust in governmental authorities leads to interpersonal trust. 4 The R2 for the formerly communist countries is negative. This is not a typographical error. Instead, R2 for two-stage least squares is not defined, so the measure reported is at best approximate. The SEE for the West is slightly below the mean, and for the countries in transition more than twice the mean value.

Bibliography Adorno, T. W., Frenkel-Brunswik, E., Levinson, D. J. and Nevitt Sanford, R. (1964) The Authoritarian Personality. Part One, New York. First published 1950. Baker, J. (1987) ‘Trust and Rationality’, Pacific Philosophical Quarterly 68: 1–13. Banfield, E. (1958) The Moral Basis of a Backward Society. New York: Free Press. Boix, C. and Posner, D. N. (1998) ‘Social Capital: Explaining Its Origins and Effects on Government Performance’, British Journal of Political Science 28: 686–93. Brehm, J. and Rahn, W. (1997) ‘Individual Level Evidence for the Causes and Consequences of Social Capital’, American Journal of Political Science 41: 888–1023.

94 Eric M. Uslaner Deininger, K. and Squire, L. (1996) ‘A New Data Set: Measuring Economic Income Inequality’, World Bank Economic Review 10: 565–92. Hoffman, D. (1996) ‘Harsh History Stymies Civil Society: Russian Activists Strive to Be Heard as Democracy Takes Hold’, Washington Post (December 26): A1, A40. Inglehart, R. F. (1999) ‘Trust, Well-Being and Democracy’, in M. Warren (ed.) Democracy and Trust. Cambridge: Cambridge University Press. Kinder, D. R. and Kiewiet, D. R. (1979) ‘Economic Discontent and Political Behavior: The Role of Personal Grievances and Collective Economic Judgments in Congressional Voting’, American Journal of Political Science 23: 495–527. Levi, M. (1999) ‘When Good Defenses Make Good Neighbors: A Transaction Cost Approach to Trust and Distrust’, unpublished paper, New York. Mishler, W. and Rose, R. (2001) ‘What are the Origins of Political Trust? Testing Institutional and Cultural Theories in Post-Communist Societies’, Comparative Political Studies 34 (1): 30–62. Mueller, J. (1999) Democracy, Capitalism, and Ralph’s Pretty Good Grocery. Princeton, NJ: Princeton University Press. Offe, C. (1999) ‘Trust and Knowledge, Rules and Decisions: Exploring a Difficult Conceptual Terrain’, in M. Warren (ed.) Democracy and Trust. Cambridge: Cambridge University Press. Putnam, R. D. (1993) Making Democracy Work: Civic Traditions in Modern Italy. Princeton, NJ: Princeton University Press. —— (1995a) ‘Bowling Alone: America’s Declining Social Capital’, Journal of Democracy 6: 65–78. —— (1995b) ‘Tuning In, Tuning Out: The Strange Disappearance of Social Capital in America’, PS: Political Science and Politics (December): 664–83. —— (2000) Bowling Alone. New York: Simon & Schuster. Rahn, W. M. and Transue, J. E. (1998) ‘Social Trust and Value Change: The Decline of Social Capital in American Youth, 1976–1995’, Political Psychology 19: 545–66. Rosenberg, M. (1956) ‘Misanthropy and Political Ideology’, American Sociological Review 21: 690–95. Rothstein, B. (in press) ‘Trust, Social Dilemmas, and Collective Memories: On the Rise and Decline of the Swedish Model’, Journal of Theoretical Politics. Seligman, A. B. (1997) The Problem of Trust. Princeton, NJ: Princeton University Press. Seligman, M. E. P. (1991) Learned Optimism. New York: Alfred A. Knopf. Silver, A. (1989) ‘Friendship and Trust as Moral Ideas: An Historical Approach’, European Journal of Sociology 30(2): 274–97. Stolle, D. (1998) ‘Bowling Together, Bowling Alone: The Development of Generalized Trust in Voluntary Associations’, Political Psychology 19: 497–526. Uslaner, E. M. (1998) ‘Social Capital, Television, and the “Mean World”: Trust, Optimism, and Civic Participation’, Political Psychology 19: 441–67. —— (2002) The Moral Foundations of Trust. New York: Cambridge University Press. Verba, S., Schlozman, K. L. and Brady, H. (1995) Voice and Equality: Civic Voluntarism in American Politics. Cambridge, MA: Harvard University Press. Verba, S., Schlozman, K. L., Brady, H. and Nie, N. H. (1993) ‘Race, Ethnicity and Political Resources: Participation in the United States’, British Journal of Political Science 23: 453–97. Wuthnow, R. (1991) Acts of Compassion. Princeton, NJ: Princeton University Press.

7

Social capital, ethnicity and support for democracy in the post-communist states Kathleen M. Dowley and Brian D. Silver

Few theoretical concepts have received more attention among social scientists in the last half-dozen years than that of ‘social capital,’ particularly among students of countries undergoing democratisation (or in some cases, redemocratisation). Putnam notes that social capital ‘refers to features of social organisation such as trust, norms and networks, that can improve the efficiency of society by facilitating coordinated action’ (Putnam 1993: 167). Thus communities with higher levels of social capital are thought to be able to cooperate more often to overcome social problems, keep their governments more responsive and more honest, and improve democratic institutional performance. Communities with low levels, in contrast, seem unable to break the vicious circle and remain with governments that are less responsive, less efficient and less honest. Much of this argument echoes an earlier literature on the role of intermediate groups in moderating political views and mediating between political elites and masses. Voluntary organisations, and especially multiple and cross-cutting group affiliations, were said to be vital to the development of democracy (Kornhauser 1959; Lipset 1960). This idea also underlies earlier discussions of political cleavage and political culture (Almond and Verba 1963). Putnam’s social capital thesis has been subjected to any number of scholarly responses, retorts and replications. The concept is attractive theoretically, as it places so much emphasis on the activities of ordinary citizens, after a decade of bringing the state back into the centre of comparative political analysis. Promoting civic engagement and an active civil society, with individuals assuming responsibility for their own government and getting the government they therefore deserve are messages that resonate with scholars and policy makers alike. If democracy needs civil society, we will support civil society, we will fund voluntary organisations, we will make social capital wherever it is lacking. Yet the criticisms have been pointed as well. Are all networks or voluntary organisations democratic? Does participation in narrowly defined organisations promote the kind of social trust that Putnam sees as so critical to a functioning democracy, or do they more often, particularly in already

96 Kathleen M. Dowley and Brian D. Silver divided societies, serve only to reinforce existing cleavages rather than help to cut across them? Might not, as Hardin (1995) notes, differential mobilisation of small groups of the population particularly along ethnic, racial or religious lines, often undermine democratic institutions as they make particularistic demands, not those that necessarily promote the public good? In her critique of the theory, Levi (1996) notes that the emergence of social trust depends on experiences with institutions outside the small groups normally identified in the scholarly literature on social capital. The expectations about the behaviour of others that would allow individuals to agree to cooperate in the first place form as a ‘result of interactions among groups defined by ethnicity, religion, or some other shared value, confidence in a backdrop of third party sanctions or sufficient costs to discourage the betrayal of trust’ (Levi 1996: 48). Thus, for Levi, francophones in Canada have come to distrust the Canadian federal government not because of a lack of civic engagement but because that engagement has often been met with broken promises from that government. Our previous research has shown large differences between groups in divided states such as Canada in evaluations of democratic institutions, subjective political competence and patriotism, and interpersonal trust (Silver and Dowley 2000). These conflicting views of the value of social (and unsocial) capital for democratic performance and overall support for democratic institutions in plural societies are particularly relevant to the study of politics in EastCentral and post-Soviet Europe. With the history of the region replete with examples of ethnic mobilisation and its devastating consequences present even today in the former Yugoslavia and Chechnya, a thorough examination of the propositions suggested above seems warranted. The usual measures of social capital seem problematic in plural societies, especially ones undergoing rapid political transitions. For example, scholars studying civic communities have generally gauged the degree of ‘civicness’ by examining mean levels of participation in voluntary organisations (excluding religious organisations and those that are said to be hierarchical), mean levels of interest in politics, and mean levels of interpersonal trust. But in plural societies undergoing transition, interest in politics and high levels of participation in voluntary associations may in fact signal the ethnic polarisation of society. With the breakdown of old institutions and the expectations that they had for so long generated, a condition of domestic anarchy may have emerged that forces ethnics to mobilise to protect themselves in an uncertain environment (Posen 1993; Snyder 1993). In these cases, higher levels of participation along ethnic lines and high interest in politics will not signal democratic social capital in the making but instead an ethnic mobilisation and countermobilisation that threaten new democratic states. Communal organisations are not necessarily community organisations. The release of data from the 1995–98 World Values Survey, which includes a large number of countries from the post-communist space, pro-

Democracy in post-communist states 97 vides an opportunity to explore the nature and consequences of social capital in this region. The countries vary enough in the quality and completeness of their democratic transitions to allow us to make some preliminary observations. First, do the usual markers of social capital (interpersonal trust, political interest, and voluntary group participation) correlate with the most and least successful cases of democratisation in East-Central Europe and the post-Soviet states? Are the countries with the highest overall levels of apparent social capital the most democratic? Are they the countries in which individuals report the highest levels of satisfaction with democratic institutions (one of Putnam’s measures of institutional performance)? Are they the same countries in which respondents are most confident in democracy? Second, are the effects the same when one takes into account the ethnic diversity of the country? Does social capital function in the same manner in plural countries, or are the usual markers instead more likely to indicate dangerous polarisation in that society instead of social engagement that keeps democracy healthy and strong? These are the questions this chapter seeks to answer by analysing the latest round of surveys from the World Values Survey project.

Democratisation and social capital in the post-communist world A survey of the vast literature on democratisation in both Eastern Europe and the former Soviet Union yields mixed assessments of its relative success, completeness, and durability. In a symposium commemorating the tenth anniversary of the fall of communism, Roeder characterised the transitions in the region as fraught with ethnic conflict and ethnoconstitutional crises. Indeed, fifty-four of the sixty-nine ethnic groups comprising at least one per cent of any state’s population have been engaged in some kind of constitutional struggle with the state since 1989 (Roeder 1999: 867). In the same symposium, however, Fish noted that ‘the region has scarcely been the seething cauldron of interethnic violence that it is often assumed to be’ (Fish 1999: 811). In Fish’s view, democratisation has been associated with a decrease in ethnic tensions in places such as Georgia, Bulgaria, Romania and Hungary. Where tensions have flared, such as in Chechnya or Serbia, the problem is the absence of real democratisation, not mass participation itself. Freedom House indicators of democratisation, measured in terms of both political and civil liberties, yield the ranking of countries identified in Table 7.1. Scores of 1 on each indicate that the highest levels of political and civil liberties are being protected, while scores of 7 indicate that virtually none are.1 In 1991, the regional mean for political rights was 3.4 and for civil liberties 3.6; in 1996 (the time of the latest World Values Surveys) the means were 3.0 for political rights and 3.4 for civil liberties, marking a

98 Kathleen M. Dowley and Brian D. Silver Table 7.1 Freedom House rating of post-communist countries in the Study, 1991–92, 1996–97, and 1999–2000.a (Most freedom  1, Least freedom  7) Political rights scale

Civil liberties scale

1991

1996

1999

1991

1996

1999

Czech Republicb Hungary Poland Estonia Slovenia Lithuania Slovak Republicb Latvia Bulgaria Macedonia Ukraine Russia Romania Moldova Georgia Armenia Azerbaijan Bosnia-Herzegovina Yugoslavia Belarus

2 2 2 2 2 2 2 2 2 3 3 3 5 5 6 5 5 6 6 4

1 1 1 1 1 1 2 2 2 4 3 3 2 3 4 5 6 5 6 6

1 1 1 1 1 2 1 1 2 3 3 4 2 2 3 4 6 5 5 6

2 2 2 3 3 3 2 3 3 4 3 3 5 4 5 5 5 6 5 4

2 2 2 2 2 1 4 2 3 3 4 4 3 4 4 4 5 5 6 6

2 2 2 2 2 2 2 2 3 3 4 5 2 4 4 4 4 5 5 6

Mean

3.4

3.0

2.7

3.6

3.4

3.2

Notes a Freedom House Annual Survey of Freedom Country Ratings, 1972–73 to 1999–2000. Includes countries that participated in World Values Survey 1990–93 or 1995–97 waves only. For Bosnia-Herzegovina and Macedonia, the first year’s scores are from 1992–93. b 1991–92 scores for Czech Republic and Slovak Republic are based on Czechoslovakia.

small improvement during the five years.2 And in 1999 the mean scores on the two indicators further improved to 2.7 and 3.2. The general trend is towards a steady improvement in the democratic quality of life in most East European states, except the former Yugoslavia; and towards a decline in the democratic quality of life in the former Soviet republics, except the Baltic countries. The indicators for social capital in the World Values Survey stem from responses to three questions, consistent with how most of the scholars in this literature have used the survey (e.g. Newton and Norris 2000). The first is a question about the respondent’s level of interest in politics: ‘How interested would you say you are in politics?’ The respondents could answer ‘Very interested, somewhat interested, not very interested or not interested at all.’ We coded the responses to allow 4 to represent those ‘very interested’ and 1 ‘not at all interested.’ Second, we constructed a variable to measure participation in voluntary

Democracy in post-communist states 99 organisations that counts whether individuals claimed to be ordinary members or active in any of these groups: • • • • • • • • •

Church or religious organisations Sport or recreation organisations Art, music or educational organisations Trade unions Political parties Environmental organisations Professional associations Charitable organisations Any other voluntary organisations.

On the basis of Putnam’s (1993) discussion of social capital in Italy, we eliminated membership in religious organisations because of their hierarchical character. This is somewhat at odds with the work of Mishler and Rose (1999: 95) who use church attendance as a measure of the civic community in East-Central Europe, though they acknowledge this is a crude indicator. Using church attendance and membership in a political party, though, they too found a positive relationship between these measures and individual responses on their political support thermometer (Mishler and Rose 1999). For a third indicator of social trust we used the dichotomous question (V27), ‘Generally speaking, would you say most people can be trusted or that you can’t be too careful in dealing with people?’ Individuals who responded ‘most people can be trusted’ were recoded as 1 and those who said ‘you can’t be too careful’ as 0. Thus, for all three sets of indicators, higher scores indicated someone with higher levels of civic engagement/ capital. Table 7.2. presents the product–moment correlations between the aggregate measures of social capital and the Freedom House scores for those countries in both 1996 and 1999.3 One might expect social capital at a given time to predict democratic outcomes at some later time, so we ran the 1996 social capital scores from the World Values Survey against the 1999 democracy scores for this reason. Social capital is not correlated with democratisation in these postcommunist countries. None of the bivariate correlations is statistically significant (at p  0.05). Indeed, the mean levels of social trust and organisational membership are consistently negatively related to levels of overall democratisation across the twenty post-communist countries included in this analysis. No one would contend that social capital is the only or a sufficient explanation of the level of democracy. But one might expect to find a positive correlation. Another possible source of the negligible (and negative) correlations may be the Freedom House measures. More recent

100

Kathleen M. Dowley and Brian D. Silver

Table 7.2 Correlations between social capital indicators and Freedom House indicators of democratisation, for twenty post-communist states in 1995–97 WVS,a based on data aggregated by country Social capital indicator

Freedom House Political rights 1996

Civil liberties 1996

Political rights 1999

Civil liberties 1999

Mean political interest


0.152 (p  0.522)


0.080 (p  0.737)


0.152 (p  0.522)

0.111 (p  0.641)

Mean interpersonal trust


0.161 (p  0.498)


0.354 (p  0.125)

0.203 (p  0.391)


0.357 (p  0.122)

Mean number of organisational memberships


0.355 (p  0.136)


0.321 (p  0.180)

0.355 (p  0.138)


0.340 (p  0.154)

Notes a WVS data on satisfaction with the government and support for democratic principles are available for only seventeen of the countries. All variables are scored so that if social capital is a cause of democratisation, there should be a positive correlation between social capital and support for democratic institutions and principles.

scholarship on political support in both established and emerging democracies has attempted to distinguish between support for the political community, support for regime principles, support for the regime’s performance and support for individual democratic institutions, all seen as independent of evaluations of a particular political leader (Norris 1999: 17–20). The social capital literature, beginning with Putnam’s work in Italy, has typically emphasised support for or confidence in democratic institutions as an indicator of regime capacity or effectiveness. Consistent with recent research, we used the World Values Survey questions that asked respondents about their confidence in democratic institutions. The questions were worded as follows: I am going to name a number of organisations. For each one, could you tell me how much confidence you have in them: is it a great deal of confidence, quite a lot of confidence, not very much confidence or none at all? We looked at evaluations of the legal system (V137) and the parliament (V144) as indicators for the efficacy of new democratic institutions.4 We recorded the confidence indicators to range from a high of 4 (great deal) to a low of 1 (no confidence at all). Table 7.3. reports the product–moment correlations between the mean confidence in democracy score and the mean scores on the three social capital indicators. To tap into what Norris (1999) refers to as support for ‘regime

Democracy in post-communist states 101 Table 7.3 Correlations between social capital indicators and support for democratic institutions, for twenty post-communist states in 1995–97 WVS,a based on data aggregated by country Confidence in legal system

Confidence in Parliament

Satisfaction with the social capital indicator

Support for democracy government

Mean political interest


0.022) (p  0.928)

0.195) (p  0.411)


0.209) (p  0.421)


0.009) (p  0.972)

Mean interpersonal trust

0.239) (p  0.310)


0.388) (p  0.093)


0.042) (p  0.872)

0.265) (p  0.303)

Mean number of organisational memberships

0.364) ( p  0.125)

0.175) (p  0.473)

0.410) (p  0.102)

0.145) (p  0.577)

Notes a WVS data on satisfaction with the government and support for democratic principles are available for only seventeen of the countries. All variables are scored so that if social capital is a cause of democratisation, there should be a positive correlation between social capital and support for democratic institutions and principles.

performance,’ we also examined an indicator of ‘satisfaction with the government’ (V165): How satisfied are you with the way the people now in national office are handling the country’s affairs? Would you say very satisfied, fairly satisfied, fairly dissatisfied or very dissatisfied? We coded the responses on a scale ranging from 4 (very satisfied) to 1 (very dissatisfied). Finally, to assess the level of support for the ‘democratic principle,’ as opposed to support for particular institutions or regime performance in a particular country, we construct a ‘pro-democracy’ scale based on answers to four questions, each with 4-point strongly agree, agree, disagree, strongly disagree response categories: • • • •

In democracy, the economic system runs badly (V160). Democracies are indecisive and have too much squabbling (V161). Democracies aren’t good at maintaining order (V162). Democracy may have problems but it’s better than any other form of government (V163).

We treat ‘disagree’ responses to the first three questions, and ‘agree’ responses to the last question, as ‘pro-democracy’ responses. The mean score of the responses to the four questions ranges from 4 (most in favour of democracy) to 1 (least in favour of democracy). We assume that responses

102

Kathleen M. Dowley and Brian D. Silver

mainly reflect the respondents’ commitment to the abstract principles of democracy, not just to the performance of the government in their country. We expected to find that countries with higher levels of social capital would have a higher percentage of people expressing support for the performance of the regime, higher confidence in its political institutions and, more broadly, a higher percentage committed to the principles of democracy. However, when we examine the relationship between the levels of confidence in new institutions and the indicators of social capital, we find no statistically significant relationships between social capital and aggregate confidence in or satisfaction with democratic institutions. We find no positive relationships even if we eliminate the non-democracies from our pool of countries. Thus, within the first ten-year post-communist transition period, we find scant evidence of a link between social capital and aggregate levels of democratisation. It is not just that other factors (such as internal war) have shaped the outcome of the transition, but that social capital does not seem to contribute to democratic development, since without a correlation the former cannot be even a partial cause of the latter. Why do the measures of social capital that have commonly been employed to predict democracy and democratic performance not work in these societies undergoing transition? Are the measures flawed in this context? Is the theory wrong or underspecified in this setting? In the extant criticisms of the theory we note a repeated concern that the survey-based indicators of social capital might have different meanings in different social contexts. In particular, high levels of interest in politics and high rates of participation in voluntary organisations in ethnically plural societies, especially but perhaps not uniquely in countries undergoing significant political transformations, might signal the ethnic polarisation of society and a consequent threat to democratic institutions, democratic values, and liberal notions of civil and political rights.

Social capital and support for democracy According to Rustow (1970), the only prior condition to successful democratisation is national unity. This does not mean that everyone has to trace his ancestry back to the same clan, nor does it mean that everyone has to speak the same language, or even practice the same religion. Instead it means that nearly everyone must believe they belong together in a single political community. Later scholars doing cross-national work on support for regime performance, institutions and principles, suggest that support for the political community is the first level of identification in a multidimensional scheme of political support (Norris 1999). It may well be that social capital operates in the expected manner at some of these levels but not others, i.e. at the level of support for regime performance, but not at that of the political community. Or it may be that national unity as meas-

Democracy in post-communist states 103 ured by general support for the political community by all relevant subnational groups is a necessary precondition for social capital to predict democratisation. Support for the political community has often been measured by means of the ‘patriotism’ question employed by the World Values Survey and countless other surveys. With this in mind, and to demonstrate the validity of our concern about the potential impact of pluralism on social capital markers, we turn to a survey question on national pride (V205) to explore differences in mean responses of significant groups within states in transition. The existence of substantially different evaluations of the country as a whole by large subnational groups would be evidence of a lack of national unity and, as such, would potentially challenge the way in which social capital affects the democratisation process. The national pride question reads as follows: How proud are you to be an [American]? Very proud, somewhat proud, not very proud, not proud at all?’ To evaluate support for the political community within significant subnational ethnic groups, we needed first to adapt the World Values Survey data to determine to which ethnic groups the respondents may have belonged. The World Values Surveys were not designed, for the most part, to study ethnic variation (Silver and Dowley 2000). Although we did our best to classify respondents using one or more indicators that are typically linked to ethnic self-identification or consciousness, we had to rely on different variables in different countries. In some cases ethnicity was derived from an ethnic self-identification question; in others it is based on the survey organisation’s or the interviewer’s identification of the respondents by ethnicity; and in still others it is based on religious identification, language used at home or the region in which the respondent resided.5 However, we were able to derive ethnic variables for all countries in which the population had at least a 10 per cent minority except Croatia and Slovenia. Using the responses to the national pride question, we calculated the mean for each ethnic group as well as the difference between the mean scores for each ethnic group and the grand mean (2.3) for all World Values Survey respondents in the region. By using the grand mean (a common metric), we can compare groups within countries as well as across countries. In Figure 7.1, in addition to reporting the mean scores by ethnic group relative to the grand mean, the cases in which the mean score of an ethnic minority population is significantly different from the mean score of the majority or titular ethnic group in the same country are marked by a white bar.6 Those for which the difference from the titular ethnic group is not statistically significant are marked by a shaded bar. Almost universally, the minority populations report lower levels of pride in the country. If we accept the common interpretation that the question captures support for

Original range 1 (not at all proud) to 4 (very proud), mean 3.20 0.85, N  27,363 Poland (all)

0.5 0.1

Hungary (all)

0.3

Slovenia (all) Croatia (all) Romania: Romanians Romania: Transylvania

0.2 0.2 0.1
0.4

Czechoslovakia: Czechs Czechoslovakia: Slovaks Bulgaria: Bulgars Bulgaria: Turks Bulgarians: Roma


0.2 0.2


0.2

Macedonia: Macedonians Macedonia: Moslems Yugoslavia: Serbs Yugoslavia: Montenegrins Yugoslavia: Moslems Yugoslavia: Hungarians Bosnia-Herzeg.: Moslems Bosnia-Herzeg.: Serbs Bosnia-Herzeg.: Croats

0.3


0.0
0.4
0.5

0.0 0.4

0.2


0.3 0.0

Armenia (all) Azerbaijan: Azeris Azerbaijan: Russians Azerbaijan: Talysh Azerbaijan: Lezgin

0.4 0.1 0.2 0.1 0.3

Georgia: Georgians Georgia: Russians Georgia: Armenians Georgia: Azeris Russia: Russians Russia: Tartars Russia: Komi Russia: Kabards-Balkars Ukraine: Ukrainians Ukraine: Russians Belarus: Belarussians Belarus: Russians Belarus: Poles Belarus: Ukrainians Moldova: Moldovans Moldova: Russians Moldova; Others Estonia: Estonians Estonia: Russians Latvia: Latvians Latvia: Russians Lithuania: Lithuanians Lithuania: Russians Lithuania: Poles

0.4


0.1

0.1 0.1
0.4
0.8


0.2
0.1


0.9


0.1


0.6
0.4
0.5


0.2 0.0


0.3 0.2


0.6

0.2


1.0
1.0


1.0

0.5


0.2


0.6


0.4


0.5

0.0

0.5

Group mean difference from the grand mean

Figure 7.1 Differences in pride in country among forty-nine ethnic groups in twenty post-communist countries Source: Calculated from World Values Survey: ‘How proud are you to be [Czechoslovak]? Very proud, quite proud, not very proud, not at all proud?’ (V205). Answers are recoded so that they range from 4 (very proud) to 1 (not at all proud). Notes Solid white bars represent minority groups whose mean level of national pride is statistically significantly different from the mean level for the majority or titular ethnic groups in the same country (at p  0.05, 2-tailed test).

Democracy in post-communist states 105 the political community, then ethnic pluralism matters, and it matters in a meaningful and predictable manner. Now that we have found evidence that people view the political community through an ethnic lens, we address again the bivariate relationships between social capital and support for regime performance, regime institutions and regime principles. This time, however, we add the ethnic component to the mix. Figures 7.2 to 7.5 report the mean scores of majority groups and minority groups in each of these societies, by expressing the means relative to the regional grand mean scores for confidence in the legal system, the parliament, the current national government and the prodemocracy scale. We find no easily apparent pattern across the political support measures. About 55 per cent of the minority ethnic groups in our sample (sixteen out of twenty-nine minority groups) are less confident than the majority population in the legal systems in their country. Only 45 per cent of the minority groups are less confident than their majority populations in the national parliament. But 63 per cent of the minority groups are less satisfied than the majority groups with the performance of the regime, and 60 per cent express less support than the majority groups do for democratic principles. The clearest cross-national tendency is that the Russian minority populations in the post-Soviet states are consistently less confident in the new institutions, less satisfied with the new national government, and less supportive of democracy as a system of governing ideals than their new majority populations. Of course, this may not be due entirely to their status as ethnic minorities in new countries, as the Russian majority in Russia itself is also the least supportive of democratic principles, the least satisfied with their current government, and among the least confident in new governing institutions of the majority populations in our sample.7 Although the differences between majority and minority groups are generally small and run in both directions, in fifty-five of 116 possible differences the gaps between the majority and minority populations were statistically significant. This is forty-nine more statistically significant differences than one would expect due to chance alone. By chance alone one would expect to find six statistically ‘significant’ differences out of 116 (at p  0.05). As noted previously, however, the sample designs in most of the World Values Survey countries either undersampled the minority populations or else failed to sample enough members of the minority groups to permit reliable estimates of inter-group differences. That despite this fifty-five of 116 of the relationships are statistically significant makes our finding of so many statistically significant differences even more notable. How does social capital fit into this complex relationship? If we first examine differences in the level of interest in politics among the ethnic groups in the countries under consideration we find that Czechs and Slovaks in the early 1990s were far more interested in politics than the rest of the

Original range 1 to 4, mean 2.38 0.86, N  27,849 Poland (all)

0.2 0.3

Hungary (all)
0.1

Slovenia (all) Croatia (all) Romania: Romanians Romania: Transylvania

0.3 0.1 0.1

Czechoslovakia: Czechs Czechoslovakia: Slovaks Bulgaria: Bulgars Bulgaria: Turks Bulgarians: Roma Macedonia: Macedonians Macedonia: Moslems Yugoslavia: Serbs Yugoslavia: Montenegrins Yugoslavia: Moslems Yugoslavia: Hungarians Bosnia-Herzeg.: Moslems Bosnia-Herzeg.: Serbs Bosnia-Herzeg.: Croats Armenia (all) Azerbaijan: Azeris Azerbaijan: Russians Azerbaijan: Talysh Azerbaijan: Lezgin Georgia: Georgians Georgia: Russians Georgia: Armenians Georgia: Azeris Russia: Russians Russia: Tartars Russia: Komi Russia: Kabards-Balkars Ukraine: Ukrainians Ukraine: Russians Belarus: Belarussians Belarus: Russians Belarus: Poles Belarus: Ukrainians Moldova: Moldovans Moldova: Russians Moldova; Others Estonia: Estonians Estonia: Russians Latvia: Latvians Latvia: Russians Lithuania: Lithuanians Lithuania: Russians Lithuania: Poles


0.5

0.0


0.1
0.2


0.4

0.1

0.3


0.3 0.0

0.3

0.2 0.2

0.5

0.2

0.6


0.3
0.0

0.0 0.2


0.1
0.1
0.1


0.0 0.3


0.1

0.2
0.0

0.3

0.1


0.2

0.1 0.1


0.1

0.2 0.1


0.0

0.0 0.3 0.1


0.1
0.2
0.3


0.2
0.2


0.3

0.0

0.3

0.5

Group mean difference from the grand mean

Figure 7.2 Differences in confidence in the legal system among forty-nine ethnic groups in twenty post-communist countries Source: Calculated from World Values Survey: ‘How much confidence do you have in the legal system? A great deal, quite a lot, not very much, or none at all?’ (V137). Answers are recoded so that they range from 4 (great confidence) to 1 (none at all). Notes Solid white bars represent minority groups whose mean level of national pride is statistically significantly different from the mean level for the majority or titular ethnic group in the same country (at p  0.05, 2-tailed test).

Original range 1 to 4, mean 2.20 0.87, N  27,190 Poland (all)

0.0 0.1

Hungary (all)
0.2

Slovenia (all) Croatia (all) Romania: Romanians Romania: Transylvania

0.3
0.3
0.3

Czechoslovakia: Czechs Czechoslovakia: Slovaks Bulgaria: Bulgars Bulgaria: Turks Bulgarians: Roma Macedonia: Macedonians Macedonia: Moslems Yugoslavia: Serbs Yugoslavia: Montenegrins Yugoslavia: Moslems Yugoslavia: Hungarians Bosnia-Herzeg.: Moslems Bosnia-Herzeg.: Serbs Bosnia-Herzeg.: Croats

0.1

Latvia: Latvians Latvia: Russians Lithuania: Lithuanians Lithuania: Russians Lithuania: Poles

0.2

0.3


0.5
0.5
0.1

0.2

0.0


0.1

0.2

0.5 0.4


0.3

Armenia (all) Azerbaijan: Azeris Azerbaijan: Russians Azerbaijan: Talysh Azerbaijan: Lezgin Georgia: Georgians Georgia: Russians Georgia: Armenians Georgia: Azeris Russia: Russians Russia: Tartars Russia: Komi Russia: Kabards-Balkars Ukraine: Ukrainians Ukraine: Russians Belarus: Belarussians Belarus: Russians Belarus: Poles Belarus: Ukrainians Moldova: Moldovans Moldova: Russians Moldova; Others Estonia: Estonians Estonia: Russians

0.2

0.0

0.7 0.6 0.7
0.1 0.1
0.3


0.2

1.0

0.1 0.5

0.1 0.1 0.1


0.2
0.1


0.2
0.2
0.1 0.0 0.0 0.2 0.1 0.1
0.1


0.4


0.0
0.1
0.1


0.5


0.3

0.0

0.3

0.5

0.8

1.0

Group mean difference from the grand mean

Figure 7.3 Differences in confidence in the Parliament among forty-nine ethnic groups in twenty post-communist countries Source: Calculated from World Values Survey: ‘How much confidence do you have in the Parliament? A great deal, quite a lot, not very much, or none at all?’ (V144). Answers are recoded so that they range from 4 (great confidence) to 1 (none at all). Notes Solid white bars represent minority groups whose mean level of national pride is statistically significantly different from the mean level for the majority or titular ethnic groups in the same country (at p  0.05, 2-tailed test).

Original range 1 (very dissatisfied) to 4 (very satisfied) mean 1.92 0.75, N  23,895 Poland (all)

0.3

Slovenia (all)

0.3 0.6

Croatia (all) Bulgaria: Bulgars Bulgaria: Turks Bulgarians: Roma


0.2

0.2


0.0


0.1
0.1

Macedonia: Macedonians Macedonia: Moslems

0.1

Yugoslavia: Serbs Yugoslavia: Montenegrins Yugoslavia: Moslems Yugoslavia: Hungarians

0.5 0.2 0.1

Bosnia-Herzeg.: Moslems Bosnia-Herzeg.: Serbs Bosnia-Herzeg.: Croats

0.6


0.1

Armenia (all) Azerbaijan: Azeris Azerbaijan: Russians Azerbaijan: Talysh Azerbaijan: Lezgin

Ukraine: Ukrainians Ukraine: Russians

0.4

0.2 0.2 0.3

Georgia: Georgians Georgia: Russians Georgia: Armenians Georgia: Azeris Russia: Russians Russia: Tartars Russia: Komi Russia: Kabards-Balkars

0.5

0.3

0.0


0.1

0.1 0.2


0.4


0.4
0.4
0.4


0.2
0.3
0.3
0.3

Belarus: Belarussians Belarus: Russians Belarus: Poles Belarus: Ukrainians


0.2


0.1


0.1
0.2

Moldova: Moldovans Moldova: Russians Moldova; Others Estonia: Estonians Estonia: Russians

0.0 0.1


0.2

Latvia: Latvians Latvia: Russians

0.2


0.2

Lithuania: Lithuanians Lithuania: Russians Lithuania: Poles


0.3


0.5


0.2


0.0

0.0

0.5

Group mean difference from the grand mean

Figure 7.4 Differences in satisfaction with government performance among fortyfour ethnic groups in seventeen post-communist countries Source: Calculated from World Values Survey: ‘How satisfied are you with the way the people now in national office are handling the country’s affairs? Very satisfied, fairly satisfied, fairly dissatisfied, or very dissatisfied?’ (V165). Answers are recoded so that they range from 4 (very satisfied) to 1 (very dissatisfied). Notes Solid white bars represent minority groups whose mean level of national pride is statistically significantly different from the mean level for the majority or titular ethnic groups in the same country (at p  0.05, 2-tailed test).

Democracy in post-communist states 109 Original range 1 to 4, mean 3.02 1.16, N  23,638
0.2

Poland (all)
0.5

Slovenia (all)

0.1

Croatia (all) Bulgaria: Bulgars Bulgaria: Turks Bulgarians: Roma

0.0


0.1
0.2
0.1

Macedonia: Macedonians Macedonia: Moslems

0.5 0.2

Yugoslavia: Serbs Yugoslavia: Montenegrins Yugoslavia: Moslems Yugoslavia: Hungarians

0.4 0.6

Bosnia-Herzeg.: Moslems Bosnia-Herzeg.: Serbs Bosnia-Herzeg.: Croats


0.0

0.3


0.2

Armenia (all)

0.5 0.5 0.6

Azerbaijan: Azeris Azerbaijan: Russians Azerbaijan: Talysh Azerbaijan: Lezgin

0.3

Georgia: Georgians Georgia: Russians Georgia: Armenians Georgia: Azeris


0.1
0.4
0.4

Russia: Russians Russia: Tartars Russia: Komi Russia: Kabards-Balkars

0.0
0.0

0.7


0.1

Ukraine: Ukrainians Ukraine: Russians

0.4 0.2


0.1

Belarus: Belarussians Belarus: Russians Belarus: Poles Belarus: Ukrainians Moldova: Moldovans Moldova: Russians Moldova; Others

0.9

0.3


0.0


0.2


0.0


0.2
0.3
0.3


0.7

Estonia: Estonians Estonia: Russians

0.3


0.2
0.0
0.1

Latvia: Latvians Latvia: Russians

0.3

Lithuania: Lithuanians Lithuania: Russians Lithuania: Poles

0.1


0.4


0.8


0.5


0.3

0.0

0.3

0.5

0.8

1.0

Group mean difference from the grand mean

Figure 7.5 Differences in support for democracy among forty-four ethnic groups in seventeen post-communist countries Source: Calculated from World Values Survey. Mean score in the ‘democratic’ direction on four 4-point agree-disagree questions: ‘In democracy the economic system runs badly’ (V160), ‘Democracies are indecisive and have too much squabbling’ (V161), ‘Democracies aren’t good at maintaining order’ (V162), and ‘Democracy may have problems but it is better than any other form of government (V163).

ethnic groups in the region. This high level of interest did not signal health of the polity, but instead polarisation and subsequent dissolution of the state. In fact, the most successful cases of democratisation are those in which levels of interest, participation, and trust were at or below the regional mean.8 Indeed, Czechs and Slovaks notwithstanding, the only other groups with expressed levels of political interest higher than the regional mean are the Montenegrins in Yugoslavia, Serbs in Bosnia, Russians and Ukrainians

110

Kathleen M. Dowley and Brian D. Silver

in Belarus, Russians and Talysh in Azerbaijan, Latvians, and Estonians. Interestingly, in the two former Soviet republics with the highest level of democratisation, Estonia and Latvia, the Russian populations express mean levels of interest in politics lower than the regional mean, and significantly lower than the majority populations in the successor states. Using interpersonal trust variables to indicate democratic social capital in this part of the world is also problematic. The most trusting groups (judging by the World Values Survey data) are the Roma in Bulgaria, Hungarians in Yugoslavia and Kabards and Balkars (small Muslim ethnic groups residing in the North Caucasus) in Russia. Above the mean are also Serbs in all settings, Muslims in Bosnia and Yugoslavia, Croats in Bosnia, and Azeris in Georgia. In Russia the Tatars and Kabards and Balkars are significantly more trusting than titular Russians, just as Roma and Turks are more trusting than Bulgars in Bulgaria. Because voluntary group membership is so central to the social capital thesis, we report in Figure 7.6 the percentage of individuals within each group that belong to at least one voluntary association. Once again we find evidence against the applicability of standard social capital indicators. The ethnic groups with the highest percentages belonging to at least one voluntary organisation are located in the states of the former Yugoslavia. A phenomenal 79 per cent of Bosnian Serbs belong to at least one voluntary organisation – compared to only 42 per cent of Hungarians (in Hungary), and 11 per cent of Poles (in Poland). The latter two countries are sometimes taken as paragons of the democratic transition in the postcommunist world. Although conceivably this evidence signals an emerging civic involvement among the Bosnian Serbs, it is more likely that it reveals high engagement in communal organisations that are far from helping to reconcile relations with non-Serbs. When we examine the individual forms of membership, we note that 164 of the 369 Bosnian Serbs who are members of voluntary associations identify these organisations as political parties, more than any other category of voluntary organisation except sports clubs, to which 176 of the 369 reported belonging. We turn now to a multivariate statistical analysis of the relationships between social capital, ethnicity, and support for democracy, using the same indicators of support for regime institutions, government performance and democratic principles. For this purpose, we rely on individuallevel rather than aggregated data. In addition, we pool the approximately 29,000 cases from the post-communist countries in the 1995–97 World Values Survey.9 Table 7.4 reports the Pearson’s r coefficients and significance levels for all respondents in the twenty countries under investigation, as well as separately for just the titular majorities, the new Russian minorities in the ‘near abroad’ (former non-Russian republics of the USSR), and all other minority groups in the post-communist space. Because Russians themselves were so consistently more negative in their evaluations of state institutions and

Mean percentage who belong to any non-religious organisation 38 48, N  27,641 Poland (all)


27% 4%

Hungary (all)

21%

Slovenia (all) Croatia (all) Romania: Romanians Romania: Transylvania Czechoslovakia: Czechs Czechoslovakia: Slovaks Bulgaria: Bulgars Bulgaria: Turks Bulgarians: Roma

9%
12%
11% 0% 0%
14%


23%
28%

5% 4%

Macedonia: Macedonians Macedonia: Moslems Yugoslavia: Serbs Yugoslavia: Montenegrins Yugoslavia: Moslems Yugoslavia: Hungarians Bosnia-Herzeg.: Moslems Bosnia-Herzeg.: Serbs Bosnia-Herzeg.: Croats

9%

Latvia: Latvians Latvia: Russians Lithuania: Lithuanians Lithuania: Russians Lithuania: Poles

26%

22%

41%

22%
22%

Armenia (all) Azerbaijan: Azeris Azerbaijan: Russians Azerbaijan: Talysh Azerbaijan: Lezgin Georgia: Georgians Georgia: Russians Georgia: Armenians Georgia: Azeris Russia: Russians Russia: Tartars Russia: Komi Russia: Kabards-Balkars Ukraine: Ukrainians Ukraine: Russians Belarus: Belarussians Belarus: Russians Belarus: Poles Belarus: Ukrainians Moldova: Moldovans Moldova: Russians Moldova; Others Estonia: Estonians Estonia: Russians

19%


8%

4%
8%


29%

27%


0%


12%
15%
21% 7%

11%

30%


9% 2% 4% 10% 10% 7%
2%
6%

11%


1% 5%


9%
28%

15%

8%


16%
11%


30%
20%
10%

0%

10%

20%

30%

40%

Group mean difference from the grand mean

Figure 7.6 Differences in percentages of those who belong to any non-religious voluntary organisation among forty-nine ethnic groups in twenty postcommunist countries Source: Calculated from World Values Survey: ‘Now I am going to read off a list of voluntary organisations: for each one, could you tell me whether you are an active member, an inactive member, or not a member of that type of organization?’ (V28–V36). Notes This diagram counts as a ‘member’ whether the person was either an active member or an inactive member. The organisational types are: church/religious, sports/recreation, art/music/educational, labour union, political party, environmental, professional, charitable, any other voluntary organisation. It excludes churches and religious organisations from the count.

112

Kathleen M. Dowley and Brian D. Silver

Table 7.4 Product–moment correlations between social capital indicators and support for institutions and democracy in post-communist states, 1995–97 WVS (pooled individual-level data) Inter-personal Interest in trust politics

Organisational member (non-church)

All respondents (max. N  27,942) Confidence in legal institutions Confidence in Parliament Satisfaction with the government Support for democracy

0.043** 0.061** 0.038** 0.041**

0.017** 0.079** 0.043**
0.032**

0.016** 0.024** 0.030** 0.002

0.039** 0.060** 0.034** 0.052**

0.013* 0.083* 0.043**
0.023

0.011 0.025** 0.017* 0.005

Titular nationalities (max. N  22,234) Confidence in legal institutions Confidence in Parliament Satisfaction with the government Support for democracy

Russians in ‘near abroad’ (max. N  2,773) Confidence in legal institutions Confidence in Parliament Satisfaction with the government Support for democracy

0.050* 0.063* 0.032 0.024

0.032 0.048* 0.021
0.053**


0.015 0.000 0.011 0.021

0.075** 0.076** 0.085*
0.035

0.012 0.051** 0.070**
0.079**

0.063** 0.014 0.129**
0.076**

Other minorities (max. N  2,661) Confidence in legal institutions Confidence in Parliament Satisfaction with the government Support for democracy

Notes * p  0.05 (two-tailed), **p  0.01 (two-tailed) Interpersonal trust: Can ‘most people be trusted?’ (1/0) Interest in politics: How interested in political affairs? (4-point scale) Organisational member: Belong to (or activist in) any non-religious voluntary organisation? (1/0) Confidence in legal institutions: How much confidence in the legal system? (4-point scale) Confidence in Parliament: How much confidence in Parliament? (4-point scale) Satisfaction with government: How satisfied with how the government is handling nation’s affairs? (4-point scale) Support for democracy: How strongly does R support democracy? (4-point scale based on four questions).

democratisation, we needed to determine whether they alone were driving some of the negative correlations between social capital and democratisation that we reported earlier in the aggregate analysis. Is minority status in states undergoing transition responsible for the negative signs, is it simply Russians who are more negative in this context, or are small or negative correlations between social capital and evaluations of regime institutions found also among the titular nationalities?

Democracy in post-communist states 113 For all respondents combined, of the three social capital indicators interpersonal trust is most consistently positively correlated with support for regime institutions, regime performance and democratic principles. The coefficients are statistically significant at the p  0.05 level for titular majorities as well on all four measures of support for the regime, government and democracy. However, for Russians in the near abroad, although higher levels of interpersonal trust are associated with greater confidence in new institutions, trust is not correlated with the evaluation of the regime performance or support for the principles of democracy. This is similar to the attitudes of all other minority populations, where higher levels of interpersonal trust are negatively associated with support for democratic ideals, though the negative correlation is not statistically significant. Interest in politics and organisational membership are the two social capital indicators that we anticipated might predict differently in plural settings. Specifically, political mobilisation might be expected to occur among those who are less satisfied with the newly established political order. Our expectations are borne out by Table 7.4. While for the titular nationalities interest in politics is positively correlated with support for institutions and democracy, for the Russian and other minority populations greater interest in politics is negatively correlated with support for democratic principles. Hence, high levels of explicitly political interest among ethnic minorities do not represent social capital in the making, but perhaps political mobilisation against democratic development. We find a less consistent pattern for organisational membership. For the titular ethnic groups, organisational membership seems to act as social capital theory predicts: involvement in voluntary organisations is associated with higher levels of confidence in new institutions, and higher levels of support for the current regime. It is also positively correlated with support for democratic ideals (though not at a statistically significant level). For Russians in the near abroad, however, none of the coefficients with organisational membership was statistically significant, suggesting that among those Russian minorities involved in voluntary organisations, such participation is not correlated with their levels of support for democratic institutions. Among other minorities in the region, while involvement is positively and significantly related to greater confidence in the legal system and satisfaction with the government’s policies, those who are involved in voluntary organisations are less likely to support democracy as an ideal form of government than those who are not involved in organisations. We fit a multivariate OLS regression model to summarise the typical effect of the social capital measures on each of the four dependent political support indicators. As predictors we include the three social capital indicators as well as dummy variables for the two minority ethnic categories.10 Furthermore, we specify interaction terms to capture the joint effects of being a member of a minority group and having an interest in

114

Kathleen M. Dowley and Brian D. Silver

politics as well as being a member of a minority group and being a member of at least one voluntary organisation. In addition, we control for the economic situation of the respondent by including responses to the following question (V64): How satisfied would you say you are with the financial situation of your household? If ‘1’ means you are completely dissatisfied and ‘10’ means you are completely satisfied, where would you put your satisfaction with your household’s financial situation? Our goal is not to account fully for the variance in support for democracy but to test for the effects of social capital and ethnicity. Because economic well-being is likely to be an important factor that differentiates the situations of ethnic groups in society, it is necessary to control for this factor in the analysis.11 The results in Table 7.5 provide support for our hypothesis that the impact of social capital differs according to whether the individuals belong to the titular or a minority ethnic group in a state of transition. First, interpersonal trust has a consistently positive association with support for democracy, regime institutions, and regime performance. The greater the interpersonal trust, the greater is the support for democracy as an ideal, the greater the confidence in the legal system and the parliament, and the greater the satisfaction with governmental performance. Second, interest in politics has a consistently positive relationship with three of the four dependent variables, but not with support for democratic principles. Indeed, interest in politics is negatively and significantly predictive of support for democracy, which implies that those who are more psychologically engaged in politics are less supportive of democratic ideals in the post-Soviet, post-communist context. Third, membership in voluntary organisations is not consistently correlated with any of the political support indicators. Bear in mind that, given the specification of variables in the equations, the effect of organisational memberships on the dependent variables (row 3) mainly reflects the relationship for the titular nationalities. Nonetheless, we find again that being a member of a voluntary organisation is not correlated with support for democratic principles. In fact, among minority populations voluntary group membership is more often associated with lower support for democratic institutions and principles. That the mobilisation of political interest and associational life among the ethnic minority populations in the post-communist countries is correlated with weaker support for democracy should be placed in context. On the whole, the minority nationalities (excluding the Russians) are likely to be substantially more positively disposed towards democratic principles than even the titular nationalities (see the positive coefficients in the fifth row of numbers in Table 7.5). But the more socially engaged of those

(13.165) (94.003)

1.539 0.104 0.70 22,536

(89.890)

0.065** (31.868)

(5.492)

3.022 0.020 1.132 22,319

0.021**

(108.068)

(6.377)


0.168** (
2.937)


0.073 (
0.991) 0.316** (3.906)
0.040 (
1.371)
0.054 (
1.678) 0.066 (1.297)

0.100** (5.597)
0.026** (
2.707)
0.006 (
0.325)

Support for democracy

Notes * Indicates coefficient significant at p  0.05 (two-tailed); ** Indicates significant at p  0.01 (two-tailed). The coefficients are unstandardised regression coefficients (bs); t ratios are given in parentheses. a Excludes membership of church or religious organisations. b Dummy variables: ‘Russian in near abroad’: 1 if a Russian living in non-Russian successor state of Soviet Union, 0 if not. ‘Other minority’: 1 if a nonRussian minority, 0 if not. The ‘omitted’ category is the titular nationality of the country (e.g., Estonians in Estonia, Bulgarians in Bulgaria). c In addition to the constant shown, the equations also included a set of dummy variables to capture the fixed effects of the countries in which no minority ethnic group members were identified.

1.862 0.035 0.86 24,403

0.036** (15.448)

(
0.304)


0.012

(1.523)

0.197**


0.113* 0.138**
0.035*
0.026 0.004

(
0.864) (1.299) (
1.161) (0.841) (
1.059)


0.047 0.073
0.025 0.019
0.041

(
2.103) (0.629) (0.413) (1.523) (
1.135)

(
2.547) (2.743) (
1.987) (
1.287) (
0.124)

0.062** (5.634) 0.043** (7.121)
0.025 (
2.243)

Satisfaction with government performance

0.106** (8.187) 0.075** (10.778)
0.049 (
0.038)

Confidence in Parliament (6.748) (2.760) (
0.857)

(114.453)

Confidence in legal system

Interpersonal trust 0.086** Interest in politics 0.019** Voluntary organisational
0.011 membera Ethnic groupb Russian in near abroad
0.112* Other minority 0.035 Social capital Russian* interest 0.009 and ethnic Other minority* interest 0.061 group jointly Russian* organisational
0.042 member Other minority*
0.061 organisational member Household Satisfaction with HH 0.030** finances finances Constantc 2.206 Adjusted R2 0.032 SEE 0.85 N 24,995

Social capital

Dependent variable

Table 7.5 Effects of social capital and ethnicity on support for the democratic system in post-communist countries (pooled individuallevel data)

116

Kathleen M. Dowley and Brian D. Silver

minorities are likely to be somewhat less favourably disposed towards regime institutions and democracy than the socially passive. Furthermore, on all four indicators of support for political institutions and democracy Russians are less supportive than the titular nationalities in the countries of the near abroad, though the negative coefficients are statistically significant only for satisfaction with governmental policies and confidence in the legal system. If we examine the joint effects of minority status and civic engagement, we find that for Russians, those who express greater interest in politics are less likely to support the regime and democracy than those who have little or no interest in politics. If we focus on just these statistically significant relationships, however, we obtain results that somewhat contradict those presented in Table 7.4. In the equations in Table 7.5 interest in politics among Russians predicts less support for the regime, while in Table 7.4 it predicted in the opposite direction (though not significantly). While we cannot readily reconcile this difference, we do not find any support for the idea that Russians who are more engaged in civic or political life are more likely to support both the new regime under which they live or the ideals of democratic governance. In contrast, being a non-Russian minority is associated with having a more positive attitude towards governmental performance and democracy. Why this difference between Russians and the other minorities? We propose that for Russians in the near abroad, political development and the process of democratisation over the past ten years has meant a loss of status as a corporate group. But for many other minorities, although there are certainly several major exceptions (including major cases of civil strife and civil war), post-communist development has opened up new opportunities for a flourishing of cultural life or even an independent political life and has provided a stake for them in achieving a further democratisation of the system to protect their civil rights.

Conclusions Our first significant finding is the weak overall relation between indicators of social capital and democratisation in the post-communist countries. An analysis of aggregate measures of social capital and democratisation failed to find a positive correlation between the aggregate amount of social capital and the level of democratisation. This remained true whether we used external evaluations of the level of democratisation or survey-based responses about the degree of confidence in and satisfaction with democracy. The analysis of individual-level data did, however, find some evidence of a positive correlation between social capital and attitudes supportive of democracy. This was especially so for social trust, though we urge caution about assuming that interpersonal trust only works in a pro-democratic direction. Viewed as a political resource, being able to rely on others can help the democrat and the anti-democrat, the individual engaged in regime-

Democracy in post-communist states 117 supportive participation or the one who adopts unconventional political methods or espouses unconventional goals (Bahry and Silver 1990). When we examined the impact of political interest and membership in voluntary associations, we found a more mixed result. By and large, for members of the titular majority, greater political involvement and social engagement were associated with greater support for democracy, the government and regime institutions. But among ethnic minorities we found that the more mobilised members of those groups are less supportive of democracy than the more passive members. Thus, social capital theory cannot be easily transported from the established democracies to ethnically plural societies in transition. In the postcommunist countries the transition unleashed the potential not only for a liberation of minorities but also for their suppression, not only for minority groups to proclaim and seek to protect their interests by endorsing a strong civil rights regime or cultural and political autonomy, but also for majority groups to want to deny or limit those claims and aspirations in the interest of preserving the physical boundaries of the state or their own newly dominant position in the system. Researchers also need to be cautious about assuming that indicators that apply to the more homogeneous societies work just as well in ethnically plural societies. Social scientists do not yet understand well the conditions under which people’s sub-national loyalties are mutually exclusive with national and state identities (Dowley and Silver 2000). Our message is not that ethnically plural societies cannot democratise but that it is more difficult for them to do so if the majority ethnic groups are exclusive in their post-communist nation-building project, or if a past history of injustice against minority populations mobilises them during periods of uncertainty in a way that makes the national unity condition impossible to satisfy. Our findings regarding the Russian minority lead us to wonder about similar tendencies among other former dominant or imperial nationals who suddenly find themselves in an inferior position. Does their undemocratic national ‘mean’ imply that some cultures never adopt democratic principles, no matter how interested or involved in organisations their peoples become? Or does their attitude stem more from their newly diminished position in these societies, which suggests parallels to other former dominant or colonial elite in parts of Africa and Asia? These are some of the questions suggested by our findings, and to which we shall direct our attention as we search for the conditions under which diverse peoples can make the transition from authoritarian rule to national unity under freer and fairer governing institutions.

Notes 1 Freedom House uses the following checklist to evaluate political rights: Is head of state/government elected through free and fair elections? Are legislative representatives elected in free and fair elections? Are there fair electoral laws,

118

2 3

4

5 6

7

8 9

10 11

Kathleen M. Dowley and Brian D. Silver

equal campaigning opportunities, fair polling and honest tabulating of results? Do people have the right to organise into political parties? Is there a significant opposition vote or de facto opposition power? Are the people free from domination by the military, foreign powers or economic oligarchies? Do ethnic, cultural and minority groups have reasonable self-determination? For civil liberties, the emphasis is on a free and independent media, free religious institutions, freedom of assembly, speech and organisation, the availability and legality of trade unions, an independent judiciary, and secure property rights. For details, see the methodology report at www.freedomhouse.org. These figures are based only on the twenty countries included in Table 7.1 – countries that participated in the World Values Survey in either the 1990–93 or 1995–97 rounds. The World Values Survey indicators are based on weighted data, which make the scores from the sample more representative of the population of each country. Except as noted, we rely on V236 from the World Values Survey as the weight variable. Initially we included confidence in the government (V142). The question was worded ‘. . . confidence in the government in [capital city]’, presumably so that respondents would think not of their regional or local leaders but instead of the national leaders. However, it appears that many respondents took the reference to the capital city literally, so that, for example, respondents in Russia might think the question referred to the leaders of Moscow city, not of the Russian Federation. Detailed recoding information is available from the authors in the form of an SPSS syntax file. Whether the scores for a given ethnic group are significantly different from average depends not only on the magnitude of the mean difference but also on the variance and on the number of cases. That the World Values Survey did not intentionally over-sample minority populations makes it more difficult to find statistically significant differences among ethnic groups. For example, the mean evaluation by Russians of the Duma of the Russian Federation is 1.88. Only the Macedonians in Macedonia have a lower mean evaluation (1.75) of parliament among titular nationalities in the third wave of the World Values Survey, although the overall samples in a few other World Values Survey countries give lower marks than the Russians to their legislature: Colombia 1.77, Argentina 1.78, Dominican Republic 1.82 and Peru 1.85. These conclusions are based on World Values Survey data that are not shown in a table or figure here. We again use V236 from the World Values Survey data to weight the cases. However, in the pooled data analysis we adjust the weights so that the total N’s from each country equal the actual N of respondents. Further information about the weighting procedures is available from the authors. The titular nationalities become the ‘omitted’ or reference category needed for specifying the equations. We assume that ‘satisfaction with household financial situation’ encapsulates a combination of factors (e.g. the respondent’s objective economic situation and frustration in meeting aspirations for economic success). Norris (1999) includes a measure of ‘satisfaction with life as a whole’ to predict support for political institutions in a regression analysis based on advanced Western democracies. Although this variable also correlates with the same dependent variables in the post-communist states, we are concerned that the answers to a question on ‘satisfaction with life as a whole’ are likely to be more contaminated by the respondent’s views of the overall political situation and thus impart a simultaneity bias to the analysis.

Democracy in post-communist states 119

Bibliography Almond, G. A. and Verba, S. (1963) The Civic Culture: Political Attitudes and Democracy in Five Nations. Princeton, NJ: Princeton University Press. Bahry, D. and Silver, B. D. (1990) ‘Soviet Citizen Participation on the Eve of Democratization’, American Political Science Review 83 (September): 821–48. Dowley, K. M. and Silver, B. D. (2000) ‘Subnational and National Loyalty: CrossNational Comparisons’, International Journal of Public Opinion Research 12 (November): 357–71. Fish, M. S. (1999) ‘Postcommunist Subversion: Social Science and Democratization in East Europe and Eurasia’, Slavic Review 58 (winter): 794–823. Hardin, R. (1995) One For All. Princeton, NJ: Princeton University Press. Kornhauser, W. (1959) The Politics of Mass Society. Glencoe, IL: The Free Press. Levi, M. (1996) ‘Social and Unsocial Capital: A Review Essay of Robert Putnam’s Making Democracy Work’, Politics and Society 24 (March): 45–55. Lipset, S. M. (1960) Political Man: The Social Bases of Politics. Garden City, NY: Doubleday. Mishler, W. and Rose, R. (1999) ‘Five Years After the Fall: Trajectories of Support for Democracy in Post-Communist Europe’, in P. Norris (ed.) Critical Citizens: Global Support for Democratic Governance. Oxford: Oxford University Press Newton, K. and Norris, P. (2000) ‘Confidence in Public Institutions: Faith, Culture or Performance?’, in S. Pharr and R. Putnam (eds) Disaffected Democracies: What’s Troubling the Trilateral Countries? Princeton, NJ: Princeton University Press: 52–73 Norris, P. (ed.) (1999) Critical Citizens: Global Support for Democratic Governance. Oxford: Oxford University Press. Pharr, S. and Putnam, R. (2000) Disaffected Democracies: What’s Troubling the Trilateral Countries? Princeton, NJ: Princeton University Press. Posen, B. (1993) ‘The Security Dilemma and Ethnic Conflict’, in M. E. Brown (ed.) Ethnic Conflict and International Security. Princeton, NJ: Princeton University Press. Putnam, R. (1993) Making Democracy Work: Civic Traditions in Modern Italy. Princeton, NJ: Princeton University Press. Roeder, P. (1999) ‘Peoples and States After 1989: The Political Costs of Incomplete National Revolutions’, Slavic Review 58 (winter): 854–82. Rose, R. (2000) ‘Uses of Social Capital in Russia: Modern, Pre-Modern, and AntiModern’, Post-Soviet Affairs 16, No. 1: 33–57. Rustow, D. (1970) ‘Transitions to Democracy’, Comparative Politics 1 (April): 337–63. Silver, B. D. and Dowley, K. M. (2000) ‘Measuring Political Culture in Multiethnic Societies: Reaggregating the World Values Survey’, Comparative Political Studies 33 (May): 517–50. Snyder, J. (1993) ‘Nationalism and the Crisis of the Post-Soviet State’, in M. E. Brown (ed.) Ethnic Conflict and International Security. Princeton, NJ: Princeton University Press.

8

Social trust and democratization in the post-communist societies Gabriel Badescu

Recent assessments of the transition processes in the ex-communist societies tend to agree on the increasing diversity in the quality and extent of their democratization (e.g. Rupnik 2000; Karatnycky 2001; Carothers 2002). One possible category of suspects that accounts for this variation relates to the level of social capital in these countries. In his 1993 book, Putnam deplores the fact that ‘proposals for strengthening market economies and democratic institutions [of developing and transitional countries] center almost exclusively on deficiencies in financial and human capital’ (Putnam 1993: 38). He considers that the deficiencies in social capital in these countries are at least as alarming, and asks, as a solution, for more efforts to encourage ‘social capital formation’ by ‘patiently reconstructing those shards of indigenous civic associations that have survived decades of totalitarian rule’ (Putnam 1993: 39). In studies that followed Putnam’s book, social trust, which in most conceptualizations of social capital is considered as one of its basic components (e.g. Ostrom and Ahn 2001; Newton 1997: 576), has been shown to have different distributions within the populations of the postcommunist societies than within those of consolidated democracies. More specifically, the proportions of people who state that ‘most people can be trusted’ are found to be systematically lower in surveys conducted in the East than in those conducted in Western Europe (e.g. Norris 2001: 11). This chapter will focus on one category of consequences that trust may have on transition within the former communist countries. The main assertion is that generalized trust is a resource for democratization processes. The fact that a high level of trust is linked to more involvement in associational life has received justification, both theoretically and, in the case of some Western nations, empirically (e.g. Brehm and Rahn 1997; Putnam 1993, 2000; Stolle 1998). It is highly debated as to what is the direction of the causal arrow between the two terms, but, in most studies, generalized trust has something to do with membership to associations, even when the effect of other factors is controlled.1 Yet finding a similar positive correlation in the case of the East European countries is less straightforward. In fact, at the macro level, West and East differ significantly with regard to

Social trust in post-communism 121 the way that trust and membership are linked. Across different data and measures, the relation is positive among Western countries and nonpositive among Eastern. My aim in this chapter is to demonstrate that a more detailed methodological treatment leads to a different image than one produced by macrolevel analysis, showing, in fact, important similarities across both West and East European countries. Thus when an analysis is performed at an individual level and the effect of contextual factors is controlled at a national level, in any of the thirteen ex-communist countries under study it is found that volunteer members in associations tend to be more trustful than ordinary citizens. Still, the relationships are generally weaker in transitional countries than in Western ones. I will analyze two possible lines of argumentation: the validity problems of the standard measure of generalized trust, and specific attributes of volunteer associations which are linked to the ethnic context, could each decrease the intensity of the observed relation between trust and membership. Finally, I will discuss the relationships among activity in associations, civil society and democracy within the former communist countries. I will argue that despite the fact that not all volunteer organizations are democratic, the overall effect of civic engagement on democracy is positive in each of these countries. By adding this result to the fact that trustful people are more likely to volunteer in associations, I conclude that generalized trust is an important ingredient for successful democratic transition.

Social trust and civic engagement Classic literature on political culture has implied that interpersonal trust promotes efficient functioning of democratic processes and stability of democratic regimes (Inglehart 1999: 103). An important part of its reasoning is based on the assertion that trust is a resource for collective action. Citizen activism is affected significantly by the capacity of citizens to identify common goals and to pursue them collectively. First, it is expected that more trustful citizens become embedded in denser and more extended social networks. Social trust, in particular, may have the potential to increase the connectives through networks that favor mobilization for participatory acts. Second, a high level of social trust seems to ease empathy towards other interests, by identification with their own. Finally, formal models and experiments showed that more trustful citizens tend to be better at overcoming collective action dilemmas (e.g. Yamagishi and Yamagishi 1994). An accurate empirical check of the causal relation from social trust to citizen activism is especially problematic because of the difficulty in measuring the latter term. One possible strategy is to focus on volunteer membership to associations, considered as an important component, but also as a proxy of a wider range of citizen participatory acts. It certainly misses a

122

Gabriel Badescu

broad category of relevant activities, consisting of ad hoc or regular but uninstitutionalized collective behavior, which are by no means negligible for democratic processes (e.g. Foley and Edwards 1996: 47; Tarrow 1996). However, it is difficult to track these activities by standardized interviews, and even more to make their measures comparable across different societies. The analyses in this chapter will rely mainly on estimations of volunteer membership in associations, as recorded in the 1990 to 1993 and 1995 to 1998 World Values Surveys, in the 1999 European Values Surveys, and in several surveys representative of the Romanian and Moldovan public.2 It is widely acknowledged that social capital can produce not only social goods but also social ‘bads’ (e.g. Portes and Landolt 1994). Social capital has the potential to empower groups of people for any kind of collective action, which sometimes has desirable consequences for their participants and bad consequences for other groups. The former communist countries tend to be characterized by the low level of social capital on some of its dimensions, and by the high level on others. Ethnic heterogeneity, more prevalent traditional social relations, the numerous cases of ineffective state institutions, have enhanced the role of personal networks (e.g. Rose 1998), but at the same time have lowered interpersonal and institutional trust (see Bartkowski, Chapter 11, this volume). In particular, the multiethnic character of a society could influence significantly the equation between trust, civic engagement, and democratization. This is because not all associations are necessarily good for democracy. A high membership in groups that stress ethnic cleavages could enhance the level of conflict in society and undermine those collective actions which cut across ethnic lines (see Dowley and Silver, Chapter 7, this volume). Therefore, the effect of trust on democracy, mediated by involvement in associations, depends on which of the following statements holds true: 1 2

Social trust is a resource for any association, including those that may be harmful to democracy. Social trust tends to influence membership only selectively, having a stronger impact for those associations that have a net positive effect on democratization.

The second point is more congruent to the assertion of a significant role of social trust than the first, especially in the case of societies where the number of undemocratic associations is significant. An accurate empirical testing should consequently be able to distinguish between associations that are ‘good’ or ‘bad’ in terms of their democratic effects. However, not only is information on activities performed by associations in the excommunist societies scarce but, as I will discuss later, assessing the net democratic effect of a certain type of activity is not straightforward. The available data will allow me to examine how the intensity of links between social trust and membership changes across different settings.

Social trust in post-communism 123 Although constrained by design, the analysis will provide some tentative interpretations regarding the mechanics of causal relations between trust and activity in associations.

Civil society, voluntary associations and democracy While virtually absent from intellectual and political life for nearly a century and a half, the concept of civil society gained wider usage during the 1970s regarding Eastern Europe. The fall of communism has brought a new situation where the term embraced additional meanings and, largely as a result of developments in the social capital field, raised new expectations concerning its explanatory power (Foley and Edwards 1998). Within and with regard to the ex-communist countries, the language of civil society has become a visible but polisemic part of the discourse of politicians, academics, business leaders, foundation executives, and citizens. For many actors that oversee democratic transition within the region, civil society refers to ‘dynamic webs of interrelated nongovernmental institutions’ (Keane 1998: 6). Thus, from this perspective, the level of development of civil society can be assessed by looking at the number of nongovernmental organizations, the number of people involved in their activities, and by various aspects regarding their activity. For others, civil society and the nongovernmental sector do not necessarily overlap. Timothy Garton Ash, for instance, considers as an open question whether in Eastern Europe nongovernmental organizations are ‘like seeds to the beautiful turf of civil society or like that thing called astroturf – artificial grass that covers the ground so the real grass cannot grow’ (2000: 400). Mark Warren considers volunteer behavior as central to a definition of civil society, in this way excluding a significant part of nongovernmental associations that are active in the East European countries. He makes the distinction between ‘pure’ associations, and associations that are ‘of’ civil society, but not ‘in’ civil society. The former consist of organizations within which voluntary associative relations are dominant, whereas the latter are organizations that mediate between ‘pure’ associations, states, and markets. Political parties and unions are two main examples in the category of mediating forms of associations (Warren 2001: 58). In this perspective, the proportion of people volunteering in associations, other than parties and unions, provides evidence on how developed civil society is in a certain country. On any of these conceptualizations, the term civil society designates very different realities from one nation to another. It is well documented that a large variance exists among the Western nations regarding the amount and type of membership in civil society organizations (e.g. Aarts 1995; Perlmutter 1991). East European countries present certain specificities but also a large variance among themselves both in quantity and quality. Clear differentiation existed during the communist period (Tismaˇneanu 2001),

124

Gabriel Badescu

and there are compelling arguments that a previously more developed civil society accounts to a large extent for a well-functioning civil society in the present. On the one hand, a significant proportion of the population in countries such as Poland and Hungary, during their communist regimes, were involved in organizations, formal and informal, that retained a high degree of autonomy from the state (Arato 2000: ch. 2). Poland stands at one extreme with up to ten million people in the Solidarity movement, and also boasting the only non-state higher education institute in communist East Europe, a Catholic university. Similarly, in Hungary there were some officially accepted and even supported nonprofit organizations, such as trade unions, Red Cross, Adult Education Society, Patriotic Front, Chamber of Commerce, and so on, and a large variety of voluntary associations (e.g. fishermen’s and hunters’ associations, sport clubs, pet fans’ societies, voluntary fire brigades, amateur theater groups, youth clubs, intellectuals’ organizations, folk dance houses) considered to be harmless and therefore tolerated by the public authorities (Kuti 2002). On the other hand, the former Soviet countries, Bulgaria, Romania, and Albania were at the other extreme, with very limited autonomous activity of groups of people (Nelson 1996; Sampson 1996). In Romania, for instance, a number of nongovernmental organizations were allowed to exist, some even with international affiliation. They included traditional craftsmen, philatelists, sport clubs, or associations for people with disabilities. A few of them, such as the Writers’ Union, were influential enough to negotiate privileges for their members, but their influence on policies was either insignificant or uncivic (Verdery 1991). A special case is that of the so-called ‘Cenaclul Flacara’, which comprised a series of gatherings taking place in all major cities, and organized by a poet close to the circles of power. The audience, consisting of young people, sometimes numbering tens of thousands, was encouraged to sing, dance, create literature. ‘Cenaclul Flacara’ may be regarded as an associative movement providing support for social and cultural exchanges, but at the same time as an instrument of mobilization by the communist regime in favor of nationalistic values (Stoiciu 2001). The type and amount of civil society in the past is relevant because some of the old organizations have survived over time and count as part of civil society in the present. In addition, some of the skills, orientations, and knowledge acquired during the past membership have been employed in the current organizations, or transmitted through family socialization. No less important is the fact that attitudes of population and of the officials toward civil society have been influenced by its degree of development and by its role under the communist regime. In Hungary, for instance, ‘governments did not trust [associations] at all, the most dictatorial ones even tried to completely eradicate them, but they were held in high esteem by citizens’ (Kuti 2002). Romania presents a case where civil society-type organizations have been facing mainly negative attitudes throughout the transition period,3 whereas in the Czech Republic the positive view that

Social trust in post-communism 125 surrounded civic associations at the beginning of the 1990s has been replaced by a climate of distrust and low esteem from both public and political representatives (Fricˇ et al. 1998: 15). One consequence of the extremely diverse picture of civil societies and of their dominant definitions across ex-communist nations is that validity of the comparative measures for both civil society and civic engagement, and, consequently, their relationships with trust can be influenced by factors that are country specific. There are three main mechanisms through which volunteer associations might produce democratic effects in a transitional society: 1

2

3

Social learning effects. Associations may contribute to forming and enhancing those attitudes, knowledge, and skills on which democratic practices are based (e.g. Putnam 1995; Hooghe 1999). While in Western nations family, school, place of work, and friends provide important instances for transmitting democratic ideas, citizens of East European nations cannot rely to the same extent on such mechanisms, which are more resistant to change over time and, in their countries, still bear undemocratic traces. Less influenced by local tradition, and frequently bringing influences of organizational practices from their Western counterparts, volunteer associations in transitional societies have an important potential to develop cognitive and deliberative skills, civic virtues, and a sense of efficacy. Public sphere effects. Associations in general, and volunteer ones in particular, provide ‘the social infrastructure of public spheres that develop agendas, test ideas, embody deliberations, and provide voice’ (Warren 2001: 61). Institutional effects. These are the more established effects of associations on democratic governance (e.g. Foley and Edwards 1996). Associations speak on behalf of groups of citizens, contributing in this way to the aggregation of individual judgment and to representation of collective decisions. In addition, associations have the capacity to affect state and market by organizing demonstrations, strikes, and civil disobedience. Civic associations in the East European countries have become known largely because of their role during the fall of communist regimes, and, later, due to their constant input into shaping the new institutions.

The strength of these effects, and the relative importance of each for the transitional countries, remain unanswered questions here. Although I fully acknowledge that not only volunteer but also non-volunteer associations have the potential to influence democratic processes, and that their influence can sometimes be negative (Schmitter 1997: 248), I will rely on the assumption that in each of the new democratic polities, benefits from the activity of volunteer associations tend to exceed losses.

126

Gabriel Badescu

Empirical testing: macro-level relationships between social trust and membership to associations The relationship between social trust and volunteerism, measured at country level, has been found consistently to be positive in Western nations (e.g. Norris 2001). In the 1999 European Values Survey, the correlation is 0.39 for all countries, and has the same value when the relation is limited to the West European countries. In the case of the former communist countries, the same relation displays a completely different pattern (Figure 8.1), which is summarized by a correlation of
0.21. What could explain this difference between the new and the established European democracies? It could be that social trust has different effects on civic involvement across the two categories of countries. Nevertheless, it is possible that specificity is only apparent, having in fact processes that are similar but reflected by distorted instruments. Measurement errors, model under-specification, statistical assumptions that are not respected, could each potentially explain why results are different between the two categories of countries at an aggregate level. Another difference between East and West is that both measures of social trust and volunteer membership tend to show less stability over time in the case of transitional countries (Table 8.1 and Table 8.2). Is this the 0.6

Proportion of volunteers (%)

0.5

0.4

0.3

0.2

West European countries Ex-communist countries

0.1

0.0 0.0

All countries 0.1

0.2

0.3

0.4

0.5

0.6

0.7

Proportion of trustful people (%)

Figure 8.1 Relationship between the proportion of people who ‘trust other people’ and the proportion of volunteer members in associations Source: Calculated using European Values Survey 1999.

Social trust in post-communism 127 Table 8.1 Correlations between aggregate measures of generalized trust in three cross-national surveys: 1990 World Values Surveys, 1995 World Values Surveys and 1999 European Values Surveys 1995 WVS

1999 EVS

West European countries

1990 WVS 1995 WVS

0.83 (16)

0.88 (25) 0.85 (18)

East European countries

1990 WVS 1995 WVS

0.27 (11)

0.61 (11) 0.44 (13)

Note Cell entries represent Pearson correlation coefficients between proportions of respondents who declare that ‘most people can be trusted’. The numbers of cases (countries) are given in parentheses.

Table 8.2 Correlations between proportion of members in volunteer associations in three cross-national surveys: 1990 World Values Surveys, 1995 World Values Surveys and 1999 European Values Surveys 1995 WVS

1999 EVS

West European countries

1990 WVS 1995 WVS

0.30 (5)

0.47 (11) 0.42 (3)

East European countries

1990 WVS 1995 WVS


0.33 (8)

0.37 (8) 0.52 (12)

Note Cell entries represent Pearson correlation coefficients between proportions of respondents who declare that they are members of volunteer associations; there are slight differences in the way the question was asked in the three comparative studies. The number of cases (countries) are given in parentheses.

result of a faster pace of change within the new democracies, or is it mostly a reflection of problems in operationalization and collecting data which are specific to these countries? The following two sections of this chapter will investigate several measurement aspects regarding the two key concepts of this analysis.

Measuring civic engagement Identifying people who did volunteer work in associations is a difficult task, particularly when standardized interviews are used as a tool. Several results of an overview of the estimates of volunteer membership among the Romanian public could also apply to some extent to other postcommunist countries. 1

Slight variations in the way questions are asked elicit wide disparities in the proportions of respondents who acknowledge their volunteer behavior. Table 8.3 shows estimates of between 3.7 percent and 23.3

128

Gabriel Badescu

Table 8.3 Proportions of volunteers in associations, as estimated by several surveys, representative for the adult population of Romania Survey

Estimated proportion of members (%)

CID October 2001* BOP November 2000 BOP May 2000 BOP November 1999 EVS 1999 BOP May 1999 WVS 1998** RPC 1996 WVS 1993**

6.4 9.2 8.0 3.9 9.6 8.0 23.3 3.7 11.4

Notes * Estimates the proportion of those who performed volunteer activities during the past twelve months. ** Estimates the proportion of active members.

2

percent. Because they are close in time, within a few years or even months, differences are not likely to be attributable to changes in actual membership. Assessments which are based on other data than national representative surveys (e.g. surveys on associations, qualitative interviews, official statistics) suggest that the proportion of Romanians who volunteer in associations is closer to the lower estimates resulting from survey data (Badescu 2002). This indicates that most measures based on national surveys inflate the real proportion of members. One possible reason is that the notion of association, as well as that of volunteer membership, has a wider meaning for some of the survey respondents than is intended.4 One of the important implications is that any correlation between membership measured by a citizen survey and another variable is likely to be underestimated.

Measuring social trust Trust is no less difficult to measure in a comparative study than membership to associations. First, cross-national surveys rely on the assumption of equivalence among translations, which may not be accurate enough in the questions used to assess the level of trust. A study on nine surveys conducted in Romania that contained various items of generalized trust, including three different translations of the same dichotomous question used in the World Values Surveys and European Values Surveys, showed significant variations of results induced by small changes in wording (Badescu 2002). We may assume from this that translations of the same item in different languages can produce even more different estimates of the proportions of trusters in those societies.

Social trust in post-communism 129 A second problem associated with the assessment of the level of trust is that of validity. It has been long assumed that survey questions of social trust, such as the standard one – ‘Generally speaking, do you believe most people can be trusted or can’t you be too careful in dealing with people?’ – which is asked in the WVS and EVS, are operationalizations of that type of trust which is praised for a broad range of benefits, especially those that involve solving collective dilemmas. Only recently, it has been argued that trust is multidimensional, and that its dimensions are distinct from each other empirically as well as analytically. Moreover, the type of trust that ‘helps us get things done’ in general, and has the potential to increase membership in associations in particular, was found to be largely similar to the one that is measured by the standard question (Uslaner 2002: ch. 5). Trust in strangers and trust in people whom the subject knows (relatives, friends), or share some attributes with (people of the same ethnicity, religion) are shown, using survey data on the American public, to have distinctive empirical manifestations (Uslaner 2002: ch. 3). However, members of the public from post-communist societies present several characteristics that may influence their representation of ‘the other people’ whom they are asked to evaluate in terms of trust. In comparison with the Western world, the former communist nations tend currently to have a much lower internal mobility and, in most cases, almost non-existent immigration. In addition, they have a larger proportion of people living in rural localities, which tend to be small, characterized by extended kinship relations, and with low interaction beyond their borders. ‘Most people’ could have a different meaning for someone whose contacts are almost exclusively with relatives and people whom they have known for a significant period of their life, than for a person who has moved several times, traveled extensively, and lives in places that undergo rapid changes in their social structure as a result of migration and immigration. The former category of people is better represented in the East, whereas the latter has a larger proportion in the West. These differences make it plausible that respondents from the former communist nations are influenced in their assessments of ‘other people’ as trustworthy or untrustworthy, not only by their level of trust in strangers, as was shown to be the case of the American public, but also by their level of particularistic trust. It is also true that people from the East have been exposed to interactions with people unlike themselves, in two categories of circumstances that are largely specific to their societies. First, most of the communist societies underwent a swift process of urbanization which entailed significant internal migration. Its extent and timing varied from one country to another, but it had in common the fact that a large number of people moved from rural areas to the fast-growing towns, replacing the traditional strong and stable ties with different types of social relations. Second, the former communist societies tend to be ethnically diverse, and, as a result, any citizen has a significant chance of knowing groups of people who share

130

Gabriel Badescu

common attributes, such as language, customs, sometimes denomination, contrasting with his or her own. Given all these attributes that are specific to the former communist countries, at the same time presenting a wide variation across them, it would be no surprise if the validity of ‘trust in other people’ question, as an indicator of social trust, fluctuates from one society to another. With this likely variance in mind, I will evaluate how social trust is estimated through different survey questions by using the data of a survey representative of the Romanian adult population, the October 1999 Romanian Barometer Survey, one of the few in the post-communist countries that asks about trusting specific groups. This survey provides indicators about trust in other people, people of other ethnicities, people of other religions, other Romanians, family members and neighbors. I will use structural equation models in order to test if the six observed measures can be considered as manifestations of two latent concepts, the notion of trust in strangers and that of particularized trust, respectively. More specifically, the questions asking about trust in family and trust in neighbors are expected to be significantly more strongly correlated with one of the latent variables than with the other. Because both questions refer to people about whom the subject has at least some degree of familiarity, the first factor would be considered as a measure of particularized trust. Similarly, I assert that the questions asking about trust in people of a different ethnicity and trust in people of a different religion are more strongly correlated with the second latent variable. Following Uslaner’s argument in his analysis of the U.S. data, this latter factor will be considered to be a measure of trust in strangers (Uslaner 2002: ch. 3). Trust in people of the same ethnicity is also expected to correlate more with the variable of particularized trust and less with trust in strangers. Because ethnicity has been a salient issue in Romanian society during the past ten years, and also since the public is well aware of its role in some neighboring countries as a source of violence, war and territorial disputes, it would be no surprise to find that subjects tend to consider the other’s ethnicity as a relevant attribute in judging level of trustworthiness. In other words, the tendency to put faith in one’s own family or neighbors, that is, people about whom the subject has previous knowledge, may work similarly in the case of people with whom the subject had no previous contact but shared a common attribute considered to be important. Table 8.4 summarizes the main results of data analysis. Consistent with expectations, five of the six measures of trust are influenced significantly (p  0.05) by only one or other of the two latent variables. The levels of trust in people of other ethnicities and in people of other religions are correlated strongly with the ‘trust in strangers’ factor (0.74 and 0.66 respectively) and not significantly correlated with particularized trust. Trust in one’s own family, in neighbors and in people of the same ethnicity are correlated positively with the ‘particularized trust’ factor (0.34, 0.56, and 0.72, respectively)

Social trust in post-communism 131 Table 8.4 Structural model with two latent variables for trust, Trust in Strangers, and Particularized Trust, in October 1999 Romanian Barometer Survey Trust measure

Particularized trust

Trust in strangers

Most people can be trusted Trust people of other ethnicities Trust people of other religions Trust people of similar ethnicities Trust your family Trust your neighbors

0.30 – – 0.72 0.34 0.56

0.19 0.74 0.66 – – –

Notes Entries are standardized coefficients (Lambda-X); maximum modification index for the coefficients constrained to zero is 1.41; Comparative Fit Index (CFI)  0.99; Adjusted Goodness of Fit Index (AGFI)  0.99; Correlation between the two latent factors  0.61.

and not significantly correlated with trust in strangers. The only measure of trust that is correlated with both factors is trust in other people. This is different from the finding of the U.S. data that particularized trust has a stronger effect on this variable than trust in strangers (0.30, compared to 0.19). Another dissimilarity is that the two factors have a stronger correlation in the Romanian than in the U.S. data (0.61, compared to 0.39). The overall result shows more similarity than dissimilarity, which is remarkable given the difference in the wording of questions and statistical technique. If the same result holds in two widely different societies, then the possibility of its generalization to other contexts is significant. However, the difference should make us cautious when comparing countries based on the standard trust questions. If trust in strangers has a stronger influence than particularized trust on the quantity and quality of social involvement of citizens in democratic societies, we should expect that the standard variable on trust is more useful in explaining democratization where it is a better measure of trust in strangers.

Relationship between social trust and volunteer membership at an individual level Performing individual level analyses, separate for each country, has several advantages over studying only one country-level model of the relation between generalized trust and volunteerism. On the one hand, it solves what statistical analysis describes as ‘ecological correlation problems’ (Huckfeldt and Sprague 1993). On the other hand, it controls for any of the country-level effects, including those induced by translation specificity. When the strength of the two variables is compared across countries, there is another statistical aspect that needs to be taken into consideration. Standard coefficients, such as the Pearson correlation, tend to underestimate the intensity of relations between variables having non-normal distributions. This is exactly the case both with membership of associations and

132

Gabriel Badescu

generalized trust. In almost every country under study, trusters and members comprise less than half of the total, but in transitional countries their proportions are, on average, closer to zero. Therefore, the relationship between the two variables has to be assessed using statistics that are not influenced by their distribution. Table 8.5 summarizes the polichoric correlations and odds ratios for each East European country, as well as for their pooled sample and for the pooled sample of West European countries. Table 8.5 shows a substantially different picture to that emerging from the country-level analysis. Thus in none of the thirteen East European countries under study is the relationship between trust and membership negative, and in eight of them it is significantly positive (p  0.05). This result suggests that the underlying social mechanism linking trust and civic engagement operates in similar ways in different contexts. The countries of East Europe, having diverse results in their democratic consolidation and having quite different structure and size of civil societies, and those of Western Europe, all display remarkable similitude from this point of view. However, the relationship tends to be weaker in transitional countries than in Western countries. Two possible explanations have been mentioned already: the validity problems of the standard measure of social trust, as well as specific attributes of volunteer associations which are linked to the ethnic context, could Table 8.5 Relationships between social trust and volunteer membership of associations Country

Relationships between social trust and volunteer membership Polichoric correlation

Odds ratioa

Bulgaria Croatia Czech Republic Estonia Hungary Latvia Lithuania Poland Romania Russia Slovakia Slovenia Ukraine

0.19* 0.25* 0.14* 0.11 0.23* 0.07 0.10 0.16* 0.05 0.15* 0.20* 0.15* 0.03

1.8* 2.1* 1.5* 1.4 2.0* 1.2 1.4 1.7* 1.2 1.8* 1.8* 1.6* 1.1

Pooled sample for Eastern Europe Pooled sample for Western Europe

0.10** 0.25**

1.4** 1.9**

Notes ** p  0.01; * p  0.05. a Odds ratio indicates, in this case, how many times the chance that a trustful person is a volunteer is higher than the chance that a non-trustful is a volunteer.

Social trust in post-communism 133 decrease the intensity of the observed relation between trust and membership. I will analyze each of these two assumptions. First, I will take again the case of Romania, one of the four countries having a positive but not statistically significant relationship at an individual level. As I showed in the previous section, the standard measure of social trust has in its case a stronger correlation with a latent variable of particularistic trust than with one of trust in strangers. I expect, then, that the correlation between volunteer membership and the latent variable of trust in strangers, which I argued above provides a better operationalization than the standard question of trust, is not only positive, but also stronger than correlations of membership with standard questions and with the latent variable of particularistic trust. Indeed, a structural equation model that includes these variables shows a statistically significant relationship from trust in strangers to membership (beta  0.10; t  3.7), and a non-significant relationship from particularistic trust to membership. This result suggests that, for Romania, and perhaps also for other postcommunist countries, better measures of social trust would lead to stronger positive relations with volunteer membership in associations. Second, I will analyze how the intensity of correlation between trust and membership is influenced by ethnic context. One possible reason for having such an influence is that the structure of civil society may not be the same in ethnically homogenous countries as in those with larger groups of ethnic minorities. Thus the proportion of associations whose functioning is based on ethnic division is likely to be higher when the proportion of ethnic minorities is higher. At the same time, I expect that these kinds of association do not benefit from a higher level of social trust; therefore the proportion of trusters among their members would not be higher than among non-members. As a result, the overall correlation between trust and membership in any association should be lower in societies with a higher proportion of ethnic minorities. The ethnic context could also influence the validity of the social trust survey question. More ethnic diversity results in more exposure of the respondents to ‘people unlike themselves’ and may increase the chance that ‘other people’ means strangers instead of people with whom the subject has regular interaction. Therefore, I believe that the standard question of trust works better when respondents live in multi-ethnic communities. A consequence would be that the observed correlation between membership and trust is more intense when ethnic diversity is higher. If these assertions are correct, then the two types of effect that ethnic context has on the relationship between volunteer membership and social trust are of opposite signs. An overall positive correlation between the ethnic diversity of a society and the strength of correlation between trust and membership would be consistent with the fact that the first effect is stronger than the second, whereas a negative sign would indicate the opposite. By considering the two effects simultaneously, it would be expected that

134

Gabriel Badescu

the weakest correlation between trust and membership is in those societies containing large minority groups that are spread unevenly among the majority group. This is because interaction with people of different ethnicities is less frequent when minorities are concentrated in certain areas than when similar-sized minority groups have a more uniform distribution. An analysis of the 1999 European Values Survey data shows a positive relation between the proportion of ethnic majorities and the intensity of the connection between trust and membership (Figure 8.2). If Moldova is added, using data from another survey,5 the correlation is 0.55 (p  0.04). When data on ethnicity are more refined, the correlation coefficient exceeds 0.65.6 This strong correlation indicates that ethnicity plays an important role in explaining variance of the link between trust and membership across East European countries. Moreover, it could also justify the overall difference between the more ethnically diverse new democracies and the more homogeneous old democracies. The positive sign shows that the influence of ethnic composition over the structure of civil society is significant and stronger than its effect on the validity of the social trust question. But does ethnicity have any effect at all on the way the trust question is understood by respondents? The Romanian data, the only data available with ethnic distribution at subCroatia Hungary

Odds ratio between trust and membership

2.00 Bulgaria Russia

1.75

Slovakia Poland Slovenia

1.50

Czech Republic Estonia

Lithuania

1.25 Romania

Latvia Ukraine 1.00 Moldova 60.00

70.00

80.00

90.00

Proportion of ethnic majority (%)

Figure 8.2 Relationship between the proportion of ethnic majority and odds ratio between trust and membership Sources: Calculated using European Values Survey 1999 and CID 2001 for Moldova.

Social trust in post-communism 135 national levels, suggests that it does. On the one hand, the correlation between trust and membership is higher among people who live in localities that are ethnically more diverse; as I argued before, higher correlation in this case could indicate lower errors in measuring social trust. On the other hand, the strength of correlation between social trust, estimated by the standard question, and trust in strangers, is higher among respondents living in ethnically diverse localities. If the earlier assertion, stating that in Romania trust in strangers is a better measure of social trust than the standard survey question is correct, then a higher correlation with this variable indicates lower error in measuring social trust. The evidence is provided by the 2001 CID survey data. Thus the odds ratio between trust and membership is 1.9 for respondents living in localities with more than 10 percent ethnic minorities, and only 1.1 for the rest of the respondents. In addition, the correlation between social trust and trust in strangers is higher among respondents living in more ethnically diverse localities. This result holds when the effect of the size of locality is controlled for (Table 8.6). In addition to the ethnic context, the level of democratization would be expected to play a role in specifying the causal links between trust and membership. This is because the structure of civil society is likely to be associated with the degree of democratization: the more democratic a country, the higher proportion of members in associations that require a higher than average level of trust. One specific reason is that democratization tends to bring a decrease in ethnic tensions (Dowley and Silver, Chapter 7, this volume), and that lower salience of ethnic issues is expected to keep down the proportion of associations based on ethnic exclusion. As a result, ethnicity would play a more important role in explaining the strength of the link between trust and membership across the less democratic countries than across the more democratic ones. However, when subjected to an empirical examination, none of these assertions is supported. Thus the 2001 Freedom House’s aggregate indicators of democratization7 do not significantly explain the variance of correlation between trust and membership: the Pearson correlation coefficient Table 8.6 Correlations between social trust and trust in strangers across several categories of population, defined by the proportion of ethnic minorities and type of locality Proportion of ethnic minorities

Type of locality Less than 200,000

More than 200,000

Capital city

Total

Less than 10% More than 10%

0.17 0.34

0.24 0.44

0.42 –.00

0.22 0.34

Total

0.25

0.33

0.42

0.27

136

Gabriel Badescu

between the Freedom House indicator (score 1 is for the most democratic countries, 7 for the least democratic) and the odds ratio coefficient between trust and membership is 0.06 (p  0.84). The level of democratization does not have the expected influence on the effect of ethnicity either: the Pearson correlation coefficient between ethnicity (the proportion of ethnic majority) and the odds ratio between trust and membership is higher among the more democratic countries8 – r  0.79 – than among the less democratic ones – r  0.42. These results shows that the level of democratization may influence the external effects of associations – a higher potential for mobilization across ethnic lines being associated with less democracy – but it is not related to the total number of people involved in organizations which stress ethnic difference.

Conclusion There is a link between social trust and civic engagement and it displays remarkable similarities across a large variety of contexts. The analysis presented here shows that in each of the thirteen post-communist societies considered, citizens who are more trustful of other people are more likely to be volunteer members in associations. The same result holds in the most successful cases of democratization, as well as in the least successful ones. Neither past attributes of civil society nor its present level of development seem to influence the strength of relationship between trust and membership. This study indicates, however, two important reasons for dissimilarity among the new democracies, and between them and the older democracies. One relates to the validity problems of the measures that are employed regularly in cross-national research. A large error of the estimate of either social trust or membership tends to decrease the observed strength of relationship between the two, and the size of the error tends to be higher in less modern societies. The second reason concerns the types of volunteer association that exist in a country. Although data analyzed in this chapter do not provide information about the structure and activities of any of the associations, it suggests that some of them do not benefit from having trustful members to the same extent as others. Moreover, it shows that a high proportion of ethnic minorities in a society elicits a larger proportion of members in the types of organization which do not value trust in other people. There is wide agreement in studies that have analyzed the causal mechanism between social trust and volunteer membership that correlation is not due exclusively to a causal connection between membership and trust, but rather to the opposite. Recent research has shown that people who have a high level trust become members of voluntary associations, and that activity in associations does not make people more trusting (Newton 1997; Stolle 1998; Uslaner 2002). It may be that in many

Social trust in post-communism 137 associations in the post-communist countries their members’ level of trust is decreasing over time as a result of exposure to an untrusty environment. But either way, as an asset that benefits or one that is depleted by participation in associational life, in transitional societies social trust appears to be one of the scarce commodities needed for sustaining civic engagement.

Notes 1 When membership is treated as undifferentiated, this result is constant across different data and analyses (e.g. Putnam 1993; Whiteley 1999). If different types of civic engagement are considered separately, trust is found to have a strong positive correlation with membership to some associations, but a weaker effect with membership to others. In a study on several associations from Sweden and Germany, Dietliend Stolle argues that ‘more diverse, more engaged and those with weak ties, accommodate more trusting people’ (1998: 521). In analyzing survey data representative of the American public, Eric Uslaner finds that trust is a strong predictor of membership to several types of organizations, such as business and cultural groups, but it has only a small positive effect on ethnic group involvement, and no impact at all on either church or children’s group membership (Uslaner 2002: ch. 5). 2 The surveys used in this chapter, and their abbreviations, are as follows: 1990–93 and 1995–98 World Values Surveys (WVS); 1998 European Values Surveys (EVS); May 1999, October 1999, May 2000, November 2000 Romanian Public Opinion Barometers (BOP); a Romanian national survey conducted in December 1996 by Paul Sum and the author (RPC); a survey conducted in October 2001 in Romania, and a survey conducted in the same month in the Republic of Moldova (CID) by Eric Uslaner, Paul Sum, Cosmin Marian, and the author. 3 In the May 2001 Public Opinion Barometer survey, respondents were asked to state their level of confidence in several institutions and types of organization. In the list of sixteen institutions, NGOs are placed in the fourteenth position, having a proportion of 73.8 percent of respondents who expressed low confidence in them. 4 For instance, mere church attendance, without any associational component, is sometimes understood by respondents as volunteer membership of an association. Similarly, affiliation to housing associations, to whom all Romanians living in condominiums belong – about one-third of the total population – is sometimes considered as volunteer membership, although in a vast majority of cases one has to join a housing association in order to receive basic services, such as heating and water supply. 5 The Romanian 2001 CID survey. 6 For instance, Moldova is represented in official statistics as having a dominant ethnicity of 64.5 percent, consisting of Moldovans and Romanians. If the two categories are distinguished, the previous proportion will be replaced by about 55 percent of Moldovans. Corrections can also be made with respect of Roma inhabitants. Thus the average estimate for Roma living in Romania is three times higher that the official one, based on the 1992 census (Gheorghe et al., 2000). 7 See Chapter 7 by Dowley and Silver for a definition of indices, or a more detailed explanation at: www.freedomhouse.org. 8 For this analysis, countries are split into two categories: more democratic – those for which the Freedom House index is less than 3 – and less democratic – the remainder.

138

Gabriel Badescu

Bibliography Aarts, K. (1995) ‘Intermediate Organizations and Interest Representation’, in H. D. Klingemann and D. Fuchs (eds) Citizens and the State. Oxford: Oxford University Press. Allport, G. W. (1954) The Nature of Prejudice. Reading, MA: Addison-Wesley. Arato, A. (2000) Civil Society, Constitution and Legitimacy. Lanham, MD: Rowman & Littlefield. Ash, T. G. (2000) ‘Conclusions’, in S. Antohi and V. Tismaˇneanu (eds) Between Past and Future. The Revolutions of 1989 and their Aftermath. Budapest: Central European University Press. Badescu, G. (2002) ‘Încredere s¸i democrat¸ie în t¸aˇrile foste comuniste’, in L. Pop (ed.) Valori ale tranzit¸iei. O perspectivaˇ empiricaˇ. Ias¸i: Polirom. Brehm, J. and Rahn, W. (1997) ‘Individual-Level Evidence for the Causes and Consequences of Social Capital’, American Journal of Political Science 41 (July): 999–1023. Carothers, T. (2002) ‘The End of the Transition’, Journal of Democracy 13 (1): 5–21 Foley, M. and Edwards, B. (1996) ‘The Paradox of Civil Society’, Journal of Democracy 7 (3): 38–52. Foley, M. and Edwards, B. (1998) ‘Beyond Tocqueville: Civil Society and Social Capital in Comparative Perspective’, American Behavioral Scientist 41 (6): 1–9. Fricˇ, P., Petr, P. and Lenka, D. (1998) ‘Defining the Nonprofit Sector: The Czech Republic’, in L. M. Salamon and H. K. Anheier (eds) Working Papers of the Johns Hopkins Comparative Nonprofit Sector Project, No. 27, Baltimore, MD: The Johns Hopkins Institute for Policy Studies. Gheorghe, N., Bitu, N., Tanaka, J. and Stefanescu, S. (2000) ‘Policy Making on Roma in Central and Eastern European Countries: Inventory, Challenges, Commitments, Good Practices and Weaknesses’, in International Expert Symposium on Roma Questions. 28–29 January, Cluj-Napoca. Hooghe, M. (1999) ‘Voluntary Associations and Social Capital. An Empirical, Survey-based Test of the Putnam Hypothesis,’ Paper presented at the 1999 Annual Meeting of the American Political Science Association. Huckfeldt, R. and Sprague, J. (1993) ‘Citizens, Contexts and Politics’, in A. W. Finifter (ed.) Political Science: The State of the Discipline, Washington, DC: American Political Science Association. Inglehart, R. (1999) ‘Trust, Well-Being and Democracy’, in M. E. Warren (ed.) Democracy and Trust. Cambridge: Cambridge University Press. Karatnycky, A. (2001) ‘Nations in Transit: Emerging Dynamics’, available at http://www.freedomhouse.org/research/nitransit/2001/cover-materials/essay1.htm. Keane, J. (1998) Civil Society. Old Images, New Visions. Stanford, CA: Stanford University Press. Kuti, É. (2002) ‘Nonprofit Organizations as Social Players in the Period of Transition: Roles and Challenges’. Unpublished manuscript. Nelson, D. N. (1996) ‘Civil Society Endangered’, Social Research 63 (2): 345–69. Newton, K. (1997) ‘Social Capital and Democracy’, American Behavioral Scientist 40: 575–86. Norris, P. (2001) ‘Making Democracies Work: Social Capital and Civic Engagement in 47 Societies’, Paper presented at the European Science Foundation

Social trust in post-communism 139 EURESCO Conference on Social Capital: Interdisciplinary Perspectives at the University of Exeter, 15–20 September. Ostrom, E. and Ahn, T. K. (2001) ‘A Social Science Perspective on Social Capital: Social Capital and Collective Action’, Paper presented at the European Science Foundation EURESCO Conference on Social Capital: Interdisciplinary Perspectives at the University of Exeter, 15–20 September. Perlmutter, T. (1991) ‘Italy: Why No Voluntary Sector?’, in R. Wuthnow (ed.) Between States and Markets: The Voluntary Sector in Comparative Perspective. Princeton, NJ: Princeton University Press. Portes, A. and Landolt, P. (1994) ‘The Downside of Social Capital’, The American Prospect 26 (May–June): 18–21, 94. Putnam, R. (1993) Making Democracy Work: Civic Traditions in Modern Italy. Princeton, NJ: Princeton University Press. Putnam, R. (1995) ‘Bowling Alone: America’s Declining Social Capital’, Journal of Democracy 6 (January): 65–78. Putnam, R. (2000) Bowling Alone. New York: Simon & Schuster. Rose, R. (1998) ‘Getting Things Done in an Anti-modern Society: Social Capital Networks in Russia’, Social Capital Initiative Working Paper No. 6. Rupnik, J. (2000) ‘On Two Models of Exit from Communism: Central Europe and the Balkans’, in S. Antohi and V. Tismaˇneanu (eds) Between Past and Future. The Revolutions of 1989 and their Aftermath. Budapest: Central European University Press. Sampson, S. (1996) ‘The Social Life of Projects. Importing Civil Society to Albania’, in C. Hann and E. Dunn (eds) Civil Society: Challenging Western Models. London: Routledge. Sandu, D. (1999) Spat¸iul social al tranzit¸ iei. Ias¸i: Polirom. Schmitter, P. C. (1997) ‘Civil Society East and West’, in L. Diamond and M. F. Plattner (eds) Consolidating the Third Wave Democracies. Baltimore, MD: Johns Hopkins University Press. Stoiciu, A. (2001) ‘Making Civil Society Work – Romania 2001’, available at http://www.idee.ro/new_page_6.htm. Stolle, D. (1998) ‘Bowling Together, Bowling Alone: The Development of Generalized Trust in Voluntary Associations’, Political Psychology 19: 497–526. Tarrow, S. (1996) ‘Making Social Science Work Across Space and Time: A critical Reflection on Robert Putnam’s Making Democracy Work’, APSR 90 (2). Tismaˇneanu, V. (2001) ‘Understanding the Balkans’, review of M. Glenny, The Balkans: Nationalism, War, and the Great Powers, 1804–1999, Tikkun 16 (2): 65–68. Uslaner, E. (2002) The Moral Foundations of Trust. New York: Cambridge University Press. Verdery, K. (1991) National Ideology Under Socialism. Identity and Cultural Politics in Ceausescu’s Romania. Berkeley, CA: University of California Press. Warren, M. E. (2001) Democracy and Association. Princeton, NJ: Princeton University Press. Whiteley, P. F. (1999) ‘The Origins of Social Capital’, in J. W. van Deth, M. Maraffi, K. Newton and P. F. Whiteley (eds) Social Capital and European Democracy. London: Routledge. Yamagishi, T. and Yamagishi, M. (1994) ‘Trust and Commitment in the United States and Japan’, Motivation and Emotion 18: 129–66.

9

Civic engagement in a postcommunist state Jeffery J. Mondak and Adam F. Gearing

During the communist era, numerous social and political constraints limited civic engagement in communities throughout Central and Eastern Europe. This chapter explores whether such constraints have left lasting marks, and whether such lingering effects potentially slow the process of democratization. Three specific questions were explored: (1) whether civic engagement in a post-communist state differs from that in an established Western democracy; (2) whether the quality of social perception suffers when civic engagement is constrained; and (3) whether a link exists between civic engagement and tangible forms of political judgment. Data were analyzed from surveys conducted in Cluj-Napoca, Romania, in 1994 and in South Bend, Indiana, in 1984. Compared with South Bend residents, the people of Cluj-Napoca discuss politics less, engage in interactive forms of political participation at lower rates, know less about their neighbors, and fail to link the interests of people in the community to broader political judgments. Given the centrality of civic engagement to democratic legitimacy, these findings justify concern regarding the prospects for full democratization in Central and Eastern European contexts in which social interaction does not yet flourish. In 1989, the nations of Central and Eastern Europe took such bold steps toward democracy that the transition from communist rule seemed tantalizingly simple. However, to depict democratization as anything less than a complex, multidimensional process is to vastly understate the obstacles present in post-communist states. At minimum, the requirements for a successful transition include the existence of appropriate institutional structures, the rise to power of leaders who truly embrace democratic practices, navigation of historical and cultural impediments, a flourishing of independent news media, and a citizenry willing and able to participate in ensuring the representation of their own individual and collective interests. This last requirement – civic engagement of the mass public – is our focus. Democracy will flounder if citizens remain detached both from their political system and from one another. Unfortunately, such detachment may be widespread in Central and Eastern Europe. We argue that the societal remnants of communist rule, coupled with more long-standing

Civic engagement and post-communism 141 cultural forces, threaten the most fundamental forms of social and political interaction. A democratic transition may stall rather than progress if the polity in question is too atomized. Consequently, that which endangers civic engagement likewise endangers systemic legitimacy. We posit that characteristic societal features in the nations of Central and Eastern Europe may limit simple face-to-face interaction in many communities. Simply put, the residents of some communities may not talk to one another, and, more to the point, they may not talk about politics. If we are correct, then have we ruled out all forms of political participation in such contexts? No. Voting, the most fundamental participatory act, requires neither civic engagement nor even social interaction. Likewise, the prevalence of news media means that even severe interpersonal isolation need not bring corresponding ignorance of national affairs. People who do not talk to one another can still remain apprised about political developments, and they can cast ballots in local and national elections. Hence we do not mean to imply that obstacles to civic engagement jeopardize all forms of mass political behavior. Although some forms of political participation can occur despite the absence of social interaction, it is difficult to overstate the pivotal role civic engagement plays in mass politics. People who do not interact with one another may fail to develop an appreciation for any form of communal good, and thus they may be limited in their capacity to see politics in terms of general rather than purely personal interests. Talk – actual face-to-face discussions about politics and society – is an essential ingredient for the emergence of an effective citizenry (e.g. Barber 1984). This point is of particular significance in transitional states where respect for democratic principles and practices is not yet firmly established. Consequently, some analysts (e.g. Nelson 1996) have warned that impediments to civic engagement in post-communist Central and Eastern Europe threaten the legitimacy, and thus the systemic survival, of democratic governance. Although we are far from willing to pronounce the demise of the democratic transition, we do propose that the possible existence of atomized polities constitutes a severe and immediate threat. Our focus on political discourse is related, but not identical, to discussions of civil society. Although we concur with Ekiert (1991) that a lack of definitional clarity plagues research on civil society, the strength of nonpolitical groups and associations has emerged as a dominant operational definition of the civic community (e.g. Putnam 1993, 1995a, 1995b; Nelson 1996; Brehm and Rahn 1997). In the case of Italy, for example, Putnam (1993) argued that the relative success of democracy in the north hinges largely on the prevalence of networks of civic engagement embodied in various groups and associations. Putnam (1995a, 1995b) has also warned that associational activity may be declining in the United States, a development of potentially substantial significance (see Schudson 1996; Mondak and Mutz 1997). Our approach differs from past research on civil

142

Jeffery J. Mondak and Adam F. Gearing

society in that we do not study participation in trade unions, literary societies, and such, but instead more fundamental forms of behavior. Associational activity may foster both civic conversation and attachment to community, but for our purposes, the bottom line in assessment of social capital is whether such conversations and attachments exist at all (for a similar view see Lake and Huckfeldt 1998). The specific questions we pursue concern the nature and significance of citizens’ interactions within the local community. In considering the nature of society in Central and Eastern Europe, we see numerous potential barriers to civic engagement; thus we question whether these barriers do, in fact, constrain social interaction, and, if so, with what political ramifications. We approach these questions by drawing on insights from research conducted in the West, and then contrasting those insights with our observations in one Eastern European municipality. Our comparative inquiry examines three interrelated questions. First, we consider whether community-level civic engagement in postcommunist Europe differs from civic engagement in the West. In particular, we focus on face-to-face interaction. That is, are political discussion and other forms of civic activity in Central and Eastern Europe adversely affected by characteristics of the local political context? Second, we examine whether the quality of people’s social perceptions suffers when civic engagement is limited. For instance, if attachments to the local community are weak, then it may be the case that basic information about the community is simply unavailable to the typical citizen. If people do not perceive themselves as being part of a broader community, then we should not expect them to attend to information about that community. Third, we explore the link between civic engagement and political judgment. If social interaction helps to foster concern with the communal good (e.g. Barber 1984; Putnam 1993, 1995a), then attitudes about politics may reflect atomization in those contexts in which civic engagement does not flourish. We make no claim that the empirical analyses reported below offer definitive answers to the questions we have raised. The simple two-city comparison conducted here represents only a first step in the analysis of important theoretical and substantive questions, and thus we wish to emphasize our hope that subsequent research tackles these and related questions with even greater rigor. With systematic research on social capital still in its infancy, works such as this and the other chapters in this volume are intended to demonstrate the utility of rigorous inquiry and to highlight important questions for future research.

Comparing two cities The questions we have raised concern the nature of mass political behavior during the post-communist transition in Central and Eastern Europe. This focus obviously required individual-level data. Two additional

Civic engagement and post-communism 143 requirements were recognized. First, because our interests centered on the extent to which individuals interact with others, it was preferable that data be drawn from community-level rather than national or cross-national surveys. All survey respondents, of course, reside in communities of some kind; the problem is how best to measure community-level effects. A national-level survey would include only a few respondents from any given municipality, and thus our ability to explore civic engagement – which takes place primarily within one’s community – would be highly constrained. In contrast, survey data from a single city would allow a much more detailed study of social and political interaction. A second requirement was data from a comparable city in the West. Such data were necessary to provide context. If we examined data only from a transitional state, we would have no way of gauging the relative quantity and quality of any civic engagement we discovered. In short, we needed a benchmark city from a stable democracy to anchor our comparison. We contrasted data from the Transylvanian city of Cluj-Napoca, Romania, with data from the U.S. city of South Bend, Indiana. The data from Cluj-Napoca are from a survey we conducted in late 1994,1 on the eve of the fifth anniversary of Romania’ s revolution; the data from South Bend are from the South Bend Survey conducted by Robert Huckfeldt and John Sprague in 1984 (see Huckfeldt and Sprague 1995).2 We chose to model many of the questions in our Cluj-Napoca survey on items from the South Bend Survey, which is an extremely rich data source for students of community-level civic engagement. Cluj-Napoca and South Bend clearly differ in that the former was in the midst of a transition from communism to democracy, whereas the latter is a city in a stable Western democracy. Several features of the two cities made them good candidates for comparative inquiry, particularly with respect to some of the societal factors that may influence patterns of social interaction. First, the cities are of similar size; about 300,000 people live in Cluj-Napoca, compared with 250,000 in the South Bend metropolitan area. Second, Cluj-Napoca’s ethnic mix (the population is 76 percent ethnic Romanian) is matched precisely by South Bend’s racial mix (76 percent white). Third, each city is the site of a major university and has surrounding areas that are largely agricultural. Fourth, presidents Ion Iliescu and Ronald Reagan had been in office for comparable periods of time at the point when data were gathered (one empirical question considered below concerns how citizens evaluate incumbent presidents). South Bend has much in common with other Western cities, and little in common with cities in Central and Eastern Europe; conversely, ClujNapoca is similar in many ways to other communities in the region, and quite different from most cities in the West. Consequently, similarities and differences in political behavior in the two cities may, in a very general sense, reflect basic contrasts between cities in established democracies and those in transitional states. Nonetheless, because the specific findings we

144

Jeffery J. Mondak and Adam F. Gearing

report below must be assumed to represent the nature of mass politics in only these two cities, replication will be needed to test the generalizability of our conclusions.3 We have suggested that the people of cities such as Cluj-Napoca face impediments to civic engagement, impediments that are rare in Western cities such as South Bend. In Cluj-Napoca, we see five specific obstacles to social interaction: 1

2

Free and open political discussion did not take place during Nicolae Ceausescu’s totalitarian rule, and thus the citizens of Cluj-Napoca simply are not practiced in civic engagement. The Ceausescu regime engaged in a pervasive infiltration of virtually all social space. The Securitate, Romania’s equivalent of the KGB, was the largest force of its kind in Eastern Europe, and arguably the most effective. Surveillance of private citizens by the Securitate was routine, including the censoring of all foreign correspondence and the monitoring of both foreign and domestic telephone conversations (Bachman 1991; Gross 1996). Likewise, print and broadcast media were controlled by the state. Ceausescu exerted such extreme control over the flow of information in Romania that the nation’s citizens stayed clear of political discussion, both because they lacked raw material (i.e., objective information) and because they feared that any expression of opinion would immediately be made known to the Securitate. The overthrow of the Ceausescu regime might have opened a floodgate of political discussion, except that the legacy of totalitarian rule manifested itself in the behavior of both national and local political leaders after Romania’s revolution. In 1990, for example, President lliescu ordered the forcible suppression of political protest in Bucharest (Nelson 1990). One cause of protest was the widely held belief that the Securitate had not been disbanded. In Cluj-Napoca, Mayor Gheorghe Funar, the leader of one of Romania’s ultranationalist political parties, attempted to block a political march in 1992, and he outlawed the display of signs in languages other than Romanian (Gallagher 1993). The people of Cluj-Napoca experienced explicit suppression of political discourse for decades under communism, and they perceived at least some signs that little had changed after 1989. Consequently, we see it as reasonable to assume that norms of political participation and civic engagement were not well established in Cluj-Napoca in the first years after Romania’s revolution, particularly in comparison with Western cities such as South Bend. Relatively few of Cluj-Napoca’s citizens have extensive roots in the community. Under Ceausescu’s ‘Romanization’ program, ethnic Romanians were provided with incentives to move to Cluj-Napoca and other Transylvanian cities in an attempt to weaken Hungarian influence in the region (Glenny 1990; Bachman 1991; Szajkowski

Civic engagement and post-communism 145

3

4

1993). The influx of people from rural areas substantially increased Cluj-Napoca’s population, but the new residents came for financial reasons rather than because of family-based or cultural attachments. Moreover, efforts to downplay all aspects of Hungarian culture created a new Cluj-Napoca that was barely recognizable to the city’s most senior residents.4 Hence, both for old residents and new of ClujNapoca, a critical legacy of the Ceausescu regime is a weak sense of attachment to community. The nature of housing in Cluj-Napoca presents an impediment to social interaction. In the West, we idealize communities in which neighbors chat over a hedge or join neighborhood associations. The physical structure of a neighborhood contributes to such prototypical forms of civic engagement. In Cluj-Napoca, as in other cities in Central and Eastern Europe, many citizens reside in rather grim lookalike cement high-rise apartment buildings, not single-family dwellings. These buildings were constructed during the Ceausescu regime, when concern with aesthetics was cast aside because of the more pragmatic need to house large numbers of new residents as quickly and efficiently as possible. Residents do refer to Cluj-Napoca’s neighborhoods by name, but the neighborhoods have little else in the way of identifying characteristics. Such an environment inherently constrains civic engagement. Cluj-Napoca’s neighborhoods are what students of neighborhood organizations label ‘diffuse’ (e.g. Warren and Warren 1977; Williams 1985). Diffuse neighborhoods tend to be homogeneous, and residents move to their dwellings in large waves, as happened in Cluj-Napoca during the Romanization program. Although some sense of identity may exist in diffuse neighborhoods, interaction within the neighborhood and linkages between the neighborhood and the larger community are minimal. The very structure of such neighborhoods discourages civic engagement. The quality of Cluj-Napoca’s local news media is poor. At best, local media, and especially local newspapers, can foster attachments to community and civic pride. In South Bend, as in many U.S. cities, the local Tribune serves precisely these functions. Professional news media also can contribute to social interaction by providing much of the raw material that fuels interpersonal political discussion (Mondak 1995a, 1995b). Unfortunately, media throughout Central and Eastern Europe suffer two crippling flaws that preclude them from contributing to civic engagement. First, local media, and especially newspapers, tend to be partisan vehicles. This is an outgrowth of the communist era, when newspapers were mechanisms to disseminate propaganda. Rather than promoting civic cohesiveness, media today still often exacerbate social and political tensions. Second, adherence to professional journalistic standards is rare because there was no professional training of journalists under communism. As a result, the capacity of

146

5

Jeffery J. Mondak and Adam F. Gearing local media to facilitate citizens’ acquisition of useful factual information is limited (see Paletz et al. (1995) on media in post-communist Central and Eastern Europe; on Romanian media, see Gross 1996; Boc and Mondak 2001; Mondak, in press). Especially in the early to mid-1990s, Cluj-Napoca’s main newspapers functioned as propaganda tools for the city’s nationalist mayor, Gheorghe Funar. Hence, rather than fostering civic engagement, local media served to intensify social divisions. Ethnic tension impedes civic engagement in Cluj-Napoca. Transylvania’s long-standing ethnic cleavage was exacerbated by Ceausescu’ s Romanization program, and this cleavage persists today. Cluj-Napoca, like Transylvania as a whole, is populated by a sizable (more than 20 percent) Hungarian minority. The presence of this ethnic minority has provided political fodder for leaders such as Funar, who, like Ceausescu, have taken up the banner of nationalism as a means to generate and maintain mass support. The consequence in Cluj-Napoca has been a deepening of the city’s ethnic divide. Indeed, numerous international observers have noted the inflammatory behavior of Funar and his Party of Romanian National Unity (e.g. Gallagher 1993; Shafir 1993; Brown 1994; Anonymous 1995). Democracy clearly cannot thrive in a context of intolerance. Citizens of a community simply will not work together when they are divided against one another along ethnic lines (Terry 1993).

Although Cluj-Napoca is one of only a few municipalities in Central and Eastern Europe with these five potential barriers to civic engagement, these and other impediments can easily be presumed to exist in various combinations throughout post-communist Europe. In Western cities, in contrast, such conditions are rare. High levels of social mobility have weakened people’s roots in some Western communities, yet voluntary relocation clearly differs from systematic state-imposed resettlement. Likewise, although ethnic (and racial, religious, and economic) tension can certainly be found in the West, such tension rarely defines the fabric of entire communities. The bottom line is that civic engagement can flourish in places like South Bend to an extent currently inconceivable in cities such as Cluj-Napoca. The five impediments described above are interconnected and mutually reinforcing, creating a context we suspect to be hostile to social interaction. Our objective is not to determine which of these factors exerts the greatest influence on mass political behavior in Cluj-Napoca. Instead, our central empirical tasks are to assess whether civic engagement is indeed muted, and what this may imply for mass political behavior in a broader sense. Accordingly, our results can best be construed as only a wide-angle snapshot, yet one that we believe illustrates the importance of close attention to the effects of structural and cultural forces on social capital.

Civic engagement and post-communism 147

Patterns of civic engagement Our initial question concerns whether face-to-face social interaction occurs less frequently in Cluj-Napoca than in South Bend. We have hypothesized that numerous forces in Cluj-Napoca combine to discourage civic engagement. As a first step toward exploring this hypothesis, we examined levels of political discussion. Interpersonal discussion is quite obviously a fundamental component of social communication (Barber 1984; Huckfeldt and Sprague 1995; Mondak and Mutz 1997). People transmit information when they talk with one another, potentially fostering the emergence of concern with communal benefits. As such, discussion may be seen as foundational to other, more formal acts of civic engagement typically addressed in research on social capital, including participation in voluntary associations. If we are correct in our assessment of the state of affairs in Cluj-Napoca, then interpersonal contact there may be infrequent. The South Bend and Cluj-Napoca surveys included items measuring the frequency of political discussion in various contexts. These data are summarized in Table 9.1, which shows the percentage of respondents who reported that they had participated in political discussion in each of three contexts. The pattern of results was similar in the two cities: such discussion is most common in the family, slightly less common with people from outside of the neighborhood,5 and least common with respondents’ neighbors.6 This similarity, however, does not obscure the striking difference in overall levels of discussion: strong majorities of South Bend respondents Table 9.1 Levels of interpersonal discussion Item

South Bend

Cluj-Napoca

%

N

%

N

Discuss politics with members of your family/Talk about politics with members of your family

93.68

1,504

58.06

949

Discuss politics with people in your neighborhood/Talk about politics with people here in this neighborhood

75.88

1,505

36.54

947

Talk about politics with people outside of your neighborhooda

90.20

1,500

54.24

944

Notes See the Appendix at the end of this chapter for complete text of South Bend items and ClujNapoca items. Cell entries indicate the percentage of respondents who ever engaged in the named form of discussion. a Data from South Bend are from three items (discuss politics with members of social or civics groups, discuss politics at work, discuss politics while having some recreational fun); the cell entry is the percentage of respondents who have ever engaged in at least one of these activities.

148

Jeffery J. Mondak and Adam F. Gearing

discuss politics in each setting, whereas the corresponding rates in ClujNapoca lag far behind. Barely half of the respondents reported ever having discussed politics with members of their family or with people from outside of the neighborhood, and fewer still have discussed politics with neighbors. Further, if anything, the data from Cluj-Napoca overstate the level of political discourse in that city. Survey items measuring discussion in Romania included four response categories. Of those respondents who do discuss politics, large majorities reported that they engage in such discussion only infrequently. For example, only 346 of 947 Cluj-Napoca respondents ever discuss politics with their neighbors, and 317 of these 346 discuss politics only ‘once in a while.’7 Circumstantially, the data in Table 9.1 support our claim that various impediments constrain civic engagement in Cluj-Napoca. However, we are unable to attribute the low level of discussion in Cluj-Napoca to the specific societal barriers described above. With data from only one Eastern European municipality, we cannot determine precisely why political discussion occurs with relative infrequency. Of the five potential obstacles discussed, we have additional leverage only on the possible impact of ethnic tension. Our Cluj-Napoca sample included 716 ethnic Romanians and 211 ethnic Hungarians; these proportions are representative of the city’s population. If ethnic tension dampens civic engagement, this effect should operate predominantly among the Hungarian minority. Ethnic Hungarians should be less likely than Romanians to discuss politics, especially outside of the family, because public expression of political beliefs risks drawing antagonistic responses from those ethnic Romanians who espouse intolerance toward Hungarians. Our data support this point. More ethnic Romanians discuss politics with their families than do ethnic Hungarians, but the gap is small (59.51 percent versus 52.61 percent) and statistically marginal ( 2  3.20, p  0.10). Outside of the family, however, the gap is much greater: 39.73 percent of ethnic Romanians have discussed politics with neighbors, versus 25.71 percent of Hungarians ( 2  13.80, p  0.001); similarly, 58.65 percent of Romanians have discussed politics with people from outside of the neighborhood, versus 38.28 percent of Hungarians ( 2  27.15, p  0.001). Most members of Cluj-Napoca’s ethnic minority stay clear of public acts of political expression; this finding provides indirect support for our contention that societal impediments are the driving force behind low levels of social interaction in the city. We next considered whether differences in levels of civic engagement in the two cities extend to conventional forms of political participation. The same forces that limit political discussion in Cluj-Napoca may restrict any form of political participation that requires face-to-face interaction. Table 9.2 shows response data for four types of participation, ranging from highly active (working for a political party or candidate) to passive (media exposure). The results conform to our expectations: Relative to respon-

Civic engagement and post-communism 149 Table 9.2 Political participation and media exposure Item

South Bend

Cluj-Napoca

%

N

%

35.89

1,488

4.23

945

39.31

1,488

20.46

948

43.85

1,487

30.10

947

94.35

1,488

92.83

949

Read the South Bend Tribune/ Read one of the local papers published here in Cluj

97.22

1,512

85.68

949

See some other daily paper/Read one of the national papers published in Bucharest

89.07

1,510

71.52

948

Campaign activity Work for a political party or candidatea Persuasive political discussion Try to convince friends to vote your way/Try to persuade another person to vote for a particular political party or candidate Attendance at political meetings Attended political meetings, rallies, dinners or similar eventsb Media exposure Listen to local reports on radio or TV/Watch the nightly news on television

N

Notes See the Appendix at the end of this chapter for complete text of South Bend items and ClujNapoca items. Cell entries indicate the percentage of respondents who ever engaged in the named form of discussion. a Data from South Bend are from two items (work for a political party, work for a candidate); cell entry is the percentage of respondents who had ever engaged in at least one of these activities. b Data from Cluj-Napoca are from two items (attend a political meeting, attend a protest rally or demonstration); cell entry is the percentage of respondents who had ever engaged in at least one of these activities.

dents in South Bend, residents of Cluj-Napoca are substantially less likely to have worked for a political party or candidate, or to have tried to sway the votes of acquaintances. A lesser gap exists for attendance at political meetings, although South Bend respondents once again outscored their Romanian counterparts. We have argued that societal impediments to civic engagement are the likely cause of low levels of face-to-face social interaction in Cluj-Napoca. Although the data we have reported are consistent with this claim, might it be that the people of Cluj-Napoca are simply apathetic with regard to any information about politics? Media exposure requires no social interaction

150

Jeffery J. Mondak and Adam F. Gearing

whatsoever; a person can be completely isolated from others and yet still follow politics in the news. Thus the data on self-reported media exposure (Table 9.2) allow us to test whether it is general political apathy or a more specific reluctance to interact with others that most distinguishes citizens of Cluj-Napoca from the people of South Bend. The results allow us to reject the possibility of widespread political apathy. Just as in South Bend, large majorities of Cluj-Napoca respondents reported exposure to news media. It is especially noteworthy that exposure to broadcast media – the most passive and atomized act we considered – is almost precisely as common in Cluj-Napoca as in South Bend. Although we were most interested in the factors that distinguish ClujNapoca and South Bend, there were similarities as well. The Cluj-Napoca respondents and their American counterparts are seemingly equal in the capacity to monitor national affairs via news media. Thus it appears that the people of Cluj-Napoca have not distanced themselves from the political world, but rather from one another. We see the observed dearth of civic engagement as being of prima facie significance, yet we also question whether the low levels of social interaction we have identified correspond with more tangible effects regarding our respondents’ perceptions of politics and society.

Social perception The importance of civic engagement largely hinges on the capacity of social interaction to transmit information. When we talk to our neighbors we learn something about them, and what we learn may influence our subsequent political judgments. Given that we have determined levels of social interaction to be low in Cluj-Napoca, we may reasonably expect to find corresponding effects on respondents’ social perceptions. Our particular focus is on perceptions about the economy: What is the nature of the economic perceptions held by the residents of Cluj-Napoca, and how do they compare with those of South Bend residents? We focus on economic perceptions for two reasons. First, particularly in transitional states, politics and the economy are inextricably linked (e.g. Przeworski 1991). In Central and Eastern Europe, there is a real possibility that prolonged economic hardship resulting from market reform may ultimately obstruct continued democratic consolidation. Indeed, linkages between perceptions of the economy and support for democratic regimes have been found in several cross-national studies in post-communist Europe (e.g. Rose and Mishler 1994; Evans and Whitefield 1995; Mishler and Rose 1996). Because economic perceptions matter for democratic stability, the possibility that levels of civic engagement partly determine the nature of those economic perceptions warrants study. A second reason to examine economic perceptions is that those perceptions can vary depending on the unit under consideration. A wealth of

Civic engagement and post-communism 151 research has explored the distinction between ‘pocketbook’ perceptions, which concern the economic well-being of a person’s own family, and ‘sociotropic’ perceptions, which center on the national-level economy (e.g. Kinder and Kiewiet 1979; Kiewiet 1983; Eulau and Lewis-Beck 1985; Lewis-Beck 1988). More recently, analysts have suggested that meaningful economic perceptions also may exist at levels between the family and the nation, including social, demographic, and class groups (e.g. Weatherford 1978, 1982; Conover 1985; Mutz and Mondak 1997); the neighborhood (e.g. Mondak et al. 1996); the state (e.g. Stein 1990; Atkeson and Partin 1995); and the region (e.g. Pattie and Johnston 1995; Lyon et al. 1997). The potential existence of multiple levels of economic perception is important for our purposes because pocketbook and national-level sociotropic perceptions can be formed even by atomized individuals, whereas some intermediate-level perceptions (e.g. those about the neighborhood) likely require social interaction. Thus, pocketbook and national-level sociotropic perceptions should not differ in form for respondents in South Bend and Cluj-Napoca, but intermediate-level perceptions should. Although we see economic perceptions as important, we do not mean to imply that other perceptions are inconsequential. In this sense, the results presented below concerning economic perceptions should be viewed as symptomatic of more general patterns of behavior. We know that our respondents do not discuss politics much at all, and especially with their neighbors. This may have important consequences for economic perceptions, and those effects, in turn, may be suggestive of broader implications for politics and society. The South Bend and Cluj-Napoca surveys both included standard measures of pocketbook and sociotropic retrospective economic perceptions, with a one-year frame of reference in South Bend and a five-year period in Cluj-Napoca. The choice of five years in Cluj-Napoca was appropriate because our survey was conducted five years after Romania’s revolution. Both surveys also included an item asking respondents to assess the economic well-being of families in the local neighborhood. We should expect residents of Cluj-Napoca to have had some difficulty in answering the neighborhood-level economic item, given both the low level of social interaction we have observed and the poor quality of local media. Descriptive statistics for the family, national, and neighborhood economic measures are reported in Table 9.3. Not surprisingly, mean scores on all items are much lower in the struggling post-communist community of Cluj-Napoca than in South Bend (all data were coded on a scale from 0  much worse to 4  much better). The levels of nonresponse for the family and nation items were low both in South Bend and Cluj-Napoca; virtually all respondents can gauge the economic well-being of their own families, and most respondents will also offer assessments of national economic conditions. This same pattern recurs for the neighborhood perception in South Bend. In stark contrast, however, more than a quarter of

152

Jeffery J. Mondak and Adam F. Gearing

Table 9.3 Retrospective economic perceptions South Bend

Family Nation Neighborhood Neighborhood – observed

Cluj-Napoca

Mean

SD

N

Nonresponse rate (%)

Mean

SD

N

Nonresponse rate (%)

2.16 2.19 2.10 2.16

0.95 0.99 0.76 0.21

1,488 1,367 1,360 1,316

0.27 8.13 8.60 –

1.56 1.09 1.50 1.53

1.12 0.95 1.02 0.14

938 885 686 9

1.26 6.84 27.79 –

Note See the Appendix at the end of this chapter for complete text of items. For both South Bend and Cluj-Napoca, data were coded 0 (much worse-off) to 4 (much better-off). Observed neighborhood scores were calculated using data for the family item in each city, with those data aggregated by neighborhood.

Cluj-Napoca respondents refused to even hazard a guess as to the economic well-being of their own neighbors. The people of Cluj-Napoca participate in little social interaction, and they are especially reluctant to engage in conversations with their neighbors. As an apparent consequence, many residents of Cluj-Napoca fail to form an important social perception. Nonresponse on the neighborhood economic item was particularly high – 33.78 percent –among those Cluj-Napoca respondents who reported that they have never discussed politics with their neighbors (the nonresponse rate was 17.05 percent for respondents who have engaged their neighbors in political discussion). Perhaps the neighborhood data in Table 9.3 should be viewed with optimism rather than concern. After all, despite the existence of numerous impediments to social interaction, more than 70 percent of Cluj-Napoca respondents willingly assessed the economic well-being of their neighbors. Before we conclude that the glass is half full, though, we should note that the simple fact that respondents answered a survey item does not ensure that the resulting data are rooted in reality. If social interaction is minimal and local media are partisan and unprofessional, then what information underlies respondents’ neighborhood-level economic perceptions? That is, might these perceptions be of poor quality? The last row in Table 9.3 reports descriptive statistics for the observed economic conditions of sixteen South Bend neighborhoods and nine ClujNapoca neighborhoods. These scores were calculated by aggregating the family-level data by neighborhood; thus the observed score for any given mean value on the family item for all respondents who live in that neighborhood.8 We can use these data to gauge the quality of respondents’ social perceptions. If perceptions of neighborhood-level economic wellbeing are accurate, then those perceptions should be correlated with the

Civic engagement and post-communism 153 observed scores. Note that the standard deviation is lower in Cluj-Napoca than in South Bend, meaning that Cluj-Napoca’s neighborhoods are more homogeneous. This fact compounds further the process of social perception: if neighborhoods are relatively indistinguishable, then the opportunity for citizens to form clear perceptions about their neighborhoods will be limited. Table 9.4 reports the results of ordinary least squares (OLS) regression models in which the dependent variable is the individual-level neighborhood perception, and the three independent variables are the family and nation economic perceptions and the observed neighborhood score. In both South Bend and Cluj-Napoca, the family and nation variables emerge as extremely strong predictors of the neighborhood-level score, indicating that the three perceptual variables are closely intertwined. This link is stronger in Cluj-Napoca than in South Bend: the family and nation coefficients are both larger by one-third in Cluj-Napoca, and model performance is also greater. Hence, it appears that respondents in Cluj-Napoca were somewhat more likely than their counterparts in South Bend to answer the neighborhood economic item by drawing guidance from their own family- and national-level perceptions. A second difference between the two cities is that observed economic conditions are linked to respondents’ perceptions in South Bend, but not in Cluj-Napoca. In other words, neighborhood-level economic perceptions in Cluj-Napoca have no clear basis in reality. Table 9.4 Accuracy and independence of neighborhood-level economic perceptions Item

Constant Perception of family’s economic well-being Perception of national economic conditions Observed neighborhood economic conditions Adjusted R2 N

South Bend

Cluj-Napoca

Coefficient

t-value

Coefficient

t-value

1,260.39* 1,260.30***

2.10 14.59

640.28 640.40***

0.60 13.24

1,260.24***

12.01

640.33***

9.18

1,260.25**

2.88

640.15

0.49

1,260.36 1,263

640.41 645

Note Cell entries are OLS regression coefficients for models in which the dependent variable is respondents’ perceptions of the economic well-being of families in their neighborhood. For both South Bend and Cluj-Napoca, data for the dependent variable and the first two independent variables were coded 0 (much worse) to 4 (much better). Values for the third independent variable, observed neighborhood scores, were calculated using data for the family in each city, with those data aggregated by neighborhood. *** p  0.001; ** p  0.01; *p  0.05.

154

Jeffery J. Mondak and Adam F. Gearing

Collectively, the results reported in Tables 9.3 and 9.4 strongly question the quality of community-level social perceptions in Cluj-Napoca. Many respondents declined to even guess at economic conditions within their own neighborhoods, whereas other respondents seemingly did little more than guess. The dearth of social interaction in Cluj-Napoca has produced an apparent causality. Our sole remaining question is whether the phenomena uncovered thus far can be linked to tangible signs of atomization in political decision-making. In short, do the residents of Cluj-Napoca exhibit concern with the well-being of others in the community when they construct political judgments, or are they prevented from doing so by their lack of social interaction and the poor quality of their social perceptions?

Political evaluation When civic engagement does not flourish, one tangible consequence for political behavior is that citizens may not even recognize, let alone attend to, the opinions and interests of their neighbors. In a fully atomized society, citizens could acquire information exclusively from impersonal sources such as national broadcast media, and then assess that information purely from the perspective of narrow self-interest. Resulting political evaluations – attitudes toward a regime, a political leader, questions of policy, and so on – would reflect this atomization. Media-informed knowledge of largescale developments (such as the condition of the national economy) might influence political evaluations, but concern with the communal good would not. If our diagnosis of social interaction in Cluj-Napoca has been accurate thus far, then political evaluations offered by our survey respondents may reflect precisely the sort of atomization described here. The political judgment we considered was support for the president – Ion Iliescu in Romania, and Ronald Reagan in the United States. At issue is which of the three economic perceptions analyzed in the previous section affect presidential support. Typically, sociotropic (i.e. nationallevel) perceptions exert strong influence on political judgments, whereas pocketbook (i.e. family-level) perceptions bring weaker effects. Although these characteristic patterns may be replicated in South Bend and ClujNapoca, we are more interested in whether neighborhood-level perceptions also influence presidential support. An effect for the neighborhood variable would not be consistent with a diagnosis of isolated, purely individualistic political behavior. Hence, exploration of a link between the neighborhood-level perception and presidential support constitutes a direct test of whether impediments to social interaction in Cluj-Napoca bring tangible political consequences. Of course, our point is not that the link between neighborhood-level economic perceptions and evaluation of a nation’s president is pivotal for democratization, but instead that the absence of such a link would be of suggestive importance regarding what it means for a polity when social interaction is rare.

Civic engagement and post-communism 155 Our presidential support dependent variables are categorical, ranging from strongly disapprove to strongly approve. In Cluj-Napoca, the variable has five categories because respondents were given the opportunity to offer neutral assessments of Iliescu. Hence approval is coded 0 (strongly disapprove) to 4 (strongly approve) in Romania (mean  1.85, SD  1.19). The South Bend question did not include a neutral option, and thus the four-category measure ranges from 0 (strongly disapprove) to 3 (strongly approve) (mean  1.74, SD  1.23). Ordered logistic regression results are reported in Table 9.5.9 Consistent with past research, we found statistically significant pocketbook and sociotropic effects both in South Bend and Cluj-Napoca. Moreover, the sociotropic effects outpaced the influence of pocketbook perceptions in both cities, albeit only modestly in Cluj-Napoca. In contrast with this similarity, a striking difference was found between the two cities for the neighborhood-level economic perception. South Bend residents exhibit a strong inclination to attend to the economic well-being of their neighbors when evaluating the president. The coefficient for the neighborhood variable is substantively large, with a value midway between the more familiar pocketbook and sociotropic indicators. In stark contrast, however, there is a complete absence of effect for the neighborhood-level economic perception in Cluj-Napoca. In a context that is not conducive to social interaction, we have seen that many residents declined to estimate the economic well-being of their neighbors, and, for those respondents who did report neighborhood-level economic perceptions, those perceptions are both inaccurate and politically inconsequential. The Cluj-Napoca respondents did evaluate their president, but they did so in an individualistic, atomized manner. In short, obstacles to civic engagement restrict the flow of information in a community, and, by doing so, they alter the nature of political judgment. Table 9.5 Economic perceptions and presidential approval Item

Family Nation Neighborhood Model 2 N

South Bend

Cluj-Napoca

Coefficient

t-value

Coefficient

t-value

1,260.33** 1,260.63** 1,260.51** 1,637.15 1,148

4.20 8.45 4.87

0.24* 0.29*
0.00 198.33 604

2.82 3.15 0.04

Notes Cell entries are ordered logistic regression coefficients from a model in which the dependent variable is support for the president, with four ordered categories in South Bend and five ordered categories in Cluj-Napoca. Models included several additional variables: age, sex, education, race (South Bend), ethnicity (Cluj-Napoca), and partisanship. ** p < 0.001, * p < 0.01

156

Jeffery J. Mondak and Adam F. Gearing

Conclusions There has been much speculation both in scholarly and popular circles regarding the possibility that civil society is giving way to increasing atomization. This topic is of particular significance for the nations of Central and Eastern Europe because the fledgling democratic regimes in those countries do not enjoy the stability of their more established Western counterparts. Unfortunately, systematic research conducted by social scientists has not kept pace with the flurry of ad hoc reports warning of the demise of civil society. The analysis presented here is offered as one step toward closing this gap. By examining the social behavior of individuals within their communities, we have been able to gain new insight regarding how civic engagement in post-communist Europe compares with that in the West. The empirical results presented here are quite straightforward. Less clear, however, are the implications of these findings for democratization in Central and Eastern Europe. We have seen that survey data from one Eastern European community differ in a substantial manner from data gathered in a U.S. city. Compared with the residents of South Bend, the people of Cluj-Napoca discuss politics less, engage in interactive forms of political participation at lower rates, know less about their neighbors, and fail to link the interests of people in the community to broader political judgments. For observers who value civic engagement, these findings must be seen as highly disconcerting. A critical question raised by our analysis is why so little civic engagement takes place in Cluj-Napoca. In contrasting Cluj-Napoca with South Bend, we have described five likely impediments to social interaction: a widespread lack of practice in civil discourse, the presence of large numbers of citizens without strong roots in the community, housing patterns that impede social communication, the poor quality of local media, and pervasive ethnic tension. These characteristics distinguish Cluj-Napoca from South Bend, and thus it seems highly plausible that the relative dearth of social interaction in Cluj-Napoca can be traced to one or more of these five factors. Nonetheless, we concede that the case we have outlined rests on circumstantial evidence. With data from only one Eastern European community, it is simply not possible for us to derive evidence that would more strongly corroborate the claim that any or all of these factors account for the disturbing pattern of empirical findings we report. It is essential that progress be made in determining which factors function as the greatest impediments to social interaction (see also Nelson 1996). Without additional evidence on this point, it will be difficult to put findings such as those reported here in perspective. Some of the obstacles we have described can be overcome much more quickly than others. The likely near-term future of civic engagement in cities such as Cluj-Napoca will differ markedly if the fundamental problem is ethnic tension or

Civic engagement and post-communism 157 housing patterns rather than citizens’ lack of practice in social discourse or the poor quality of local media. It is also the case that the potential obstacles to civic engagement do not affect all Central and Eastern European communities equally, and thus there may be considerable variance across communities in their prospects for fostering civic engagement. For example, ethnicity is not an issue at all in some cities, but it is an issue of extraordinary salience in others. Our methodological strategy of focusing on only two cities has enabled us to study civic engagement where it takes place, namely among individuals interacting within their communities. Without such data, for example, we would have been unable to explore the link between respondents’ neighborhood-level economic perceptions and the actual economic climate of each neighborhood. Still, the next step must be to determine what these findings imply for political behavior outside of the communities we have examined. Social interaction in Cluj-Napoca likely lags behind that in many post-communist municipalities, but it may also outpace that in many others. This is a question of vast importance, because democratic legitimacy may be undermined if civic engagement fails to flourish in the transitional regimes of Central and Eastern Europe. Likewise, research must be extended to explore additional possible consequences of low levels of social interaction. It is vital that we gain a new understanding of the nature and significance of civic engagement in post-communist Europe, but to do so will require creative, rigorous analyses based on diverse methodological approaches.

Acknowledgments This research was supported by a United States Information Agency (USIA) Affiliations Grant; by the Social Science Curriculum Development Program at the Faculty of Political Science and Public Administration of Babes-Bolyai University, a program sponsored by the USIA and administered by the International Research & Exchanges Board in partnership with the American Council of Learned Societies; and by a Hewlett International Grant, a John G. Bowman Faculty Grant, and a Central Research Development Fund Grant to the first author.

Appendix Sampling information and survey items The 1984 South Bend Study This project was headed by Robert Huckfeldt and John Sprague (see Huckfeldt and Sprague 1995: ch. 2). The survey included three waves: The a-wave was conducted early in the summer of 1984, the b-wave was

158

Jeffery J. Mondak and Adam F. Gearing

conducted from early September until the November election, and the c-wave was conducted after the election. Interviews were conducted by telephone. Respondents were drawn from sixteen South Bend neighborhoods, with neighborhoods defined by city and county planning departments. The neighborhoods are not merely political divisions, since they are recognized as distinct neighborhoods by their residents. Neighborhoods were selected to maximize social-status variation. About 1,500 interviews were completed in each wave, with 926 respondents participating in all three waves. The 1994 Cluj-Napoca Survey This project was headed by the first author. Interviewers were conducted in November 1994 in person by a team of thirty-seven native Romanian speakers. Interviewers and six field supervisors were political science and sociology students at Babes-Bolyai University. Interviewers and field supervisors underwent two days of training before surveying commenced. Since virtually all Cluj-Napoca residents speak Romanian, no serious problems were encountered concerning language or ethnicity.10 A native Romanian who is a professional survey researcher designed the sampling procedure. The sample was drawn using a random cluster design, and it is representative of the population of Cluj-Napoca. The full sample included 1,621 households and yielded 950 completed interviews. Comparison of survey data with census data indicates that the samples match the city’s population on ethnicity (76 percent ethnic Romania, 22 percent ethnic Hungarian, and 2 percent other). The sampling procedures used in the South Bend and Cluj-Napoca surveys differ in one respect. The South Bend survey was conducted in sixteen distinct neighborhoods, with approximately equal numbers of respondents drawn from each neighborhood. That is, the primary sampling unit was the neighborhood, and thus the data technically are not representative of the population of the South Bend area. In ClujNapoca, neighborhood markers were included to identify each respondent’s neighborhood, but interviews were conducted in all city neighborhoods, the number of interviews varies across neighborhoods as a function of neighborhood size, and the data are representative of the city’s population. Political discussion items South Bend ‘How often do you discuss politics with members of your family – often, once in a while, rarely, or never?’ (b-wave); ‘How often do you discuss politics with people in your neighborhood – would you say you do this often, once in a while, rarely, or never?’ (b-wave); ‘How often do you

Civic engagement and post-communism 159 discuss politics with the other members of your group[s] – often, once in a while, rarely, or never?’ (b-wave); ‘How often do you discuss politics at work – would you say you do this often, once in a while, rarely, or never?’ (b-wave); ‘While you are having some recreational fun, how often do you talk about politics – often, once in a while, rarely, or never?’ (b-wave). Cluj-Napoca ‘How often do you talk about politics with members of your family – would you say every day, most days, once in a while, or never?’; ‘How often do you talk about politics with people here in this neighborhood – every day, most days, once in a while, or never?’; ‘And how often do you talk about politics with people outside of your neighborhood, such as with people at work, at church, etc. – every day, most days, once in a while, or never?’ Political participation and media exposure items South Bend ‘Some people are more involved in politics than others, and we would like to find out about your involvement. I’m going to read you a list of things that some people do in politics. Could you tell me whether you do these things regularly, sometimes, rarely, or never? Work for candidates in elections? Work for a party? Try to convince friends to vote your way? Attend political meetings, rallies, dinners, or things like that? Listen to political reports on radio or TV?’ (a-wave; this battery includes items concerning several other actions, none of which was asked on the Cluj survey); ‘How often do you read the South Bend Tribune – would you say every day, most days, every now and then, seldom, or never?’ (a-wave); ‘How often do you see some other daily paper – every day, most days, every now and then, seldom, or never?’ (c-wave). Cluj-Napoca ‘People participate in politics in various ways. Please tell me if you have done each of the following: Work for a political party or candidate? Try to persuade another person to vote for a particular political party or candidate? Attend a political meeting? Attend a protest rally or demonstration?’; ‘How often do you watch the nightly news on television – every day, most days, once in a while, or never?’; ‘How often do you read one of the local papers published here in Cluj – every day, most days, once in a while, or never?’; ‘How often do you read one of the papers published in Bucharest – would you say every day, most days, once in a while, or never?’

160

Jeffery J. Mondak and Adam F. Gearing

Retrospective economic perceptions items South Bend ‘In general, would you say that you and your family are better off, worse off, or about the same financially compared with a year ago?’ (South Bend items are from the a-wave; respondents who answered better off or worse off were asked a follow-up question, ‘Much or somewhat better/worse?’); ‘How about your neighborhood, would you say that most families in your neighborhood are better off, worse off, or about the same financially compared with a year ago?’; ‘Now let’s talk about the country as a whole. Would you say that most families in the country are better off, worse off, or about the same financially compared with a year ago?’ Cluj-Napoca ‘Think about the changes in Romania in the past five years. In general, would you say that you and your family are much better off, somewhat better off, somewhat worse, or much worse financially compared with five years ago, or would you say that things are about the same?’; ‘How about your neighborhood? Would you say that most families in your neighborhood are much better off, somewhat better, somewhat worse, or much worse financially compared with five years ago, or about the same?’; ‘Now let’s talk about the country as a whole. Would you say that most families in the country are much better off, somewhat better, somewhat worse, or much worse financially compared with five years ago, or about the same?’ (The five-year period corresponds with the nation’s post-communist era.) Presidential approval items South Bend ‘Do you approve or disapprove of the way Ronald Reagan is handling his job as president?’ followed up with ‘Do you approve/disapprove strongly or not strongly?’ (a-wave). Cluj-Napoca ‘For each of the following, tell me if you strongly approve, approve, disapprove, or strongly disapprove: The job Ion Iliescu is doing as president.’ (This battery includes additional items concerning several other national and local political actors and institutions.)

Civic engagement and post-communism 161

Notes 1 The Cluj-Napoca survey includes sixty-nine items concerning social and political attitudes and behavior. A random cluster design was used to draw 950 respondents from within the city of Cluj-Napoca. Markers were included to indicate the neighborhood in which each respondent resided. Interviews were conducted in person by a team of thirty-seven native Romanian speakers. 2 The South Bend Study was a three-wave panel survey conducted in 1984. The South Bend Study used a purposive sampling procedure; data were drawn from respondents in sixteen specific neighborhoods. 3 The use of data from the South Bend Study as a benchmark for assessment of political behavior in Cluj-Napoca brings considerable efficiency in that only one new survey, not two, was needed. However, our reliance on an extant survey for benchmark data limits comparability versus what we could have achieved had we ourselves administered both surveys. Although the specific items we analyze here are comparable, the two surveys differ in many other respects, and these differences may have unknown influences on our results. That said, we have no reason to believe that our conclusions systematically misrepresent the nature of political behavior in the two cities. We believe that similarities and differences between South Bend and Cluj-Napoca in patterns of civic engagement, social perception, and political evaluation are genuine. 4 Even the name was new: Cluj was changed to Cluj-Napoca as a way to emphasize the city’s Romanian roots. 5 In South Bend, this category includes data from items concerning social groups, the workplace, and recreational activities. A single item was used in Cluj-Napoca. 6 In both South Bend and Cluj-Napoca, a great deal of civic interaction takes place outside of the neighborhood; as Table 9.1 shows, levels of discussion in both cities are higher outside than inside the neighborhood. We recognize that neighborhoods are only one context for social interaction, and not necessarily the most important (the workplace, the church, the social group, and so on, may be of greater significance). Nonetheless, the neighborhood is a useful context for the present study because we can aggregate survey data by neighborhood to obtain contextual-level indicators. In future research, we hope to obtain similar data for other contextual units. 7 It is possible that residents of Cluj-Napoca discuss politics quite actively but were hesitant to admit such interaction to our survey interviewers. However, we see this as unlikely. Respondents were not shy about voicing their opinions on the survey as a whole. In response to other questions, the vast majority of respondents expressed an opinion on the state of Romania’s economy, and most of these assessments were critical. Likewise, nearly 70 percent of respondents reported that they disapproved of the nation’s parliament. Given such results, we believe that the data given in Table 9.1 reveal a true absence of political discussion, not a reluctance to admit to the existence of such conversations. 8 We have at least fifty cases per neighborhood. For additional discussion of the use of aggregate data in multilevel analysis, see Iversen (1991) and Huckfeldt and Sprague (1995). 9 We use ordered logistic regression because presidential approval is measured using ordinal scales. We could simply assume the ordered categories on our approval measures to be spaced at equal distances from one another, and then estimate approval using OLS regression. However, the space between any two categories on our variables is not equal. In South Bend, the move from disapproval (scale value  1) to approval (2) represents a shift of one unit on the scale; in Cluj-Napoca the same move brings a shift of two units because it

162

Jeffery J. Mondak and Adam F. Gearing

entails an advance from disapproval (1) to neutral (2), and then to approval (3). Hence it is inappropriate to treat our dependent variables as interval-level scales, and we must use a nonlinear estimation procedure such as ordered logit. For additional discussion of ordered (or cumulative) logit, see Agresti (1990), Greene (1990), Demaris (1992), and Liao (1994). 10 Since most interviewers were ethnic Romanians, a risk existed that ethnic Hungarians might decline to participate in the survey, and that those Hungarians who did participate might be guarded in answering politically sensitive items. Fortunately these problems were not encountered. Refusal rates were no higher in areas with larger concentrations than elsewhere, and the proportion of Hungarian respondents matches precisely the census data for Cluj-Napoca. Moreover, Hungarian respondents were uniformly vehement in their disdain for Cluj-Napoca’s controversial mayor, Gheorghe Funar, and the ultranationalist party he heads; they would not have expressed themselves so openly if they had been suspicious of the survey or the interviewers.

Bibliography Agresti, A. (1990) Categorical Data Analysis. New York: Wiley. Anonymous (1995) ‘Romania: Getting Nastier’, The Economist. 4 February. Atkeson, L. R. and Partin, R. W. (1995) ‘Economic and referendum voting: a comparison of gubernatorial and senatorial elections’, American Political Science Review 89: 99–107. Bachman, R. D. (1991) Romania: A Country Study. Washington, DC: Federal Research Division, Library of Congress. Barber, B. (1984) Strong Democracy: Participatory Politics for a New Age. Berkeley, CA: University of California Press. Boc, E. and Mondak, J. J. (2001) ‘Characteristics of the Romanian press’, Current Politics and Economics of Russian, Eastern and Central Europe 16: 109–23. Brehm, I. and Rahn, W. M. (1997) ‘Individual-level evidence for the causes and consequences of social capital’, American Journal of Political Science 41: 999–1023. Brown, I. F. (1994) ‘Extremism in Eastern Europe and the former Soviet Union’. RFEIRL Research Report 3 (16): 1–4. Conover, P. I. (1985) ‘The impact of group economic interests on political evaluations’, American Politics Quarterly 13: 139–66. Demaris, A. (1992) Logit Modeling: Practical Applications (Sage University Paper series on Quantitative Applications in the Social Sciences, 07-086), Thousand Oaks, CA: Sage. Ekiert, G. (1991) ‘Democratization processes in Eastern Europe: a theoretical reconsideration’, British Journal of Political Science 21: 285–313. Eulau, H. and Lewis-Beck, M. S. (eds) (1985) Economic Conditions and Electoral Outcomes: The United States and Western Europe. New York: Agathon. Evans, G. and Whitefield, S. (1995) ‘The politics and economics of democratic commitment: support for democracy in transition societies’, British Journal of Political Science 25: 485–514. Gallagher, T. (1993) ‘Ethnic tension in Cluj’. RFEIRL Research Report 2 (9): 27–33. Glenny, M. (1990) The Rebirth of History: Eastern Europe in the Age of Democracy. New York: Penguin.

Civic engagement and post-communism 163 Greene, W. H. (1990) Econometric Analysis. New York: Macmillan. Gross, P. (1996) Mass Media in Revolution and National Development: The Romanian Laboratory. Ames, IA: Iowa State University Press. Huckfeldt, R. and Sprague, J. (1995) Citizens, Politics and Social Communication: Information and influence in an election campaign. New York: Cambridge University Press. Iversen, G. R. (1991) Contextual Analysis (Sage University Paper series on Quantitative Applications in the Social Sciences, 07-081), Thousand Oaks, CA: Sage. Kiewiet, D. R. (1983) Macroeconomics and Micro Politics. Chicago, IL: University of Chicago Press. Kinder, D. R. and Kiewiet, D. R. (1979) ‘Economic discontent and political behavior: the role of personal grievances and collective economic judgments in congressional voting’, American Journal of Political Science 23: 495–527. Lake, R. L. and Huckfeldt, R. (1998) ‘Social capital, social networks, and political participation’, Political Psychology 19: 567–84. Lewis-Beck, M. S. (1988) Economics and Elections: The major Western Democracies. Ann Arbor, MI: University of Michigan Press. Liao, T. F. (1994) Interpreting Probability Models: Logit, Probit, and Other Generalized Linear Models (Sage University Paper series on Quantitative Applications in the Social Sciences, 07-101), Thousand Oaks, CA: Sage. Lyon, C., Grosskopf, A. and Mondak, J. J. (1997) ‘Mass support for European integration: extending the pocketbook–sociotropic dichotomy?’ Paper presented at the annual meeting of the Midwest Political Science Association, Chicago. Mishler, W. and Rose, R. (1996) ‘Trajectories of fear and hope: support for democracy in post-communist Europe’, Comparative Political Studies 28: 553–81. Mondak. J. J. (1995a) ‘Media exposure and political discussion in U.S. elections’, Journal of Politics 57: 62–85. Mondak, J. J. (1995b) Nothing to Read: Newspapers and Elections in a Social Experiment. Ann Arbor, MI: University of Michigan Press. Mondak, J. J. (in press) ‘Media and tolerance: empirical evidence of mixed effects’, in D. L. Paletz and K. Jakubowicz (eds) Relapse: Media and Government in Central and Eastern Europe. Cresskill, NJ: Hampton. Mondak, J. J. and Mutz, D. C. (1997) ‘What’s so great about league bowling?’ Paper presented at the annual meeting of the Midwest Political Science Association, Chicago. Mondak, J. J., Mutz, D. C. and Huckfeldt, R. (1996) ‘Persuasion in context: the multilevel structure of economic evaluations’, in D. C. Mutz, P. M. Sniderman, and R. A. Brody (eds) Political Persuasion and Attitude Change. Ann Arbor, MI: University of Michigan Press, pp. 249–66. Mutz, D. C. and Mondak, J. J. (1997) ‘Dimensions of sociotropic behavior: groupbased judgments of fairness and well-being’, American Journal of Political Science 41: 284–308. Nelson, D. N. (1990) ‘Romania’, Electoral Studies 9: 355–66. Nelson. D. N. (1996) ‘Civil society endangered’, Social Research 63: 345–68. Paletz, D. L., Jakubowicz, K. and Novosel, P. (eds) (1995) Glasnost and After: Media and Change in Central and Eastern Europe. Cresskill, NJ: Hampton. Pattie, C. and Johnston, R. (1995) ‘It’s not like that “round here”: region, economic evaluations, and voting at the 1992 British general election’, European Journal of Political Research 28: 1–32.

164

Jeffery J. Mondak and Adam F. Gearing

Przeworski, A. (1991) Democracy and the Market: Political and Economic Reform in Eastern Europe and Latin America. Cambridge: Cambridge University Press. Putnam, R. D. (1993) Making Democracy Work: Civic Traditions in Modern Italy. Princeton, NJ: Princeton University Press. Putnam, R. D. (1995a) ‘Bowling alone: America’s declining social capital’, Journal of Democracy 6: 65–78. Putnam, R. D. (1995b) ‘Tuning in, tuning out: the strange disappearance of social capital in America’, PS: Political Science and Politics 28: 664–83. Rose, R. and Mishler, W. (1994) ‘Mass reaction to regime change in Eastern Europe: polarization of leaders and laggards?’, British Journal of Political Science 24: 159–82. Schudson, M. (1996) ‘What if civic life didn’t die?’, The American Prospect 25: 17–28. Shafir, M. (1993) ‘The HDFR Congress: Confrontations postponed’. In RFEIRL Research Report 2(9): 34–39. Stein, R. M. (1990) ‘Economic voting for governor and U.S. senator: the electoral consequences of federalism’, Journal of Politics 52: 29–53. Szajkowski, B. (1993) Encyclopedia of Conflict, Disputes, and Flashpoints in Eastern Europe, Russia and the Successor States. Harlow: Longman. Terry. S. M. (1993) ‘Thinking about post-communist transitions: how different are they?’, Slavic Review 52: 333–37. Warren, R. B. and Warren, D. I. (1977) The Neighborhood Organizer’s Handbook. Notre Dame, IN: University of Notre Dame Press. Weatherford, M. S. (1978) ‘Economic conditions and electoral outcomes: Class differences in the political response to recession’, American Journal of Political Science 22: 917–38. Weatherford, M. S. (1982) ‘Recessions and social classes: economic impacts and political opinions’, Political Behavior 4: 7–32. Williams, M. R. (1985) Neighborhood organizations: Seeds of a new urban life. Westport, CT: Greenwood.

10 Why post-communist citizens do not join voluntary organizations Marc Morjé Howard

Introduction The events in Eastern Europe in the late 1980s contributed greatly to the revival of the concept of civil society in comparative politics. Yet, ironically, when compared to other regions of the world today, civil society remains extremely weak throughout post-communist Europe (Smolar 1996; Lomax 1997; Padgett 2000). Moreover, analysis of the 1995 to 1997 World Values Survey (WVS) results in thirty-one democratic and democratizing countries shows that even when controlling for levels of economic well-being, political rights and civil liberties, and ‘civilization,’ a country’s prior communist experience is the most important factor in explaining variation in levels of civil society participation across countries (Howard 2003). Although significant within a cross-national perspective, the establishment of the importance of a prior communist experience still does not tell us very much about why this is the case. What is it about the communist experience that makes its citizens, almost ten years later, much less likely to join organizations than citizens of countries from comparable economic backgrounds, living in democratic systems with comparable political rights and civil liberties, and even for a comparable number of years? This puzzle of non-participation is all the more surprising since many analysts claim that the actual number of voluntary organizations in post-communist Europe has been increasing considerably over the past decade (Ekiert and Kubik 1999). This chapter seeks to provide a causal argument that accounts for the particularly low levels of organizational membership in post-communist Europe, by applying quantitative analysis of the Post-Communist Organizational Membership Study (PCOMS) survey, a specially designed representative survey conducted in 1999 in Germany and Russia. Although it develops an argument that applies to much of post-communist Europe, this chapter focuses primarily on Russia and Eastern Germany, which may be viewed as the two ‘most different’ cases within contemporary post-communist Europe (Przeworski and Teune 1973). In other words,

166

Marc Morjé Howard

these two cases differ greatly on most factors that would constitute alternative explanations for variation in levels of participation in voluntary organizations: economic levels are very high in Eastern Germany, but extremely low in Russia; political and legal institutions are very well developed in Eastern Germany, while the Russian political and legal systems are still very weak and undeveloped; and both countries belong to different cultural and religious ‘civilizations.’ The high variation between these two cases on these alternative explanatory factors provides maximal analytic leverage to explain the causes of their similarly weak civil societies. Moreover, the inclusion of Western Germany for several of the survey questions provides an optimal test case to evaluate the extent to which, in terms of participation in civil society, East Germans are closer to their new ‘compatriots’ or to their former ‘comrades.’ I start with a brief discussion of the definition of civil society, and how the concept can be applied and operationalized empirically. I then summarize the results of the 1995 to 1997 WVS, which show the consistent weakness of civil society across post-communist Europe. The next two sections of the chapter incorporate individual-level statistical analysis of the PCOMS survey, focusing on general socio-economic factors as well as three ‘experiential’ factors that are specific to post-communist citizens – a legacy of mistrust of formal organizations, the persistence of private friendship networks, and a deep disappointment with post-communism. I conclude by discussing the theoretical implications of the ways in which post-communist citizens link their current behavior to their prior experiences, and I mention the empirical consequences of this lasting weakness of civil society on the future of democracy in post-communist Europe.

Defining and operationalizing civil society Broadly speaking, I view civil society as the community of citizens who come together and associate within the public ‘space’ that is distinct from the individual, family, and friendship networks, on the one hand, and the state and market, on the other.1 This space consists of intermediary groups, organizations, and associations that are formally established, legally protected, autonomously run, and voluntarily joined by ordinary citizens. Civil society is deemed by many scholars to be an important element both of democratizing countries and established democracies. Although skeptics have rightly objected to exaggerated claims about the beneficial effects of civil society, and to the way in which it is sometimes treated as if it were a panacea, or the only or the most important solution to a country’s problems (Berman 1997; Edwards and Foley 1998), few would dispute the more modest claim that civil society is one of several factors that are important for establishing and sustaining a vibrant democratic system.

Post-communist voluntary organizations

167

The scholarship on the relationship between civil society and democracy may be broken down into two general approaches. The first, most closely associated with Robert Putnam’s work, emphasizes the positive effects participation in civil society has on individual people, which in turn contributes to a more peaceful, harmonious, and democratic society (Putnam 1993, 2000; Verba et al. 1995). The second approach is generally associated with the work of historical institutionalists such as Theda Skocpol, who argue that the organizations of civil society provide a direct source of popular influence on political or economic developments, thus benefiting individuals and society (Skocpol 1999). Of course, these two types of argument are not mutually exclusive, and proponents of one often recognize the importance of the other, but the difference in emphasis – either on the value of participation or on the power of institutional leverage – is analytically and substantively significant. Overall, scholars who study civil society view it as an important attribute of, and factor in, democratic performance and the quality of democracy. While there is no perfect measure of civil society across different countries – and any attempt to capture such a rich concept across many different societies will obviously have to sacrifice some of its complexity – representative surveys provide the best starting point for such research. The alternative approach, which counts the number and type of registered organizations, is often based on inconsistent sampling methods that are not conducive to cross-national comparison, and it usually fails to take into account that many organizations ‘die’ or cease to exist over time.2 In the post-communist context this point is described vividly by Miszlivetz and Jensen (1998), who refer to the ‘emerging paradox’ of a ‘statistically strong civil society,’ but one that is primarily organized ‘from above,’ with only limited participation by ordinary citizens. In other words, although many organizations may exist on paper (or even on the internet), the ‘pseudo existence’ of many of them does not necessarily indicate corresponding levels of and increases in participation of ordinary citizens. In contrast, by measuring the percentage of respondents who are actually members of certain voluntary associations using the identical question replicated across countries, surveys not only give a better approximation of the development of a country’s civil society than a potentially hollow list of registered organizations, but they also allow for extensive comparisons, both with other countries and between different social strata within a country.

The weakness of civil society in post-communist Europe Before turning to the two case studies of East Germany and Russia, and to the specific factors that explain the particular weakness of post-communist civil society, this section summarizes the cross-national results from the 1995 to 1997 WVS and the 1999 PCOMS surveys, briefly demonstrating

168

Marc Morjé Howard

the relatively low, and perhaps even declining, levels of membership in voluntary organizations. The question on membership in voluntary organizations in the WVS questionnaire (which was also replicated in the PCOMS survey) specifies nine different types of groups: (1) church or religious organizations, (2) sports or recreational clubs, (3) educational, cultural, or artistic organizations, (4) labor unions, (5) political parties or movements, (6) environmental organizations, (7) professional associations, (8) charitable organizations, and (9) any other voluntary organization. While this list is by no means comprehensive or exhaustive – and one could certainly argue that other types of organizations, such as women’s groups, student organizations, veterans’ associations, animal rights groups, or many others should have been included as well – it does capture a wide enough range of organizations, both traditional and contemporary, that are central to civil society so as to allow us to investigate the comparative levels of participation in them across countries. Moreover, the final category of ‘other’ organizations should capture, albeit less explicitly and directly, the remaining types of organizations that were not included in the question list. Figure 10.1 shows the average number of organizational memberships per person for each of the thirty-one democratic and democratizing countries included in the WVS. The countries are grouped into three categories based on their prior regime type:3 (1) the ‘older democracies’ category refers to countries that have been democratic since at least the early postSecond World War period; (2) the ‘post-authoritarian’ category consists of countries that recently experienced a period of (non-communist) nondemocratic rule; and (3) the ‘post-communist’ category consists of most of Number of organizational memberships per person

4 3.59

3.5 3.07

3 2.68

2.62 2.48 2.47

2.5

2.44 2.22

2.32 2.13

2.12

2

1.84 1.53

1.5

1.50 1.44

1.39 1.39

0.92

1

1.30 1.14 1.12 1.07

1.10 1.03

0.82 0.70 0.65 0.64

0.60 0.46

0.5

0.35

Older democracies mean2.39

Bulgaria

Lithuania

Estonia

Ukraine

Latvia

Russia

Hungary

Slovakia

Czech Rep.

Slovenia

Romania

Macedonia

E, Germany

Argentina

Post-authoritarian mean1.82

Philippines

Spain

Uruguay

Venezuela

Bangladesh

Chile

Brazil

S. Africa

S. Korea

Japan

Switzerland

W. Germany

Finland

Norway

Sweden

USA

Australia

0

Post-communist mean0.91

Figure 10.1 Average number of organizational memberships per person, by country Source: World Values Survey 1995–97.

Post-communist voluntary organizations

169

the countries in the region of post-communist Europe.4 Out of a maximum total of nine organizations, the highest score is in the United States, where the average person is a member of 3.59 organizations, ranging down to Bulgaria’s score of 0.35. Overall, this figure shows that membership levels are related closely to the country groupings based on prior regime type. Although there are certainly differences among the post-communist countries when placed in a larger comparative perspective, they are best viewed as ‘differences in degree,’ not ‘differences in kind.’5 The PCOMS survey, conducted in 1999, replicates the WVS question on organizational membership in Russia, East Germany, and West Germany. When compared with the WVS scores, the overall results of the PCOMS survey indicate a continued decline in organizational membership in the two post-communist societies, compared to only a very slight decrease in West Germany. Table 10.1 presents these changes, showing that in Russia the levels of participation dropped from an already very low level of 0.66 organizational memberships per person in 1995 to 1997 to 0.45 in 1999. In East Germany, average membership appears to have dropped quite sharply in that short time span, as the average East German citizen belonged to 1.44 organizations in 1995–97, but only 0.78 in 1999.6 Finally, in West Germany there appears to have been a small decline, from 2.12 to 2.00 memberships per person. This finding of low, and perhaps even declining, post-communist organizational membership presents a puzzle that needs to be explained. The following two sections do so by incorporating statistical analysis of the PCOMS survey, first by focusing on the main socio-economic variables in Russia, East Germany, and West Germany, and then by incorporating specifically post-communist factors within the Russian and East German samples. Table 10.1 Changes in levels of membership, 1995–97 to 1999: Russia, East Germany, and West Germany Russia

■

East Germany

■

West Germany

1995– 1999 Change 1995– 1999 Change 1995– 1999 Change 97 97 97 Average 0.65 number of organizations of which each person is a member

0.45
0.20

1.44

Source: WVS (1995–97) and PCOMS (1999).

0.78
0.66

2.12

2.00
0.12

170

Marc Morjé Howard

Socio-economic factors in West Germany, East Germany, and Russia This section incorporates the results from the 1999 PCOMS survey of West Germany, East Germany, and Russia, and tests a set of applicable individual-level hypotheses, using the average number of organizational memberships per person as the dependent variable. The goal is to estimate the relative significance of several different independent variables, both within each society separately, as well as in the pooled dataset. In this section, I focus especially on socio-economic status (SES), which has been identified as a strong predictor of participation in numerous studies.7 The literature on individual attributes and participation in civil society is already well established, as scholars have developed and tested a series of hypotheses about what types of people are more or less likely to participate. Many studies have shown that people with greater economic resources and those who have higher levels of education are more likely to participate in voluntary organizations.8 A third element of SES is age, and the general expectation is that, since participation in organizations is a habit and skill that develops over the course of an individual’s lifespan, older people will participate more than younger people. A fourth hypothesis considers gender, and the differences between men and women. Although there is no consistent agreement across studies as to whether men or women participate more, Verba and colleagues (1995: 256) found that ‘Contrary to stereotype, women are slightly less likely than men to be affiliated with any organization, political or non-political’. In addition to these important socio-economic factors, Eric Oliver (2000) has shown that city size has a strong effect on several different types of participation, and that people who reside in larger cities are less likely to participate than those who live in smaller places. These basic hypotheses allow us to test the main parts of the SES model, along with the variable for city size.9 In addition to these, I also test a sixth hypothesis, which incorporates the concept of an individual’s prior communist experience. The expectation is that people who lived under communist rule are less likely to participate in voluntary organizations. This hypothesis allows us to measure the causal effect of the communist experience, holding socio-economic factors constant. I operationalize the prior communist experience within the pooled sample in two ways. First, I substitute a dummy variable for prior communist experience, with Russians and East Germans coded 1, and West Germans coded 0. Second, I include a dummy variable for East Germans and a dummy variable for Russians in order to compare the direction and strength of the coefficients in both countries, and to ensure that the same causal mechanism is taking place in each. When I analyze the countries separately, however, I cannot include these variables, since they are constant for all respondents in each country.

Post-communist voluntary organizations

171

Tables 10.2 and 10.3 present multivariate regression results. In Table 10.2 which includes the pooled respondents from all three country samples, Model A tests the SES and city size variables together, showing that, as predicted, income, education, and age have a positive effect on organizational membership, men participate more than women, and city size has a negative effect. Of these five variables, income, age, gender, and city size are all statistically significant at p  0.001, while education is significant only at p  0.05. Despite these significant coefficients – which should not be surprising given the very large sample size – Model A explains only 4 percent of the variability in this pooled sample. Model B adds the individual-level variable for prior communist experience. In this model, with the exception of age, all of the variables are statistically significant, most at the 0.001 level. When looking at the coefficients, however, it becomes clear that most of the explanatory power Table 10.2 OLS regression analysis of individual-level membership in voluntary organizations: Russia, East Germany, and West Germany, combined Variable

A

B

C

Income

0.14*** (0.02)

0.12*** (0.02)

0.10*** (0.02)

Education

0.06* (0.03)

0.12*** (0.03)

0.13*** (0.03)

Age

0.006*** (0.002)

0.002# (0.001)

Gender (male)

0.25*** (0.05)

0.22*** (0.04)


0.07*** (0.02)


0.08*** (0.02)

City size


1.44*** (0.04)


0.0006 (0.001) 0.21*** (0.04)
0.05** (0.02)

Prior communist experience

–

East German dummy

–

–


1.26*** (0.05)

Russian dummy

–

–


1.61*** (0.05)

Constant Adjusted R2 SEE N

0.27* (0.13) 0.04 1.18 2,644

1.44*** (0.11) 0.34 0.98 2,644

–

1.49*** (0.11) 0.35 0.97 2,644

Notes Table entries are unstandardized regression coefficients, with standard errors in parentheses. The dependent variable is the average number of organizational memberships per person. Data come from PCOMS (1999). *** p  0.001; ** p  0.01; * p  0.05; # p  0.10 (all one-tailed).

172

Marc Morjé Howard

Table 10.3 OLS regression analysis of individual-level membership in voluntary organizations: Russia, East Germany, and West Germany, separately Variable

Russia

East Germany

West Germany

Income

0.05** (0.02)

0.09*** (0.03)

0.15*** (0.03)

Education

0.06* (0.03)

0.15** (0.05)

0.21** (0.07)


0.003* (0.002)

0.002 (0.003)

0.0007 (0.003)

Gender (male)

0.04 (0.05)

0.18** (0.07)

0.44*** (0.09)

City size

0.02 (0.02)

0.004 (0.04)


0.15*** (0.04)

Constant

0.24* (0.12) 0.03 0.72 1,009

0.007 (0.20) 0.05 0.96 842

1.24*** (0.26) 0.10 1.21 793

Age

Adjusted R2 SEE N

Notes Table entries are unstandardized regression coefficients, with standard errors in parentheses. The dependent variable is the average number of organizational memberships per person in each country. Data come from PCOMS (1999). *** p  0.001; ** p  0.01; * p  0.05 (all one-tailed).

comes from the prior communist experience variable, which is associated with a very large decrease of 1.44 in the predicted organizational membership. Moreover, Model B explains 34 percent of the variability in the sample, a major improvement over the 4 percent explained by Model A. Model C substitutes East German and Russian dummy variables for prior communist experience, and the results show that the two variables have similarly strong negative effects, both highly statistically significant (p  0.001), and with similarly low standard errors. The effect of being East German is associated with a decrease in 1.26 in the predicted organizational membership, while being Russian is associated with a slightly larger decrease of 1.61. The fact that these coefficients are so similar suggests a strong commonality between East Germans and Russians, at least in terms of organizational membership. Table 10.3 presents a multivariate analysis of the three country samples, showing the most fully specified model for each country separately. The results show that income and education have positive and statistically significant effects in all three countries. Age is only significant in Russia, showing a negative effect on participation that is significant at the 0.05 level, which is surprising given that the effect of age is positive both in East and West Germany. The effect of gender is positive in all three coun-

Post-communist voluntary organizations

173

tries, showing that men participate more than women, and it is statistically significant in East Germany (p  0.01) and in West Germany (p  0.001), but not in Russia. Finally, city size is only statistically significant in the West German sample, where it has a strong negative effect (p  0.001), unlike the slightly positive, but not statistically significant, effects of city size in the Russian and East German samples. Overall, the fully specified model of the SES variables and city size combine to explain 3 percent of the variability in the Russian sample, 5 percent in the East German sample, and 10 percent in the West German sample. Given that in the pooled sample from Table 10.2 prior communist experience is far more powerful than any of the other independent variables, this suggests that there may be other specifically post-communist factors that are missed by the SES model. While important in itself, the establishment of the prior communist experience’s strong negative effect on participation in the organizations of civil society does not tell us very much about why this is the case. We still need to understand much more about the nature of that communist experience, and why its legacy seems to have such a discouraging effect on participation in the organizations of civil society today.

Post-communist factors in Russia and East Germany This section focuses on Russia and East Germany in order to achieve a better understanding of the prior communist experience, and to specify why its negative effect on organizational membership is so powerful. I start by presenting several hypotheses, and I then turn to the regression results for the pooled sample of Russian and East German respondents, as well as for each society separately. Overall, the analysis shows that three main ‘experiential’ factors – people’s prior experiences with organizations in the communist period, the persistence of private friendship networks, and the disappointment with recent post-communist developments – have both substantive and statistical significance for explaining their current behavior and organizational membership today. Mistrust of communist organizations The first factor connects people’s mistrust of organizations in communist times to their current participation. The assumption of the argument is that the collapse of communism did not create a tabula rasa by erasing people’s prior experiences, but rather that those very experiences influence people’s current behavior. This is not to suggest that the communist organizations were analytically equivalent to civil society organizations; they were neither voluntary nor autonomous, and participation in them was often forced or coerced. Whether people participated willingly or not – which, of course, depended on the type of organization, as well as each

174

Marc Morjé Howard

person’s individual situation and choices – they did have extensive experience with these organizations, which were omnipresent in communist societies. This hypothesis suggests that the more people mistrusted communist organizations, and the more they avoided them during the communist time period, the less likely they will be to participate in voluntary organizations today. The logic of this argument goes against a common view, articulated optimistically by many in the early post-communist period, that those people who avoided or opposed communist organizations should be eager to participate now that the communist system no longer exists, and that people who were active in the communist system should be more resigned and passive today. To measure this hypothesis, the PCOMS survey included questions about the same nine categories of organizations from the World Values Survey, only this time referring to membership in state-run communist organizations. Table 10.4 presents the level of mistrust of communist organizations by both Russians and East Germans – measured by the extent to which they avoided communist organizations in the communist time period10 – as well as their mean organizational membership today. The results show great similarities between the two countries, in that increasing mistrust of communist organizations seems to be associated strongly with decreasing levels of organizational membership today. The persistence of friendship networks The second argument I incorporate into this analysis involves the private networks that developed under the communist regime. These networks arose because of the shortage economy, where people needed connections and help to acquire many goods and services, and because the public domain was so controlled and politicized that people formed close and trusting ties to friends and family (Ledeneva 1998; Völker and Flap 1999). This hypothesis predicts that those people who have maintained their Table 10.4 Mistrust of communist organizations and organizational membership today Mistrust of communist organizations

% of responses

Average number of organizational memberships per person

Low Medium High Valid N

01,13.9 01,14.7 01,71.5 1,009

0.99 0.55 0.32 –

Source: PCOMS (1999).

Russia

East Germany

19.6 24.2 56.2 941

Russia

East Germany

1.39 0.87 0.53 –

Post-communist voluntary organizations

175

close ties and vibrant friendship circles will feel less of a need or desire to participate in voluntary organizations. In order to operationalize friendship networks and to measure change in networks over time, the PCOMS survey included questions, referring (at different parts in the questionnaire) to both the communist period and then to the contemporary period, asking how often people used to rely on, or currently rely on, friends for help fixing their house or car, assisting someone in their family, or in difficult situations in general. I then constructed two indices of reliance on friends, one for the communist period and one for the contemporary period. Finally, in order to measure the persistence of, or change in, these networks, I subtracted the current index from the communist index, thus yielding three basic categories: less reliance, more reliance, and no change. Table 10.5 shows the aggregate results, with the last two categories grouped together under the label ‘the same or more’ reliance on friends, since the important cleavage is between the respondents whose use of personal networks has decreased, and those for whom it has not. The results show interesting differences between Russians and East Germans. For Russians, there have been fewer changes in the frequency with which people rely on friends for different kinds of help. This is not surprising, given that, although almost all goods and services can now be purchased in Russia, for most people prices are prohibitively expensive, and they still have to rely on friends for assistance. In East Germany, however, the advantage of living in a stable market economy and a more free and safe society is that most people now have enough economic means and peace of mind to get things done through official channels, rather than relying on friends as much as they used to (Völker and Flap 1996). Turning to the relationship between personal networks and organizational membership, Table 10.5 shows that for Russians, levels of participation in voluntary organizations are virtually identical, whether or not they currently rely on friends as much as they used to. For East Germans, however, those people whose personal networks have been disrupted seem Table 10.5 Changes in reliance on friends from the communist period to today, and organizational membership People who now rely on friends

Less The same or more Valid N Source: PCOMS (1999).

Russia

East Germany

Russia

East Germany

% of responses

Average number of organizational memberships per person

1,040.6 1,059.4 1,009

0.44 0.45 –

58.2 41.8 942

0.84 0.70 –

176

Marc Morjé Howard

to participate in voluntary organizations today at higher levels than those people whose friendship networks have stayed the same or grown closer. In other words, it does appear that to the extent that they have not changed, the vibrant personal networks which developed under communism remain an impediment or an alternative to organizational membership today. Post-communist disappointment The third and final factor I consider here has to do with the basic concept of disappointment, which has become a common theme in many postcommunist countries, as many people’s high expectations about the future have not been fulfilled (Geremek 1996; Rupnik 1996). Although one could conceivably argue that disappointed people would become more active in protesting against the current system, the general expectation of this hypothesis is that people who are disappointed are less likely to participate in voluntary organizations, and that their disenchantment or frustration will be associated with increased passivity and withdrawal. The PCOMS survey measured this concept with a question about people’s evaluation of the changes of the past decade, compared to their initial expectations. Table 10.6 presents the results, showing that disappointment in Russia is extraordinarily high, much higher than in East Germany. Over 80 percent of Russians claimed that the changes have been worse than they expected, compared to only 25 percent of East Germans. In both countries, however, the effect of disappointment on participation in voluntary organizations is negative, as the average number of organizational memberships declines with increasing disappointment. Multivariate regression analysis Having specified these three main explanatory factors – the mistrust of communist organizations, the persistence of friendship networks, and the Table 10.6 Disappointment and organizational membership ‘Compared to what you expected . . . the changes in your life over the past decade have been . . .’: Better than expected About what expected Worse than expected Valid N Source: PCOMS (1999).

Russia

East Germany

Russia

East Germany

% of responses

Average number of organizational memberships per person

1,010.1 1,0 9.5 1,080.4 1,005

0.63 0.53 0.41 –

43.3 31.1 25.6 968

0.86 0.76 0.67 –

Post-communist voluntary organizations

177

feeling of disappointment – that are particularly applicable to postcommunist countries, we can now put them to the test of multivariate regression analysis in order to see how they improve on the relatively low explanatory power of the SES model, as shown in the previous section. Table 10.7 presents the regression results for the pooled sample of Russian and East German respondents, alongside the two samples taken separately. For each of the three sets of regressions, the first model shows the five variables from the SES model and city size, the second model shows my three post-communist variables, and the third model shows all eight variables combined. The results show clearly that the addition of the post-communist variables significantly improves our understanding of organizational membership. In all three samples, the adjusted r-squared scores for the post-communist variables are much higher than for the SES factors alone, and they rise slightly when all eight variables are included together. Of the SES factors, income remains the most statistically significant. Education is also statistically significant in the SES models, but loses its significance when included in the fully specified models. Age has a significant negative effect in the fully specified model in Russia, but less so in East Germany. As noted above, men participate more than women, particularly in East Germany where gender is statistically significant in the complete model. Finally, city size has little to no statistical significance in most of the models. Looking more closely at the effects of the post-communist variables, we see that the mistrust of communist organizations is by far the most powerful explanatory factor. In the fully specified model of the pooled sample, an increase in one unit on the three-unit scale of mistrust is associated with a predicated decrease in organizational membership of 0.36, a very large figure considering that the mean of organizational membership is only 0.45 in Russia and 0.78 in East Germany. Generally speaking, there are two ways of interpreting the strong effect of people’s mistrust of communist organizations. First, focusing on those people who used to participate at high levels in the communist organizations, many did so because they wanted to. While almost all citizens of communist countries had little choice about being members of the trade union and perhaps one other organization, it is very likely that those people who participated in four organizations or more did so by choice – whether for instrumental reasons or simply because they enjoyed participating in the various activities. In other words, the active participants in the communist organizations probably developed positive associations with organizations, and had positive experiences while participating in them, which have translated into higher levels of participation in voluntary organizations today. In contrast, the people who did not join any communist organizations, or who were perhaps members of one or two, but no more, most likely had unpleasant experiences and developed negative


0.13*** (0.02)


0.14*** (0.02) 1.36*** (0.14) 0.15 0.81 1,806


0.07* (0.04)


0.05 (0.04)

1.51*** (0.06) 0.13 0.81 1,934


0.36*** (0.03)


0.01 (0.02)

0.10** (0.04)

0.24* (0.12) 0.03 0.72 1,009

–

–

–

0.02 (0.02)

0.04 (0.05)


0.003* (0.002)


0.002# (0.001)


0.37*** (0.03)

–

0.05** (0.02) 0.06* (0.03)

R1

0.02 (0.03)

0.05*** (0.02)

REG3

–

–

–

–

–

1.18*** (0.11) 0.11 0.70 1,005


0.10** (0.03)

0.05 (0.05)


0.32*** (0.03)

R2

0.03 (0.03)

0.03* (0.02)

1.01*** (0.17) 0.12 0.69 1,005


0.05 (0.04)

0.03 (0.05)


0.32*** (0.03)

0.02 (0.02)

0.03 (0.04)


0.004** (0.001)

R3

0.007 (0.20) 0.05 0.96 842

–

–

–

0.004 (0.04)

0.18** (0.07)

0.002 (0.003)

0.15** (0.05)

0.09*** (0.03)

EG1

1.54*** (0.09) 0.11 0.92 929


0.09** (0.04)


0.10* (0.06)


0.40*** (0.04)

–

–

–

–

–

EG2

1.31*** (0.25) 0.13 0.92 801


0.07* (0.04)


0.13* (0.07)


0.37*** (0.04)

0.003 (0.03)

0.16** (0.07)


0.003 (0.003)

0.02 (0.05)

0.07** (0.03)

EG3

Notes Table entries are unstandardized regression coefficients, with standard errors in parentheses. The dependent variable is the number of organizational memberships per person in each country. *** p  0.001; ** p  0.01; * p  0.05; # p  0.10 (all one-tailed).

Source: PCOMS (1999)

Adjusted R2 SEE N

.07 (0.11) 0.03 0.86 1,851

–

Disappointment

Constant

–

Persistence of friendship networks


0.04* (0.02)

City size

–

–

0.12** (0.04)

Gender (male)

Mistrust of communist organizations

–

0.002* (0.001)

Age

–

0.09*** (0.03)

Education

–

REG2

0.09*** (0.02)

REG1

Income

Variable

Table 10.7 OLS regression analysis of individual-level membership in voluntary organizations (Russia and Eastern Germany, combined and separately)

Post-communist voluntary organizations

179

opinions about organizations. Today, even though the circumstances surrounding participation have changed entirely, since people are now free to form and join organizations that are not state-controlled, the behavioral patterns of many of these formerly inactive participants or nonparticipants do not appear to have changed very much. Their disinterest in, and perhaps even antipathy toward, all kinds of organizations that developed during their experience with communist organizations seems to have carried over into the current period. In short, whereas during the communist period people were forced to join organizations, today they are free to choose not to participate (cf. Berlin 1969; Rose 1995a).11 As described above, friendship networks have lost their salience more in East Germany than in Russia. Moreover, those East Germans whose friendship networks have persisted or even intensified are less likely to join voluntary organizations, whereas those who have experienced more disruption in their own networks are more likely to participate. In the fully specified model for the East German sample, the persistence of friendship networks has a statistically significant (p  0.05) negative effect on current organizational membership. These results show that, at least in East Germany, there is often a trade-off between the vibrant social networks that developed during the communist period and the propensity to participate in voluntary organizations today. Finally, the disappointment variable is statistically significant (p  0.001) in the pooled sample, where an increase in one unit on the three-unit scale of disappointment is associated with a predicted 0.08 decrease in organizational membership. While not as powerfully as for the other variables, these results show that disappointment with postcommunism is an inhibiting factor to participation in voluntary organizations. In short, as expected, the more people are disappointed, the less they will participate.

Conclusion The analysis presented in this chapter has important theoretical and empirical implications. Theoretically, the significance of the three specifically post-communist factors suggests that attitudinal norms and behavioral patterns can have a powerful effect that lasts beyond the institutional context within which they first arose. This indicates the need for an ‘experiential’ approach to societal continuity and change, based on an understanding of human beings as agents who make choices that are rooted in their prior experiences and their interpretations of these experiences. In other words, this chapter has shown that while institutions do constrain and restrict the available options for social activity, they do not shape or change people directly or automatically. Moreover, although the period of late childhood and early adult socialization may be the most formative time period in many people’s lives, they do not become ‘programmed,’

180

Marc Morjé Howard

destined to remain the same throughout their entire adult lives. Only such a dynamic and interactive theoretical approach that focuses on people’s actual experiences can lead to a more grounded and realistic understanding of human behavior and societal change. More specifically, in terms of the low levels of organizational membership in post-communist Europe, the main argument of this chapter is that – as a result of the institutional experience of communism, with its forced mobilization and strict separation of public and private spheres (Jowitt 1992; Bunce 1999) – three main causal factors are responsible, and all three involve people’s ongoing reinterpretations of prior and present experiences. These factors consist of (1) people’s prior experiences with organizations, and particularly the legacy of mistrust of formal organizations that result from the forced participation in communist organizations, (2) the persistence of informal private networks, which function as a substitute or alternative for formal and public organizations, and (3) disappointment with the new democratic and capitalist systems today, which has led many people to avoid the public sphere. Together, these three factors present an account of the causal link between people’s interpretations of their prior experiences and their social behavior and activities today. Empirically, the lasting weakness of post-communist civil society indicates a troubling, although not necessarily disastrous, future for the type and quality of democracy as it continues to develop throughout the region. As we now enter the second decade of post-communism, it appears that democracy in post-communist Europe is developing in a way that makes it distinct from many post-authoritarian countries, as it is characterized by what Richard Rose (1995b: 34–42) calls an ‘hour-glass society,’ with few organized links between the state and its citizens.12 On the one hand, this weakness of post-communist civil society is troubling, since it means that post-communist citizens will have fewer opportunities to develop the ‘civic skills’ that many scholars view as being important for consolidating and legitimating a democratic system, while also lacking the crucial institutional representation and leverage that could be otherwise provided by active voluntary organizations. On the other hand, a weak civil society does not necessarily mean that democracy is doomed or in danger of collapse, since such a development would depend largely on the personalities, ideologies, and strategies of political elites, and the functioning of political and economic institutions. What is most likely is that post-communist democracy will continue to ‘muddle through,’ with elites and institutions that vary widely in their style and performance, but a citizenry that remains disengaged from the public sphere.

Post-communist voluntary organizations

181

Notes 1 For comprehensive accounts of the history of the concept of civil society see especially Cohen and Arato (1992). Also see Howard (2003) for a more complete discussion of the role of civil society in democratization. 2 For a trenchant critique of this method see Putnam 2000: 416. 3 The theoretical and empirical justification for the typology of prior regime type comes from the work of Juan Linz (2000). For a summary of Linz’s main argument see Linz and Stepan 1996: 38–54. For a full discussion of the application of Linz’s typology and the countries for which data are available on all three indicators see Howard 2003. 4 Unfortunately, although the WVS was conducted in Poland, the question on membership in voluntary organizations was not asked there in its entirety. Only two of the nine different types of organizations were mentioned. This omission is notable and regrettable, since – as a result of the noncollectivization of agriculture, the influence of the Catholic church, as well as mass mobilization in the Solidarity movement – Poland has generally been viewed as the ‘exception’ within post-communist Europe, with a more active civil society (see, e.g., Ekiert and Kubik 1999). However, it is worth pointing out that a cross-national study of several post-communist countries conducted by Samuel Barnes and his colleagues in 1990 to 1992 found that Poland actually had considerably lower levels of organizational membership than every other post-communist country in the study (even in religious organizations), and Barnes actually uses the term ‘Polish exceptionalism’ to refer to Poland’s unusually weak civil society (see Barnes 1998: 127). Moreover, in the two categories in which the 1995 to 1997 WVS question on membership in voluntary organizations was asked in Poland – political parties and unions – the results confirmed Barnes’ findings, showing exceptionally low levels of membership. Unfortunately, due to missing data on the other seven types of organization, this chapter will not be able to answer any questions about Poland definitively, although certainly the preliminary evidence suggests that Poland may fit in with other post-communist countries more than has been acknowledged previously. 5 For much more detailed and explicit discussion and analysis of the crossnational differences, including the WVS data on each of the nine types of voluntary organizations, see Howard 2003. 6 For a compelling discussion of the continuing decline of group membership in East Germany, see Padgett 2000: 87–95. 7 See, e.g., Milbraith and Goel 1977; Verba et al. 1978; Wolfinger and Rosenstone 1980; see also the discussion of the strengths and weaknesses of SES in Verba et al. 1995. 8 Education is generally viewed as the most important element of the SES model. See, e.g., Wolfinger and Rosenstone 1980. 9 For a full explanation of the coding of these variables see Howard 2003. 10 The category of ‘low’ mistrust applies to those who were members of between four and nine communist organizations; ‘medium’ mistrust to those who belonged to three communist organizations; and ‘high’ mistrust to those who were members of only up to two communist organizations. 11 In Howard (2003), I explore these issues in much greater detail using qualitative and interpretive methods, bringing in the life stories of in-depth interview respondents in East Germany and Russia. 12 Incidentally, this discovery calls into question the way in which Western governments, organizations, and foundations have been supporting the many ‘civil society initiatives’ throughout the region, which are based generally on

182

Marc Morjé Howard neo-liberal ideals and conceptions of self-initiative that are perceived as foreign, alien, and even suspicious by many post-communist citizens. For an extremely critical view of these aid programs in post-communist Europe, see Wedel 1998; for a balanced approach to such programs in general, see Carothers 1999.

Bibliography Barnes, S. H. (1998) ‘The Mobilization of Political Identity in New Democracies’, in S. H. Barnes and J. Simon (eds) The Postcommunist Citizen. Budapest: Erasmus Foundation. Berlin, I. (1969) ‘Two Concepts of Liberty’, Four Essays on Liberty. London: Oxford University Press. Berman, S. (1997) ‘Civil Society and the Collapse of the Weimar Republic’, World Politics 49, 3: 401–29. Bunce, V. (1999) Subversive Institutions: The Design and the Destruction of Socialism and the State. Cambridge: Cambridge University Press. Carothers, T. (1999) Aiding Democracy Abroad: The Learning Curve. Washington, DC: Carnegie Endowment for International Peace. Cohen, J. L. and Arato, A. (1992) Civil Society and Political Theory. Cambridge, MA: MIT Press. Edwards, B. and Foley, M. W. (1998) ‘Civil Society and Social Capital Beyond Putnam’, American Behavioral Scientist 42, 1: 124–40. Ekiert, G. and Kubik, J. (1999) Rebellious Civil Society: Popular Protest and Democratic Consolidation in Poland, 1989–1993. Ann Arbor: University of Michigan Press. Geremek, B. (1996) ‘Civil Society Then and Now’, in L. Diamond and M. F. Plattner (eds) The Global Resurgence of Democracy. Baltimore, MD: Johns Hopkins University Press. Howard, M. M. (2003) The Weakness of Civil Society in Post-Communist Europe Cambridge: Cambridge University Press. Jowitt, K. (1992) New World Disorder: The Leninist Extinction. Berkeley, CA: University of California Press. Ledeneva, A. V. (1998) Russia’s Economy of Favours: Blat, Networking, and Informal Exchange. Cambridge: Cambridge University Press. Linz, J. J. (2000) Totalitarian and Authoritarian Regimes, Boulder, CO: Lynne Rienner. First published 1975. Linz, J. J. and Stepan, A. (1996) Problems of Democratic Transition and Consolidation: Southern Europe, South America, and Post-Communist Europe. Baltimore, MD: Johns Hopkins University Press. Lomax, B. (1997) ‘The Strange Death of Civil Society in Post-communist Hungary’, Journal of Communist Studies and Transition Politics 13, 1: 41–63. Milbraith, L. W. and Goel, M. L. (1977) Political Participation: How and Why Do People Get Involved in Politics. Chicago, IL: Rand McNally. Miszlivetz, F. (1999) Illusions and Realities: The Metamorphosis of Civil Society in a New European Space. Szombathely: Savaria University Press. Miszlivetz, F. and Jensen, J. (1998) ‘An Emerging Paradox: Civil Society from Above?’, in D. Rueschemeyer, M. Rueschemeyer and B. Wittrock (eds) Partici-

Post-communist voluntary organizations

183

pation and Democracy East and West: Comparisons and Interpretations. Armonk, NY: M. E. Sharpe. Oliver, J. E. (2000) ‘City Size and Civic Involvement in Metropolitan America’, American Political Science Review 94, 2: 361–73. Padgett, S. (2000) Organizing Democracy in Eastern Germany: Interest Groups in Post-Communist Society. Cambridge: Cambridge University Press. Przeworski, A. and Teune, H. (1973) The Logic of Comparative Social Inquiry. New York: John Wiley & Sons. Putnam, R. D. (1993) Making Democracy Work: Civic Traditions in Modern Italy. Princeton, NJ: Princeton University Press. Putnam, R. D. (2000) Bowling Alone: The Collapse and Revival of American Community. New York: Simon & Schuster. Rose, R. (1995a) ‘Freedom as a Fundamental Value’, International Social Science Journal 145: 454–71. Rose, R. (1995b) ‘Russia as an Hour-Glass Society: A Constitution without Citizens’, East European Constitutional Review 4, 3: 34–42. Rupnik, J. (1996) ‘The Post-Totalitarian Blues’, in L. Diamond and M. F. Plattner (eds) The Global Resurgence of Democracy. Baltimore, MD: Johns Hopkins University Press. Skocpol, T. (1999) ‘How Americans Became Civic’, in T. Skocpol and M. P. Fiorina (eds) Civic Engagement in American Democracy. Washington, DC: Brookings Institution Press and the Russell Sage Foundation. Smolar, A. (1996) ‘From Opposition to Atomization’, Journal of Democracy 7, 1: 24–38. Verba, S., Nie, N. H. and Kim, J. (1978) Participation and Political Equality: A Seven-Nation Comparison. Cambridge: Cambridge University Press. Verba, S., Schlozman, K. L. and Brady, H. E. (1995) Voice and Equality: Civic Voluntarism in American Politics. Cambridge, MA: Harvard University Press. Völker, B. and Flap, H. (1996) ‘Changes in Personal Networks in the Former GDR’, Netherlands Journal of Social Sciences 2: 87–110. Völker, B. and Flap, H. (1999) ‘Getting Ahead in the GDR: Social Capital and Status Attainment under Communism’, Acta Sociologica 42, 1: 17–34. Wedel, J. R. (1998) Collision and Collusion: The Strange Case of Western Aid to Eastern Europe 1989–1998. New York: St Martin’s Press. Wolfinger, R. E. and Rosenstone, S. J. (1980) Who Votes?. New Haven, CT: Yale University Press.

11 Social capital in Poland Jerzy Bartkowski

Introduction Weakness of social capital is one of the most important obstacles to the current development of Eastern and Central European countries. Lack of social capital is one of the main reasons for the under-utilization of new opportunities by them, after poverty and lack of resources. The source of the low level of social capital in this part of Europe is a mixture of communist legacy and backwardness in social development. Progress in transformation means not only the introduction of new economic, social, and political institutions, but also the enriching and strengthening of social capital. On the other hand, the process of transition is in some ways contradictory and may affect negatively some forms of social capital. Also the main political danger to transformation, namely political populism, is often a result of the combination of low social capital and negative social effects of the transformation. Low social capital manifests itself in many areas of social and political life in Poland. This is visible in low interpersonal trust, the weakness of community spirit at a local level, the low level of political participation, low membership in voluntary associations, weak trust in political institutions, the poor image of politicians and public administration, and business practices that affect the social image of the private sector. Communism devastated civil society. The high level of social fear under the dictatorship destroyed social solidarity among many groups and mutual trust among the people even within families. The communist monopoly of organizational life resulted in a lack of skills to work together among the people. The ritualization of public life corrupted the social prestige of activism. The strong position of public administration vis-à-vis society affected the quality of services in public administration because the majority of civil servants had started their jobs before 1989. The same is also visible in commerce. Attitudes of salesmen toward customers have not changed much since the times of the economy of shortage. The communist way of getting legitimacy caused a high level of social expectations from the state. This resulted in passivity among many people

Social capital in Poland

185

who continue to wait for upward help. This phenomenon is common in many post-communist societies. It has been given such labels as ‘homo sovieticus,’ ‘the soviet mind,’ or ‘claiming consciousness.’ For example, according to the Polish General Social Survey (PGSS), the opinion ‘differences in income are too high in Poland’ is accepted by 83.2 percent of respondents. Similarly, 82 percent held that ‘the state should make the differences in income lower’; 87.5 percent agreed that ‘the government should guarantee minimal income to everybody’; and 87.5 percent said ‘the government should guarantee a job for everybody’ (PGSS, 1992–98). Another recent survey completed by CBOS had similar results, with 95 percent saying that the state should provide a minimum income, university education to those who want it (90 percent), basic shelter (89 percent), a job consistent with one’s qualification (85 percent) or any job (80 percent). These opinions are not diminishing, but growing. By comparison, in 1990, the state’s responsibility for basic shelter for everyone was accepted by 79.1 percent of respondents, and the state’s responsibility for providing a job for everyone by 69.4 percent (OBOP, XI 1990; CBOS, XI 1996, XII 1999). Another factor that affects the level of social capital in Poland is the legacy of the major crisis of the 1980s. The individual response to this was a high level of activity to keep up a family’s level of consumption through illegal practices, corruption, and participation in the informal economy. This undermined the rules of game in economic life. Many of the pathologies of the private sector in Poland have their roots in the way the economy worked under communism. This chapter is an overview of the main binding forces within Polish society: trust in social and political institutions, trust in politicians, the social image of the private economic sector, participation in voluntary organizations, interpersonal relations, community involvement, and the level of trust in different circles around the individual. Although the chapter mainly reports the current level of social capital, it also looks into the dynamics in the past ten years, and the possible effects of transformation on social capital. In the interest of space, references to specific Polish sources are limited to materials in print. A list of surveying organizations and the social surveys used is provided at the end of the chapter.

Trust in the main social and political institutions The analysis of different aspects of social capital in Poland starts from societal trust and extends to the main social and political institutions (Tables 11.1 and 11.2). Trust is not only a sign of their legitimacy but also a condition needed for their effective work. Social trust is especially important for police and the judiciary. There is a rather wide gap of public trust in political bodies compared to social institutions. Political institutions are evaluated rather poorly by

186

Jerzy Bartkowski

Table 11.1 Trust in institutions Institution

Evaluation

1996–97

VI/2000

Mean 1996–2000

President

Good Wrong Good Wrong Good Wrong Good Wrong Good Wrong Good Wrong Good Wrong Good Wrong Good Wrong Good Wrong Good Wrong

58 30 41 45 35 47 35 51 58 32 87 5 72 20 55 16 65 8 60 34 39 32

76 15 15 76 25 56 18 71 52 34 84 6 77 17 58 15 48 18 49 42 22 59

74.32 16.77 37.25 48.33 35.18 44.59 30.45 55.95 57.60 29.80 84.75 5.50 72.33 19.83 62.44 13.22 56.83 13.25 54.42 37.00 32.17 44.75

The cabinet Senate Parliament (Sejm) Local authorities Public radio Public television National Bank Ombudsman Police The judiciary Source: CBOS

Note Percentages are computed without answers ‘difficult to say’. The question is ‘How do you evaluate the activity of ______ in terms of the good of the country?’ Trust in the main political institutions was surveyed in 1997, in the main societal institutions in 1996. If the data are unavailable, the closest survey is used.

Table 11.2 Corruption as social problem What do you think; is corruption a serious problem or not in Poland?

VII 1991 (%)

II 1992 (%)

VII 2000 (%)

VIII 2001 (%)

Very big Rather big Rather low Very low Difficult to say

33 38 15 2 12

49 37 8 1 5

46 40 6 0 8

68 25 2 1 5

Source: CBOS.

respondents. This is due mainly to the present political situation in Poland. The ruling coalition had been losing its support since the elections of 1997. This resulted in its being trounced in the elections of 2001. The level of dissatisfaction is so high that it has led to electoral results that are quite uncommon in democracies. Both main parties, which had possessed more than 50 percent of seats in Parliament after the 1997 elections, have no

Social capital in Poland

187

seats in the new Parliament elected in September 2001. This has produced strong and broad effects in the evaluation of all political institutions with the exception of the president, who is supported by the main opposition party. On the other hand, these results reinforce the opinion that there is strong support for democracy in Poland. In the eyes of the public, the evaluation of political leaders differs from the legitimacy of the political system. In sharp contrast to political institutions, social institutions are positively evaluated. The level of trust in radio and television is very high. In addition, local governments have a better reputation than do national institutions. Yet there is a low level of trust in some important institutions that are not so strictly connected to current politics, such as the ombudsman, and especially the judiciary. This pattern of attitudes toward political and social institutions is typical in post-communist countries. In spite of the change of political system, there has been no dramatic change in attitudes toward political institutions.

Social image of politicians and public administration The main reason for low trust in the main political institutions in Poland is the dissatisfaction with the ‘Solidarity’ government that ruled the country from 1997 to 2001. Yet the low trust is also rooted in a very poor social image of public administration and politicians. They are perceived as corrupt, selfish, and incompetent. The trends in public opinion in perceived corruption and dishonesty of the political class and public administration are negative. A similar trend is to be found in the evaluation of civil servants. The view that ‘many top civil servants in the government abuse their posts for private profits’ was chosen by 51 percent in 1995, 55 percent in 1997, 61 percent in 1999, 59 percent in 2000, and 70 percent in 2001. The answer ‘none’ was chosen by only 1 to 2 percent of respondents in all these surveys. Practices perceived as common among public servants are: public contracts for family and friends (78 percent), bribes (79 percent), nepotism (87 percent), and using a public post in ways expected by their political parties (65 percent). Similar levels of corruption are perceived at both local and national level (58 percent) (2000). These opinions seem to be supported by the experience of many people. In 2000, 19.2 percent of those surveyed admitted they had given a bribe. The main places where people encountered corruption are hospitals (52 percent), public administration (14 percent), employment (8 percent), and economic activity (8 percent). Surveys completed in 1997, 1999, and 2000 yielded similar results (CBOS, 1995, 1997, 1999, 2000, VIII 2001). This high level of perceived corruption is supported by other sources. This may be traced in the scores of corruption in Poland given by Transparency International on a 10-point scale starting from 1 – very high

188

Jerzy Bartkowski

corruption to ten – null corruption. Poland was scored as follows: 3.64 in 1980–85, 5.20 in 1988–92, 5.57 in 1997, 4.6 in 1998, 4.2 in 1999, and 4.1 in 2000. There has been a growing trend toward public honesty since 1989. The change in the political system resulted in higher honesty of public life. The trend changed in 1997, since when there has been a visible worsening in the standards of public administration. The social image of politicians and the civil service is rather negative in Poland. The perceived honesty of the political class is low. Among those surveyed in 1999, low and very low honesty was attributed to top civil servants by 38 percent, to Members of Parliament by 45 percent, to politicians by 46 percent, and to local public administration by 35 percent. This image has worsened since 1997: for top civil servants from 18 percent to 38 percent, for politicians from 32 percent to 46 percent (CBOS, XI 1997, I 1999). According to recent surveys, politicians are perceived as untrustworthy (74 percent). The public perceive them as not keeping electoral promises (2 percent), not thinking about the common people (6 percent), but about their own interests (82 percent). Motivations attributed to the politicians were money (45 percent in 2001 compared to 31 percent in 1993), and power (36 percent compared to 24 percent in 1993). Only 6 percent of those surveyed in 2001 (compared to 22 percent in 1993) believed in politicians’ willingness to work for others (CBOS, IV 2001). This social image of political institutions, politicians and public administration has produced a high level of political alienation. The higher the level of government, the lower the perceived influence of the common man. Although the feeling of efficacy in politics on the national level has increased since 1988, it is still low (Table 11.3). This is computed as a percentage of those employed. The feeling of no representation of interest is based on the percentage of the answers ‘None and difficult to say’ among those employed, to the question ‘Which trade union, in your opinion, is the best in representing the interests of people like you?’ Table 11.3 Political alienation Do people like you have any influence on how things are going on in:

V 1988 (%)

III 1990 (%)

VI 1992 (%)

VII 1993 (%)

X 1997 (%)

V 1999 (%)

XII 1999 (%)

The country? The voivodship they live in? The county (powiat), city or municipality they live in? The parish they belong to? Their neighborhood? The factory they work in?*

6.1 –0.0 19.6

–0.0 –0.0 13.5

7 – 16

10 – –

19 – 26

11 – 25

16 17 31

–0.0 –0.0 22.0

–0.0 –0.0 27.1

– – 37

– – –

– – –

– – –

36 43 47

Source: CBOS Note * Percentage of employed.

Social capital in Poland

189

This high level of alienation among the public is also indicated by the large number of people who feel their interests lack representation in politics. The percentage of those expressing satisfaction with the degree of representativeness decreases with each successive level of government. This is shown in the pattern of the answers to the question ‘Do you think that there are people who care about the interests and needs of people like you?’ At the local level, 47 percent answered yes; at the regional level, 33 percent; in Parliament, 32 percent; in the national government, 31 percent (CBOS, XII 1999). A similar pattern is visible in other important areas of social life. Of special interest is the representation of workers’ interests. A very high percentage of people do not see any organization that acts on their behalf. This is accompanied by falling membership in trade unions (Table 11.4). The current level of trust in political institutions is very low in Poland. The prestige of politicians and civil servants is also low. There is a high level of political alienation and feeling of lack of representation in the political system and an alarmingly high level of mistrust in the institutions that are important for the current state activities and for the legitimacy of the system.

The social image of the private economic sector The development of a private sector in the economy is one of the most impressive successes of the systemic transformation. In 2000, the private sector accounted for 75 percent of active economic entities, 72.3 percent of employment, 62.1 percent of current investments, 71.3 percent of sold production in industry, 95 percent of sales of construction, 90.5 percent of retail sale, 46.5 percent of gross value of fixed assets, and 99.4 percent of all shops. Small businesses (enterprises below 500 employees) have been contributing to a growing percentage of the gross national product since the 1980s: 18 percent in 1985, 19 percent in 1989, 25 percent in 1990, and 40 percent in 1996 (Zienkowski 1996; GUS 2000). The public treats the private sector in Poland with suspicion. Its image Table 11.4 Trade union membership

1990 1992 1993 1994 1995 1998 2000

Yes

No

Feeling of no interest representation (%)

19.90 17.65 18.19 14.58 13.53 8.91 –0.00

80.90 82.35 81.81 85.42 86.47 91.09 –0.00

–0.00 64.00 46.00 –0.00 –0.00 72.50 64.00

Sources: Union membership: OBOP XI ’90. PGSS, 1992–98. Other: CBOS.

190

Jerzy Bartkowski

Table 11.5 Perceived sources of wealth Why, in your opinion, have some people become rich in a short time? Please tick up to three causes.

X 1992

XII 1994

I 1997

I 1999

Courage, readiness to take risks Hard work, devotion to their work Abilities Illegal practices Imperfect laws Ruthlessness Luck Bribery Family help Other Difficult to say

38 26 29 44 39 27 15 32 10 2 3

36 31 32 42 39 23 17 28 8 1 4

49 46 40 40 34 29 20 19 13 1 2

46 41 39 38 34 35 17 21 13 3 3

Source: CBOS.

weakens its position in public life and therefore its ability to defend its interests against economic populism. It also affects people’s motivation to start their own business or to work in this sector. Partially, it is due to jealousy of high income and level of consumption, but it is also rooted in widespread business practices. Criminal activities have long been perceived as a main source of wealth (Table 11.5). It is not surprising, then, that the public has a low opinion of private businessmen. Many of those surveyed (56 percent) expressed dislike of them (CBOS, X 1992). The number of respondents who are ready to give the rich a high degree of social respect is low (30 percent in 1994, 32 percent in 1997, and 33 percent in 1999) (CBOS). Fortunately this has been changing. The percentage of people who see economic success as a result of ability and hard work is growing (Table 11.5).

Participation Participation is the core of civil society. This section of the chapter discusses the scope of participation and the number of associations to which people belong, as well as the level of individual involvement in civil society associations. Basic trends in participation and the impact of the transformation on them are also reviewed. There are data to show the scope of such activities. In 1994, 29,580 associations were registered and another 12,216 had local branches in Poland. In addition, 5,900 foundations were registered, including 2,681 trade unions and chambers of vocational groups (Wygnan´ski 1996). In 1999, 19,328 cooperatives, 4,906 foundations, and 33,493 associations (GUS) were registered in Poland. There are also some informal associations that will be discussed below in the section devoted to community involvement. The general level of individual activity is shown in Table 11.6. The data

13.2

–

Outside parties, unions, and youth organizations

Informal activity

–

24.8

74.2

1975

–

29.9

70.1

1977

11.4

–

74.5

1980

05.9

–

85.1

1981

Sources: OBOP 1972–75; Polacy 1980–81, 1990–95; CBOS 1986–89, 1997–98.

44.1

No membership

1972

07.0

–

60.1

1986

13.1

–

84.7

1988a

Table 11.6 Activity in associations according to public opinion surveys (1972–98)

11.8

–

86.1

1988b

08.6

–

85.4

1989

6.4

–

–

1990

–

7.8

–

1995

5.0

–

–

1997

6.7

–

–

1998

192

Jerzy Bartkowski

also allow us to look into the dynamic of participation. Measurement is difficult because of the impact of artificial activity under communism and changes in activity of informal groups. In addition, the results are strongly dependent on how questions are asked. The data rather supported the perception of low membership and activism in Poland. There is falling membership and activity in organizations. The membership was also affected by state–society conflict in Poland in the 1980s. This resulted in a reluctance to join organizations and keeping only passive membership in organizations supported by the state. During martial law in the 1980s there was a shift to informal activity. The transformation resulted in falling membership in officially supported bodies such as parties, trade unions, and youth organizations. During the 1990s there was a trend toward lower membership in trade unions (Table 11.4), due mainly to the transformation of the ‘Solidarity’ movement as well as the impact of privatization in many industries. The level of activity in associations is low in Poland. Of those who are active, the most common memberships are in trade unions, religious organizations, and professional associations, as well as parents’ committees in schools, associations of hobbyists, and sport, religious, and cultural associations (Table 11.7). Some information about the motivation of activists may be drawn from the CBOS survey taken in February 1997 which found differences in Table 11.7 Activity in associations (1975–99)

Welfare organizations Religious organizations Cultural organizations Trade unions Political parties Local community action Self-government Environment Professional associations Youth work Sports, recreation associations Women’s groups Health organizations Parents’ school organizations Voluntary fire brigade Hobbyist group Other groups None

1975

1990

1997

1998

1999a

–0.0 –0.0 –0.0 –0.0 5.3 –0.0 –0.0 –0.0 1.3 2.0 –0.0

4.8 9.5 1.9 5.7 0.8 2.2 –0.0 1.6 1.9 1.9 2.7

0.9 3.1 1.3 –0.0 –0.0 2.0 1.7 –0.0 –0.0 1.2 –0.0

1.5 3.6 1.5 2.2 0.3 0.6 2.1 0.9 0.7 1.5 3.8

2.3 3.6 1.7 2.4 0.7 1.3 –0.0 0.7 1.2 0.7 2.2

2.8 3.9 3.2 3.1 0.8 1.1 2.9 1.5 0.8 2.0 5.1

–0.0 1.3 –0.0

2.5 1.5 –

0.8 1.3 5.9

1.0 0.7 4.5

0.5 0.7 –0.0

1.5 1.2 5.5

–0.0 –0.0 –0.0 93.7

–0.0 –0.0 5.1 71.9

–0.0 1.5 1.3 85.1

3.0 4.0 5.8 77.0

–0.0 –0.0 2.6 86.5

2.0 4.9 8.6 78.0

Sources: EVS 1990–99 (1999a); OBOP 1975; CBOS 1997; II 1998; XI 1999 (1999b).

1999b

Social capital in Poland

193

values among passive and active respondents. Those who were active declared more attachment to such values as being useful to others (10.1 percent), seeking love and friendship (8.2 percent), having an interesting job (5.2 percent), but less attachment to living a peaceful life without trouble (14 percent). These values are a combination of social orientation, social sensitivity, and desire to be active. The motives for activity most frequently mentioned in another survey were a feeling of compassion for the poor and people in need, and a feeling of moral obligation (EVS 1990). The most difficult issue is measuring the impact of systemic change on associations, the majority of which have lower membership now than before 1989. This is the case even for politically neutral sports associations. This trend has been consistent throughout the 1990s and is probably attributable to a combination of factors. First, loss of membership is most visible among those organizations which had been based on state support (e.g. the Polish–Soviet Friendship Society). Organizations such as the Union of Rural Youth now has only 4.7 percent of its membership of the 1970s. Second, private industry has fewer opportunities to support sport and tourist organizations than the old system of state ownership. Third, changes in participation are less significant among those associations that were based in real need of the people or have strong local roots (e.g. voluntary fire brigades). In many cases, the decrease in numbers represents the loss of artificial members, with the resulting levels equal to the membership before the artificial increases of the 1970s.

Local communities and community-oriented activity Community is the most visible manifestation of social capital. A feeling of community is a very important motivation for people to become active. Acquiring social capital often starts from successful action toward local public goals. This kind of capital is unevenly distributed. This is an important asset of many groups and is present even in poor communities. The feeling of community can be lost and is very difficult to recover. There are three groups of important local institutions that maintain community spirit, especially in the countryside (Table 11.8). The first group consists of the local media. About 31 percent of municipalities and 19.5 percent of rural communes maintain a local newspaper. The second group is formed from regional cultural associations and societies of friends of the locality. There were 1,170 regional cultural associations in 1993. They cultivate local folklore and create arenas for local integration. There were also 509 societies of the friends of the locality spread across 20 percent of the communes (Regionalne 1993). Their membership can be estimated as 2.4 percent among the people according to CBOS (IV 2000). Some are active even abroad, especially among Polish immigrants in the U.S.A.

194

Jerzy Bartkowski

Table 11.8 Activity at community level (1999) (percentage of municipalities) Number of registered associations Registered associations in rural commune Foundation in rural commune Local newspaper Local newspaper in rural commune Regional association (1993) Regional association in rural commune (1993) Educational association Educational association in urban commune

14.8 3.7 9.7 31.0 19.5 43.0 23.0 7.8 47.2

Sources: GUS. Regionalne 1993; KKP 1995.

The third group consists of such local associations as volunteer fire brigades, parent school committees, and sports clubs. They are especially important in the countryside. First, they are often centers of other local social life. Sports clubs and volunteer fire brigades organize a lot of local cultural activity. Their buildings often serve as local community centers. Activity in them is a first step toward obtaining local respect and getting a chance for further advancement in local politics. This is important even for politicians at the national level. Members of volunteer fire brigades form the biggest interest group in the Polish Parliament that exerts some influence on legislation. All these organizations are active in local politics. Their support is an important factor for success in local elections. In the last local elections (1998) the parents’ committees supported 2,293 candidates (440 of whom were elected), sports clubs – 1,059 (225 elected), and volunteer fire brigades 3,250 (1,153 elected). The associations are dispersed according to a center–periphery dimension. Although rural communities preserve a lot of traditional informal activity, they are not so ready to adapt to new circumstances. The number of formal associations and foundations in the countryside is very low. There are practically no educational associations in rural communes (Table 11.8). Participation in local community action is often the first step toward the creation of social capital. About one-fifth or one-quarter of the people worked with others on local goals (Table 11.9). The nature of the goals for which people are ready to act together is significant. According to the CBOS survey in December 2000, the most frequent goals were: clean neighborhood – 25 percent, communication (roads, garages, parking lots) – 17 percent, social and cultural activity – 14 percent, education – 12 percent, local parish – 11 percent, charity – 10 percent, environment – 9 percent (as a percentage of those involved). A similar hierarchy was obtained earlier in a survey conducted in December 1986 also by CBOS. People were ready to take part in building schools – 30.3 percent, health centers – 25.3 percent, church – 15.6 percent, and kindergartens – 14 percent. Community activity is not developed broadly in Poland. Lack of it is

Social capital in Poland

195

Table 11.9 Participation in community action Have you ever worked with others in this community to try to solve some community problems?

Percentage of ‘often’ and ‘sometimes’

I 1991 VI 1992 1993 II 1997 II 1998 XII 1999

20 29.6 26 16 21 22

Sources: 1991 from Bruszt and Simon (1992); 1993 from Giza-Poleszczuk (2000: 164); other data after CBOS.

especially visible in new urban neighborhoods. Some are called simply a social desert. This resulted in a high level of social pathology and crime in these areas because of weak social control. In some rural communities, living traditions of voluntary work for a common purpose have been preserved. Such actions were abused under the communist regime, so the public’s memory of them is now very negative. Nevertheless, this is still important in rural communities in which the inhabitants have little money, but are able to contribute with their personal work. For many communities it is the only way to build such local improvements as roads or sewage systems. This kind of action is more common in the traditional rural communities of south-eastern Poland.

Networks of personal ties: family, friends, neighbors and acquaintances A network of personal contact is not only a set of social contacts, or a source of information, but also a network of solidarity. This works especially in the case of personal bad luck. A personal network manifests itself in the frequency of contacts. The main close contacts of an individual take place within the nuclear family among parents, children, and mother-in-law or father-in-law. But even contacts among siblings are not so common. The extended family does not count for much in contemporary Poland. There is a wide gap even between one’s own family and the family of the wife or husband (Table 11.10). The close personal circle is formed mainly from family, and friends from school and the workplace. Contacts with people outside of the immediate family are rare. The number of people with whom close ties are kept is low. Close contact with one or two persons is admitted by 14 percent of the respondents; three to four persons, 19 percent; five to six persons, 19 percent; seven to ten persons, 24 percent; more than ten persons, 19 percent. ‘Nobody’ was chosen by 5 percent of respondents (CBOS, VII 1999). The frequency of personal contact with friends and acquaintances is

196

Jerzy Bartkowski

Table 11.10 Frequency of personal contacts (relatives) How often do you have contact with:

At least once a week

At least once a month

At least once a year

Once a year or rarely

Never

Parents? Grown-up children that do not live with you? Friends, colleagues, acquaintances? Mother-in-law or father-in-law? Siblings? Siblings of wife or husband? Relatives? Relatives of wife or husband?

72 66

16 22

8 7

4 4

0 0

62

37

16

14

8

52

26

13

6

2

45 27

22 26

20 26

12 15

0 6

12 8

21 16

35 33

25 32

9 11

Source: CBOS VIII 1999.

Table 11.11 Frequency of personal contacts (non-relatives) Almost One or Several Once a Several Once Never Don’t every two times times month times a year know day weekly a month a year Friends 8.12 Neighbors 20.60 Acquaintances 7.46

17.55 13.48 13.29

22.47 12.78 20.20

17.90 10.62 18.09

25.09 14.41 22.91

6.73 6.74 8.02

1.83 0.31 20.80 0.57 9.17 0.86

Source: PGSS; cumulative data from 1992–98.

not high, around once a month (Table 11.11). Home visitors comprise mainly friends from school or work (70.1 percent). Including neighbors among visitors is rare, only 18.2 percent (Town study, 1998). Of special interest are contacts with neighbors. Although they are admitted as frequent, in most cases they are only casual contacts. This was studied in a survey held in two small towns. The respondents were asked about character of their contacts with their neighbors. In the majority of cases, contact was limited to saying ‘hello’ (43.4 percent). Closer contact was kept by about 45.7 percent of respondents. In most cases they comprise such supports as mutual borrowing of small items or food (32.5 percent of all respondents). Help in shopping or repairs are mentioned by 19.4 percent, help in childcare by 8.6 percent, mutual visiting by 22 percent, borrowing small amounts of money by 9.8 percent (Town study, 1998). Although personal ties are weak and limited, they are important in the case of bad luck. Asked whom they can count on, those surveyed mentioned mainly close family and relatives (77.7 percent). Other sources of support

Social capital in Poland

197

are rarely indicated: employer – 24.8 percent, trade union – 8.6 percent. The answer ‘none to support’ was chosen by 26.1 percent (Polacy 1995). A similar question asked ten years earlier yielded the following answers: family and relatives – 78.9 percent, supervisors in the workplace – 35.8 percent, trade union – 23 percent (CBOS, 1984 workers and intelligentsia). Personal ties play an important role in looking for a job in the case of unemployment. The most frequent factor which people would use in the case of unemployment is family, friends, and acquaintances – 77 percent. Of those who are now employed, 30 percent found their current job in this way (CBOS, X 1999). Personal ties (family, friends, and neighbors) will be discussed in comparison to the attachment to and trust of other groups. The level of interpersonal trust is very low in Poland. Asked about their feelings, only 10 to 20 percent of respondents said they trusted other people. During the 1990s, the perception of other people has become worse. Independent of the survey, the percentage of people who trust is currently lower than at the beginning of the transition (Table 11.12). The same mechanisms affect comparison of trust in different social groups (Table 11.13). Particularly striking is a strong gap among family Table 11.12 Trust in people (%) Generally speaking, would you say that most people can be trusted or that you cannot be too careful in dealing with them?

Most people can be trusted

Cannot be too careful

Difficult to say

No answer

1989 WVS 1990 EVS 1992 PGSS 1993 PGSS 1994 PGSS 1995 PGSS 1998 PGSS 1999 EVS

31.3 25.7 11.2 9.3 8.5 8.7 10.6 17.8

59.5 62.3 86.9 88.9 89.9 89.1 87.1 78.9

9.2 11.9 0.9 0.7 0.8 1.1 1.1 3.2

– 0.1 1.1 1.1 1.0 1.1 1.3 0.1

Table 11.13 Trust in different social groups (%) Trust

Total

Some

Little

None

Difficult to say

Family Poles Colleagues in workplace Neighbors Superiors in workplace Police Trade union

39.6 14.1 10.8 7.3 5.4 7.4 2.4

47.4 43.4 40.8 40.7 24.8 23.9 11.9

11.0 33.5 23.2 39.8 26.8 40.9 39.7

0.5 2.1 2.3 6.6 9.2 16.0 22.3

1.4 6.8 22.9 5.5 33.8 11.7 23.8

Source: Bruszt and Simon (1992).

198

Jerzy Bartkowski

circles and other social groups. This was observed long before 1990. It was called ‘a social vacuum’ (Nowak 1981). This observation is still valid. The general statement of trust in ‘a Pole’ is more acceptable than trust in neighbors or colleagues. Although the survey was held in 1991, there is no reason to believe that there was a big change later.

Summary The level of social capital in Poland is currently low. This pattern consists of scant trust among people and in institutions, a high degree of perceived pathology in economic and political institutions, high passivity of people, poor participation in associations, weak social ties limited to nuclear family, and weakness of local communities. The high level of social pathology that affects the social and personal life of people does not result in their activity. Many people keep looking for external action or escape into passivity. This is a legacy of the modern communist political dictatorship combined with its paternalism. The longitudinal effect of communism is a manifestation of an asymmetry of losing and rebuilding social capital. After 1989, slow reconstruction and an enriching of social capital may be observed. The process of transition does not help greatly in the creation of social capital. There is some contradiction in post-communist development if social capital is considered. Poverty, unemployment, competition in the workplace, and strong primary concern for the family do not create a good environment for mutual trust among people, rebuilding social ties, and networks of cooperation. It is also not a good time to look for volunteers to do unpaid work for others. The culture of the political elite is also not supportive of building bridges between society and its political institutions. The weakness of social capital is a significant impediment to the further development of Eastern Europe. The problem is that relatively developed societies lost their traditional rural communal roots, but were not elevated to civil societies. What is needed are new ways not only to build an appropriate institutional framework, but also to find stimuli to reinvigorate the willingness of people and communities to seek new opportunities.

Appendix Sources of data CBOS – Centre of Public Opinion Research. EVS – European and World Value Survey. The studies were conducted in 1989, 1990, and 1999. Principal investigators were: A. Jasin´ska-Kania and R. Siemien´ska. GUS – National Statistical Office. KKP95 – Krajobraz Kultury Polskiej. Survey of all municipalities in Poland by The Institute of Culture, Warsaw, 1995.

Social capital in Poland

199

OBOP – Public Opinion Research Organization. PGSS – Polish General Social Survey. Cumulated data 1992 to 1998. N  8,895. Principal investigator was B. Cichomski. Polacy (The Poles) – Series of surveys conducted from 1980 to 1995. 1995: ‘Strategies and Actors of Transformation and Modernisation’. The principal investigator was W. Adamski. Some of the surveys were conducted by S. Gebethner. Town Study – survey of two small towns. Principal investigator was I. Krzemin´ski, 1998.

Bibliography Bruszt L. and Simon, J. (1992) Political Culture, Political and Economic Orientations in Central and Eastern Europe During the Transition to Democracy. Budapest: Institute for Political Science of the Hungarian Academy of Sciences. Chmielewski, P. (1995) ‘Mountain Commons in the Tatras’, New Institutional Approach, Polish Sociological Review, 3. Czapin´ski, J. and Panek, T. (eds) (2001) Diagnoza spol-eczna. Warszawa. Dyczewski L., Wadowski, D. and Szwed, R. (1997) Kapital- ludzki czynnikiem rozwoju Euroregionu Bug. Lublin. Giza-Poleszczuk, A., Marody, M. and Rychard, A. (2000) Strategie i system. Wyd. IFiS PAN, Warszawa. Gorzelak, G. and Jal-owiecki B. (1998) Koniunktura gospodarcza i mobilizacja spol-eczna w gminach. Warszawa. GUS (2000) Rocznik statystyczny rzeczpospolity Polskiej. Warszawa. Hausner, J. et al. (2000) Jakos´c´ rza˛dzenia: Polska bliz˙ej Eunii Europejskiej – monitoring IV. Kraków. Hryniewicz, J. (1996) ‘Czynniki rozwoju lokalnego’, in B. Jal-owiecki (ed.) Oblicza polskich regionów. Warszawa. Nowak, S. (1981) ‘Values and Attitudes of the Polish People’, Scientific American 245 (1). Regionalne (1993) Regionalne Towarzystwa Kultury w Polsce. Ciechanów. Wygnan´ski, J. J. (1996) ‘Dlaczego warko mówwe o organizacjach pozarz’dowych?’, in A. Kunicka (ed.) Miejsce dla kaz˙dego. Warszawa. Zienkowski, L. (1996) Wkl-ad mal-ych i s´rednich przedsie˛ biorstw w rozwój gospodarki polskiej. Warszawa.

12 What does social capital add to democratic values?1 Richard Rose and Craig Weller

To what extent, if any, does social capital increase support for democratic values of citizens? Before the recent explosion of interest in social capital, it was usual to explain support for democratic values as a consequence of individual attributes such as education; individual economic circumstances or a country’s economic development; and characteristics of political institutions and performance. Because social capital is only one element of the political system, its influence is likely to be limited rather than the primary or exclusive cause of democratic values (cf. Boix and Posner 1998). The question invites empirical examination of the extent to which social capital, however measured, is the primary influence, one among several major influences, or of only minor importance in forming political attitudes. In answering this question, we have the advantage of using a questionnaire specially designed to measure social capital in a multiplicity of forms, generic and specific to politics. The 1998 New Russia Barometer (NRB) survey was developed by the first-named author as part of an interdisciplinary World Bank programme designed to determine the importance of social capital in many settings around the world.2 The concept of social capital fits well in Russia, inasmuch as in Soviet times people developed a wide variety of informal networks to compensate for, or even subvert, the formal commands of a repressive society, and many networks have remained in use following the collapse of communism (cf. Shlapentokh 1989; Rose 2000a). Moreover, examining the impact of social capital on democratic values in a regime variously described as an ‘incomplete democracy’ or ‘partly free’ (cf. Freedom House 2001; Rose and Shin 2001) provides a robust test of the importance of national context on individual commitment to democratic values.

Multiple forms of capital require a multi-causal model In economic analysis, capital is the stock of resources used to satisfy wants by contributing to the production of goods and services (Black 1997: 47). It is consistent with this usage to define social capital instrumentally as pro-

Social capital and democratic values

201

ducing democracy, jobs or bread (cf. Coleman 1990). But the influence of social capital is additional to three conventional measures of capital: money, land and labour (Serageldin and Steer 1994). Robert Putnam’s (1997: 31) widely invoked definition of social capital as ‘features of social life – networks, norms and trust – that facilitate cooperation and coordination for mutual benefit’ conflates individual attitudes and behavioural networks, making it impossible to separate the influence of one on the other. Inglehart (1997: 188) avoids this confusion by adopting a strictly social psychological definition of social capital – ‘a culture of trust and tolerance’ in which ‘extensive networks of voluntary associations emerge’ – but it begs the question: What creates cultural attitudes? Insofar as attitudes are derived from experience, then social capital networks are the cause rather than the consequence of trust (Dasgupta 1988). Social capital networks may be informal or formal. Informal associations are face-to-face networks, and this is true of those with a modicum of formal organisation, such as a bowling league or a choir, as well as of groups that meet informally in a pub or picnic in a park. A modern society is defined by having many impersonal formal organisations too – joint stock companies, central banks, social welfare ministries and political parties. Informal networks support or supplement the activities of formal organisations (cf. Weber 1947; North 1990). Informal networks are the building blocks of Putnam’s theory of Making Democracy Work (1993). He postulates that the norms and practice of cooperation in face-to-face networks ‘spill up’ into the formation of formal organisations such as political parties that aggregate the preferences of individuals. The paradigm example of spilling up is the preamble to the American Constitution, in which ‘we the people of the United States’ pledge cooperation to satisfy their collective wants. In Tocqueville’s time, more than 90 per cent of Americans could engage only in face-to-face voluntaristic networks, for in the 1830s census figures record few communities with more than 2,500 people. However, the government of a country of 150 or 250 million people necessarily requires formal organisations. The principal concern of Putnam (2000) is whether involvement in informal and formal networks is increasing or decreasing in the United States today. Putnam takes it as axiomatic that any change in social capital networks will influence democracy in America. The existence of multiple resources requires a multi-causal model to test to what extent different forms of capital influence democratic values. For example, educated persons are expected to be more committed to democracy but those with postmodern values may be more dissatisfied if there is a coincidental deterioration in the natural environment. Economic approaches assume that people with a higher standard of living will be more committed to a democratic regime. These influences are expected to operate net of any impact of social capital. It is possible for support for democratic values to be high or rising if the influence of a decline in social

202

Richard Rose and Craig Weller

capital is offset by the positive influence on democratic values of other forms of capital in society.

Measuring democratic values While social capital is about networks, democratic values are attitudes of individuals. One approach to surveying values is to ask individuals the extent to which they regard democracy as the ideal way to govern a country; the concrete meaning of the symbol is left open to each individual respondent. More detailed survey questions show a convergence in meanings attributed to democracy – individual liberty; the right to hold government accountable through elections; and, outside the United States, welfare state values – and Russians too associate these three sets of attributes with the idea of democracy (cf. Simon 1998; Bratton and Mattes 2001; Rose 2001: 22f.). An alternative approach is to ask individuals how satisfied they are with the way democracy works in their country today (Klingemann 1999: 36ff.). However, this approach ignores Robert Dahl’s central argument (1971) about the gap between democracy as an unattainable ideal and the ‘second best’ alternative of polyarchy. Given this gap, idealistic democrats who are dissatisfied with the way their polyarchy is working ought to register dissatisfaction. The question also ignores the distrust of democratically elected governors shown by James Madison and his co-authors in the construction of the American Constitution, and by Winston Churchill in his argument for democracy as the lesser evil compared to every other form of government (cf. Rose et al. 1998: ch. 5). Given that the Russian Federation has a partial or incomplete democratic regime, the NRB Social Capital Survey asked people to evaluate democracy as an ideal, adapting a question pioneered by Doh Chull Shin (1999) in the Republic of Korea, which has likewise not yet completed democratisation. Russians were asked to say where they would like their country to be placed on a scale ranging between complete democracy (point 10) and complete dictatorship (point 1). The largest group, 36 per cent, said they would like their system to be completely democratic. Almost three-quarters chose a position in the top half of the scale arithmetically and an additional 13 per cent chose 5, the psychological midpoint of the scale (Figure 12.1). Since Russians have lived longer under undemocratic rule than under a regime based on free elections, the Social Capital Survey also asked whether people would prefer to return to a communist regime, have the army rule or be governed by a tough dictator (Figure 12.2). Each alternative is opposed to the cooperative democratic values postulated in theories of social capital. Although the median Russian is in favour of democracy as an ideal, 61 per cent were prepared to endorse at least one undemocratic alternative. However, Russians disagreed about which

Social capital and democratic values

203

Q. Here is a scale ranging from a low of 1 to a high of 10. On this scale, point 1 means complete dictatorship and 10 means complete democracy. Where would you personally like our country to be placed? 50% 40%

36%

30% 20%

16%

13% 10%

3%

4%

4%

1 2 Complete dictatorship

3

4

4%

0%

5

5%

6%

6

7

9%

8

9

10 Complete democracy

Figure 12.1 Endorsement of democracy as an ideal Source: Centre for the Study of Public Policy, New Russia Barometer Social Capital Survey. Fieldwork: 6 March to 13 April 1998. Number of respondents: 1,908.

undemocratic alternative would be preferable: 41 per cent endorsed a return to communist rule; 36 per cent a tough dictatorship; and 15 per cent military rule. Among those with sympathy for undemocratic rule, 38 per cent were selectively undemocratic, favouring only one undemocratic alternative; 18 per cent favoured two alternatives; and 5 per cent were uncritically anti-democratic, endorsing all three alternatives. Q. There are different opinions about the nature of the state. To what extent would you agree with the following statements: It would be better to restore the Communist regime 41% A tough dictatorship is the only way out 36% The army should govern the country 15% Rejects all three alternatives 39% 0%

20%

40%

60%

80%

100%

Figure 12.2 Attitudes toward undemocratic alternatives Source: Centre for the Study of Public Policy, New Russia Barometer Social Capital Survey, 1998.

204

Richard Rose and Craig Weller

Measures of social capital Many studies of social capital depend on the secondary analysis of data collected for other purposes, a situation that encourages the uncritical acceptance of any available quantitative indicator as a proxy for individual social capital. A number of studies cite aggregate figures of organisational membership or any other group activity for which aggregate data are readily available. The association between democratic values and social capital is inferred from trends in national aggregate data or from crossnational aggregate correlations. It is not tested by the conventional social science method of designing and fielding a survey that provides individuallevel data that can be subject to multivariate analysis determining the extent to which social capital, as against other forms of capital, influences individual values. While Robert Putnam’s study of social capital in the United States (2000) cites an enormous variety of social capital indicators, he does not support his argument with systematic multivariate analysis of American survey data. The NRB questionnaire offers multiple measures of social capital in many forms, as well as measuring democratic and undemocratic values. It was developed in a seven-year process in which questionnaires were written in order to translate anecdotes about networking endemic in ethnographic writing about Russia into valid and quantified data about an individual’s involvement in different networks relevant to the concept of social capital (see www.socialcapital.strath.ac.uk). Instead of assuming a single meaning the questionnaire incorporated alternative approaches, thus making it possible to test the influence on democratic values of competing concepts of social capital. Trust in people is used typically as the prime social psychological indicator of social capital, and this is the case whether trust is conceived as a consequence or a cause of participation in social networks. The NRB Social Capital Survey repeated the World Values Survey measure of social trust: Would you say that most people can be trusted or you can’t be too careful in dealing with people? By defining the ‘radius of trust’ (Fukuyama 1995) as extending well beyond face-to-face ties, the question focuses on conditions essential for trust to ‘spill up’ to large-scale formal organisations necessary in a modern society. Given socialisation into a totalitarian or post-totalitarian environment, Russians tend to be distrustful of others. Only 7 per cent say one can usually trust most people and an additional 27 per cent think this is sometimes possible. Two-thirds say that one sometimes or usually needs to be careful when dealing with other people.3 Organisational involvement, as distinct from nominal membership, offers a behavioural approach to social capital networks. Participation in organisations with local branches and a national headquarters (e.g. a trade union or a church) links face-to-face groups with large, formal organisations stating national policies, a necessary condition for national, as dis-

Social capital and democratic values

205

tinct from ‘town meeting’ democracy. In the Soviet Union, the Communist Party controlled social, cultural and civic organisations directly or indirectly, and its totalitarian efforts to mobilise the population as compulsory members of these organisations generated widespread disaffection. When asked about membership in six different types of named organisations – a sport or recreation group; a musical, literary or art group; a community association; a political party; a housing association; a charitable organisation – only 9 per cent reported belonging to any type of organisation and more than nine-tenths of Russians belong to none. Many long-established sociological concerns with social integration or its opposite, anomie or social exclusion, are often embraced in the concept of social capital. Thus the NRB Social Capital questionnaire included a variety of familiar behavioural measures of involvement in informal networks such as groups of friends who would look after you when ill or from whom one could borrow money if in need, and major formal organisations such as churches and trade unions. In Russia church attendance is very low; only 5 per cent report going to church at least once a month and 11 per cent attend a few times a year. As for trade unions, upwards of twofifths of the electorate are outside the labour force. Half those in employment do belong to a union, but less than one in five union members trusts both local and national union leaders to look after their interests. James Coleman’s (1990: 302ff.) instrumental approach to social capital emphasises that social capital networks are situation-specific. The network most appropriate for one situation (e.g. getting a job) may not be the most appropriate for childcare or protection against crime. Trust in government is an example of situation-specific social capital. In post-communist countries it is far lower than trust in most people or people whom you know. In the NRB Social Capital Survey, only 7 per cent expressed positive trust in political parties, 13 per cent trust in representatives elected to the Duma (Parliament), 14 per cent trust in the president, and 18 per cent trust in local government. The Social Capital questionnaire therefore asked people what type of network they would turn to in order to get something done. The specifics varied from situation to situation, but the conceptual significance of alternative networks was normally the same. The options were: a formal, modern network (e.g. going to a private hospital if a public hospital was inadequate); informal networks (e.g. protecting against theft by making sure someone was always at home or keeping a fierce dog); anti-modern strategies (e.g. offering a bribe to get a municipally owned house or keeping a knife or gun to deal with burglars). If an individual has no network to turn to, this indicates social exclusion. Many theories of social capital treat it as a diffuse asset productive in many different situations. Therefore, the networks invoked in each of eight different situations have been summed for each individual. Russians do not rely on a single type of network in all situations. Informal networks of friends and neighbours are relied on most often. Given low incomes, the

206

Richard Rose and Craig Weller

least used alternative was buying goods and services in the market. Antimodern networks involving corruption or using connections to bend or break rules to get things done are second in frequency of use. The average Russian was very rarely socially excluded, that is, without any network to turn to (see Rose 1999: 154ff.). As would be expected, those who score high on social exclusion score low in their involvement in other networks,4 but there are no substantial correlations among the other three scales.5 The Social Capital Survey also asked about involvement in politically specific networks. Factor analysis showed that trust takes two forms: trust in authoritative institutions of government (police, courts, army), and trust in representative institutions (parties, president, Duma). Being an opinion leader when discussing politics with friends and trusting political information from friends are politically specific informal networks. Membership in the Communist Party and reliance on government programmes involve formal political organisations.

Hypotheses of the impact of social capital Before the term social capital erupted in political science discourse there was an established paradigm about the determinants of democratic values. There was disagreement about the relative importance of economic influences; education and other forms of human capital; and civic political attitudes. Nonetheless, there was a consensus that together these resources were sufficient to account for democratic values. Hypothesis 1: Established paradigm. The more human capital, economic resources and civic attitudes individuals have, the more they support democratic values. In the first flush of enthusiasm for social capital, established influences on democratic values were treated as secondary or dependent on social capital. This interpretation could be supported by showing simple correlations between individual or aggregate-level social capital indicators and democratic values or by excluding long-established influences from a multivariate statistical equation. Hypothesis 2: Social capital paradigm. The more social capital individuals have, the more they support democratic values. There is no necessary contradiction between the two paradigms; each can increase support for democratic values without supplanting the other. A

Social capital and democratic values

207

multi-causal approach avoids the false antithesis between the social capital and established paradigms. Hypothesis 3: Integrated paradigm. The more human capital, social capital, economic resources and civic attitudes individuals have, the more they support democratic values. The integrative paradigm leaves open to empirical investigation the relative importance of social capital as against other resources and the particular pathways whereby these influences interact with each other. Testing hypotheses The three hypotheses may be tested in turn by three sequential multiple regression analyses. Given a multiplicity of definitions and indicators, initially regressions were run with several dozen indicators. Those that failed to register statistical significance at the 0.05 level were then dropped, except for instances of theoretical relevance, for example, the failure of socio-economic class to influence democratic values. For the most part, the assignment of indicators to categories consistent with each hypothesis is straightforward. For example, education is a prime indicator of human capital and membership in the Communist Party of politically specific social capital. Given that Russians have had more experience with undemocratic than with democratic government, each hypothesis is tested with two dependent variables: endorsement of democracy as an ideal (Figure 12.1) and an addition scale showing the number of times individuals reject undemocratic alternatives (Figure 12.2). Following the literature, we would expect social capital to be particularly strong in influencing support for democracy as an ideal, even if it is not a major influence on the rejection of undemocratic alternatives. Hypothesis 1 receives support, since those who have more human capital, economic and civic resources are more likely to reject undemocratic forms of government. Altogether, the measures explain 22 per cent of the variance in the rejection of undemocratic regimes (Table 12.1). A negative view of the former Soviet regime (Beta: 0.24) and an appreciation of greater freedom under the new regime (0.13) make people more likely to reject undemocratic alternatives. Economic conditions are also important: those with a higher income and a more positive outlook about the country’s current economic system and their household’s future situation are more likely to reject undemocratic alternatives. Human capital is also of substantial importance. The more educated and younger people are, the more likely they are to reject undemocratic alternatives. Women are more likely than men to reject undemocratic alternatives.

208

Richard Rose and Craig Weller

Table 12.1 Impact of human capital, economic and civic resources Supports democracy as ideal

Rejects undemocratic alternatives

4.6% Beta*

22.0% Beta*

Civic political attitudes Negative rating of communist system Feels freer than in old system

11 8

24 13

Economic resources Household income Household economic situation in five years Positive current economic system Positive economic system in five years

–0 –0 –0 10

7 8 8 –0

Human capital Age Education Control over what happens to me Gender Socio-economic status

–0 –0 5 –0 –0


10 6 8 8 –0

Variance explained: adjusted R2

Source: New Russia Barometer Social Capital Survey (1998) Note *OLS regression beta values significant at  0.05.

The Soviet Union abolished class distinctions rooted in a capitalist society while substituting a politically based hierarchy of power and influence. It is therefore not surprising that a conventional 10-point scale measuring subjective socio-economic status is not a significant influence. In a society in transformation, Russians are not so much struggling against other classes as trying to cope with rapid and unpredictable change in a turbulent, even anomic society. The NRB survey therefore asks people to say how much or how little control they have over their own lives as against being dependent on fate and the actions of others. Those who feel most self-control have developed a distinctive form of human capital: confidence in their ability to overcome adversity; they are also more likely to reject undemocratic alternatives. The established paradigm accounts for only 4.6 per cent of the variance in attitudes toward democracy as an ideal form of government; only four of eleven indicators are statistically significant. A greater sense of freedom and a negative view of the communist regime are each significant, but the Beta statistics are less than for the rejection of undemocratic regimes. One economic indicator – a positive expectation of what the national economic system will be like in five years’ time – is significant, and one human capital measure, gender. The absence of any influence from most human

Social capital and democratic values

209

capital and economic indicators is consistent with the expectation that social capital is the key to making citizens more positive about democratic values. Social capital, however, has even less influence on commitment to the democratic ideal: altogether, fourteen different social capital indicators explain only 1.3 per cent of the variance (Table 12.2). None of the six politically specific measures of social capital is significant. Whether individuals trust authoritative institutions (police, courts, army) or elected representatives (Duma, president, parties) makes no difference, nor does relying on government to help with a problem or, an alternative form of organisational capital, having a family member in the Communist Party in Soviet times. Nor are informal networks significant, such as being an opinion leader or relying on friends for political information. Two measures of diffuse social capital are significant. Russians who rely on the market to solve everyday problems and on anti-modern methods, such as offering bribes or using connections, are more likely to support the democratic ideal. This is a comment on the perception of ‘Yeltsin-style democracy’ (cf. Carnaghan 2001). Only one of the four measures of traditional Table 12.2 Impact of social capital on democratic values Supports democracy as ideal

Rejects undemocratic alternatives

1.3% Beta*

5.5% Beta*

Social integration Trusts most people Village resident Belongs to organisations Church attendance

6 – – –

5
7 –0 –0

Diffuse social capital networks Anti-modern Market Informal Socially excluded

6 6 – –

–0 10 –0
10

Politically specific capital Communist Party member in family Trusts government authorities Trusts representatives Opinion leader Friends as source of information Relies on government when has a problem

– – – – – –


10
6 5
6 –0 –0

Variance explained: adjusted R2

Source: New Russia Barometer Social Capital Survey (1998) Note *OLS regression beta values significant at  0.05.

210

Richard Rose and Craig Weller

social integration, namely trusting most people, is positively significant for endorsing democracy as an ideal. Individuals with social capital are a little more likely to reject undemocratic alternatives. Altogether, eight indicators account for 5.5 per cent of the variance – but often in different ways than predicted by theories based on the role of social capital in established democracies. Persons who talk about politics with friends and trust them as sources of information are more likely to endorse undemocratic forms of government. The ambiguity of trust is illustrated by those trusting government authorities being more likely to endorse undemocratic alternatives, while Russians who trust elected representatives are more likely to reject undemocratic alternatives. Even though the Beta values for these two influences are almost the same, their size is not. An average of 25 per cent of Russians trust the army, courts and police, more than twice as many as the 11 per cent on average trusting Duma representatives, the president or parties. Having family ties to the old Communist Party is more likely to make people endorse undemocratic alternatives. Individuals most involved in market networks are more likely to support democratic rule, and persons who are socially excluded or live in villages are more likely to support undemocratic forms of government. The fall-back case for social capital, stated in Hypothesis 3, above, is that even if it is not the dominant influence it should contribute substantially to democratic ideals. But in Russia this is not so. When social capital influences are integrated with civic attitudes, economic resources and human capital, the variance explained in support of the democratic ideal remains exactly the same (4.6 per cent), as the established paradigm does without taking social capital into account.6 Of the few significant influences, the most important are civic attitudes about the former communist regime and about gains in freedom in the new Russian Federation. Economic evaluations are also significant. Net of the established paradigm’s influence on support for democracy as an ideal, only one of the fourteen indicators of social capital, namely trust in most people, manages to be statistically significant. Concurrently, two diffuse scales of social capital become insignificant when influences from the established paradigm are taken into account (cf. Tables 12.1, 12.2. and 12.3). Integrating social capital along with civic, economic and human capital adds 1.2 per cent to the 22 per cent of the variance that the established paradigm explains in the rejection of undemocratic alternatives (cf. Tables 12.2 and 12.3). Civic attitudes remain the most important: the Beta for an individual’s evaluation of the communist regime is 0.24, and appreciation of gains in freedom has a Beta of 0.13. Taken together, age, education and gender are of substantial influence, and control over one’s own life remains significant too, and the same economic resources are also significant. Taking into account influences in the established paradigm reduces the number of statistically significant social capital indicators from eight to six (cf. Tables 12.2 and 12.3).

Social capital and democratic values

211

Table 12.3 Integrating influences on democratic values Supports democracy as ideal Total variance explained: adjusted R2

Rejects undemocratic alternatives

4.6% b Beta*

23.2% b

Beta*

Established paradigm Civic attitudes Negative rating of communist system Feels freer than in old system Economic resources Income Household economic situation in five years Positive current economic system Positive economic system in five years

0.005 0.175

11 9

0.004 0.088

24 13


0.003 0.005

–0 –0
6 9

0.0007 0.078 0.001

8 9 7 –0

Human capital Age Education Gender Control over own life Socio-economic status

–0 –0 –0 –0 –0


0.005 0.026 0.153 0.032


10 8 9 8 –0

Social capital Social integration Trusts most people Church attendance Village resident Belongs to organisations

0.157

5 –0 –0 –0


0.005

–0
5 –0 –0

Diffuse social capital Anti-modern Informal Socially excluded Market

–0 –0 –0 –0


0.044
0.043
0.038


7
6
6 –0

Politically specific capital Communist Party member in family Trusts government authorities Trusts representatives Opinion leader Friends as source of information Relies on government when has a problem

–0 –0 –0 –0 –0 –0


0.151
0.011


8
5 –0 –0 –0 –0

Source: New Russia Barometer Social Capital Survey (1998) Note * b value is unstandardised regression coefficient. Beta is standardised coefficient.

212

Richard Rose and Craig Weller

The integrated model shows that in Russia neither trust nor organizational membership influences democratic values. When established influences are introduced, trust in other people is no longer statistically significant, and the same is true of trust in elected representatives.7 The sole measure that does remain significant is trust in authoritarian institutions – the army, courts and the police – that are far from the cooperative ideal envisioned in social capital theories. The integrated model also shows that membership in face-to-face and formal organisations is not a source of democratic values in Russia. At least to a limited degree, social capital networks actually increase support for undemocratic forms of governance, such as a member of one’s family belonging to the Communist Party of the Soviet Union. It is also striking that two measures of diffuse social capital – involvement in informal networks and reliance on anti-modern networks – are significant, but not in the expected direction. Net of all other influences, Russians who are involved in informal networks are actually more likely to support undemocratic alternatives than those who are not, and the same is true for those who are socially excluded. Altogether, the three regression models show most support for Hypothesis 1, above. The established civic, economic and human capital characteristics of individuals have a big influence on the rejection of undemocratic alternatives. Their influence on support for the democratic ideal, while low in absolute terms, nonetheless explains more than three times the variance explained by social capital measures. While social capital measures have some influence on the rejection of undemocratic values, the variance thus explained is less than a quarter that explained without reference to social capital. The cumulative impact We can calculate the impact of significant influences on democratic values by making use of the unstandardised regression coefficients (b) of each independent variable significant in Table 12.3. The b value shows how much change is likely to occur in an individual’s support for the democratic ideal or rejection of undemocratic values as the result of one unit of change in an independent variable. Independent variables differ in their metrics; for example, gender is a dichotomous variable, while trust in other people is a four-point scale and income a continuous variable. To compare the impact of different influences, we calculate how much democratic values are likely to change if a person’s position on a given independent variable alters by one standard deviation from its mean in the 1998 New Russia Barometer Survey. Civic variables have the greatest impact on support for democracy as an ideal (Table 12.4). If a person is one standard deviation above the mean in feeling freer than under communist rule and is similarly more negative

Social capital and democratic values

213

Table 12.4 Impact of significant influences on democratic values Democracy as an ideal Civic: Negative communist system Feels freer than in old system

Impact* 30 23

Mean 7.41 7.71 7.94

Economic:

Positive current economic system Positive economic system in five years

17 27

8.11 8.38

Social capital:

Trusts most people

13

8.51

Impact 21 11

Mean 2.11 2.32 2.43

Rejection of undemocratic alternatives Civic: Negative communist system Feels freer now than before Economic:

Income Household economic situation in five years Positive current economic system

Human capital: Younger Control over own life More educated Female Social capital:

Communist Party member in family Trusts government authorities Anti-modern networks Informal networks Socially excluded Church attendance

8 8 6

2.51 2.59 2.65

9 8 7 7

2.74 2.82 2.89 2.96


7
5
6
5
5
4

2.89 2.84 2.78 2.73 2.68 2.64

Note * Calculated by multiplying the unstandardised regression coefficient (b) by one standard deviation from its mean.

about that regime, he or she is likely to be half a point higher on the tenpoint scale of commitment to democracy as an ideal. Similarly, those above the mean in viewing the economic system positively and in optimism about the future are more than two-fifths of a point higher in endorsing democracy as an ideal. While trust in other people also has a positive impact, it changes attitudes by only 0.13 of a point. The very poor fit between social capital influences and endorsement of democracy as an ideal confirms its low influence. Many sources have some impact on the rejection of undemocratic alternatives. The biggest impact is again registered by civic attitudes: Russians one standard deviation above the mean in negative feeling about the old regime and in feeling freer now are almost one-third of a point closer to rejecting all three undemocratic alternatives. Human capital has almost as strong an impact. Russians who are younger, more educated, female and have a greater sense of control over their lives are an additional third of a point closer to rejecting all undemocratic alternatives. An increase in income and economic optimism moves Russians an additional fifth of a point closer to rejecting all three undemocratic alternatives.

214

Richard Rose and Craig Weller

While the cumulative impact of established influences implies that a person one standard deviation above the mean on all nine variables will almost certainly reject all undemocratic values, the probability of these attributes all being conjoined in the same person is very low, as those one standard deviation above the mean on any one measure are by definition a minority of the population. Nonetheless, many influences are likely to go together. For example, more educated people are likely to be younger, people with an above-average income are more likely to be positive about the current and future economic system, and those most negative about the communist system are likely to feel very strongly that they are freer today than before. Contrary to Hypothesis 2, above, Russians involved in social capital networks tend to be less supportive of democratic values. This is true of people who trust government authorities today and of people who were members of the Communist Party of the Soviet Union or had a family member in the Party. Church-goers are also more likely to support undemocratic alternatives. Contrary to Putnam’s theory, the spill-over and spill-up effects of being in both informal and anti-modern networks lead Russians to support undemocratic alternatives. In addition, exclusion from any social network encourages support for undemocratic alternatives. Although half a dozen social capital indicators have some impact on undemocratic alternatives, their total impact is much less than that of the established paradigm. Moreover, those involved in social capital networks, such as church-goers or Party members, are a small minority of the Russian population and far fewer than those feeling optimistic about future economic conditions or freer now than in communist times. Hence the cumulative impact of the established paradigm has more than two-anda-half times the impact on undemocratic values as does social capital.8 Taken at face value, the evidence implies, contrary to Tocqueville-inspired paradigms, that a decrease in social capital would increase support for democratic values. The Russian evidence is best interpreted as a caution that any positive merits in social capital are contingent. In Russia, people who trust authority are more likely to favour undemocratic alternatives, because it is communist-style rather than democratic institutions that they trust. Similarly, organisational membership is neutral. Furthermore, a link with the Communist Party does not encourage democratic commitment, and attendance at a church subservient to Stalin does not breed the commitment to liberal democratic values of belonging to a Quaker meeting.

Broader implications The strength of the established paradigm and the weakness of social capital in Russia goes so much against the grain of current writing that it could be argued that it is due to inadequate measurement of social capital.

Social capital and democratic values

215

If this were the case, then the social capital indicators in the NRB Survey would not add anything to the explanation of other phenomena, such as health (cf. Putnam 2000: section 4). But this is not the case. Similar regression analyses with the same set of data and different dependent variables show that social capital networks have as much influence on the physical and emotional health of Russians as do measures of human capital and income (Rose 2000b). They also have more influence on a sense of safety on the street; as much influence on income security; and a noteworthy influence on getting enough food (Rose 2000a). Moreover, these additional analyses show that anti-modern forms of social capital are more important than the friendly forms of social capital emphasised by Putnam. Since more than two dozen countries have been subject to generations of communist rule, the NRB approach is suitable for comparative analysis and the focus on using networks to compensate for deficiencies in the economy and/or the polity makes it potentially generalisable to developing countries across many continents. Analysis of data from a single country requires the caution that it may be different. Russia is certainly very different from the United States, for socialisation into a totalitarian or post-totalitarian regime has left Russians with a legacy of distrust very different from American predispositions to collective action. However, it may be argued that it is the United States, not Russia, that is exceptional – as many Americanists have done (cf. Hartz 1964; Shafer 1991). From a global perspective, the post-totalitarian legacy of Russians is shared with more than a billion people from East Berlin to Beijing. For example, the social capital networks that the Chinese use to get things done in the People’s Republic of China are similar to the reciprocal exploitation practices of the Soviet Union (cf. Shi 1997; Ledeneva 1998). A justifiable concern with putting limits to generalisation from Russian experience implies a more important proposition: context matters both for social capital and support for democratic values (cf. Newton and Norris 2000; Mishler and Rose 2001). In an undemocratic regime, a commitment to democratic values can lead to a rejection of existing political institutions. In a regime that is incompletely democratic the situation is more complex, depending on whether the missing element in the regime is accountability to the electorate, as in Singapore, or the rule of law, as in Russia (cf. Rose and Mishler 2002). The importance of civic attitudes is a reminder that politics matters, and politics changes with national context. The influences with the greatest impact on Russians are derived from living under very different regimes, a post-totalitarian party state and an electoralist democracy. The latter is judged in the light of the former. However, in an established democracy politics cannot reflect comparisons between regimes, and choices between parties and personalities are far less great than between democratic and undemocratic alternatives. The experience of regime change places Russia

216

Richard Rose and Craig Weller

in the mainstream of global politics, for today established democracies are the exception rather than the rule. Notwithstanding the above qualifications, the great majority of the significant influences on democratic values in Russia are common to social science paradigms generally. This is most obviously the case for education, age and gender. While the mean level of income differs between OECD countries and even more worldwide, every country has a greater or lesser degree of income inequality. Within-country differences in economic satisfaction and future economic expectations are also found everywhere. Moreover, the weak or non-existent influence of trust on democratic values is not restricted to Russia; it can be found across Europe too (Rose et al. 1998: Table 8.4; Newton 1999: 180ff.). The meteoric rise of social capital as an all-purpose explanation threatens a meteoric fall if the expectations raised are not qualified. Prudent proponents of the concept as well as sceptics have good reason to adopt the old Russian proverb that was a favourite of both Lenin and Ronald Reagan: ‘Trust but verify.’ The multi-causal integrative model set out in this chapter is a generally applicable method for determining empirically to what extent, under which circumstances, in which national contexts and to what ends social capital does or does not add to our understanding of individual and collective welfare in society.

Notes 1 This chapter has been written as part of a project financed by the Leverhulme Trust, London, on ‘Coping with Organizations: Networks of Social Capital’. The survey data were collected with help from a grant to the World Bank Social Capital Initiative by the Development Fund of the Danish government. The authors are solely responsible for the interpretation presented here. 2 In the Barometer Survey, VTsIOM, Russia’s oldest not-for-profit survey organisation, interviewed 1,908 Russian adults in 191 different primary sampling units representative of the population of the Russian Federation nationwide. Fieldwork took place between 6 March and 13 April 1998. For full details of questions and answers, including a report of the sample, see Rose 1998. 3 When the tenth NRB Survey asked separate questions about trust in ‘most people in this country’ and ‘most people you know’, 32 per cent said they trusted most people, while 55 per cent said they trusted most people they knew (Rose 2001: 36). This confirms that the measure of social trust used in this chapter concerns most Russians rather than people whom one knows. 4 The Pearson r correlation between social exclusion and market networks is
0.33; anti-modern networks,
0.43, and informal networks,
0.29. All three correlations are statistically significant at the 0.000 level. 5 Market networks correlate
0.16 with informal networks, and
0.05 with antimodern networks, and anti-modern networks correlate –0.0 with informal networks. Only the first of these three correlations is significant at the 0.000 level. 6 Given a degree of skewness in the distribution of endorsement for democracy as an ideal, a number of additional regressions were run with the dependent variable collapsed to reduce or eliminate this, and the independent variables were kept the same. This made no difference to the total proportion of variance

Social capital and democratic values

217

explained or to the classification of independent variables as insignificant or significant. 7 The significance level of trust in other people is 0.10 (Beta 0.03) and for trust in representatives it is 0.46 (Beta 0.02). 8 The mean value of the zero to three scale of rejection of undemocratic alternatives is raised by 0.85 points by established paradigm influences, while being lowered by only 0.32 points by social capital influences.

Bibliography Black, J. (1997) A Dictionary of Economics. Oxford: Oxford University Press. Boix, C. and Posner, D. N. (1998) ‘Social Capital: Explaining its Origins and Effects on Government Performance’, British Journal of Political Science 28, 4: 686–93. Bratton, M. and Mattes, R. (2001) ‘Africans’ Surprising Universalism’, Journal of Democracy 12, 1: 107–21. Carnaghan, E. (2001) ‘Thinking about Democracy: Interviews with Russian Citizens’, Slavic Review 60, 2: 336–66. Coleman, J. S. (1990) Foundations of Social Theory. Cambridge, MA: Harvard University Press. Dahl, R. A. (1971) Polyarchy: Participation and Opposition. New Haven, CT: Yale University Press. Dasgupta, P. (1988) ‘Trust as a Commodity’, in D. Gambetta (ed.) Trust: Making and Breaking Cooperative Relations. Oxford: Blackwell, pp. 49–72. Freedom House (2001) ‘The 2000 Freedom House Survey’, Journal of Democracy 12, 1: 87–92. Fukuyama, F. (1995) Trust: The Social Virtues and the Creation of Prosperity. New York: Free Press. Hartz, L. (1964) The Founding of New Societies. New York: Harcourt, Brace. Inglehart, R. (1997) Modernization and Postmodernization: Cultural, Economic and Political Change in 41 Societies. Princeton, NJ: Princeton University Press. Klingemann, H. D. (1999) ‘Mapping Political Support in the 1990s: A Global Analysis’, in P. Norris (ed.) Critical Citizens: Global Support for Democratic Governance. Oxford: Oxford University Press, pp. 78–99. Ledeneva, A. V. (1998) Russia’s Economy of Favours. Cambridge: Cambridge University Press. Mishler, W. and Rose, R. (2001) ‘Political Support for Incomplete Democracies: Realist vs. Idealist Theories and Measures’, International Political Science Review 22, 4: 303–20. Newton, K. (1999) ‘Social and Political Trust in Established Democracies’, in P. Norris (ed.) Critical Citizens: Global Support for Democratic Governance. Oxford: Oxford University Press, pp. 169–87. Newton, K. and Norris, P. (2000) ‘Confidence in Public Institutions: Faith, Culture or Performance?’, in S. Pharr and R. Putnam (eds) Disaffected Democracies: What’s Troubling the Trilateral Countries? Princeton, NJ: Princeton University Press, pp. 52–73. North, D. C. (1990) Institutions, Institutional Change and Economic Performance. New York: Cambridge University Press. Putnam, R. D. (1997) ‘Democracy in America at Century’s End’, in A. Hadenius

218

Richard Rose and Craig Weller

(ed.) Democracy’s Victory and Crisis. New York: Cambridge University Press, pp. 27–70. Putnam, R. D. (2000) Bowling Alone: The Collapse and Revival of American Community. New York: Simon & Schuster. Putnam, R. D. with Leonardi, R. and Nanetti, R. Y. (1993) Making Democracy Work. Princeton, NJ: Princeton University Press. Rose, R. (1998) Getting Things Done with Social Capital: New Russia Barometer VII, Glasgow: University of Strathclyde Studies in Public Policy No. 303. Rose, R. (1999) ‘Getting Things Done in an Anti-Modern Society: Social Capital Networks in Russia’, in P. Dasgupta and I. Serageldin (eds) Social Capital: A Multi-faceted Perspective. Washington, DC: The World Bank, pp. 147–71. Rose, R. (2000a) ‘Uses of Social Capital in Russia: Modern, Pre-Modern and AntiModern’, Post-Soviet Affairs 16, 1: 33–57. Rose, R. (2000b) ‘How Much Does Social Capital Add to Individual Health? A Survey Study of Russia’, Social Science and Medicine 51: 1421–35. Rose, R. (2001) Russians Under Putin: New Russia Barometer 10, Glasgow: University of Strathclyde Studies in Public Policy No. 350. Rose, R. and Mishler, W. (2002) ‘Comparing Regime Support in Non-Democratic and Democratic Countries’, Democratization 9, 2: 1–21. Rose, R. and Shin, D. C. (2001) ‘Democratization Backwards: The Problem of Third-Wave Democracies’, British Journal of Political Science 34, 1: 30–62. Rose, R., Mishler, W. and Haerpfer, C. (eds) (1998) Democracy and Its Alternatives: Understanding Post-Communist Societies. Oxford: Polity Press, and Baltimore, MD: Johns Hopkins University Press. Serageldin, I. and Steer, A. (eds) (1994) Making Development Sustainable: From Concepts to Action. Washington, DC: World Bank Environmentally Sustainable Development Occasional Paper 2. Shafer, B. (ed.) (1991) Is America Different? New York: Oxford University Press. Shi, T. (1997) Political Participation in Beijing. Cambridge, MA: Harvard University Press. Shin, D. C. (1999) Mass Politics and Culture in Democratizing Korea. New York: Cambridge University Press. Shlapentokh, V. (1989) Public and Private Life of the Soviet People. New York: Oxford University Press. Simon, J. (1998) ‘Popular Conceptions of Democracy in Postcommunist Europe’, in S. H. Barnes and J. Simon (eds) The Postcommunist Citizen. Budapest: Erasmus Foundation and Institute for Political Science, Hungarian Academy of Sciences, pp. 79–116. Weber, M. (1947) The Theory of Social and Economic Organization. Glencoe, IL: The Free Press.

13 Legacies and conflicts The challenges to social capital in the democratic transition Eric M. Uslaner and Gabriel Badescu

Almost fifteen years ago communism came tumbling down in most of Central and Eastern Europe. A new democratic age was upon us. As Francis Fukuyama (1992) celebrated in his famous essay, The End of History, Western civilization had triumphed over all types of authoritarian systems. Most notorious was the communist system that destroyed civil society in much of Europe (and elsewhere). The new democratic era would lead to a rebirth of civic engagement. Not only would the nations of Central and Eastern Europe become politically aligned with the West, there would also be a rapid social transformation of daily life. The West and the East would become one culture. The transition to democracy has not been so smooth. Citizens in states in transition did not rush out to join voluntary organizations. They did not become more trusting of each other. State repression ended, but the culture left by more than half a century of authoritarian government endured. People had been socialized not to trust their neighbors. They had few opportunities to participate in civic life. The only forms of participation permitted tended to be activities that reinforced, rather than challenged, the regime. People could join the Communist Party or unions, but civic engagement that was divorced from political authority or that could in any way threaten governmental authority was generally prohibited. There were, of course, exceptions, such as the Catholic Church in Poland. But even religious authorities in most countries were coopted by communist leaders (on Romania, see Stan and Turcescu 2000). The bumpy transition to democracy in Central and Eastern Europe challenged three of the core tenets of social capital theory. First, democratic regimes should stimulate participation (Vanhanen 1997). Second, democratic regimes should breed trust in others (Putnam 1993: 111–15; Offe 1997: 26; Levi 1998: 96; Rahn et al. 1999: 24; Stolle 1999: 9; Rothstein 2000). Third, civic engagement leads to more trust (Putnam 2000). Civic engagement is important for at least two reasons. First, an active citizenry makes it easier for the public to hold government accountable for its decisions. Authoritarian governments pay no attention to what the public wants. Second, participating in civic life can be transforming.

220

Eric M. Uslaner and Gabriel Badescu

Political theorists from John Stuart Mill to Alexis de Tocqueville have argued that participation in civic life makes one a better citizen. Tocqueville (1945: 108–09) offers the most famous statement on how socializing builds trust: Feelings and opinions are recruited, the heart is enlarged, and the human mind is developed only by the reciprocal influence of men upon one another . . . these influences are almost null in democratic countries; they must therefore be artificially created, and this can only be accomplished by associations. Putnam (1993: 90) writes: ‘Participation in civic organizations inculcates skills of cooperation as well as a sense of shared responsibility for collective endeavors.’ And Stolle elaborates: [M]embership in voluntary associations should increase face-to-face interactions between people and create a setting for the development of trust. . . . [T]he development of interpersonal trust and cooperative experiences between members tends to be generalized to the society as a whole. (cf. Levi 1998; Stolle 1998: 500) Trust, in turn, is important because it ‘lubricates cooperation’ (Putnam 1993: 170). Trust in people who are different from oneself is an essential building block of a well-functioning society, where people care about those who are less fortunate or who are minorities who have faced discrimination. Trusting societies have legal systems that are fair and they have lower crime rates. They spend more on education and on income transfers from the rich to the poor. Overall, they have stronger and more open economies and governments that function better (Putnam 1993: 111; Knack and Keefer 1997; LaPorta et al. 1997, 1998; Uslaner 2002: ch. 8). Not surprisingly, both civic engagement and trust were far lower in communist countries than in the West (Uslaner, Chapter 6, this volume). Fewer than twenty years ago, communist dictatorships were replaced by democratic institutions. However, the reluctance to trust others and to take an active part in one’s community did not give way to a new era of civic engagement and faith in others. Many new organizations were founded, but only a handful of activists participated in them. In many countries in Central and Eastern Europe there were also deepseated ethnic conflicts that had long been repressed by communist regimes. The fall of communism led to the rise of ethnic tensions that had long been bottled up by authoritarian regimes. Now, the nations making the transition to democracy were hardly distinctive in being marked by ethnic tensions. However, these ethnic tensions arose in a political vacuum. Communist states did not permit debate over macroeconomic

Challenges to social capital 221 policy, social policy, or even church–state relations. The new democratic regimes were chaotic – lots of political parties gaining tiny shares of the vote (Agh 1994). Under such democratic disarray, groups and parties that made the most noise had a better chance of success. And groups and parties that stirred ethnic tensions often made the most noise. When ethnic tensions are strong, trust is weak. If people trust only people like themselves (from their own group), they will fight attempts to redistribute a society’s resources to those who have less or have faced persistent discrimination. This ‘particularized’ trust (in one’s own kind) can make the transition to democracy more difficult. When particularized trusters form civic associations, they may band together to fight, or even oppress, people who are different from oneself. Such organizations may lead to less generalized trust, faith in people who are different from oneself. As group tensions rise, it may become more difficult to govern diverse societies. Particularized trusters will also fight to close markets, to treat immigrants differently, and essentially to close the borders to the outside world. Self-contained economies, however, cannot continue to grow (Woolcock 1998). Societies marked by particularized rather than generalized trust are doomed to low growth and corrupt governments. All democratic regimes depend upon an engaged and trusting citizenry, but new democracies may need it even more than established democratic regimes for two reasons. First, public support is essential in countries where authoritarian regimes have repressed (and alienated) their citizens. When things go bad, as they do some of the time everywhere, democratic regimes are buffeted by a reservoir of public support. Formerly authoritarian states do not have that automatic safety-valve of goodwill. Second, the problems of countries in transition are more severe than those of longer standing democracies. Trust can make a bigger difference in political and economic outcomes where the need is greater. Yet, where we need social capital more, we find it in shorter supply. We review the evidence presented in the chapters in this book to assess the state of social capital in the formerly communist countries in Europe.

Whom do you trust? Weak ties, relations between people who do not know each other well, are essential to civil society, Gibson (Chapter 5, this volume) argues. Yet, across most formerly communist countries, there is a strong reluctance to put faith in people whom you do not know. Under communism, it was dangerous to place confidence in strangers. They could be agents of the state and trust could lead to betrayal. In democratic societies, the stakes are much smaller. Strangers generally cannot get you into trouble with the law. People living under communism compartmentalized their lives into small social networks made up of people whom they know well. Most

222

Eric M. Uslaner and Gabriel Badescu

people had two types of social networks: their very close confidants and acquaintances who helped them cope with the difficulties of daily life under communism. The more expansive type of social network found in voluntary associations simply did not exist under communism. People felt safe dealing with their small, intimate group of close friends and family members. They were free to speak their mind to a handful of people (Badescu, Chapter 8; Gibson, Chapter 5; Howard, Chapter 10; Bartkowski, Chapter 11; and Flap and Völker, Chapter 3, this volume). But these small, closed circles provided the social support needed to cope with a repressive state. They were anything but public. New acquaintances could not simply join in. People selected their ‘strong ties’ (their closest confidants) very carefully (Flap and Völker, Chapter 3, this volume). Beyond these small private circles of friendship people had second, and often very different, networks. Flap and Völker call them ‘provision networks.’ Each country had its own word for this ‘economy of favors’ (Bartkowski, Chapter 11, this volume; Ledeneva 1998): blat in Russian, relat¸ii in Romanian.1 Daily life was marked by shortages and rationing, as well as difficult dealings with bureaucrats. Because goods were often in short supply, shopping was an onerous burden, so people took turns waiting in different lines. These ‘provision networks’ were much wider than the small, tight networks of one’s closest friends. Neighbors would help each other out. The economy of favors is clearly an example of reciprocal assistance: You help me, I’ll help you. It is essential in a highly bureaucratic society with shortages and rations. But it is a poor substitute for either close ties or the generalized trust that we find in societies with a greater sense of social cohesion. Provision networks had few emotional bonds. They were connections of convenience. Trust was easy to establish in these networks. This is experienced-based trust. It is based upon mutual advantage, not on any deep commitment to other people. Generalized trust, on the other hand, has moral roots. Putnam (2000: 21) equates trust with ‘generalized reciprocity,’ where people do favors for others based upon an expectation that others will come to their aid if necessary. You don’t need to tally up when Ivan helped you get potatoes and when Ivana helped you get tomatoes and what you did for Elena in return. ‘If we don’t have to balance every exchange instantly, we can get a lot more accomplished,’ Putnam (2000: 21) argued. But some sort of tally sheets do seem in order in societies marked by persistent shortages and rations. It is easy for others to exploit you if you fail to keep accurate records – especially if there are no strong bonds between members of provision networks. Uslaner’s (2002: ch. 2) more expansive view of generalized trust is even more of an anathema to a world of shortages, rations, and provision networks. His argument is that generalized trust is a moral sentiment that is not based upon experience. Instead, it reflects a moral commitment to treat others as you would have them treat you (as in the Golden Rule),

Challenges to social capital 223 regardless of how they actually treat you. This is a very risky strategy in a society where social bonds are so weak. It is also hazardous in societies marked by shortages and rations. The fundamental bases of generalized trust are optimism and control, the beliefs that the world is a good place, that it is going to get better, and that you can make it better (Uslaner 2002: chs 2, 4). In a world of shortages and rationing, it would be foolhardy to be too optimistic. There is little evidence that the world is a good place and rationing is prima facie evidence that you have little power to shape your own fate (see Uslaner, Chapter 6, this volume). It is hardly surprising that neither close personal friendships nor the provision networks are related to generalized trust. Trust in your closest associates is, not surprisingly, very high, while overall generalized trust is low. Crucially, trust in one’s close friends and family does not translate into a more generalized sense of trust in strangers (see Badescu, Chapter 8, this volume on Romania; and Gibson, Chapter 5, this volume on Russia). This may hardly seem surprising since life under communism was based on very small, intimate groups of friends, and there was little reason to trust strangers (Flap and Völker, Chapter 3, this volume; Uslaner 2002: ch. 8). The missing link between trust in people whom we know well and trust in strangers is not a peculiar feature of communist or post-communist societies. The two types of trust are fundamentally different. Trust in people whom you know is based upon experience, while trust in strangers cannot be based upon experience. Thus more generally there is no connection between the two types of trust, neither in Russia, Romania, nor the United States (Uslaner 2002: chs 2 and 5). Generalized trust is much lower in formerly communist states, and it has not increased during the transition to democracy. Muller and Seligson (1994) argued that democracy breeds trust, but the experience of the states of Central and Eastern Europe gives little support to this claim. The correlations between democratization and social trust in formerly communist states are minuscule. Some are even negative. Of the eight Eastern bloc countries for which we have measurements on trust in 1990 and 1995 to 1996, only one nation had an increase in faith in others (Latvia, 6 percent), while seven had decreases, four of which were substantial.2 The constitutions of the formerly communist nations have become increasingly democratic over time, as measured by Freedom House indices.3 As democratization proceeded apace, trust lagged behind and the correlation between trust and democratization became increasingly (and significantly) negative.4 Eastern bloc countries with more trusting citizenries did not become democratic sooner than nations whose populations had less faith in others. Formerly communist countries with higher levels of trust did not create polities with more political or property rights. There is even a perverse positive relationship between trust and corruption in these nations as late as 1998.5 The citizens of Central and Eastern European states yearned for a

224

Eric M. Uslaner and Gabriel Badescu

quick fix to the problems of communism. Democratic institutions would lead to greater social capital – more trust in others, trust in government, and greater civic engagement. Levi (1999: 182) argues that ‘democratic states may be even better at producing generalized trust than are nondemocratic institutions . . . because they are better at restricting the use of coercion to tasks that enhance rather than undermine trust.’ Democratic institutions would lead to trust and people who did not have the opportunity to participate in governing themselves would rush out to take part in civic life. We often put democracy on a pedestal and expect it to perform political miracles. Social science has played a role in inflating these expectations. As Mueller (1999: 199) argued: Democracy does indeed correlate with various social and economic characteristics – wealth, capitalism, literacy, and so forth – but these correlations are . . . essentially spurious. . . . The McDonald hamburger sold first, and continues to sell best, in rich capitalistic, Protestant countries, but it doesn’t follow that you have to be rich or capitalistic or Protestant or well-prepared or sophisticated or middle class or industrialized or cosmopolitan or uncontentious to buy one. So the presumption that democratizing will increase all sorts of capital, both economic and social, was based upon fallacious reasoning. Changing one set of institutions for another is not sufficient to reshape a country’s political culture. An Indian journalist commented on the sharp cleavages that led to a cycle of unstable coalitions, none of which could form a government: ‘We have the hardware of democracy, but not the software, and that can’t be borrowed or mimicked’ (Constable 1999: A19). Trust in others and tolerance of people from different backgrounds are not products of governmental institutions. Yes, some governmental policies can lead to greater trust – particularly policies that reduce income disparities and open up markets. But we have put far too much emphasis on structures and not enough on culture. In addition, more than half a century of communism has torn the optimism and sense of personal control from the citizenry, making trust, tolerance, and civic engagement scarce. There was no easy answer to the problems inflicted by communism. Democracy seemed to make things worse at times. It is a messy system of governing. Mueller (1999: 9) called it ‘undisciplined, chaotic, and [marked by the] essentially unequal interplay of “special interests”.’ The German leader Otto von Bismarck said that ‘if you like sausages and legislation, you should never watch either being made.’ The more American citizens know about their legislature, the less faith they have in it (Hibbing and Thiess-Morse 1995). The idealized view that people, even citizens with long democratic traditions, have of democratic government clashes with the reality of the contentious, often difficult problems of self-governance.

Challenges to social capital 225 Democracy raises expectations, but is insufficient to reshape a society, at least in the short term. Giving formerly communist countries parliamentary institutions will not make them little Swedens, as foretold in the Yiddish expression, ‘If my grandmother had wheels, she’d be a bus.’ The disorder inherent in democracy is particularly problematic in formerly communist countries. Communism was anything but disorderly. Daily life was regimented and all forms of participation were controlled by the state. Adjusting to a more open, often anarchic, political life was not easy. In the World Values Survey of 1995 to 1997, 71 percent of people in formerly communist countries said that maintaining order was a very important value, compared to 54 percent in the West. Forty-six percent of respondents in formerly communist states said that democracy did not have enough order, compared to 26 percent in the West. The disorder of democratic government may mean that no clear policy emerges among the din of voices. Or it may mean that the most popular policy is adopted, rather than the one that is best for a particular policy problem. The transition to democratic governance has gone hand-in-hand with the transition to a market economy. The enforced equality under communism gave way to greater inequality, and inequality leads to less generalized trust (Uslaner 2002: chs 2, 6, and 8). A growing income gap leads to greater pessimism for the future, especially among those who have become worse off under democratic capitalism. And pessimism also leads to distrust (Uslaner, Chapter 6, this volume; 2002: chs 2, 4, 6, and 8). The democratic transition seemed doomed to subvert a rise in social capital.

Associating with whom? The democratic transition also unleashed ethnic tensions that had been kept under wraps by totalitarian governments. Civil war broke out in many parts of the former Yugoslavia; ethnic tensions led to separatist movements in many formerly communist states, from Russia to what used to be Czechoslovakia. In Romania, 67 percent of respondents to a 2001 survey conducted by Badescu and Uslaner said that they had little or very little confidence in people from different ethnic backgrounds to themselves; and 61 percent had little confidence in people from different religious backgrounds. In Moldova, our companion survey found that 76 percent had little faith in people with either a different ethnic or religious heritage. In Romania, only 25 percent of the population are fully trusting (by religion and ethnicity); in Moldova, it is just 13 percent. In most of the formerly communist countries represented in this volume, particularized trust seems more common than generalized trust. People trust others like themselves, rather than people in general. Since social networks centered around family and a few close friends, it is hardly surprising that people would distrust people who were not very much like

226

Eric M. Uslaner and Gabriel Badescu

themselves. In Russia (Gibson, Chapter 5, this volume), Serbia (Iglicˇ, Chapter 2, this volume) and Romania (Badescu, Chapter 8, this volume; Mondak and Gearing, Chapter 9, this volume), people associate primarily with people of the same background. Russian social ties are also highly oligarchical, making it difficult for people to reach out to people who are different from themselves (Hayoz and Sergeyev, Chapter 4, this volume). The nature of social ties in the formerly communist states constrains civic engagement and civil society in several important ways. First, the limited social ties people have means that civic engagement is very low. People join few voluntary associations (Howard, Bartkowski, and Uslaner, all in this volume). Membership of unions was one of the only outlets for what Western scholars measure as ‘civic engagement.’ But unions were hardly voluntary organizations in the Western sense. Unions served the regime, and membership was generally mandatory and the source of social benefits and vacation packages (Howard 2002: ch. 4). Only in Poland was there an independent union, Solidarity. After the fall of communism, union membership has fallen. Participation in religious organizations was always important in Poland and it remains so, and people now also join professional associations (Bartkowski, Chapter 11, this volume). However, there is little evidence of more extensive civic engagement in Poland, and not much elsewhere in other former communist states. Second, there does seem to be an increase in civic engagement in political life in several countries. However, political activity is dominated by the majority ethnic group and it is not geared toward increasing social trust. In all of the formerly communist countries except for Russia, majority ethnic groups are far more participatory than minority groups (Dowley and Silver, Chapter 7, this volume). And political participation is often highly contentious, exacerbating societal tensions rather than relieving them (Iglicˇ, Chapter 2, this volume). Civic engagement may create a less civil, and less trusting, society. Minority groups are marginalized, both by their lower rates of participation and by the further tensions that arise from political movements that threaten their place in society. Participation becomes the property of a few, mostly belonging to the majority ethnic group, and hardly aimed at resolving the key problems of the transition to democracy. Third, people substitute the close ties with a small number of people for membership in voluntary associations (Howard, Chapter 10, this volume). Fourth, many people simply fail to see the point of joining groups. There is a widespread, and continuing, mistrust of communist organizations and a doubt that other organizations are much different. Disappointment with the workings of democracy is widespread. More than half of the respondents to the 1995 to 1997 World Values Survey from the formerly communist states (as opposed to 42 percent in the West) had a negative view of their political system. People who distrust communist organizations and

Challenges to social capital 227 who are dissatisfied with their current government are less likely to become civically engaged. Gibson sees the plentiful but small groups that dominate social life in Russia as enhancing the prospects for democratization. Howard (Chapter 10, this volume), and Rose and Weller (Chapter 12, this volume) are less convinced that the sorts of social network which the Russians develop will lead to a greater acceptance of democracy. Both Gibson and Rose and Weller find minuscule relationships between generalized trust and support for democratic government (see also Uslaner, Chapter 6, this volume). Howard finds Russia still mired in the political culture of mistrust of authority and particularized trust of others that marked it during the communist era. He finds that distrust of authority leads people away from organizational membership; Rose and Weller find that even those who do join civic groups are no longer supportive of democratic values. Much of the disagreement may stem from how we conceptualize strong and weak ties. Howard and Rose and Weller, on the one hand, treat both family and friends as strong ties. Gibson considers only family to reflect strong ties and sees friendship networks as weaker links.

Conclusion By one account, the transition to democracy is complete. As Mueller (1999: 189–91) argues: [m]ost of the postcommunist countries of central and eastern Europe as well as many of the new democracies elsewhere have essentially completed their transition to democracy; they are already full-fledged democracies . . . what they have now is, pretty much, it. It will, in all likelihood, never get much better . . . like the older democracies that are their model, the new democracies are unlikely ever to achieve orderly deliberation, political equality, or wide and enlightened participation by the mass of the public. Mueller is right that we may demand too much of democracy as a political system. Governance is messy and a free society is more disorderly than a totalitarian one. Yet there is more to governance than the right to complain without being harassed. Democratic government ought to be responsive government and if no one is listening, then the right to complain does not amount to much. And good government, whatever the institutional structure, has to be more than legalized anarchy, with thousands of voices shouting and no one trying to impose any sort of order. Communism was all about order, but order itself is not bad; it depends upon where it comes from. Order dictated from above is unhealthy. Order as negotiated voluntarily among competing groups is not totalitarianism; it is governing. And

228

Eric M. Uslaner and Gabriel Badescu

Western societies marked by trust seem to govern better than less trusting nations (LaPorta et al. 1997, 1998; Uslaner 2002: ch. 8), especially formerly communist states. The question is not so much whether the transition brings democratic institutions. Mueller is right: that job is largely done. The velvet revolutions made democracy. Now, can we (in Putnam’s felicitous words) make democracy work? Whether you are an optimist, like Gibson, or more reserved in expectations for the future, there is little doubt that social capital is considerably lower in the formerly communist states than in the West. People don’t trust strangers – and there is much less trust in anyone of a different background. Communism taught people to distrust each other. It did not permit most forms of civic engagement. The bad news is that the legacy of communism persists. A decade and a half after the fall of communism, there is still widespread distrust and civic disengagement. As long as the economic performance lags and inequality grows, there may not be a sufficient foundation for generalized trust to grow. Our surveys of Romania and Moldova also do not give us grounds for optimism. In 2001 we asked respondents a range of questions including generalized trust, trust in people of different ethnic groups and different religions, whether people believe that they would be treated equally to others under the law, whether life would be better for the next generation, how satisfied they are with their lives, how many group memberships people have, and whether they trust a range of governmental institutions (Parliament, the courts, the police, politicians, and the European Union). If democratization is enhancing social capital, we would expect that the younger generation, whose socialization was less shaped by the communist regime, would be more trusting – both of other people and of governmental institutions – and participatory and would be more optimistic for the future. We divided the respondents into three cohorts: pre-communist (born before 1945), communist (born between 1945 and 1974), and postcommunist (born since 1975). We present the results in Table 13.1. There is little to cheer about in this table. In Romania, the statistically significant generational differences almost always indicate that the precommunist generation is the most trusting, the most tolerant, and the most optimistic (see also Badescu 2001: 182). The younger respondents are the least trusting and the least tolerant. We expected that they would be more optimistic for the future – yet they are not. And we expected that they might look Westward to the European Union – yet they do not. The postcommunist generation is more satisfied with life overall, but the roots of their satisfaction are unclear. They have less faith in others and in their political system, and they are less tolerant of people who are different from themselves. They are no more optimistic for the future than are other Romanians. They do participate more in civic organizations than the pre-communist generation, but less so than the communist generation.

0.398*** 2.171*** 2.271*** 2.234 2.955 5.129 0.221*** 0.143*** 5.127*** 4.148*** 4.299 5.592*** 3.398*** 5.653

Notes  Range from 0 to 10. *** p  0.01; ** p  0.05; * p  0.10.

0.298 2.065 2.115 2.209 3.064 5.249 0.373 0.248 4.572 3.936 4.321 5.202 3.185 5.593

0.270 1.931 2.555 2.159 2.950 5.678 0.243 0.164 3.789 3.225 3.966 4.605 2.510 5.074

0.239 2.063 2.017 2.307** 2.776 4.058*** 0.163*** 0.108*** 3.956** 3.507* 3.410 3.515 2.730 4.437***

Pre-communist

Post-communist

Pre-communist

Communist

Moldova

Romania

Source: CID Survey of Romanian and Moldovan publics, 2001

Generalized trust (0–1) Trust different ethnic groups (1–4) Trust different religion (1–4) Equal treatment under law (1–3) Life better for next generation (1–5) Life satisfaction (1–10) Memberships of groups (0–10) Member of any group (0–1) Trust cabinet Trust Parliament Trust courts Trust police Trust politicians Trust European Union

Variable (range)

0.217 2.033 1.959 2.201 2.750 4.768 0.437 0.221 3.639 3.291 3.462 3.358 2.660 4.811

Communist

0.164 1.897 1.937 2.176 2.730 5.428 0.464 0.265 3.281 2.977 3.491 3.182 2.605 5.291

Post-communist

Table 13.1 Generational differences in social capital in Romania and Moldova from CID surveys of Romania and Moldova

230

Eric M. Uslaner and Gabriel Badescu

Most of the younger people’s participation is in youth groups – and the ‘big’ advantage for the communist generation comes from membership in unions (9.7 percent compared to 1.8 percent for the other two generations). The pattern is largely similar for the Moldovan sample, though the differences across generations are often not significant. Younger Moldovans are not significantly less trusting of other people, different groups, or government (though the averages are always lower than the other cohorts). Like younger Romanians, they are more satisfied with their lives, but they are no more optimistic for the future and they are less likely to believe that they would receive equal treatment under the law. They are more participatory, especially in sports, youth, and hobby groups, and they have not withdrawn from unions to the extent of their Romanian counterparts. They are considerably less likely to participate in political groups (1.3 percent) compared to the communist cohort (3.8 percent). In Romania, there are no significant differences in membership in political groups and indeed the youngest cohort is slightly more participatory. Yet in both Romania and Moldova membership of a political association leads to small (if any) increments in trust in governmental institutions. There is some good news from our research as well. What separates East and West is experience, not psychology. The roots of trust and membership in voluntary organizations are similar in the formerly communist nations and the West – a sense of optimism for the future and control over one’s own life (Uslaner, Chapter 6, this volume). What distinguishes East and West is the experience people have had under communism, and thereafter (Howard, Chapter 10, this volume). Two formerly communist states have made some strides toward greater civic engagement. Small, closed social networks have become less important in the former East Germany (Howard, Chapter 10, this volume). On a more pessimistic note, these personalistic ties have not been replaced by wider participation in voluntary associations. The only ‘major success’ is in the wealthiest former communist state, Slovenia, where social networks are as expansive as those in the West (Iglicˇ, Chapter 2, this volume). Perhaps Slovenia offers more hope than is warranted. No other formerly communist nation is as well-off as Slovenia. But Poland and the Czech Republic are within reach of wealthier, more equal societies and each had a more engaged citizenry, at least during some periods, before the communist governments fell. The task ahead will be difficult. Ethnic divisions will remain, though perhaps reduced if the economies become stronger. A tolerant and engaged citizenry is not just around the corner, but it is not out of sight.

Challenges to social capital 231

Notes 1 This economy of favors was not restricted to Central and Eastern Europe (including Russia) or even to communist countries. The practice of guanxi in China is well established. In Israel, the same set of mutual assistance is called protektsia. 2 The rise in Latvia was 32 percent. The drops are: Estonia (6 percent, 21.4 percent of 1990 base), Lithuania (9 percent, 29 percent of the base), Russia (14 percent, 37 percent of the base), and Poland (17 percent, 49 percent of the base). Belarus, East Germany, and Slovenia each had a drop of 1 percent (4, 4, and 6 percent of the 1990 bases). The 1996 measure on Hungary was not available, but there is a 1981 survey and it has more of the same bad news: trust fell 8 percent from 1981 to 1990, or one-third of the base. 3 The mean value on the freedom index ranges from 2 (completely free) to 14 (not at all free) and it moved from 12.167 in 1978 to 11.056 a decade later to 7.222 in 1993–94 and to 6.047 in 1998–99. By 1998–99, 44 percent of the nations of Eastern and Central Europe had scores of 3 or lower. Just 50 percent of countries with no legacy of communism had scores of 3 or lower. Belgium, Italy, Japan, Northern Ireland, South Africa, Spain, Uruguay, the United Kingdom, and the former West Germany each had scores of 3, as did the Czech Republic, the former East Germany, Estonia, Hungary, Latvia, Lithuania, Poland, and Slovenia. 4 The correlation is
0.466 in 1998–99, p 0 .02, one-tailed, p  0.04, two-tailed. 5 The correlation is 0.274. The measures of political and property rights come from LaPorta et al. (1997). The correlation with corruption (the Transparency International Index for 1998) is reflected since higher scores on the corruption perceptions index indicate more honesty. The trust measure used is not an average, but rather the most recent available figure for social trust. I am grateful to Ronald Inglehart for providing some available data from the third wave of the World Values Survey.

Bibliography Agh, A. (1994) ‘The Hungarian Party System and Party Theory in the Transition of Central Europe’, Journal of Theoretical Politics 6: 217–38. Badescu, G. (2001) Participare politica si capital social. Cluj-Napoca: Accent Constable, P. (1999) ‘India’s Democracy In Uncertain Health,’ Washington Post (April 21): A17, A19. Fukuyama, F. (1992) The End of History and the Last Man. New York: Free Press. Hibbing, J. and Thiess-Morse, E. (1995) Congress as Public Enemy. New York: Cambridge University Press. Howard, M. M. (2002) The Weakness of Civil Society in Post-Communist Europe. New York: Cambridge University Press. Knack, S. and Keefer, P. (1997) ‘Does Social Capital Have An Economic Payoff? A Cross-Country Investigation,’ Quarterly Journal of Economics 112: 1251–88. LaPorta, R. et al. (1997) ‘Trust in Large Organizations,’ American Economic Review Papers and Proceedings 87: 333–8. Ledeneva, E. (1998) Russia’s Economy of Favours. Cambridge: Cambridge University Press. —— (1998) ‘The Quality of Government’, unpublished manuscript, Harvard University.

232

Eric M. Uslaner and Gabriel Badescu

Levi, M. (1998) ‘A State of Trust’, in M. Levi and V. Braithwaite (eds) Trust and Governance. New York: Cambridge University Press. —— (1999) ‘When Good Defenses Make Good Neighbors: A Transaction Cost Approach to Trust and Distrust’, Russell Sage Foundation Working Paper No. 140. Mueller, J. (1999) Democracy, Capitalism, and Ralph’s Pretty Good Grocery. Princeton, NJ: Princeton University Press. Muller, E. N. and Seligson, M. A. (1994) ‘Civic Culture and Democracy: The Question of Causal Relationships’, American Political Science Review 88: 635–52. Offe, C. (1997) ‘How Can We Trust Our Fellow Citizens?’, presented at the John F. Kennedy School of Government Conference on Confidence in Democratic Institutions, August, Washington, DC. Putnam, R. D. (1993) Making Democracy Work: Civic Traditions in Modern Italy. Princeton, NJ: Princeton University Press. —— (2000) Bowling Alone. New York: Simon & Schuster. Rahn, W., Brehm, J. and Carlson, N. (1999) ‘National Elections as Institutions for Generating Social Capital’, in T. Skocpol and M. Fiorina (eds) Civic Engagement in American Democracy. Washington, DC: Brookings Institute. Rothstein, B. (2000) ‘Trust, Social Dilemmas, and Collective Memories: On the Rise and Decline of the Swedish Model’, Journal of Theoretical Politics 12: 477–501. Stan, L. and Turcescu, L. (2000) ‘The Romanian Orthodox Church and Postcommunist Democratisation’, Europe-Asia Studies 52: 1467–88. Stolle, D. (1998) ‘Bowling Together, Bowling Alone: The Development of Generalized Trust in Voluntary Associations’, Political Psychology 19: 497–526. —— (1999) ‘Communities, Citizens, and Local Government: Generalized Trust and the Impact of Regional Factors’, presented at the Annual Meeting of the American Political Science Association, Atlanta, September. Tocqueville, A. (1945) Democracy in America, vol. 2. New York: Alfred A. Knopf. First published 1940. Uslaner, E. (2002) The Moral Foundations of Trust. New York: Cambridge University Press. Vanhanen, T. (1997) Prospects of Democracy. New York: Routledge. Woolcock, M. (1998) ‘Social Capital and Economic Development: Toward a Theoretical Synthesis and Policy Framework’, Theory and Society 27: 151–208.

Index

Aberg, Martin 46 age: and generational differences in social capital, in Romania and Moldova 228–30; and organizational participation, in Russia and Germany 170, 171, 172, 177, 178; and social capital in Russia 207, 208, 211, 213, 214, 216 Albania 124 Almond, G. A. and Verba, S., The Civic Culture 2 Ash, Timothy Garton 123 Banfield, Edward 83–4 Bernhard, M. 63 Bismarck, Otto von 224 Bosnian Serbs 110 Brehm, J. 83 Bulgaria: civil society 124; ethnic tensions 97; organizational participation in 169; social networks 67, 68, 69 Burt, R. S. 19 Calhoun, C. J. 10–11 Canada 96 CBOS survey, on social capital in Poland 185–9, 190, 192–3, 194–5 Ceausescu, Nicholae 144 Chechnya 96, 97 China, social capital networks 215 Churchill, Winston 202 citizen activism, and undemocratic outcomes 4–5 city size, and organizational participation in Russia and Germany 170, 171, 172, 173, 178 civic attitudes, and social capital in Russia 210, 211, 212–13

civic engagement 5, 6, 81–2, 81–93, 89, 140–62; and democratization 95, 140–1, 219, 224, 226–7; determinants and effects of 6–7; and generalized trust 86–7, 89–92, 221; generational differences, in Romania and Moldova 228–30; importance of 219–20; increase in 226–7; and the legacy of communism 219, 220, 224, 228, 230; measuring 127–8; in Poland 7, 190–5, 226; and political participation 141; in Romania and the United States, comparing two cities 140, 142–60; and social capital 47; and social trust 121–3, 125, 136–7; see also voluntary organization involvement civil liberties, Freedom House rating of 97–8, 100 civil servants, attitudes to in Poland 187, 188, 189 civil society 61–78; and communism 219; defining and operationalizing 166–7; and democracy 2, 61, 95, 166–7 and voluntary associations 123–5 increasing atomization of 156; and interpersonal trust 61–2; in Poland 124, 184–5, 198; and political mobilization in Yugoslavia 12; in Russia 48, 63–78; and social networks 62; weakness of post-communist 165, 167–9, 180; see also voluntary organization involvement clientelist networks, in Russia 50–1, 56 Cluj-Napoca, Romania 140, 143–6; ethnicity and ethnic tension 143, 144–5, 146, 148, 156–7; housing 145, 157; local news media 145–6, 149; and political discussion 140, 144, 147–8,

234

Index

Cluj-Napoca continued 149, 156; and political evaluation 154–5, 156; and political participation 140, 148–50, 156; reasons for lack of civic engagement in 156–7; sampling information and survey items 157–60; size 143; social and economic perception 140, 150–4 Coleman, James 4, 47, 205 Committee for the Protection of Human Rights 14 communist networks, East Germany 28–32 communist organizations, mistrust of 173–4, 177–9 Communist Party membership, and social capital in Russia 207, 209, 210, 211, 213, 214 communitarian social capital 49 communities, local communities and community-oriented activity in Poland 193–5 community mobilization 15–16, 18 corruption: in Poland 187–8; and trust 223 cultural variables, and democratization 2–3 Czech Republic 124–5, 230 Czechs, and political interest 105–9 Dahl, Robert 202 democracy: and civil society 2, 61, 95, 166–7 and voluntary organizations 123–5 disorderly nature of 221, 224–5; making it work 227–30; measuring democratic values 202–3; and organizational participation 95–6; social capital and support for 102–16, 117 in Poland 187 in Russia 105, 110–12, 200–16 and trust 87–8, 91, 93, 122, 224, 228 democratic attitudes, and Russian social networks 74–6 democratization: and civic engagement 95, 140–1, 219, 224, 226–7; in former communist countries 1–2; and generalized trust 120, 223, 227; and social capital 3–4, 97–102, 120 disappointment, post-communist, in East Germany and Russia 176–7, 178, 179

disjointed political identities, in Yugoslavia 15, 16 East Germany 5, 28–44; communist networks 28–32; democratization process in 1; and the fall of the Berlin Wall 28; networks after the political turn 32–42, 230 heterogeneity and density 34, 38–9, 42 multiplexity 36–8, 42, 43 niche networks 29–30, 31, 34, 36–9, 40–2 provision networks 30, 34, 35–6, 38–40, 41 tie strength 34, 36–8, 39, 41, 42, 43 organizational participation in 43–4, 165–6 changes in membership levels 169 post-communist factors in 173–9 and socio-economic status 170–3, 177, 178 economic inequality, and trust 6, 85–6, 93, 225, 228 economic perception, in Cluj-Napoca, Romania and South Bend, Indiana 150–4 economic resources, and social capital in Russia 207, 208, 210, 211 education levels: and organizational participation in Russia and Germany 170, 171, 172, 177, 178; and social capital in Russia 207, 208, 211, 213, 214, 216 Ekiert, G. 141 embedded political identities, in Yugoslavia 15, 16 equality, and trust 6, 87 ethnicity: and civic engagement 226 in Cluj-Napoca, Romania 143, 144–5, 146, 148, 156–7 ethnic mobilization 96; ethnic tensions and democratization 97, 220–1, 225, 230; and political community 103–5; social capital and support for democracy 105–16, 117; and social trust 6, 96, 110, 122 and organizational participation 133–5 in Romania and Moldova 228, 229 European Union 228 European Values Surveys 6; on social trust 128, 129

Index 235 and organizational participation 126–7, 134–5 family networks 227; in communist societies 222; in Poland 195; Russia compared with other countries 69; in Yugoslavia 20 Fish, M. S. 97 formal networks, and social capital in Russia 201, 205, 209 France: interpersonal trust 71, 72; social networks 67, 68, 69 Freedom House indicators of democratization 97–8, 99–100, 223; and social trust 135–6 friendship networks 7, 227; in Russia and East Germany 174–6, 178, 179 Fukuyama, Francis, The End of History 219 Funar, Gheorghe, mayor of Cluj-Napoca, Romania 144, 146 gender: and organizational participation in Russia and Germany 170, 171, 172–3, 177, 178; and social capital in Russia 207, 208, 211, 213, 216 generalized trust 6, 17; and civic engagement 86–7, 89–92, 221; and democratization 120, 223, 227; and inequality 225, 228; measures of 130–1; moral basis of 222–3; and optimism 83–5, 93; and organizational participation 83; and particularized trust 223, 225–6; and personal networks 50; and political power 49; and provision networks 222, 223; in Russia 48; and support for democratic government 227; in Yugoslavia 19 Georgia 97 Germany see East Germany; West Germany Gini index of inequality 85–6 government performance, satisfaction with 105, 108, 115 Granovetter, M. S. 63–4, 69 Halebsky, S. 16 ‘hour-glass’ societies 180 Huckfeldt, Robert 143, 157 Hungarian minority in Romania 144–5, 146, 148

Hungary; civil society 124; democratization 1, 2; ethnic tensions 97; social networks 67, 68, 69 Ilescu, Ion 143, 144, 154, 155 incomes: attitudes to in Poland 185; and organizational participation in Russia and Germany 170, 171, 172, 177, 178 inequality see economic inequality informal networks, and social capital in Russia 201, 205–6, 209, 212 Inglehart, R. 62, 201 institutional effects, of voluntary organization involvement 125 institutional trust, in Russia 48 interpersonal trust: and civil society 61–2; and community mobilization 15; dyadic 17–18; measures of 130–1; and optimism 84–5; particularized trust and generalized trust 223, 225–6; in Poland 197–8; Russia 70–4; and social capital 110, 112, 184; societies marked by particularized trust 221; see also social trust Italy 141 Jensen, J. 167 leadership: and mobilization 16–17; populist political leadership in Serbia 12–13 legal systems, confidence in 105, 106, 115 Lenin, V. I. 216 Levi, M. 49, 87, 96, 224 Madison, James 202 Marxism, in East Germany 29, 43 materialism, trust and civic engagement 88, 90, 92 media: and civic engagement 154 in Cluj-Napoca, Romania 145–6, 149–50 in South Bend, Indiana 149–50 and civic initiatives in Slovenia 14 micro mobilization contexts 15 Mili´c, A. 20 Mill, John Stuart 220 Milosˇevi´c, Slobodan 12, 13 Mishler, W. 63, 70, 99 Miszlivetz, F. 167 Moldova: attitudes to trust in 225; generational differences in social capital in 228–30

236

Index

Mondak, J. 4 Montenegro 13 Mueller, J. 224, 227, 228 Muller, E. N. 223 multiple collective identities, in Yugoslavia 11 Mutz, D. 64 national parliaments, confidence in 105, 107, 115 national unity, and democratization 102–5 nationalist ideology, Yugoslavia 11 nationalist mobilization in Serbia 12–13 new social movements: in Slovenia 13–14, 23; and social networks 18 Newton, K. 3 niche networks, East Germany 29–30, 31, 34, 36–9, 40–2 non-communitarian social capital 49, 50 norms 4 Norris, P. 100 Oliver, Eric 170 optimism: and generalized trust 83–5, 93; and generational differences in social capital, in Romania and Moldova 228–30; and social capital 7, 46, 47 parliaments (national), confidence in 105, 107, 115 particularized trust see interpersonal trust PCOMS Survey: on organizational participation 7, 167, 168 in East Germany and Russia 165, 166, 170–3, 174, 175, 176 personal networks 5; East Germany 5, 28 after the political turn 32–42 and communist society 28–32 and generalized trust 50; in Poland 195–8; and political networks 15–17; and social capital theory 28–9; Yugoslavia 10, 11–12, 16, 23; see also social networks personal trust see interpersonal trust Petro, Nicolai 49 Poland 7, 184–99, 230; Catholic Church in 87, 219; civil society in 124, 184–5, 198; democratization 1, 2, 61; electoral results 186–7; interpersonal

trust in 197–8; local communities and community-oriented activity 193–5; organizational participation in 190–3; personal networks in 195–8; social images in politicians and public administration 187–9 the private sector 189–90 Solidarity 61, 124, 192, 226; support for democracy in 187; trade union membership 189, 192, 226; trust in social and political institutions 185–7; weakness of civil society in 184–5, 198 political alienation, in Poland 188–9 political community, support for 102–5 political discussion: in Cluj-Napoca, Romania 140, 144, 147–8, 149, 156; in South Bend, Indiana 140, 147–8, 149 political evaluation, in Cluj-Napoca, Romania and South Bend, Indiana 154–5, 156 political mobilization: ethnicity and support for democracy 113; Yugoslavia 11, 12–14, 23, 24 political networks 7; East Germany 29, 31, 37, 41–2; Moldova 230; Russia 52–6, 206, 230; and social networks 14–18, 23; Yugoslavia 5, 11 political participation: and civic engagement 141; in Cluj-Napoca, Romania and South Bend, Indiana 140, 148–50, 156; and low levels of social capital 184 political power, and generalized trust 49 political rights, Freedom House rating of 97–8, 100 politics, and social capital in Russia 215–16 Portugal, interpersonal trust in 71, 72 Post-Communist Organizational Membership Survey see PCOMS survey postmaterialists, and trust 87, 90 presidential support, in Cluj-Napoca, Romania and South Bend, Indiana 154–5 prior communist experience, and organizational participation in Russia and Germany 170, 171–2, 173, 179–80 private sector, in Poland 189–90, 193

Index 237 provision networks 222, 223; East Germany 30, 34, 35–6, 38–40, 41 public sphere effects, of voluntary organization involvement 125 Putnam, R. 49, 64, 85, 141; on civic engagement 82–3, 91–2, 220; on civil society 167; on democracy 228; on social capital 3, 4, 47, 95, 99, 201, 204, 214, 215; on trust 222 Rahn, W. 83 Reagan, Ronald 143, 154, 216 religious backgrounds, attitudes to different 225 religious organization membership, in Poland 192, 219 religious people, trust and civic engagement 87, 88, 90–1, 99 Roeder, P. 97 Romania: attitudes to trust in 223, 225, 226; Ceausescu regime 144–5, 146; ‘Cenaclul Flacara’ 124; civil society 124; compared with the United States 7; democratization process in 1; ethnic tensions 97; generational differences in social capital in 228–30; legacy of totalitarian rule in 144; measuring civic engagement in 127–8; measuring social trust in 128, 130–1 and organizational participation 132, 133, 134–5 see also Cluj-Napoca, Romania Rose, R. 48, 63, 70, 99 Rose-Ackerman, S. 50 Rothstein, on trust and coercion 87 rural areas: and community-oriented activity in Poland 193–5; trust in 91 Russia 5–6, 46–57; attitudes to trust in 223, 226 interpersonal trust 70–4, 75, 76 church attendance 205; civil society 48, 63, 76–8; clientelist networks 50–1, 56; Democratic Russia movement 52, 53–4, 55; democratization process 1, 6; Gaidar government 52, 54–5, 56; ‘kitchen circles’ 64; networks of trust 46, 47, 48, 56–7 as negative social capital 51–2 as a tool for the survival of political groups 52–6 Nijny Novgorod group 52, 55–6; organizational participation in 165–6

changes in membership levels 169 post-communist factors in 173–9 and socio-economic status 170–3, 177, 178 social atomization 62, 63, 67, 70; social capital in 44–7 and support for democracy 105, 110–12, 200–16 social networks 5–6, 46–57, 63–70 and civil society 48, 63, 76–8 consequences of 74–6 and interpersonal trust 72, 73, 74, 76, 77 and political capacity 67, 69, 70, 74–6 politicization of 66, 68–9, 70, 78 size of 66 and ‘weak’ network ties 67, 69, 77, 78 trade union membership 205; voluntary organizations, non-participation in 63, 77–8 Russian minorities, social capital and support for democracy 105, 110–13, 114, 116, 117 Rustow, D. 102 Sacharov, Andrei 53, 54 Sajo, Andras 50 Seligman, M. E. P. 85 Seligson, M. A. 223 Serbia: ethnic violence 97; mass rallies and populist political leadership 12–13; social networks 20–3, 23–4, 226 Skocpol, Theda 167 Slovaks, and political interest 105–9 Slovenia: democratization in 1; new social movements and civic initiatives 13–14, 23; social networks 20–3, 24, 230 Smith, Adam 17 social capital: and civic engagement 122; defining 2–3, 200–1; and democratic values 200–16; and democratization 3–4, 97–102, 120; ethnicity and support for democracy 105–16, 117; forms of in Eastern Europe 49; levels of 1, 95, 120, 122, 184; measures of 204–6; negative aspects of 46, 47–8 and networks of trust 51–2 networks 201, 205–6; optimistic view

238

Index

social capital continued of 7, 46, 47; in Poland 7, 184–99; in Russia 44–7, 105, 110–12, 200–16; statist and non-statist views of 49; and support for democracy 102–16, 200–16 social capital theory: and personal networks 28–9; and post-communist transition 8 social learning effects, of voluntary organization involvement 125 social mobilization 17, 18 social networks 5; and civic engagement 121; and civil society 62; in communist societies 222; and political networks 14–18, 23; Russia 5–6, 46–57, 63–70, 226 and civil society 48, 63, 76–8 consequences of 74–6 cross-country comparisons with 66–70 and interpersonal trust 72, 73, 74, 77 of power 46, 47, 48–9, 56 of trust 46, 47, 48, 51–7 Slovenia 20–3, 24, 230; Yugoslavia 11, 19–23, 23–4; see also personal networks social perception, in Cluj-Napoca, Romania and South Bend, Indiana 150–4 social trust: and civic engagement 121–3; and democratization 223; in the East and West 129–30; measuring 128–31; and organizational participation 95–6, 99, 126–7, 136–7 at an individual level 131–6 see also interpersonal trust socio-economic status: and organizational participation in Russia and Germany 170–3, 177, 178; and social capital in Russia 208, 211 South Bend, Indiana 140, 143; ethnicity 143; knowledge of neighbours 140; and political discussion 140, 147–8; and political evaluation 154–5; and political participation 140, 148–50; sampling information and survey items 157–60; size 143; social and economic perception 140, 154–5 Spain: interpersonal trust 71, 72; social networks 69 Sprague, John 143, 157

Stamboli´c, Ivan 12 Stolle, D. 83, 220 strong network ties 227; and personal networks 222 in East Germany 30, 39 in Russia 6 television-watching, and civic engagement 88, 91–2 ‘third wave of democracy’ 1 Tocqueville, Alexis de 220 tolerance, and communist societies 224 trade union membership: and civic engagement 226; in Poland 189, 192, 226 trust 4, 81, 221–5, 225–6; characteristics of trusting societies 220; and civic engagement 6, 81–93; and clientelist networks 50–1; and corruption 223; and democracy 87–8, 91, 93, 122, 224, 228; and ethnic tensions 221; and generational differences in social capital, in Romania and Moldova 228–30; and inequality 85–6, 93, 225, 228; institutional trust in Poland 185–7; and the legacy of communism 219, 220, 221–2; networks 5, 23; networks of trust as negative social capital 51–2; and organizational participation 6, 17–18, 19, 83, 88, 89–92, 121, 121–3; and political mobilization 24; and religious people 87, 88, 90–1; in rural and urban areas 91; and social capital 47 measures of 204 in Russia 206, 209, 210, 211, 212, 213 see also generalized trust; institutional trust; interpersonal trust United States: American Constitution 201; and democracy 201, 224; and Russia 215; social capital 204; social networks 67, 68, 70; and trust 223; voluntary organization involvement 141, 169; see also South Bend, Indiana universal sociability 17–18 Uslaner, E. M. 83, 222–3 Verba, S. 2, 170 Vojvodina 13 Volkenstein, Masha 81, 83

Index 239 Völker, B. 20 voluntary organization involvement 165–82, 226–7; and democracy and civil society 123–5 ethnicity and support for 110, 111, 113, 114, 117 and democratization 95–7, 98–9, 101, 219, 226–7; East Germany 43–4; and optimists 84; Poland 190–3; Russia 63, 77–8, 212; and social capital 95–7, 98–9, 101, 184, 204–5; and trust 6, 17–18, 19, 83, 88, 89–92 social trust 95–6, 99, 121–3, 126–7, 131–7 in Western and Eastern Europe 120–1, 126–7; see also civic engagement Warren, Mark 123 weak network ties 227; and civil society

62, 63–4; in East Germany 30–1, 33, 39, 41, 42, 43; in Russia 67, 69, 77, 78; in Slovenia 23 West Germany 166, 169, 170–3 Western model of democracy 1 World Values Surveys 2, 6, 7, 70–2; on democratization and social capital 96–7, 98, 100–2; on organizational participation 122, 166, 167, 168–9, 174; and support for democracy 103–16, 225; on trust 128, 129 and civic engagement 86, 88 Yeltsin, Boris 53, 54, 55 Yugoslavia 5, 10–25, 96; democratization and social capital 98; political mobilization 11, 12–14; social and personal networks 5, 11, 19–23, 23–4; see also Serbia