125 44 37MB
English Pages 384 [370] Year 2006
THE CRISIS OF DEMOCRATIC REPRESENTATION IN THE ANDES
THE CRISIS OF DEMOCRATIC REPRESENTATION IN THE ANDES ••• ~
Edited by
Scott Mainwaring, Ana Mar{a Bejarano, and Eduardo Pizarro Leong6mez
••• ~
Stanford University Press Stanford, California 2006
Stanford University Press Stanford, California
© 2006 by the Board ofTrustees of the Leland Stanford Junior University. All rights reserved. No part of this book may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying and recording, or in any information storage or retrieval system without the prior written permission of Stanford University Press.
Library of Congress Cataloging-in-Publication Data The crisis of democratic representation in the Andes I edited by Scott Mainwaring, Ana Maria Bejarano, and Eduardo Pizarro Leong6mez. p. cm. Includes bibliographical references and index. ISBN-13: 978-0-8047-5278-7 (cloth: alk. paper) ISBN-10: 0-8047-5278-8 (cloth: alk. paper) 1. Representative government and representation-Andes Region. 2. Democracy-Andes Region. 3. Political culture-Andes Region. 4. Andes Region-Politics and government. I. Mainwaring, Scott, 1954II. Bejarano, Ana Maria. Ill. Pizarro Leong6mez, Eduardo. JL1866.C76 2006 320.98-dc22 2006005159
Typeset by G&S Typesetters, Inc. in 10/12 Bembo
CONTENTS List ofTables and Figures vn Acknowledgments x1 List of Contributors xm 1. The Crisis of Democratic Representation in the Andes:
An Overview 1 Scott Mainwaring, Ana Maria Brjarano, and Eduardo Pizarro Leong6mez PART I. PARTY SYSTEMS, POLITICAL OUTSIDERS, AND THE CRISIS OF DEMOCRATIC REPRESENTATION 2. From Crisis to Collapse of the Party Systems and Dilemmas of Democratic Representation: Peru and Venezuela 47 Martin Tanaka 3. Giants with Feet of Clay: Political Parties in Colombia Eduardo Pizarro Leong6mez 4. Ecuador: The Provincialization of Representation Sim6n Pachano
78
100
5. Outsiders and Neopopulism: The Road to Plebiscitary Democracy 132 Rene Antonio Mayorga PART II. DECENTRALIZATION, LEGISLATURES, AND DEMOCRATIC REPRESENTATION 6. Decentralized Politics and Political Outcomes in the Andes Kathleen O'Neill
171
7. The Nature of Representation in Andean Legislatures and Attempts at Institutional Reengineering 204 Brian F. Crisp
Vl
Contents PART Ill. POPULAR POLITICS AND THE CRISIS OF DEMOCRATIC REPRESENTATION 8. Urban Citizen Movements and Disempowerment in Peru and Venezuela 227 Daniel H. Levine and Catalina Romero 9. Indigenous Politics in the Andes: Changing Patterns of Recognition, Reform, and Representation 257 Deborah]. Yashar PART IV. CONCLUSION 10. State Deficiencies, Party Competition, and Confidence in Democratic Representation in the Andes 295 Scott Mainwaring Index
347
TABLES AND FIGURES
Tables Table 1.1. Table 1.2. Table 1.3. Table 1.4. Table 1.5. Table 1.6.
Table 1.7. Table 2.1. Table 2.2. Table 2.3. Table 2.4. Table 2.5. Table 2.6. Table 3.1.
Support for Democracy, Latin America, 2005 9 GDP per Capita in the Andes, 1960-2002 10 Citizen Trust in Representative Institutions, Andean Countries 17 Confidence in Parties and Parliament, Select Countries, World Values Survey 18 Electoral Volatility and Share ofVote for New Parties in Lower-Chamber Elections, Andean Countries 19 Average Share of Vote Won by Outsider Presidential Candidates in Five Most Recent Presidential Elections, Andean Countries 22 Programmatic Representation in Latin America and Spain 26 Peru: Vote Percentages for the Major Political Parties, 1978-2000 49 Venezuela: Vote Percentages for Presidential Elections, 1973-1998 50 Peru and Venezuela: Inflation, GNP, and Subversive Acts, 1980-1999 51 Venezuela: Trends in Party Identification 52 Venezuela: Percentage of Seats in the Chamber ofDeputies, 1973-2000 60 Venezuela: Number of Governors Elected, by Party, 1989-2000 62 Number of Lists that Registered for Senate and LowerChamber Elections, 19 58-2002 84
Tables and Figures
Vlll
Table 3.2. Table 3.3. Table 4.1. Table 4.2. Table 4.3. Table 4.4. Table 4.5. Table 4.6. Table 4.7. Table 4.8. Table 5.1. Table Table Table Table
6.1. 6.2. 7.1. 7.2.
Table 7.3. Table 7.4. Table 7.5. Table 10.1. Table 10.2.
Table 10.3. Table 10.4. Table 10.5. Table 10.6.
Electoral Performance of Lists That Elected More Than One Senator, 1991-2002 86 Senate Seats Won by Political Parties and Movements, 1991-2002 88 Ecuador: Main Political-Electoral Reforms, 1983-2003 105 Share of Congressional Vote Won by Four Major Parties, 1979-2002 106 Share of National Electorate and Number ofDeputies per Province, 2002 107 Share ofPresidential Vote in First Round, 1978-2002 117 Size ofLegislative Delegations, 1979-2002 118 Electoral Strongholds of the Main Political Parties, 1979-2002 122 Territorial Distribution Index (TDI) of Main Parties, 1979-2002 124 Regional Distribution of Origin ofDeputies, by Party, 1979-2002 125 Bolivia: Party Votes and Seats, Lower-Chamber Elections, June 30, 2002 156 Expenditure Decentralization in Latin America 17 4 Popular Election of Subnational Executives 175 Pre-Reform Intraparty Characteristics 208 Partisan Composition of Andean Legislatures Prior to Constitutional Reform 21 0 Partisan Composition of Andean Legislatures after Constitutional Reform 214 Bill Targets in the Colombian Senate before and after Electoral Reform 217 Public Confidence in Congress, 1996 219 State Performance and Perceptions Thereof in the Andes, 1996-2005 299 Pearson Correlation Coefficients between State Performance and Confidence in Parties and Parliaments, Country-Level Indicators 300 Determinants of Confidence in Political Parties in the Andes 302 The Impact of Assessment of the National Economic Situation and Interpersonal Trust on Confidence in Parties 303 Determinants of Confidence in Congress in the Andes 304 Attitudes about Democracy and Representation in the Andes 307
Tables and Figures Table 10.7. Table 10.8.
lX
Citizen Confidence in Institutions in the Andes 309 Secondary Education and Urbanization in the Andes 315
Figures Figure 1.1.
Presidential turnout as a percentage of the eligible electorate in the Andes 24 Figure 1.2. Turnout as a percentage of the eligible electorate in lowerchamber elections in the Andes 25 Figure 6.1. Turnout in millions of voters, by level 181 Figure 6.2. Percent of the vote won by traditional parties, by level 187 Figure 7.1. Interparty and intraparty incentives oflegislators in the pre-reform Andean countries 211 Figure 7.2. Public satisfaction with democracy 220 Figure 10.1. Votes cast in presidential elections as a percentage of total population, Andean countries, 1950s to 2004 314 Figure 10.2. Effective number of parties, lower chambers in the Andean countries 318 Figure 10.3. Effective number of parties in the Senate: Peru, Venezuela, Bolivia, and Colombia 319
ACKNOWLEDGMENTS
Our
first and foremost gratitude is to three institutions without which this book would not have come to fruition: the Kellogg Institute for International Studies at the University ofNotre Dame, the Ford Foundation, and the CocaCola Company. The distant background to the book is that in the fall of 1998 Colombia's foremost human rights lawyer, Gustavo Gallon of the Colombia Commission of Jurists, came to spend what was originally intended to be a semester at the Kellogg Institute as a Visiting Fellow. The deteriorating human rights situation in Colombia made it unwise for Gallon to return to Colombia as planned and instead he stayed at Notre Dame for an additional semester. During his year at Notre Dame, the Colombian Commission ofJurists, the Kellogg Institute, the Kroc Institute for International Peace Studies at Notre Dame, and the Center for Civil and Human Rights, also at Notre Dame, hatched a collaborative project on "Democracy, Human Rights, and Peace in Colombia." The Ford Foundation's Santiago office generously supported this project (Grant 1000-0727). This project enabled Colombian scholars at risk to spend time in residence at the Kellogg Institute, thus allowing them to work productively and safely at a time when they could not live safely in Colombia. Through this program, Ana Maria Bejarano, Eduardo Pizarro Leongomez, and four other distinguished Colombian scholars came to the Kellogg Institute for one-year periods between 2000 and 2003. The three editors of this volume became friends. We spent many hours discussing the Colombian situation and the deterioration of democracy in other Andean countries-issues of deep normative concern and intellectual interest to all three of us. Along with Miriam Kornblith, Rene Antonio Mayorga, Simon Pachano, and Martin Tanaka, we conceived a broader project on the Andean crisis. The Ford Foundation once again provided support for this project (Grant 980-0350-3). Its support enabled Kellogg, CEBEM (Centra Boliviano de Estudios Multidisciplinarios),
Xll
Acknowledgments
FLACSO-Ecuador, IESA (Instituto de Estudios Superiores de Administraci6n, Caracas), the Instituto de Estudios Peruanos (IEP), and the Universidad Los Andes (Bogota) to have some scholarly exchanges within the Andean region and to sponsor a scholarly workshop in Quito in July 2003. A generous gift from the Coca-Cola Company to the Kellogg Institute has funded a series of conferences on some of the outstanding problems confronting Latin America since the 1990s, and it enabled us to host an eponymous conference at the Kellogg Institute in May 2002. We are grateful for this support. The staff of the Kellogg Institute spent long hours arranging the travel of participants, designing posters, and hosting the participants. At the conference, we benefited from the stimulating comments of Michael Coppedge, Humberto de la Calle, Paul Drake, Myles Frechette, Luis Gallegos-Chiriboga, Prances Hagopian, Miriam Kornblith, Curtis Kamman, Ricardo Luna, Alejandro Reyes, and Samuel Valenzuela. Since its inception in 1982, the Kellogg Institute has promoted research on democracy and other important normative issues that confront humanity. We benefited from the rich intellectual debate at the Kellogg Institute on this theme. During the years of gestation of this project, among the contributors to this volume, Ana Maria Bejarano, Rene Mayorga, Sim6n Pachano, Eduardo Pizarro Leong6mez, Catalina Romero, and Martin Tanaka were Visiting Fellows at the Kellogg Institute. Brian Crisp, Daniel Levine, and Deborah Yashar, who also contributed to this volume, were Visiting Fellows at the Kellogg Institute in earlier years. A grant by the Fulbright Educational Partnerships of the U.S. Department of State enabled Kellogg to host several Visiting Fellows from Venezuela and Peru starting in academic year 2003-2004 (Grant SPE50003GR057). This grant helped the Kellogg Institute to continue developing its ties with the Andean countries, especially with our institutional partners, the Instituto de Estudios Peruanos (IEP) and the Institute de Estudios Superiores de Administraci6n (IESA), as we brought this book to fruition. Through this program, Martin Tanaka was a Visiting Fellow at Kellogg as we worked on this book; he wrote Chapter 2 of this volume. We are grateful to our editors at Stanford University Press, Amanda Moran and Jared Smith, who gave their steadfast support to this project. Anna Eberhard Friedlander was a capable project manager, and Ruth Steinberg did the copy editing. Elizabeth Rankin ably prepared the manuscript for publication and created the index. Julia Smith and Elizabeth Station translated several chapters from Spanish to English. The comments of Paul Drake and an anonymous reviewer helped all of us sharpen our arguments.
CONTRIBUTORS
Ana Maria Bejarano is Assistant Professor of Political Science at the University of Toronto. She holds a Ph.D. in Political Science from Columbia University. She previously was Professor of Political Science at the Universidad de Los Andes in Bogota, where she also served as Director of the Center for Social and Legal Research (CIJUS). She co-edited Elecciones y democracia en Colombia, 1997-1998 (Universidad de los Andes, Fundaci6n Social, Veeduria Elecci6n Presidencial, 1998), and co-authored the chapter on Colombia in Frances Hagopian and Scott Mainwaring, eds., The Third Wave cif Democratization in Latin America: Advances and Setbacks (Cambridge University Press, 2005). Recent publications include articles in Constellations and the Canadian Journal of Latin American and Caribbean Studies. She is finishing a book on the historical origins and divergent trajectories of democracy in Colombia and Venezuela. Her current research deals with regime change, institution-building, and constitutionmaking in the Andes. Brian F. Crisp received his Ph.D. from the University of Michigan and is currently an Associate Professor in the Department of Political Science at Washington University in St. Louis. His work on the institutional mechanisms constructed to formalize state-civil society relations and the impact of these relations on policy choices has appeared in the American Political Science Review, American Journal cif Political Science, and Journal of Politics. His book, Democratic Institutional Design: The Powers and Incentives of Venezuelan Politicians and Interest Groups, was published by Stanford University Press (2000). Daniel H. Levine is the James Orin Murfin Professor of Political Science at the University of Michigan. He has published extensively on issues of democracy, democratization, social movements, Venezuelan politics, and religion and politics
XlV
Contributors
in Latin America. His books include Conflict and Political Change in Venezuela (Princeton University Press, 1973), Religion and Politics in Latin America: The Catholic Church in Venezuela and Colombia (Princeton University Press, 1981), and Popular Voices in Latin American Catholicism (Princeton University Press, 1992).
Scott Mainwaring is Eugene Conley Professor of Political Science and Director of the Kellogg Institute for International Studies at the University of Notre Dame. His books include The Third Wave cif Democratization in Latin America: Advances and Setbacks (Cambridge University Press, co-edited, 2005), Democratic Accountability in Latin America (Oxford University Press, co-edited, 2003), Christian Democracy in Latin America (Stanford University Press, co-edited, 2003), Rethinking Party Systems in the Third Wave of Democratization: The Case cif Brazil (Stanford University Press, 1999), Presidentialism and Democracy in Latin America (Cambridge University Press, co-edited, 1997); Building Democratic Institutions: Party Systems in Latin America (Stanford University Press, co-edited, 1995), and many others. He received a John Simon Guggenheim Memorial Foundation fellowship in 2000 for work on a project on authoritarianism and democracy in Latin America, 1945-2000. Rene Antonio Mayorga is a Senior Researcher at the Centro Boliviano de Estudios Multidisciplinarios (CEBEM). He is also Professor ofPolitical Science at the Facultad Latinoamericana de Ciencias Sociales (FLACSO) in Ecuador, and in the Joint Master's Program of CEBEM, FLACSO, and the Universidad Complutense of Madrid. He has been a visiting professor at Brown University and at the Universities of Salamanca, Berlin, and Notre Dame, among others. His books include De la anomia poUtica al orden democratico (CEBEM, 1991); AntipoUtica y neopopulismo (CEBEM, 1995); and La cuesti6n militar en cuesti6n: Democracia y fuerzas armadas (CEBEM, 1994). He is editor of Democracia y gobernabilidad en America Latin a (ILD IS, CEBEM, 1992). He also contributed chapters to The Third Wave of Democratization in Latin America (Cambridge University Press, 2005), Transitional Justice and the Rule of Law in New Democracies (Notre Dame University Press, 1997), and Mixed-Member Electoral Systems: The Best cif Both Worlds? (Oxford University Press, 2001). Kathleen O'Neill is studying law at New York University. She received a Ph.D. in Political Economy and Government from Harvard in 1999, and she was an Assistant Professor of Government at Cornell University from 2000 to 2003. Sim6n Pachano is a Researcher at FLACSO in Quito, Ecuador. He has also taught in Spain and Bolivia, and he has published extensively on Ecuadorian politics. Among his many publications are Democracia sin sociedad (ILDS, 1996) and La representaci6n ca6tica: Analisis del sistema electoral ecuatoriano (FLACSO, 1998).
Contributors
XV
Eduardo Pizarro Leong6mez is Professor at the Universidad Nacional de Colombia in Bogota and served as Director of its Instituto de Estudios Politicos y Relaciones Internacionales (IEPRI). He was a Visiting Fellow at the Kellogg Institute for International Studies in 2000-2001. He has published several scholarly analyses of the conflict in Colombia, including Una democracia asediada: Balance y perspectivas del conjlicto armado en Colombia (Editorial Norma, 2004), and he is a weekly contributor to the national newspaper, El Tiempo. Catalina Romero is Professor of Sociology at the Department of Social Sciences and Dean of the Social Sciences at the Pontificia Universidad Cat6lica del Peru. She obtained her Ph.D. at the New School for Social Research in 1989. She was Director of the Instituto Bartolome de Las Casas in Lima from 1982 to 1992. She has published several articles and books about the Catholic Church, social movements, and civil society in Peru. Martin Tanaka holds a Ph.D. in Political Science from the Facultad Latinoamericana de Ciencias Sociales (FLACSO) in Mexico City. He is Director of the Instituto de Estudios Peruanos in Lima. Tanaka has written about Peruvian politics, Latin American politics, and social movements and participation in Peru. He is author of Los espejismos de la democracia: El colapso del sistema de partidos en el Peru (Instituto de Estudios Peruanos, 1998). He contributed chapters to Frances Hagopian and Scott Mainwaring, eds., The Third Wave of Democratization in Latin America: Advances and Setbacks (Cambridge University Press, 2005), and to Carol Wise, Riordan Roett, and Guadalupe Paz, eds., Post-Stabilization Politics in Latin America (Brookings Institution Press, 2003). DeborahJ. Yashar is Associate Professor ofPolitics and International Affairs and Director of the Program in Latin American Studies at Princeton University. She is the author of Contesting Citizenship in Latin America: The Rise ofIndigenous Movements and the Postliberal Challenge (Cambridge University Press, 2005) and Demanding Democracy: Reform and Reaction in Costa Rica and Guatemala, 1870s-1950s (Stanford University Press, 1997). Her current research focuses on the intersection of civil wars, reconstruction, and democratization, and on inequality.
THE CRISIS OF DEMOCRATIC REPRESENTATION IN THE ANDES
1
The Crisis of Democratic Representation in the Andes: An Overview
....• Scott Mainwaring, Ana Mar£a Bejarano, and Eduardo Pizarro Leong6mez
This book analyzes and explains the crisis of democratic representation in five Andean countries: Bolivia, Colombia, Ecuador, Peru, and Venezuela. In this region, disaffection with democracy, political parties, and legislatures has spread to an alarming degree. In Bolivia (2003), Ecuador (1997, 2000, and 2005), Peru (2000), and Venezuela (1993), democratically elected presidents were not able to finish their terms of office because of popular and elite discontent. In Peru and Venezuela, massive discontent with existing party options gave rise to surprising collapses of the party system in the 1990s. Political outsiders with anti-establishment discourses have flourished as traditional parties have faded. In Ecuador, a successful military coup removed a democratically elected president in 2000; in Peru, a successful palace coup led to a democratic breakdown in 1992; and in Venezuela, a coup in 2002 removed the democratically elected president for one day, although he made a rapid comeback to the presidency. Some parties that in the recent past were major electoral contenders and won the presidency have seemingly suffered their terminal demise. The crisis of democratic representation in the Andes is important both intellectually and politically. Understanding what has gone wrong with democracy in Latin America and many other "third wave" democracies has become one of the outstanding intellectual challenges of our day. The widespread dissatisfaction with democratic representation is a core ingredient in the crisis of democracy in the Andes and throughout much of Latin America. In recent years, as the wave of transitions to democracy and semi-democracy in Latin America has ebbed, intellectual and political attention has turned to how to build more robust democracies that satisfY the aspirations of more citizens-and how to comprehend the grave shortcomings of most existing democracies and semidemocracies in the region. Because the Andean countries provide clear examples
2
Mainwaring, Bejarano, Pizarro Leong6mez
of the weakness of mechanisms of democratic representation, they are an excellent set of countries for examining this problem. Politically, this subject is important because the Andean countries have the potential to be negative role models in a region (Latin America) that has historically had strong demonstration and diffusion effects in terms of regime changes (Mainwaring and Perez-Lifian, forthcoming). Moreover, a deep discrediting of mechanisms of democratic representation can have grave implications for democracy. In Peru, disenchantment with traditional mechanisms of democratic representation helped pave the way to a democratic breakdown in 1992. In Venezuela, the growing disaffection with conventional vehicles of democratic representation led Hugo Chavez to the presidency in 1998. Under his leadership, democracy in Venezuela has eroded, and the country has polarized sharply between his followers and foes. We hope that the book makes five main contributions to political science and to understanding Latin American politics. First and foremost, we hope to contribute to the broadening of theoretical and empirical horizons about democratic representation by studying a region in crisis. Our work shifts the mainstream thinking about representation in three ways. Most of the work on democratic representation focuses on the advanced industrial democracies, and almost all of it analyzes how representation works. Analyzing the Andes suggests a more innovative (in relation to the existing literature) question that is more important for our region and some other parts of the world: why representation sometimes fails to work. 1 This issue is paramount because in the Andes as well as some other parts of the developing world the perceived failures of democratic representation are widespread and profound. In the extensive literature on political representation, to the best of our knowledge this is only the second book to focus on a crisis of democratic representation (see Novaro 1994). Many previous works have dealt with a related subject, namely, a decline of political parties, but most of this literature has focused on the advanced industrial democracies, where (perhaps excluding Italy) there is nothing resembling the crisis of parties and of democratic representation that has plagued the Andean region. Much of the existing literature assumes that programmatic convergence between voters and legislators is at the core of democratic representation and exclusively analyzes such convergence. In contrast, such programmatic or ideological representation is very weak in the Andes. To understand representation in this region, it is essential to look beyond programmatic and ideological convergence between voters and their representatives. Most of the literature on the advanced industrial democracies posits that patterns of political representation remain relatively stable over time (Bartolini and Mair 1990; Converse 1969; Lipset and Rokkan 1967). When we turn to many post-1978 democracies, however, it is important to think about a range in patterns of democratic representation, running from more to less legitimate and
The Crisis of Democratic Representation in the Andes
3
stable. We pose new theoretical questions about why patterns of democratic representation in many countries do not achieve the stability that most of the theoretical literature posits. 2 Second, we aspire to enrich empirical knowledge about democratic representation in the Andes. New work on this type of representation in Latin America is needed because of its importance in democratic theory and the paucity of empirical work on it for Latin America (see Chalmers, Martin, and Piester 1997; Hagopian 1998; Luna and Zechmeister 2005; Roberts, forthcoming). For Latin America, there is extensive literature on legislators and legislatures and on parties and party systems, and a growing literature on voters. There is little, however, on democratic representation, which involves the relationship between voters and parties or elected politicians. Our third contribution revolves around explaining why a crisis of democratic representation occurs. When voters have free choices from an ample array of party options, why do they remain deeply dissatisfied? Why can't they find a party option that satisfies them? We do not definitively resolve why a crisis of democratic representation erupted in the Andes. This is a new research question that demands further examination, and disagreement over it is intractable. Two authors, Brian Crisp and Sim6n Pachano, focus on institutional arrangements to explain deficiencies in democratic representation. Crisp (Chapter 7) argues that in the Andes institutional incentives foster either too much or too little focus on national programmatic issues as opposed to district-level constituency demands. Pachano (Chapter 4) claims that many deficiencies of representation in Ecuador stem from institutional rules of the game. These rules of the game favor party system fragmentation, impede the formation of stable ruling coalitions, and encourage a focus on provincial and local constituency service rather than programmatic national issues. Both chapters are emblematic of institutionalist approaches to understanding the shortcomings of democratic representation. Both authors imply that citizens' deep dissatisfaction with democratic representation could be attenuated with well-conceived institutional reforms. In their chapters, in contrast, Rene Antonio Mayorga (Chapter 5) and Scott Mainwaring (Chapter 10) see the crisis of democratic representation as stemming from governance (Mayorga's term) or state deficiencies (Mainwaring's focus). Mayorga asserts that deep dissatisfaction with democratic performance underlies the crisis of democratic representation and the rise of political outsiders. In the book's Conclusion, Mainwaring argues that institutional rules of the game are not generally at the core of the dissatisfaction with democratic representation. He asserts that the main cause of the crisis of democratic representation in the Andes has been state deficiencies in many arenas, ranging from citizen security to corruption and economic performance. For both Mayorga and Mainwaring, the rise of political outsiders, declining confidence in parties, high electoral
4
Mainwaring, Bejarano, Pizarro Leong6mez
volatility, and the other manifestations of a crisis of democratic representation are products of bad performance by democratic regimes. Bad performance has bred dissatisfaction with politicians and parties. To this performance-based argument, Mainwaring adds one other explanation for the deep dissatisfaction with democratic representation. Mainwaring argues that the zero-sum nature of party competition and the media focus on negative images of parties and assemblies help account for the discrediting of these agents of democratic representation. This explanation resonates with constructivist approaches to social science because it calls attention to the way in which political competition and media images help construct citizen conceptions of politics, and specifically of parties and assemblies. Although neither this volume nor any other can definitively resolve what has caused the crisis of democratic representation in the Andes, our volume makes a contribution by explicitly putting this question on the intellectual agenda and by staking out three of the most important explanations: institutionalist, performance-based, and constructivist. Our fourth contribution is to advance understanding of the consequences of a crisis of democratic representation. This issue comes to the fore in the chapters by Martin Tanaka (Chapter 2), Rene Antonio Mayorga (Chapter 5), and Daniel Levine and Catalina Romero (Chapter 8). Tanaka and Mayorga examine the consequences of the discrediting of democratic representation for democratic regimes. In many post-1978 democracies and semi-democracies, citizens have become disillusioned with the mechanisms of democratic representation. As both authors show, the discrediting of the conventional mechanisms of democratic representation in several Andean countries had ominous consequences for democratic regimes. In Peru and Venezuela, the perceived failures of traditional mechanisms of democratic representation, including most dramatically the collapse of the old party systems, paved the way for an erosion (Venezuela) and breakdown (Peru) of democracy. In both cases, political outsiders took advantage of the discrediting of the old parties, won the presidency, and began to attack and dismantle some key democratic institutions. In Bolivia, the decay of the major parties and the discrediting of conventional mechanisms of representation led to the forced ouster of President Gonzalo Sanchez de Lozada in 2003, closing a chapter in Bolivian history during which the prospects for democracy in a poor, ethnically divided country temporarily improved. One lesson ofMayorga's and Tanaka's chapters is that the deep discrediting of agents of democratic representation is often dangerous for democracies. In this respect, Tanaka's and Mayorga's analyses are relevant for the large number of countries where there is a crisis of democratic representation. In this sense, the problems that we address have implications for democratic and semi-democratic regimes in Mrica, Asia, the post-Communist world, and elsewhere in Latin America.
The Crisis of Democratic Representation in the Andes
5
Levine and Romero ask a different question about the consequences of a crisis of democratic representation. While Mayorga and Tanaka examine the consequences of a crisis of democratic representation for democratic regimes, Levine and Romero analyze the consequences for how poor urban citizens pursue their interests. In both Peru and Venezuela, politicians and parties failed to deliver what citizens wanted. Levine and Romero argue that the "discredit and decay of established leaders and parties combined with institutional failure and sustained economic crisis opened the way ... for a wide range of movements to emerge and claim a voice as 'civil society.'" Stated more generally, a crisis of democratic representation has profound consequences for citizen politics. Our fifth contribution is conceptual. We define and operationalize a crisis of democratic representation in this introductory chapter. In our definition, democratic representation is the relationship by which voters authorize representatives to govern. We argue that citizen satisfaction with the agents of democratic representation (politicians, parties, and assemblies) varies widely, and that this variance is expressed in both subjective and behavioral indicators. At the subjective level, citizens express more or less confidence in parties, politicians, and assemblies, and they view these agents as having more or less legitimacy. At the behavioral level, citizens vote or withdraw from electoral participation. They remain loyal to the same party over time, or they might switch party preference with frequency in order to find a more satisfactory agent to represent them. They continue to vote for establishment parties or search for anti-system candidates because of their dissatisfaction with the existing party options. We use the term "crisis of democratic representation" to refer to contexts in which, at the subjective level, citizens do not trust or confer legitimacy to agents of democratic representation. At the behavioral level, they are more likely to support anti-system candidates and parties, to turn to new parties, to switch electoral preferences with frequency, and to withdraw from electoral participation. The book also addresses other important questions. Can innovations in representation at the subnational level offset deficiencies at the national level? To what extent can institutional reforms of the formal mechanisms of democratic representation overcome perceived deficiencies in the system? Have the mechanisms designed to enhance representation of indigenous groups been good or harmful to democracy? Deborah Yashar (Chapter 9) sees these new mechanisms as advancing democracy, whereas Rene Mayorga (Chapter 5) argues that they have made democracy more inclusionary but that some indigenous groups have a utopian, anti-liberal-democracy discourse and practice. This book primarily addresses the literatures on political representation and democracy. It is one of the first books in English to analyze democratic representation in Latin America. By examining the crisis of democratic representation, we hope to contribute to the literature that seeks to understand why many competitively elected regimes around the world have huge deficiencies.
6
Mainwaring, Bejarano, Pizarro Leong6mez
In the vast literature on democracy, this book contributes to recent work on the severe shortcomings of many competitively elected regimes (e.g., O'Donnell2003). In most of Latin America and in some post-Communist countries, competitively elected governments have failed to deliver the goods, generating widespread dissatisfaction with democracy and concern about its future. We contribute to this literature by looking at the deficiencies of democratic representation. One of the fundamental arguments of this book involves the relationship between these two literatures, and in particular between a crisis of democratic representation and regime or state deficiencies. Tanaka, Mayorga, and Mainwaring argue that the crisis of democratic representation resulted largely from regime or state deficiencies. The profound delegitimation and repudiation of parties and politicians has paved the way for democratic breakdowns and erosions. Traditional agents of democratic representation -above all, political parties-may be deeply flawed, but democracy without parties is at best severely deficient, and at worst, as Schattschneider (1942, 1) wrote long ago, simply unthinkable. In the remainder of this introductory chapter, we undertake four tasks. First, we explain our reason for focusing on the Andes. Second, we examine why it has become meaningful to think of the five Andean countries as facing some common challenges. Third, we define the concept of democratic representation. Finally, we explain what a crisis of democratic representation is and examine empirical manifestations thereof
Case Selection: Why Focus on the Andes? If the problems that we are addressing are common throughout the world today, why focus on one specific region within Latin America rather than adopting a cross-regional research strategy such as that successfully pursued by Beissinger and Young (2002) in their book on state failure? We have two reasons. First, the Andean region is widely perceived as being in crisis, and its international importance has grown as a result of the crisis. The perceived deficiencies of democratic representation and the discrediting of parties are more acute in the Andes than in most of the rest ofLatin America. Therefore, it is a particularly good region for examining the subject at hand. Yet the Andean region, as a region, has not been studied in much detail. In contrast to the situation with the Southern Cone and Central America, there are few works on the Andes as a region (Burt and Mauceri 2004; Conaghan and Malloy 1994; Crandall et al. 2005; Drake and Hershberg, forthcoming; O'Neill2005). An attempt to fill this gap is important. This is not to claim, however, that the Andean region is unique in experiencing a crisis of democratic representation or in experiencing severely deficient competitive regimes. On this score, the Andean region is illustrative of many struggling democracies in and beyond Latin America.
The Crisis
cif Democratic Representation in the Andes
7
Second, regions of the world, such as Latin America, and, within them, subregions, such as the Andes, are important in world politics (Gleditsch 2002; Mainwaring and Perez-Lifian, forthcoming). Within the Andean region, there are powerful cross-national influences and demonstration effects. The rise of President Hugo Chavez in Venezuela (1998-present), for example, influenced the electoral victory of President Lucio Gutierrez in Ecuador (2003 -present), as well as the emergence ofEvo Morales, Bolivia's most famous leader of coca growers, as a viable presidential candidate in 2002. Morales was subsequently elected president in 2005. These five countries have created some regional organizations that have reinforced common influences and the sense of a regional identity. On May 26, 1969, the governments of these five countries signed the Cartagena Agreement, thus beginning an early process of regional integration. The current Andean Community consists of a set of organizations known as the Andean Integration System, which includes the Andean Parliament, the Andean Tribunal ofJustice, the Andean Presidential Council, the Andean Council of Foreign Ministers, and the Andean Corporation ofPromotion. 3 Common institutions and some common problems make a focus on the Andes a reasonable way to geographically delimit our study. Although our focus is the Andean region, an important part of our research design, especially when we explain a crisis of democratic representation, involves comparing the Andes to a broader set of countries. Without such comparison involving variance in the dependent variable (i.e., the extent to which the democratic representation is in crisis), it would be impossible to explain the outcome.
The Andean Region in Crisis The Andean region as understood in this book includes Bolivia, Peru, Ecuador, Colombia, and Venezuela. It does not include Chile and Argentina, although both countries have borders along the Andean range. These two countries are part of the Southern Cone. We include Venezuela as part of the Andean region even though most of its inhabitants see themselves as more geographically, ethnically, and culturally aligned with the Caribbean than with the Andean region. The same is true of the Colombian population living along the Atlantic Coast. The reason for including them is that both countries were part of the set of republics whose independence was established by Sim6n Bolivar, and they have long been part of the set of countries with common membership in Andean regional organizations. Until the late 1980s or the early 1990s, a book on democratic representation in the Andes would have made little intellectual sense. There would have been no compelling grounds for grouping these five countries together in terms of
8
Mainwaring, Bejarano, Pizarro Leong6mez
their systems of representation. The five countries faced very different political challenges from 1958 until the late 1980s. During these decades, Venezuela and to a lesser degree Colombia were among the most successful democracies in Latin America. In 1976-77, they were exceptional cases in the region; along with Costa Rica, they were two of three islands of democracy in a sea of authoritarianism. In contrast, the other three countries analyzed in this volume had only short-lived experiments with democracy before 1978. Bolivia had a semi-democratic regime from 1956 until 1964, followed by a string of mostly harsh military dictators from 1964 until 1982, interrupted only by two very short-lived efforts to install democracy in 1979 and 1980. Ecuador had semi-democratic regimes from 1948 until 1962 and from 1968 to 1970, but they were punctuated by military coups. Until1980, Peru's only experience with democracy was short-lived, lasting only from 1963 until1968. Peru also had semi-democratic regimes from 1945 to 1948 and from 1956 to 1962. Economically, too, there was a sharp contrast among these five countries until the late 1980s. In most of the post-1945 period, Venezuela had the highest per capita income in Latin America. Colombia was well behind Venezuela, but had a per capita income and a standard of living far higher than that found in Bolivia and Ecuador. In contrast, Bolivia has been one of the poorest countries in Latin America since the early twentieth century. Ecuador was also much poorer than Venezuela. These economic differences underpinned contrasts in democratic representation during those interludes in which democracy existed in Bolivia, Ecuador, and Peru. Venezuela was a wealthier, more educated, and more urban society than the other four countries, with corresponding differences for democratic representation. That today these five countries face some important common challenges in democratic representation is remarkable in view of their very divergent histories. The emergence of an intellectually interesting common puzzle about a crisis of democratic representation in the 1990s reflects the confluence ofstriking changes in these five countries. One change is positive. With the exception of the Peruvian breakdown of democracy in 1992, all five countries have had democratic or semi-democratic regimes since the Bolivian transition to democracy in 1982. While Colombia and Venezuela underwent early transitions to competitive political regimes in 1958 and 1959, respectively, 4 Ecuador, Peru, and Bolivia restored such regimes during what Huntington (1991) called the "third wave of democratization" at the end of the 1970s and the beginning of the 1980s. Other changes have been inauspicious, and a shared sense of crisis that has roiled the five countries has contributed to the relevance of analyzing them as a subregion within Latin America. Venezuela's once solid democracy began to face serious challenges in 1989 with the outbreak of massive popular protests against President Carlos Andres Perez. In 1992, a military coup led by future president Hugo Ch:ivez failed, but it nonetheless signaled the growing disenchantment
The Crisis
cif Democratic Representation in the Andes TABLE
9
1.1
Support for Democracy, Latin America, 2005 % of respondents who unconditionally
Country or region
Bolivia Colombia Ecuador Peru Venezuela Average-Andean Region Total-Latin America SOURCE:
favor democracy
49 46
43 40 76 51 53
1Atinobar6metro survey, 2005.
with the existing political system. Deepening repudiation of the establishment parties made possible Chavez's election in 1998. Colombia's democracy also eroded in the 1990s, victim of an armed conflict between drug lords, paramilitary forces, left-wing guerrillas, and of a weakened state in the rural areas (Bejarano and Pizarro Leong6mez 2005). Once Venezuela and Colombia stood out as more democratic than their Andean counterparts; by the 1990s, the challenges they faced were more similar to those of their Andean neighbors than had been the case in the previous four decades. Thus, a region that was once characterized by profound contrasts in terms of democratic representation, running the gamut from relatively stable and legitimate patterns of democratic representation in Venezuela and Colombia to dictatorships in the other three countries during much of the 1960s and 1970s, started to acquire some similarities. In all five countries, political outsiders burst onto the scene and challenged for the presidency-successfully in Bolivia (2005), Ecuador (2002), Peru (1990), and Venezuela (1998). 5 In all five countries, electoral volatility escalated, reflecting citizen discontent with existing party options. In all but Venezuela, 6 public opinion reflected poor evaluations of parties and Congress, two pillars of democratic representation. Support for democracy is fairly low in all the countries except Venezuela, as Table 1.1 shows. The table gives the percentage of survey respondents who agreed that "Democracy is always the best form of government." Respondents were given two other options: (1) "For people like me, the form of government does not matter"; and (2) "Under some conditions, an authoritarian regime is better." Economically, too, there has been some convergence among these five countries, mainly as result of Venezuela's protracted economic demise. A country that once prospered relative to its Andean neighbors no longer stands out so distinctively. In 1960, Venezuela's per capita income was 3.4 times greater than Colombia's; by 2002, its per capita income was only 31 percent higher. During these four decades, Venezuela's per capita income slid from $3,720 to $2,979,
Mainwaring, Bejarano, Pizarro Leong6mez
IO
TABLE
1.2
GDP per Capita in the Andes, 1960-2002 (constant 1995 U.S. dollars)
Bolivia Colombia Ecuador Peru Venezuela SOURCE:
%change,
%change,
1960
1980
2002
1960-2002
1980-2002
$ 848 1,104 1,090 1,875 3,720
$1,014 1,868 1,816 2,569 3,991
$ 940 2,282 1,796 2,380 2,979
11 107 65 26 -20
-7 22 -2 -7 -25
World Bank, World Development Indicators Database.
while the per capita income of Colombia and Ecuador increased substantially, and that of Bolivia and Peru increased modestly (11 percent and 26 percent, respectively) (Table 1.2). From 1980 to 2002, per capita GDP fell in four of the five countries, all but Colombia, with a particularly protracted and steep decline in Venezuela. Colombia, which enjoyed modest economic growth during the 1980s and first half of the 1990s, has experienced economic stagnation since the mid-1990s. The negative per capita economic growth for the region, coupled with poor job generation, has led to increasing poverty. According to data of the Economic Commission for Latin America and the Caribbean, in 2001, 61 percent ofBolivians, 60 percent ofEcuadorians, 55 percent of Colombians, 49 percent ofPeruvians, and 49 percent of Venezuelans lived in poverty. Poor economic growth and increased poverty have bred dissatisfaction with democracy, resulting in people's deteriorated image of two of the main pillars of representative democracy: parties and parliament. A third factor that has fostered convergence across these five countries in terms of representation has been the social dislocation caused by a marketoriented model of economic development. The industrial crisis due to the demise of import substitution industrialization and the turn toward marketoriented policies in the 1980s and 1990s was a turning point in Latin America's political development. Government withdrawal, fiscal crises, and policies favoring economic austerity limited the flow of resources needed to sustain parties founded on clientelistic (Colombia) and corporatist (Venezuela) networks (Hagopian 1998; Roberts, forthcoming). Some parties and party systems in Latin America (Chile, Uruguay, and Costa Rica) have adapted to the new challenges ushered in by the era of market-oriented economic policies, while others have not. Market-oriented models of economic growth and the decline in living standards for large sectors of society deepened the social chasm in most Latin American countries, especially in the Andean region, between groups either incorporated or unincorporated into the formal economy, social security, stable
The Crisis of Democratic Representation in the Andes
II
employment, unions, public services, and legalized neighborhoods. The unincorporated sectors form Hugo Chavez's electoral base in Venezuela. They are also the source of widespread social and political movements responsible for popular protests in Bolivia, Ecuador, and Peru. Since the mid-1990s, the Andean region has been the most volatile in Latin America. It has also been the region within Latin America of greatest concern to U.S. policy makers. Insufficient economic growth, rising poverty, increased economic inequality, disillusionment with the results of the democratic process, drug trafficking, and the risk that the armed conflict in Colombia will overflow the country's borders highlight the gravity of the situation throughout the region. 7
The Concept
of Democratic Representation
Because democratic representation is the central subject of this book, it is crucial to be clear about what we mean by this term. We use the term "representation" to denote a principal-agent relationship whereby A (the principal) authorizes B (the agent) to act on her behalf The clearest relationships of representation are those in which a clearly defined principal (an individual, a group, an association, the electorate, etc.) explicitly delegates a clearly defined agent to undertake a task. 8 Examples of explicit acts of delegation include voting for someone to represent one's interests or formally designating them to do so, creating a union to represent workers' interests, and hiring a lawyer to represent someone legally. Our definition of representation is narrower and more clearly delineated than some. Some prominent definitions of representation are impossible to operationalize. For example, Manin, Przeworski, and Stokes (1999a, 2) define representation as "acting in the interest of the represented," or as "acting in the best interest of the public." We downplay whether the agent is acting in the interest of the public or the represented and instead focus on whether a principal authorizes an agent to act on her behalf. It is extremely difficult to establish whether elected representatives are acting in the best interest of the public or of the represented. 9 Hence, by Manin et al.'s definition, it is very difficult to establish when a relationship of representation exists. 10 In a similar vein, Pitkin (1967, 209) argues that representing "means acting in the interests of the represented in a manner responsive to them." The definitions in Manin et al. and Pitkin do not stipulate that representation requires a principal-agent relationship. By their definitions, a vast and ill-defined range of actions purportedly undertaken on behalf of some people or of the public good could be understood as "representation." In contrast, our definition is easy to operationalize. We make no claim that elected representatives actually act on behalf of their constituents or the public good because of the difficulties in judging such claims.
I2
Mainwaring, Bejarano, Pizarro Leong6mez
Our understanding of "representation" also diverges from a usage found in many empirical studies in the advanced industrial democracies. Many such studies (Barnes 1977; Converse and Pierce 1986; Hill and Hurley 1999; McCrone and Kuklinski 1979; Pierce 1999; Thomassen 1999; Weissberg 1978; Wlezien 2004) measure representation by the degree of programmatic convergence between voters and representatives. 11 They implicitly assume that representation must be programmatic. In contrast, and in agreement with Kitschelt (2000), our definition allows for the possibility of clientelistic forms of representation, a possibility that is widely practiced in the Andean region (see the chapters by Pachano and Pizarro Leong6mez). The degree of programmatic convergence between voters and representatives might be important for the quality of representation, but it is an empirical issue rather than a defining characteristic of a relationship of representation. Although some definitions and uses of "representation" diverge from the one we employ, others are close to it (e.g., Manin 1997, 6-7). Democratic representation refers to the specifically democratic form of representation that is established when a voter (the principal) chooses an agent (a politician or a party) to represent her interests in a democratic regime. The core of democratic representation lies in the relationship between citizens, on the one hand, and elected politicians, parties, and assemblies, on the other. In this relationship, voters are the principals, and elected politicians, parties, and assemblies are the agents. Elections are the mechanism through which the relationship of democratic representation is produced and reproduced (Manin 1997). They not only provide the means to elect the representatives. They are also a mechanism through which citizens make their preferences known, sending a message-albeit a blunt oneto their representatives about their policy preferences. In principle, elections should guarantee that electoral accountability will be a periodic event. 12 For this reason, the institutions that regulate the way elections are structured and carried out, that is, electoral systems, are key in structuring representative institutions and relationships (] ackisch 1997). To qualify as democratic, a relationship of representation requires that free and fair elections take place. In addition, in the contemporary period, democratic representation implies a political system that affords nearly universal adult suffrage, respect for human rights and traditional civil liberties, and the subordination of the military to elected officials. Of course, political parties and politicians are not the only vehicles that express citizens' interests in democracies. Citizens also pursue their interests through social organizations that aggregate, articulate, and express interests, as well as through intermediaries and movements that translate those interests into the political arena and formulate public policy (see the chapters by Levine and Romero and by Yashar). In a strict sense, however, democratic representation occurs only between voters and their elected representatives in a democracy.
The Crisis
cif Democratic Representation in the Andes
13
Neighborhood associations, social movements, and non-governmental organizations do not fall within our definition of democratic representation in the narrow sense-that is, that form of representation specific to democracies-because they can function under a democratic or a non-democratic regime. Vibrant social movements can exist under authoritarian regimes. Their form of representation is not specifically democratic. Indeed, in their classic work on transitions to democracy, O'Donnell and Schmitter (1986) argued that social movements are frequently less dynamic under democracy than in the waning phases of authoritarianism. In addition, direct forms of collective action do not involve representation. With direct collective action, individuals mobilize to work for some outcome rather than authorizing an agent to do so. The web of representation under democratic regimes involves a wide range of different kinds of mechanisms. Democratic representation, however, implies a more specific relationship, namely, that between voters and their elected representatives in a democratic regime. This is the form of representation specific to modern mass democracies.
Shirking, Accountability, and Democratic Representation Democratic representation has been a perennial political problem. In all principal-agent relationships, agents inevitably acquire some autonomy with respect to the principal. The problem of democratic representation is the difficulty of ensuring that this autonomy is somewhat limited so as to promote some responsiveness of politicians and parties to citizen interests or to the public good -that is, to ensure that a formal relationship of representation works for the represented, as democratic theorists and citizens hope will be the case. Curbing agents' autonomy is more problematic in the relationship between voters and elected politicians than in many principal-agent relationships because of a huge information asymmetry between voters and politicians, the blunt character of the preferences transmitted from voters to their representatives, the relative infrequency of elections, and the difficulties of sanctioning agents who do not perform well until the next elections. A rational citizen could willingly cede some autonomy in decision making to her representative on the grounds of the representative's superior expertise on many issues (Dahl1970, 28-40; Pitkin 1967, 145; Rogowski 1981). Nevertheless, a rational citizen would desire some responsiveness to her interests by the representative. Through elections, the represented delegate to their representatives the power to make binding decisions, presumably in exchange for some measure of responsiveness and electoral accountability (Fearon 1999; Ferejohn 1999; Manin 1997; Maravall 1999; Pitkin 1967, 209; Powell 2000). We understand responsiveness as the policy congruence between citizens and their representatives (Stimson, MacKuen, and Erikson 1995) Y By electoral
I4
Mainwaring, Bljarano, Pizarro Leong6mez
accountability, we mean that voters periodically have a chance to choose different representatives. Elections do not guarantee that elected representatives will represent their constituents well (Downs 1957; Dunn 1999, 338-39; Fearon 1999; Ferejohn 1999; Przeworski et al. 1999; Manin 1997; Manin, Przeworski, and Stokes 1999b; Maravall1999; Schumpeter 1946; Stokes 1999). Elected politicians have great opportunities to shirk. Elections occur intermittently, and nothing assures that elected representatives will behave according to voters' preferences between elections. Tremendous information asymmetries between elected officials and the average voter give the former ample opportunities to behave with autonomy. As Ferejohn (1999, 137) succinctly summarizes, "Electoral punishment ... is a fairly blunt instrument, and incumbent officials will be, at best, only moderately responsive to public wishes." 14 If all relationships of democratic representation afford opportunities for shirking, this problem is particularly acute in countries with more pronounced information gaps between voters and politicians. In the Andes, the main agents of democratic representation (political parties, elected politicians, and assemblies) until recently had too much autonomy with respect to most principals-in particular, the large contingent of poor voters, who are formally represented by parties, politicians, and assemblies but whose capacity to influence political outcomes was seemingly marginal. 15 By "autonomy," we do not mean merely the independence that representatives need in order to make good decisions for the public good on issues where their technical expertise exceeds that of the common citizen. In the popular perception, representatives in the Andean countries enjoy another, far more pernicious autonomy-the ability to turn their backs on the electorate and function as a freewheeling, self-serving political class. In Latin America and in the Andean region in particular, until recently there was a chronic lack of political responsiveness to the masses. The information asymmetries between voters and representatives are much greater in the Andes, where most voters have limited education and little information about politics, than in the advanced industrial democracies.
What Is a Crisis of Democratic Representation? In this section, we explain what we mean by "a crisis of representation," provide empirical measures thereof, and indicate the theoretical underpinnings of these empirical measures. The idea of a crisis of representation has gained currency in Latin America (Cheresky 2003; Grompone 1996; Novaro 1994, 1995; Peruzzotti 2004), and it finds faint echoes in the extensive literature on the decline of trust in institutions and the weakening of parties in the advanced industrial democracies. Yet the social science literature has not clarified what the concept "crisis of democratic representation" means or how it can be empirically assessed. Without
The Crisis of Democratic Representation in the Andes
15
explicit comparative benchmarks, the notion of a crisis of democratic representation is underspecified. The legitimacy of democratic representation is a continuous variable. We use the term "crisis of democratic representation" to refer to one end of this continuum, at which citizens do not believe they are well represented. A crisis of democratic representation has an attitudinal!subjective and a behavioral component. The subjective component involves citizen perceptions: large numbers of citizens are dissatisfied with the way in which they are represented, or they may not feel represented at all. The represented (or those who should in principle be entitled to be represented but are not) believe that the putative terms of the principal-agent relationship of delegation are being broken. They do not believe that the representatives are effectively acting on behalf of the represented or of some common good. The existence of a relationship of democratic representation does not depend on whether the representatives are acting on behalf of the public good or of their own constituents, but citizens' perceptions of being adequately represented hinge on whether they believe the representatives are acting on behalf of some vision of the public good or of the citizens' interests. If citizens do not believe that representatives are acting on behalf of their constituents or of some vision of a public good, they have no reason to feel adequately represented. When such a perception of not being adequately represented is pervasive and more than transitory, it constitutes a crisis of democratic representation. A citizen may feel adequately represented either because she believes her agents of representation are attempting to further some public good or because she believes they are acting appropriately on behalf of her interests. A relationship of representation exists when citizens elect representatives, but citizen satisfaction with their representatives depends on their perception of how well the latter perform their duties. Representatives in two countries could carry out their jobs in the same way and obtain the same results in terms of government output, but citizen evaluations of their representatives could differ markedly in the two countries. 16 The continuum from greater to lesser satisfaction with and legitimacy of agents of democratic representation also has a behavioral aspect. Even if there were widespread citizen disaffection with existing agents of democratic representation, we would be reluctant to affirm that a crisis existed unless there were behavioral indications of repudiation of those agents. Citizens must reject existing mechanisms of democratic representation. They can do so by withdrawing from electoral participation, voting for new parties (especially anti-establishment ones), voting for political outsiders, turning to anti-system popular mobilizations, or joining revolutionary struggles, among other possibilities. "Crisis" is a useful heuristic concept, although not one that has a precise social scientific demarcation. The degree of disaffection with and rejection of
r6
Mainwaring, Bejarano, Pizarro Leong6mez
democratic representation is best thought of as a continuum rather than a dichotomy, with no precise cut point that enables one to categorize case A as a crisis and case B as a non-crisis. If there is no such precise cut point, how can we claim that there is a crisis of democratic representation in the Andes? Some empirical measures indicate where different country cases fit on the continuum, and these empirical indicators all locate the Andean cases toward the crisis end of the continuum. The concept "crisis of democratic representation" is not useful for intermediate cases, but it is useful for the unambiguous cases at this end of the continuum found in the Andes.
Confidence in Representative Institutions This section begins our analysis of different measures of disaffection with and rejection of democratic representation, that is, of our dependent variable. Our first measure is based on survey data on confidence in the agents of democratic representation-specifically, parties and nationallegislaturesY Trust in parties and parliament is a proxy for, not identical to, a judgment about whether a citizen feels properly represented. Nevertheless, in the absence of direct survey questions about whether citizens feel properly represented, information about trust in the agents of democratic representation is a good proxy. A crisis of democratic representation involves a situation in which citizens have very low trust or confidence in these agents of democratic representation. In surveys, in almost all democracies, citizens express low trust in parties and the National Congress. Hence, it is important to look at the data in comparative terms rather than absolute terms. How do citizen evaluations in country X compare with those in country Y, or how do citizen evaluations of institution A compare with those of institution B, or how do evaluations of institution Z change over time? Data from the Latinobar6metro surveys (Table 1.3) show a profound lack of trust in representative institutions in the Andes, especially in Ecuador, Colombia, and Bolivia. Even in the context of a region (Latin America) where democratic institutions were performing poorly, the Andean subregion stood out in terms oflow citizen trust in parties and the national assembly. Table 1.3 is based on responses to the question, "Please tell me how much trust [con.fianza] you have in each of the following groups, institutions, or persons mentioned on the list: a lot, some, a little, or no confidence?" We summed the percentage of positive answers ("a lot" and "some"). In 1996, the five Andean countries scored much worse than the Latin American average. Among the seventeen Latin American countries and Spain, Venezuelans and Colombians expressed the least trust in parties, and Bolivians were the fifth most negative. A similar story obtained in citizen evaluations of the National Congress. In 1996, citizen trust in Congress was worst among the eighteen countries in Colombia, third worst in Venezuela, and sixth worst in Bolivia.
The Crisis of Democratic Representation in the Andes TABLE
I7
1.3
Citizen Trust in Representative Institutions, Andean Countries % of respondents who express some or a lot of trust
1996
Bolivia Colombia Ecuador Peru Venezuela Average, five Andean countries Average, twelve other Latin American countriesa SOURCE: a
1997
2002
2003
political
national
political
national
political
national
political
parties
assembly
parties
assembly
parties
assembly
parties
16.3 11.3 18.3 18.5 11.3
21.5 14.8 26.9 32.9 18.8
20.4 21.1 15.5 20.6 20.8
31.3 28.9 19.5 26.1 29.6
9 10 7 13 19
16.0 14.0 9.0 23.0 37.0
6.0 9.0 5.0 8.0 14.0
15.1
23.0
19.7
27.1
11.6
19.8
8.4
23.5
29.3
31.4
38.1
NA
NA
11.8
1996, 1997, 2002, and 2003 Latinobar6metro.
Argentina, Brazil, Chile, Costa Rica, El Salvador, Guatemala, Honduras, Mexico, Nicaragua, Panama,
Paraguay, and Uruguay.
Trust in parties and Congress increased significantly in Venezuela after Hugo Chavez's election in 1998, but in the other four countries it has been chronically low. The upsurge in trust in parties and Congress in Venezuela presumably stems primarily from pro-Chavez voters who were disaffected under the pre1998 regime. In Bolivia, Ecuador, and Peru, between 1996 and 2003 trust in parties and the National Congress decreased. Many authors have asserted that confidence in public institutions in the advanced industrial democracies has eroded in recent decades (Dalton 1999; Dogan 1997; Hetherington 1998; Lipset and Schneider 1983; Pharr and Putnam 2000). How do levels of confidence in parties and parliament compare in Western Europe, the Andes, and the rest of Latin America? Table 1.4 looks at this question based on data in the World Values Surveys, which included Colombia, Peru, and Venezuela, but unfortunately not Bolivia and Ecuador. The crossregional comparison puts the Andean countries in a broader comparative perspective.18 Confidence in parties and parliament was much higher in Western Europe than in Colombia, Peru, and Venezuela. The average percentage of those expressing confidence in parliament for the seventeen Western European countries was 47 percent, and the average for the three Andean countries was only 23 percent. The average percentage who expressed trust in political parties was 25 percent for five Western European countries, 22 percent for the seven Latin American countries outside the Andes, and only 15 percent for the three Andean countries. Citizens in the Andes do not trust the institutions that are designed to represent them. Even if trust in parties and parliament has declined in the advanced industrial democracies, it remains far higher than in the Andes. 19 The low level of trust in
r8
Mainwaring, Bejarano, Pizarro Leong6mez TABLE
1.4
Confidence in Parties and Parliament, Select Countries, World Values Survey
Average for Western European countries Average for seven other Latin American countries Venezuela Colombia Peru
Confidence in parties(%)
parliament (%)
24.6
46.9
22.0 20.1 17.3 7.9
Confidence in
26.9 34.4
24.8 9.6
SOURCE: 1995-97 and 1999-2001 waves of the World Values Surveys. Figures are for 1999-2001 when a country was included in both waves. Peru and Venezuela are from the 1999-2001 wave; Colombia is from 1997. NOTE: Cell figures are the% of respondents who had a great deal or quite a lot of confidence in institutions. For confidence in parliament, we used all the Western European and Latin American countries included in the World Values Surveys of 1995-97 and 1999-2001. The averages for Western Europe and Latin America are unweighted. The seventeen Western European countries are Austria, Belgium, Denmark, Finland, France, Germany, Great Britain, Iceland, Ireland, Italy, Luxembourg, the Netherlands, Norway, Portugal, Spain, Sweden, and Switzerland. The seven other Latin American countries are Argentina, Brazil, Chile, the Dominican Republic, El Salvador, Mexico, and Uruguay. For confidence in parties, we used all seven Latin American countries; for Western Europe, we used Norway, Spain, Sweden, Switzerland, and West Germany.
the Andean parties and parliaments is troubling. We agree with Newton and Norris (2000, 52) that "an erosion of confidence in the major institutions of ... representative democracy is a far more serious threat to democracy than a loss of trust in other citizens or politicians."
Electoral Volatility A cns1s of democratic representation should also manifest itself in concrete, measurable behavioral results. In the next four sections we examine several such behavioral indicators using aggregate data about elections and patterns in party systems. Our first aggregate indicator is electoral volatility, the net share of votes that shifts from one party to any other party from one election to the next (Pedersen 1983; Przeworski 1975; Roberts and Wibbels 1999). High electoral volatility shows large numbers offloating voters, that is, voters who do not support the same party in most elections. Persistently high volatility-high volatility in at least two consecutive electoral periods-is a possible sign of a crisis of democratic representation. It shows that large numbers of voters are repeatedly seeking alternative representative vehicles, and hence suggests dissatisfaction with the quality of representation. Widespread dissatisfaction with representation could occur with
The Crisis of Democratic Representation in the Andes TABLE
19
1.5
Electoral Volatility and Share rf Vote for New Parties in Lower-Chamber Elections, Andean Countries
Colombia Venezuela Ecuador Bolivia Peru
volatility
Elections included for vote for new parties
Share of vote for new parties
1978-2002 1978-2001 1979-1998 1980-2002 1980-2001
1991-2002 1993-2001 1996-1998 1993-2002 1990-2001
27.3 39.0 17.0 32.4 60.0
Mean electoral volatility, lower
included for
chamber
22.1 31.3 36.4 39.8 51.9
Elections
SOURCE: Electronic dataset available from Scott Mainwaring. NOTE: New parties are operationalized as those that first competed in lower-chamber elections within the
last ten years.
low or moderate electoral volatility in the context of an oligopolistic electoral market,Z0 but usually when there is massive dissatisfaction with existing party options, it opens the door to high volatility. Table 1.5 provides data on volatility in lower-chamber elections in the five Andean countries. The data reinforce the idea that there is a crisis of democratic representation in the Andes, especially in Peru, which has had one of the highest levels of electoral volatility in the world. Since 1978, mean volatility in the lower chamber has been 22.1 in Colombia, 31.3 in Venezuela, 36.4 in Ecuador, 39.8 in Bolivia, and 51.9 in Peru. Among these five countries, only Colombia has had moderate volatility, and in Colombia volatility increased sharply in the elections of 1998 and 2002. Mainwaring and Torcal (2006) calculated electoral volatility for thirty-nine countries, including the five Andean countries, some advanced industrial democracies, some other Latin American countries, and some post-Communist cases. In terms of rank order, Colombia was almost exactly in the middle, with the 19th lowest volatility; Venezuela was 27th; Ecuador was 28th; Bolivia was 31st; and Peru was 36th. Among the nine Latin American cases included in the analysis, Peru, Bolivia, Ecuador, and Venezuela had the highest volatility. Bartolini and Mair (1990) analyzed electoral volatility in 303 electoral periods in thirteen Western European countries, and the mean volatility was 8.6, a small fraction of what it has been in the Andean countries. Thus, the data on electoral volatility support the argument that representative institutions in the Andes are undergoing intense citizen questioning. Table 1.5 also provides data on the share of the lower-chamber vote won by new parties. We operationalize new parties as those that competed for the first time within the last ten years. A high share of the vote allocated to new parties reflects dissatisfaction with all of the traditional parties (Zoco, forthcoming). In the advanced industrial democracies, with rare exceptions, such as the Italian
20
Mainwaring, Bejarano, Pizarro Leong6mez
elections of1993, even if citizens shift their votes away from one party to the next, most continue to vote for a party within the existing system. This measure is therefore a useful complement to the widely used data on electoral volatility. Indeed, for the purpose of assessing disgruntlement with the existing parties, it is a more useful measure. As Table 1.5 shows, new parties have been able to burst on the scene and become successful electoral contenders in the Andes. The data are especially dramatic for Peru, where on average 60 percent of the lower-chamber vote went to new parties (with a high of 93 percent in 1995), and Venezuela, where 39 percent did. Consistent with Sim6n Pachano's argument (Chapter 4) on the relative stability of the main party contenders in Ecuador, it is the Andean country where new parties have on average registered the lowest share of the vote. The flip side of the dramatic rise of new parties is the withering or disappearance of some of the traditionally major parties in these systems. In Venezuela, Acci6n Democratica (AD, or Democratic Action) is a shadow of the party that won the presidency five of seven times between 1958 and 1988. COPEI, which won the presidency the other two times between 1958 and 1988, no longer exists. In Peru, three of the four main parties of the 1980s-IU (Izquierda Unida, or the United Left), AP (Acci6n Popular, or Popular Action), and the PPC (Partido Popular Cristiano, or the Popular Christian Party)-have disappeared. In Bolivia, the Acci6n Democratica Nacionalista (ADN, or Nationalist Democratic Action), one of the three main contenders from 1982 until 2002, has been reduced to irrelevance. The two other main parties, the MNR (Movimiento Nacionalista Revolucionario, or Nationalist Revolutionary Movement) and the MIR (Movimiento de Izquierda Revolucionaria, or Movement of the Revolutionary Left) also suffered huge setbacks in 2005. The traditional parties that have survived bear a faint resemblance to what they once were. Bolivia's MNR led the 1952 revolution and inspired deep passion. The MNR, AD in Venezuela, and Peru's APRA (Alianza Popular Revolucionaria Americana, or American Popular Revolutionary Alliance) integrated the masses politically and forged strong loyalties and identities. Little if any of that fervor remains. In Colombia, traditional parties have experienced an electoral erosion, and independents and minor parties have occupied growing political space (see Chapter 3, by Pizarro Leong6mez).
Collapses of Party Systems The collapse of a party system is a dramatic and unusual expression of a crisis of democratic representation. It evinces repudiation not only of individual parties, but also of most of the existing parties. Citizens prefer to risk the unknown rather than sticking with the existing options. A party system collapse means a profound rejection of existing agents of democratic representation.
The Crisis of Democratic Representation in the Andes
2I
Zoco (forthcoming) operationalizes a collapse of a party system as a situation in which new parties gain more than 45 percent of the votes over the course of two consecutive lower-chamber elections. For operational purposes, "new parties" are those that won less than 5 percent of the lower-chamber vote in the previous election and did not have candidates for national political office (Congress or the presidency) in any election prior to that. We exclude the first two elections after the inauguration of a democratic or semi-democratic regime because it does not make sense to think of a system collapsing before it forms. According to this definition, there have been only three party-system collapses in recent decades: those ofitaly (1993), Peru (1995), and Venezuela (1998-2000). The Peruvian party system remains in disarray more than a decade after its collapse. In keeping with our argument that the magnitude of the crisis of representation is distinctive in the Andes, two of the three recent party-system collapses in the world's set of democracies have taken place in this region. Consistent with our earlier argument that there should be some congruence between individuallevel lack of trust in parties and aggregate behavioral indicators, we hypothesize that countries in which citizens have very low trust in parties should be more vulnerable to party-system collapse. This is in fact the case. Of the eighteen countries in the 1996 Latinobar6metro survey, Venezuela was the country with the lowest trust in parties, and it was second lowest to Colombia in trust in the national assembly. In the 1995 World Values Survey (WVS), Venezuela was among the countries with the lowest confidence in parties and parliament. In the WVS, Peru registered the lowest of any Latin American or Western European country in confidence in parties and parliament. In the 1980s and early 1990s, Italy was consistently the country in Western Europe in which citizens expressed the least trust in parties and parliament (Dogan 1997, 26; Listhaug 1995, 304). Thus, all three countries that experienced a party-system collapse in the 1990s had been characterized by extremely low trust in the institutions of representative democracy. This repudiation of parties and the national assembly was a key factor in the collapse of these three party systems. Even with a small number of party-system collapses, the evidence supports the hypothesis that collapse is more likely where trust in parties is low.
Outsider Presidential Candidates We use a behavioral indicator to assess dissatisfaction with democratic representation as expressed in presidential elections. Widespread dissatisfaction with representative institutions might affect presidential elections through the emergence of electorally competitive outsider candidates. Substantial support for outsider candidates expresses citizen dissatisfaction with conventional party options. We define an outsider candidate as someone who runs as an independent or on a new party label. Electorally competitive independent presidential candidates and
Mainwaring, Bejarano, Pizarro Leong6mez
22
TABLE
Average Share
Country
United States Ecuador Bolivia Venezuela Colombia Peru
1.6
of Vote
Won by Outsider Presidential Candidates in Five Recent Presidential Elections, Andean Countries
Elections included
1984-2000 1988-2002 1985-2002 1983-2000 1986-2002 1985-2001
Share of vote won by outsider candidates, most recent election
0.3% 58.9 51.3 40.2 66.5 27.9
Average share won by outsider candidates, last five elections
6.0% 17.5 22.1 26.5 28.5 32.7
SOURCE: Electronic dataset available from Scott Mainwaring.
candidates from new parties reflect disaffection with existing party options. As noted above, we operationalize a new party as one that won less than 5 percent of the lower-chamber vote in the previous election and did not have candidates for national political office in any election prior to the previous one. A candidate who runs as an independent or on a new party is more of an outsider than one who runs on an established party label. We exclude the first election after the inauguration of a democratic or semi-democratic regime. Table 1.6 presents data on the share of the vote won by outsider presidential candidates in the five Andean countries and, for a baseline comparison, the United States. 21 On average, outsiders have won between three (Ecuador) and five and one-half (Peru) times the share of the vote that they won in the United States-and this includes a U.S. election, 1992, with the most successful outsider candidate in recent U.S. history (Ross Perot). Outsiders won the election in Peru in 1990, Venezuela in 1993 and 1998, and Colombia and Ecuador in 2002. 22 This is an extraordinary political occurrence that has happened in few other Latin American countries. It manifests a repudiation of the existing system of democratic representation. Another outsider (Evo Morales) made it to the runoff round in the presidential election in Bolivia in 2002, and subsequently won in 2005 (but he does not meet our operational definition of an outsider in 2005 because of his party's success in 2002). The results ofBolivia's 2002 and 2005 elections signaled a profound erosion of the parties that had dominated Bolivian politics from 1982 until2002 (Mayorga 2005). In 1990, in Peru, Alberta Fujimori created a new party and easily defeated renowned author Mario Vargas Llosa in the presidential runoff. In 1993, in Venezuela, Rafael Caldera was the first presidential winner from outside the two parties (Acci6n Democratica and COPE I) that had dominated presidential elections
The Crisis of Democratic Representation in the Andes
23
from 1958 until 1988. The founder of COPEI and an ex-president, Caldera broke with his own party, formed an independent political movement, and won the election. As an ex-president, he was not a political outsider, but he was a party outsider because he ran outside the established parties. Moreover, he ran and railed against them (Crisp, Levine, and Molina 2003). His victory marked the beginning of the end of the two-and-a-half party system (COPEI and AD) that dominated Venezuela from 1973 until1988. Then, in 1998, Hugo Chavez, who had led a failed 1992 military coup, won the presidential election in Venezuela as a political outsider. Alvaro Uribe Velez, the winner of the Colombian election of 2002, was the first winning presidential candidate from outside the Liberal or Conservative Parties since the nineteenth century. Like Caldera in Venezuela, Uribe defected from his party (the Liberals) when he failed to win the presidential nomination. Although Uribe Velez came from the ranks of the Liberal Party, he ran as an independent backed by the Conservative Party, dissident liberals, and independent sectors. For the first time in the lengthy history of the Liberal Party, a dissident candidate defeated the official candidate (Horacio Serpa Uribe). It was also the first time since 1942 (except during the Frente Nacional of 1958-7 4) that the Conservative Party did not present a presidential candidate. Finally, in 2002, another ex-golpista military leader, Lucio Gutierrez, won the presidential runoff in Ecuador. Gutierrez led the 2000 coup that deposed President Jamil Mahuad. When Gutierrez was himself overthrown in April 2005, his vice president, Alfredo Palacio, also a political outsider, assumed the presidency. It is not only at the presidential level that political outsiders have displaced political parties. Peru is the most extreme example among these five countries in terms of the ability of outsiders to displace parties (Conaghan 2000). In Peru, in 2004, independent regional movements controlled 13 of the 25 regional governments and 1,634 of the 2,281 jurisdictions. 23 Among the Andean countries, only Venezuela and Colombia had even moderately institutionalized party systems before the 1980s. Nevertheless, the decay of parties and party systems across the Andes in the 1990s and in the first halfdecade of the twenty-first century is notable.
Electoral Participation Widespread dissatisfaction with democratic representation might lead to depressed electoral participation and/or increased numbers of spoiled ballots. If citizens lose their confidence that voting makes a difference in how well they are represented, they are presumably less likely to vote (Dalton 1988). On the other hand, the relationship between a crisis of representation and diminished electoral turnout and/or more spoiled ballots might be less clear than is the case with the other indicators we have used in this chapter, especially for cross-national
Mainwaring, Bejarano, Pizarro Leong6mez
24
fiGURE
1.1
Presidential turnout as a percentage of the eligible electorate in the Andes
100
80
bD
60
"u... "
40
"
~
p...
20
0 1960
1970
1980
1990
2000
Presidential Elections ----Peru ---- Venezuela
Ecuador Bolivia
-A-
Colombia
Sources: Nohlen (1993); International Institute for Democracy and Electoral Assistance (IDEA), http:/ /www.idea.int; Oficina Nacional de Procesos Electorales (ONPE), http:/ /www.onpe.gob.pe; International Foundation for Electoral Systems, http:/ /www.ifes.org; Latinamerica Press, http:/ I www.latinamericapress.org, based on UNICEF 2001 and INEI 2002 reports; Political Database of the Americas, http:/ /www.georgetown.edu/pdba; Consejo Nacional Electoral (CNE); World Factbook 2002, http:/ /www.odci.gov/cia/publications/factbook/; and Elections Around the World, http:/ /www.electionworld.org.
comparisons. Conclusions about cross-national differences in turnout are not straightforward because of differences in incentives to vote. Voting is obligatory in Bolivia, Ecuador, and Peru, and was obligatory in Venezuela from 1961 until 1999, but it has not been obligatory in Colombia or in Venezuela since 1999. One would expect substantially higher turnout with obligatory voting, other things being equal. Longitudinal within-country comparisons should still be useful, except perhaps for comparing Venezuela before and after 1999, because the switch from obligatory to optional voting could explain a decline in the first few years after 1999. Figures 1.1 (presidential elections) and 1.2 (lower-chamber elections) provide data on electoral participation. The trends are very similar in presidential and lower-chamber elections, and they show one more dimension of the crisis of democratic representation in the Andes: declining voting. In Bolivia, Ecuador,
The Crisis of Democratic Representation in the Andes FIGURE
25
1.2
Turnout as a percentage of the eligible electorate in lower-chamber elections in the Andes
100
•
.,;m:
,/
80
'''d\:c chi2 Pseudo R2 N
1996 Latinobar6metro. Standard error appears within parentheses.
SOURCE: NOTE:
'p < 'p < 'p < 'p