You are on page 1of 379

THE CRISIS OF DEMOCRATIC REPRESENTATION IN THE ANDES

THE CRISIS OF DEMOCRATIC REPRESENTATION IN THE ANDES

Edited by

Scott Mainwaring, Ana Mara Bejarano, and Eduardo Pizarro Leongmez

Stanford University Press Stanford, California 2006

Stanford University Press Stanford, California 2006 by the Board of Trustees 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. Printed in the United States of America on acid-free, archival-quality paper Library of Congress Cataloging-in-Publication Data The crisis of democratic representation in the Andes / edited by Scott Mainwaring, Ana Mara Bejarano, and Eduardo Pizarro Leongmez. 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 representationAndes Region. 2. DemocracyAndes Region. 3. Political cultureAndes Region. 4. Andes RegionPolitics and government. I. Mainwaring, Scott, 1954II. Bejarano, Ana Mara. III. Pizarro Leongmez, Eduardo. JL1866.C76 2006 320.98 dc22 2006005159 Typeset by G&S Typesetters, Inc. in 10/12 Bembo

CONTENTS
List of Tables and Figures vii Acknowledgments xi List of Contributors xiii 1. The Crisis of Democratic Representation in the Andes: An Overview 1 Scott Mainwaring, Ana Mara Bejarano, and Eduardo Pizarro Leongmez 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 Martn Tanaka 3. Giants with Feet of Clay: Political Parties in Colombia Eduardo Pizarro Leongmez 4. Ecuador: The Provincialization of Representation Simn Pachano 78

100

5. Outsiders and Neopopulism: The Road to Plebiscitary Democracy 132 Ren Antonio Mayorga PART II. DECENTRALIZATION, LEGISLATURES, AND DEMOCRATIC REPRESENTATION 6. Decentralized Politics and Political Outcomes in the Andes Kathleen ONeill 171

7. The Nature of Representation in Andean Legislatures and Attempts at Institutional Reengineering 204 Brian F. Crisp

vi

Contents PART III. 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 J. Yashar PART IV. CONCLUSION 10. State Deciencies, Party Competition, and Condence 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. Support for Democracy, Latin America, 2005 9 GDP per Capita in the Andes, 1960 2002 10 Citizen Trust in Representative Institutions, Andean Countries 17 Condence in Parties and Parliament, Select Countries, World Values Survey 18 Electoral Volatility and Share of Vote 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: Ination, GNP, and Subversive Acts, 1980 1999 51 Venezuela: Trends in Party Identication 52 Venezuela: Percentage of Seats in the Chamber of Deputies, 1973 2000 60 Venezuela: Number of Governors Elected, by Party, 1989 2000 62 Number of Lists that Registered for Senate and LowerChamber Elections, 1958 2002 84

Table 1.7. Table 2.1. Table 2.2. Table 2.3. Table 2.4. Table 2.5. Table 2.6. Table 3.1.

viii 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 6.1. Table 6.2. Table 7.1. Table 7.2. Table 7.3. Table 7.4. Table 7.5. Table 10.1. Table 10.2.

Tables and Figures Electoral Performance of Lists That Elected More Than One Senator, 19912002 86 Senate Seats Won by Political Parties and Movements, 19912002 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 of Deputies per Province, 2002 107 Share of Presidential Vote in First Round, 1978 2002 117 Size of Legislative 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 of Deputies, by Party, 1979 2002 125 Bolivia: Party Votes and Seats, Lower-Chamber Elections, June 30, 2002 156 Expenditure Decentralization in Latin America 174 Popular Election of Subnational Executives 175 Pre-Reform Intraparty Characteristics 208 Partisan Composition of Andean Legislatures Prior to Constitutional Reform 210 Partisan Composition of Andean Legislatures after Constitutional Reform 214 Bill Targets in the Colombian Senate before and after Electoral Reform 217 Public Condence in Congress, 1996 219 State Performance and Perceptions Thereof in the Andes, 1996 2005 299 Pearson Correlation Coefcients between State Performance and Condence in Parties and Parliaments, Country-Level Indicators 300 Determinants of Condence in Political Parties in the Andes 302 The Impact of Assessment of the National Economic Situation and Interpersonal Trust on Condence in Parties 303 Determinants of Condence in Congress in the Andes 304 Attitudes about Democracy and Representation in the Andes 307

Table 10.3. Table 10.4. Table 10.5. Table 10.6.

Tables and Figures Table 10.7. Table 10.8. Citizen Condence in Institutions in the Andes 309 Secondary Education and Urbanization in the Andes 315

ix

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 of legislators 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

ur rst 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 of Notre Dame, the Ford Foundation, and the CocaCola Company. The distant background to the book is that in the fall of 1998 Colombias foremost human rights lawyer, Gustavo Galln 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 Galln 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 of Jurists, 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 Foundations Santiago ofce 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 Mara Bejarano, Eduardo Pizarro Leongmez, 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 countriesissues of deep normative concern and intellectual interest to all three of us. Along with Miriam Kornblith, Ren Antonio Mayorga, Simn Pachano, and Martn 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 (Centro Boliviano de Estudios Multidisciplinarios),

xii

Acknowledgments

FLACSO-Ecuador, IESA (Instituto de Estudios Superiores de Administracin, Caracas), the Instituto de Estudios Peruanos (IEP), and the Universidad Los Andes (Bogot) 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 beneted from the stimulating comments of Michael Coppedge, Humberto de la Calle, Paul Drake, Myles Frechette, Luis Gallegos-Chiriboga, Frances 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 beneted 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 Mara Bejarano, Ren Mayorga, Simn Pachano, Eduardo Pizarro Leongmez, Catalina Romero, and Martn 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 Instituto de Estudios Superiores de Administracin (IESA), as we brought this book to fruition. Through this program, Martn 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 Mara 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 Bogot, where she also served as Director of the Center for Social and Legal Research (CIJUS). She co-edited Elecciones y democracia en Colombia, 19971998 (Universidad de los Andes, Fundacin Social, Veedura Eleccin Presidencial, 1998), and co-authored the chapter on Colombia in Frances Hagopian and Scott Mainwaring, eds., The Third Wave of 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 nishing 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 of 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 Murn 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

xiv

Contributors

in Latin America. His books include Conict 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 of 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 of 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. Ren Antonio Mayorga is a Senior Researcher at the Centro Boliviano de Estudios Multidisciplinarios (CEBEM). He is also Professor of Political Science at the Facultad Latinoamericana de Ciencias Sociales (FLACSO) in Ecuador, and in the Joint Masters 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 poltica al orden democrtico (CEBEM, 1991); Antipoltica y neopopulismo (CEBEM, 1995); and La cuestin militar en cuestin: Democracia y fuerzas armadas (CEBEM, 1994). He is editor of Democracia y gobernabilidad en Amrica Latina (ILDIS, 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 of Both Worlds? (Oxford University Press, 2001). Kathleen ONeill 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. Simn 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 representacin catica: Anlisis del sistema electoral ecuatoriano (FLACSO, 1998).

Contributors

xv

Eduardo Pizarro Leongmez is Professor at the Universidad Nacional de Colombia in Bogot and served as Director of its Instituto de Estudios Polticos 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 conict in Colombia, including Una democracia asediada: Balance y perspectivas del conicto 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 Ponticia Universidad Catlica del Peru. She obtained her Ph.D. at the New School for Social Research in 1989. She was Director of the Instituto Bartolom 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. Martn 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 Per (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). Deborah J. Yashar is Associate Professor of Politics 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 of Indigenous Movements and the Postliberal Challenge (Cambridge University Press, 2005) and Demanding Democracy: Reform and Reaction in Costa Rica and Guatemala, 1870s1950s (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 Mara Bejarano, and Eduardo Pizarro Leongmez

his book analyzes and explains the crisis of democratic representation in ve 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 nish their terms of ofce 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 ourished 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 citizensand how to comprehend the grave shortcomings of most existing democracies and semidemocracies in the region. Because the Andean countries provide clear examples

Mainwaring, Bejarano, Pizarro Leongmez

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 Prez-Lin, 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 Chvez 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 ve 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

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 dissatised? Why cant they nd a party option that satises them? We do not denitively 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 Simn Pachano, focus on institutional arrangements to explain deciencies 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 deciencies 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, Ren Antonio Mayorga (Chapter 5) and Scott Mainwaring (Chapter 10) see the crisis of democratic representation as stemming from governance (Mayorgas term) or state deciencies (Mainwarings focus). Mayorga asserts that deep dissatisfaction with democratic performance underlies the crisis of democratic representation and the rise of political outsiders. In the books 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 deciencies in many arenas, ranging from citizen security to corruption and economic performance. For both Mayorga and Mainwaring, the rise of political outsiders, declining condence in parties, high electoral

Mainwaring, Bejarano, Pizarro Leongmez

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 specically of parties and assemblies. Although neither this volume nor any other can denitively 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 Martn Tanaka (Chapter 2), Ren 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 Snchez 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 of Mayorgas and Tanakas chapters is that the deep discrediting of agents of democratic representation is often dangerous for democracies. In this respect, Tanakas and Mayorgas 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 Africa, Asia, the post-Communist world, and elsewhere in Latin America.

The Crisis of Democratic Representation in the Andes

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 fth contribution is conceptual. We dene and operationalize a crisis of democratic representation in this introductory chapter. In our denition, 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 condence 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 nd 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 deciencies at the national level? To what extent can institutional reforms of the formal mechanisms of democratic representation overcome perceived deciencies 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 Ren 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 rst 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 deciencies.

Mainwaring, Bejarano, Pizarro Leongmez

In the vast literature on democracy, this book contributes to recent work on the severe shortcomings of many competitively elected regimes (e.g., ODonnell 2003). 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 deciencies 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 deciencies. Tanaka, Mayorga, and Mainwaring argue that the crisis of democratic representation resulted largely from regime or state deciencies. The profound delegitimation and repudiation of parties and politicians has paved the way for democratic breakdowns and erosions. Traditional agents of democratic representationabove all, political partiesmay be deeply awed, but democracy without parties is at best severely decient, 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 ve Andean countries as facing some common challenges. Third, we dene 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 specic 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 deciencies of democratic representation and the discrediting of parties are more acute in the Andes than in most of the rest of Latin 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; ONeill 2005). An attempt to ll 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 decient competitive regimes. On this score, the Andean region is illustrative of many struggling democracies in and beyond Latin America.

The Crisis of Democratic Representation in the Andes

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 Prez-Lin, forthcoming). Within the Andean region, there are powerful cross-national inuences and demonstration effects. The rise of President Hugo Chvez in Venezuela (1998 present), for example, inuenced the electoral victory of President Lucio Gutirrez in Ecuador (2003 present), as well as the emergence of Evo Morales, Bolivias most famous leader of coca growers, as a viable presidential candidate in 2002. Morales was subsequently elected president in 2005. These ve countries have created some regional organizations that have reinforced common inuences and the sense of a regional identity. On May 26, 1969, the governments of these ve 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 of Justice, the Andean Presidential Council, the Andean Council of Foreign Ministers, and the Andean Corporation of Promotion.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 Simn Bolvar, 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 ve countries together in terms of

Mainwaring, Bejarano, Pizarro Leongmez

their systems of representation. The ve 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. Until 1980, Perus only experience with democracy was short-lived, lasting only from 1963 until 1968. Peru also had semi-democratic regimes from 1945 to 1948 and from 1956 to 1962. Economically, too, there was a sharp contrast among these ve 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 ve 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 reects the conuence of striking changes in these ve countries. One change is positive. With the exception of the Peruvian breakdown of democracy in 1992, all ve 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 ve countries has contributed to the relevance of analyzing them as a subregion within Latin America. Venezuelas once solid democracy began to face serious challenges in 1989 with the outbreak of massive popular protests against President Carlos Andrs Prez. In 1992, a military coup led by future president Hugo Chvez failed, but it nonetheless signaled the growing disenchantment

The Crisis of Democratic Representation in the Andes


Table 1.1 Support for Democracy, Latin America, 2005
% of respondents who unconditionally favor democracy

Country or region

Bolivia Colombia Ecuador Peru Venezuela AverageAndean Region TotalLatin America


SOURCE :

49 46 43 40 76 51 53

Latinobarmetro survey, 2005.

with the existing political system. Deepening repudiation of the establishment parties made possible Chvezs election in 1998. Colombias democracy also eroded in the 1990s, victim of an armed conict between drug lords, paramilitary forces, left-wing guerrillas, and of a weakened state in the rural areas (Bejarano and Pizarro Leongmez 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 ve countries, political outsiders burst onto the scene and challenged for the presidencysuccessfully in Bolivia (2005), Ecuador (2002), Peru (1990), and Venezuela (1998).5 In all ve countries, electoral volatility escalated, reecting citizen discontent with existing party options. In all but Venezuela,6 public opinion reected 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 ve countries, mainly as result of Venezuelas protracted economic demise. A country that once prospered relative to its Andean neighbors no longer stands out so distinctively. In 1960, Venezuelas per capita income was 3.4 times greater than Colombias; by 2002, its per capita income was only 31 percent higher. During these four decades, Venezuelas per capita income slid from $3,720 to $2,979,

10

Mainwaring, Bejarano, Pizarro Leongmez


Table 1.2 GDP per Capita in the Andes, 1960 2002
(constant 1995 U.S. dollars) 1960 1980 2002 % change, 1960 2002 % change, 1980 2002

Bolivia Colombia Ecuador Peru Venezuela


SOURCE :

$ 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 ve countries, all but Colombia, with a particularly protracted and steep decline in Venezuela. Colombia, which enjoyed modest economic growth during the 1980s and rst 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 of Bolivians, 60 percent of Ecuadorians, 55 percent of Colombians, 49 percent of Peruvians, and 49 percent of Venezuelans lived in poverty. Poor economic growth and increased poverty have bred dissatisfaction with democracy, resulting in peoples deteriorated image of two of the main pillars of representative democracy: parties and parliament. A third factor that has fostered convergence across these ve 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 Americas political development. Government withdrawal, scal crises, and policies favoring economic austerity limited the ow 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

11

employment, unions, public services, and legalized neighborhoods. The unincorporated sectors form Hugo Chvezs 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. Insufcient economic growth, rising poverty, increased economic inequality, disillusionment with the results of the democratic process, drug trafcking, and the risk that the armed conict in Colombia will overow the countrys 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 principalagent 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 dened principal (an individual, a group, an association, the electorate, etc.) explicitly delegates a clearly dened agent to undertake a task.8 Examples of explicit acts of delegation include voting for someone to represent ones interests or formally designating them to do so, creating a union to represent workers interests, and hiring a lawyer to represent someone legally. Our denition of representation is narrower and more clearly delineated than some. Some prominent denitions of representation are impossible to operationalize. For example, Manin, Przeworski, and Stokes (1999a, 2) dene 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 difcult 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 denition, it is very difcult 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 denitions in Manin et al. and Pitkin do not stipulate that representation requires a principalagent relationship. By their denitions, a vast and ill-dened range of actions purportedly undertaken on behalf of some people or of the public good could be understood as representation. In contrast, our denition 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 difculties in judging such claims.

12

Mainwaring, Bejarano, Pizarro Leongmez

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 denition 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 Leongmez). 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 dening characteristic of a relationship of representation. Although some denitions 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 specically 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 messagealbeit a blunt one to 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 ( Jackisch 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 ofcials. 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 of Democratic Representation in the Andes

13

Neighborhood associations, social movements, and non-governmental organizations do not fall within our denition of democratic representation in the narrow sensethat is, that form of representation specic to democraciesbecause 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 specically democratic. Indeed, in their classic work on transitions to democracy, ODonnell 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 specic relationship, namely, that between voters and their elected representatives in a democratic regime. This is the form of representation specic to modern mass democracies.

Shirking, Accountability, and Democratic Representation


Democratic representation has been a perennial political problem. In all principalagent relationships, agents inevitably acquire some autonomy with respect to the principal. The problem of democratic representation is the difculty 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 goodthat 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 principalagent 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 difculties 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 representatives superior expertise on many issues (Dahl 1970, 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).13 By electoral

14

Mainwaring, Bejarano, Pizarro Leongmez

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; Maravall 1999; 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 ofcials 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 ofcials 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 principalsin particular, the large contingent of poor voters, who are formally represented by parties, politicians, and assemblies but whose capacity to inuence 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 autonomythe 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 nds 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 claried 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 underspecied. 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 dissatised 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 principalagent 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 afrm 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 scientic demarcation. The degree of disaffection with and rejection of

16

Mainwaring, Bejarano, Pizarro Leongmez

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 t 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 rst measure is based on survey data on condence in the agents of democratic representationspecically, parties and national legislatures.17 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 condence 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 Latinobarmetro 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 of low citizen trust in parties and the national assembly. Table 1.3 is based on responses to the question, Please tell me how much trust [conanza] you have in each of the following groups, institutions, or persons mentioned on the list: a lot, some, a little, or no condence? We summed the percentage of positive answers (a lot and some). In 1996, the ve 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 fth 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 1.3 Citizen Trust in Representative Institutions, Andean Countries
% of respondents who express some or a lot of trust 1996 political parties national assembly political parties 1997 national assembly 2002 political parties national assembly

17

2003 political parties

Bolivia Colombia Ecuador Peru Venezuela Average, ve Andean countries Average, twelve other Latin American countriesa
SOURCE :
a

16.3 11.3 18.3 18.5 11.3 15.1 23.5

21.5 14.8 26.9 32.9 18.8 23.0 29.3

20.4 21.1 15.5 20.6 20.8 19.7 31.4

31.3 28.9 19.5 26.1 29.6 27.1 38.1

9 10 7 13 19 11.6 NA

16.0 14.0 9.0 23.0 37.0 19.8 NA

6.0 9.0 5.0 8.0 14.0 8.4 11.8

1996, 1997, 2002, and 2003 Latinobarmetro. Argentina, Brazil, Chile, Costa Rica, El Salvador, Guatemala, Honduras, Mexico, Nicaragua, Panama, Paraguay, and Uruguay.

Trust in parties and Congress increased signicantly in Venezuela after Hugo Chvezs 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-Chvez 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 condence 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 condence 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 Condence in parties and parliament was much higher in Western Europe than in Colombia, Peru, and Venezuela. The average percentage of those expressing condence 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 ve 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

18

Mainwaring, Bejarano, Pizarro Leongmez


Table 1.4 Condence in Parties and Parliament, Select Countries, World Values Survey
Condence in parties (%) Condence in parliament (%)

Average for Western European countries Average for seven other Latin American countries Venezuela Colombia Peru

24.6 22.0 20.1 17.3 7.9

46.9 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 gures are the % of respondents who had a great deal or quite a lot of condence in institutions. For condence 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 condence 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 condence 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 crisis 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 rst 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 of oating voters, that is, voters who do not support the same party in most elections. Persistently high volatilityhigh volatility in at least two consecutive electoral periodsis 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 1.5 Electoral Volatility and Share of Vote for New Parties in Lower-Chamber Elections, Andean Countries
Mean electoral volatility, lower chamber Elections included for volatility Elections included for vote for new parties

19

Share of vote for new parties

Colombia Venezuela Ecuador Bolivia Peru


SOURCE : NOTE :

22.1 31.3 36.4 39.8 51.9

1978 2002 1978 2001 1979 1998 1980 2002 1980 2001

19912002 1993 2001 1996 1998 1993 2002 1990 2001

27.3 39.0 17.0 32.4 60.0

Electronic dataset available from Scott Mainwaring. New parties are operationalized as those that rst competed in lower-chamber elections within the last ten years.

low or moderate electoral volatility in the context of an oligopolistic electoral market,20 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 ve 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 ve 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 ve 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 rst time within the last ten years. A high share of the vote allocated to new parties reects 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 Leongmez

elections of 1993, 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 Simn Pachanos 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 ip 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, Accin Democrtica (AD, or Democratic Action) is a shadow of the party that won the presidency ve 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 1980sIU (Izquierda Unida, or the United Left), AP (Accin Popular, or Popular Action), and the PPC (Partido Popular Cristiano, or the Popular Christian Party)have disappeared. In Bolivia, the Accin Democrtica 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. Bolivias MNR led the 1952 revolution and inspired deep passion. The MNR, AD in Venezuela, and Perus 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 Leongmez).

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

21

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 ofce (Congress or the presidency) in any election prior to that. We exclude the rst 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 denition, there have been only three party-system collapses in recent decades: those of Italy (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 worlds 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 Latinobarmetro 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 condence in parties and parliament. In the WVS, Peru registered the lowest of any Latin American or Western European country in condence 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 dene an outsider candidate as someone who runs as an independent or on a new party label. Electorally competitive independent presidential candidates and

22

Mainwaring, Bejarano, Pizarro Leongmez


Table 1.6 Average Share of Vote Won by Outsider Presidential Candidates in Five Recent Presidential Elections, Andean Countries
Share of vote won by outsider candidates, most recent election Average share won by outsider candidates, last ve elections

Country

Elections included

United States Ecuador Bolivia Venezuela Colombia Peru


SOURCE :

1984 2000 1988 2002 1985 2002 1983 2000 1986 2002 1985 2001

0.3% 58.9 51.3 40.2 66.5 27.9

6.0% 17.5 22.1 26.5 28.5 32.7

Electronic dataset available from Scott Mainwaring.

candidates from new parties reect 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 ofce 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 rst 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 ve Andean countries and, for a baseline comparison, the United States.21 On average, outsiders have won between three (Ecuador) and ve and one-half (Peru) times the share of the vote that they won in the United Statesand 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 denition of an outsider in 2005 because of his partys success in 2002). The results of Bolivias 2002 and 2005 elections signaled a profound erosion of the parties that had dominated Bolivian politics from 1982 until 2002 (Mayorga 2005). In 1990, in Peru, Alberto 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 rst presidential winner from outside the two parties (Accin Democrtica and COPEI) 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 until 1988. Then, in 1998, Hugo Chvez, who had led a failed 1992 military coup, won the presidential election in Venezuela as a political outsider. lvaro Uribe Vlez, the winner of the Colombian election of 2002, was the rst 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 Vlez 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 rst time in the lengthy history of the Liberal Party, a dissident candidate defeated the ofcial candidate (Horacio Serpa Uribe). It was also the rst time since 1942 (except during the Frente Nacional of 1958 74) that the Conservative Party did not present a presidential candidate. Finally, in 2002, another ex-golpista military leader, Lucio Gutirrez, won the presidential runoff in Ecuador. Gutirrez led the 2000 coup that deposed President Jamil Mahuad. When Gutirrez 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 ve 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 rst halfdecade of the twenty-rst 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 condence 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

24

Mainwaring, Bejarano, Pizarro Leongmez


Figure 1.1

Presidential turnout as a percentage of the eligible electorate in the Andes


100

80

Percentage

60

40

20

0 1960 1970 Peru Venezuela 1980 Presidential Elections Ecuador Bolivia 1990 Colombia 2000

Sources: Nohlen (1993); International Institute for Democracy and Electoral Assistance (IDEA), http://www.idea.int; Ocina Nacional de Procesos Electorales (ONPE), http://www.onpe.gob.pe; International Foundation for Electoral Systems, http://www.ifes.org; Latinamerica Press, http:// 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 rst 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 1.2

25

Turnout as a percentage of the eligible electorate in lower-chamber elections in the Andes


100

80

Percentage

60

40

20

0 1960 1970 Peru Venezuela 1980 Parliamentary Elections Ecuador Bolivia 1990 Colombia 2000

Sources: Nohlen (1993); International Institute for Democracy and Electoral Assistance (IDEA), http://www.idea.int; Ocina Nacional de Procesos Electorales (ONPE), http://www.onpe.gob.pe; International Foundation for Electoral Systems, http://www.ifes.org; Latinamerica Press, http:// 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.

and Peru there has been a modest decline in turnout over the extended period of time shown in Figures 1.1 and 1.2. The data also show signs of disenchantment with democratic representation in Venezuela and Colombia, though of a different nature in the two countries. Venezuela, for decades characterized by very high electoral participation, has experienced a sharp decline in turnout. Venezuelas turnout dropped somewhat in 1978 and 1988, and then plunged in 1993. Colombia has had chronically low turnout rates but without a clear negative trend since 1962. Thus, in Colombia there are indications of a chronic citizen lack of enthusiasm about democratic representation. None of the ve countries evinces a clear upward trend in the percentage of spoiled ballots (data not shown).

Table 1.7 Programmatic Representation in Latin America and Spain

Country

Nagelkerke R 2

Statistical signicance, Party 1 vs. Reference Party

Statistical signicance, Party 2 vs. Reference Party

Standardized coefcient, Party 1 standardized coefcient, Party 2

Mean leftright representation gap between parties and their median voters (110 scale)

Spain Uruguay Chile Paraguay Colombia Argentina El Salvador Mexico Costa Rica Brazil Venezuela Ecuador Honduras Peru Guatemala Nicaragua Panama Bolivia

.56 .55 .18 .25 .04 .14 .23 .21 .09 .19 .09 .02 .02 .04 .02 .01 .01 .01

.000 Not signicant at .10 .000 .000 .031 .087 .000 .000 .000 Not signicant at .10 .073 Not signicant at .10 .024 Not signicant at .10 Not signicant at .10 Not signicant at .10 Not signicant at .10 Not signicant at .10

.000 .000 .005 .024 .021 .000 .000 .000 .005 .000 .003 .004 Not signicant at .10 Not signicant at .10 Not signicant at .10 Not signicant at .10 Not signicant at .10 Not signicant at .10

4.66 2.86 1.84 1.72 1.68 1.64 1.57 1.54 1.48 1.32 1.27 1.12 1.09 1.00 1.00 1.00 1.00 1.00

NA 1.93 1.39 2.15 2.55 1.83 2.11 2.64 2.90 2.83 2.63 2.56 2.32 2.11 3.08 2.45 1.75 2.52

SOURCES : 1996 Latinobarmetro, Questions 38 and 40, for the multinomial logistic regression. Sources for mean gap between parties and their median voter: for voters leftright positions, World Values Survey 1995 97 for Argentina, Brazil, Chile, Mexico, Uruguay, and Venezuela; World Values Survey 1999 2001 for Peru; Latinobarmetro 1998 for Bolivia, Colombia, Costa Rica, Honduras, Ecuador, Guatemala, Nicaragua, and El Salvador. Sources for mean score of parties as perceived by deputies of all other parties: Manuel Alcntara, Director, Proyecto de Elites Latinoamericanas, Universidad de Salamanca (1994 2005). NOTES : Columns 2 to 5 are based on results of multinomial logistic regressions. Dependent variable: individuals party choice. Independent variable: individuals leftright location. Andean countries in italics.

The Crisis of Democratic Representation in the Andes

27

Programmatic Representation in the Andes


An extensive literature has analyzed programmatic linkages between voters and their representatives (Barnes 1977; Converse and Pierce 1986; Hill and Hurley 1999; Luna and Zechmeister 2005; Kitschelt et al. 1999, 309 44; McCrone and Kuklinski 1979; Miller and Stokes 1963; Weissberg 1978; Wlezien 1996; Wood and Andersson 1998).24 Programmatic representation is not the only form of democratic representation. Nevertheless, programmatic convergence between voters and their agents is an important ingredient of democratic representation. For democracy to function well, many elected politicians must be concerned about the success of public policy and hence about programmatic issues. If elected politicians focus only on supplying selective (clientelistic) goods for their constituents, democracy cannot function well (Guevara Mann 2001). We presume that in countries in which citizens are deeply dissatised with democratic representation, programmatic representation tends to be weak. If programmatic representation is strong, that is, if representatives and voters converge in their preferred ideological or programmatic positions, voters are probably more likely to be satised with parties and politicians. Conversely, voters are probably more likely to be dissatised if programmatic representation is weak. Programmatic representation occurs along broad ideological lines more than according to specic issues (Converse and Pierce 1986; Hinich and Munger 1994; Kitschelt et al. 1999, 336 39; Thomassen 1999). Therefore, if programmatic representation is functioning well, and if the leftright dimension effectively captures most of the salient issues in party competition (Sani and Sartori 1983), there should be a correspondence between voters ideological position and their party choice. For this reason, one way of assessing programmatic representation is to look at the extent to which voters ideological positions predict their preferred party. If voters ideological position is a poor indicator of their party choice, programmatic representation is probably weak. Columns 2 through 5 of Table 1.7 assess the variance in ideological voting for seventeen Latin American countries and Spain, based on a multinomial regression analysis using the 1996 Latinobarmetro. The dependent variable was party choice among the three largest parties (according to the survey responses) in each country. For each country, the reference category was the party that, based on binary logistic regressions for the three possible pairs of the three largest parties (results not shown), had the lowest standardized coefcient. The only independent variable in the equations was voters leftright position along a 0 10 scale.25 We rank-ordered the countries by multiplying the standardized coefcient for Party 1 by the standardized coefcient for Party 2.26 (The third party in each country is the reference category, so there is no coefcient for it.) The resulting product provides one rough indicator of how important ideological voting is in determining party choice.

28

Mainwaring, Bejarano, Pizarro Leongmez

Ideological voting, and by extension programmatic representation, were weak throughout Latin America except Uruguay and Chile.27 Ideological voting was extraordinarily weak in Ecuador, Bolivia, and Peru. Colombias rank in the penultimate column is deceptive because a small leftist party, with only ve party supporters in the survey, drove it up. In four of the ve Andean countries, all but Venezuela, the Nagelkerke Pseudo R2 was extremely low, under .05, and Venezuela was not much higher at .09. In Bolivia and Peru, in the competition among the three largest parties, voters ideological positions provided no leverage in predicting their party choice; the coefcients for voters ideological positions are not signicant at p .10. The nal column of Table 1.7 presents a second empirical indicator of programmatic representation in Latin America. For each party, we measured the mean distance between each of its voters positions on a leftright scale from 1 to 10 and its mean position on this scale as evaluated by deputies of other parties.28 We took each voters distance from the elite mean because calculating the representation gap by taking a mean score for all voters of the party could be misleading. In principle, scores for each party range from 0 (all voters of a given party position themselves at exactly the same point on the 1 to 10 scale as the mean for the deputies) to 9 (all voters locate themselves at one extreme end of the scale; all elites locate the party at the other extreme end). We then weighted each party by its share of voters in the survey to generate a country-level score. With the partial exception of Peru, the Andean countries had high representation gaps. In Venezuela, which had the largest representation gap in the Andean region, the mean distance between where a voter located herself on the 110 leftright scale and the mean elite position for her party was 2.63 points. A programmatic representation gap of 2.63 for an individual party indicates that if the party were located at 5.00 at the elite level, its median voter in terms of the mean distance from the 5.00 score at the elite level would locate herself at 7.63 or at 2.37a veritable chasm. The scores for Ecuador (2.56), Colombia (2.55), and Bolivia (2.52) also reect very large programmatic representation gaps between parties at the elite level and their voters. Programmatic representation has increased in Venezuela since 1998 and probably in Bolivia since 2002, when Evo Morales rst ran for president. The emergence of strong leftist presidential candidates produced political polarization. The sharp polarization has claried political options and raised the stakes of politics, and presumably therefore intensied programmatic competition. Nevertheless, programmatic competition was historically fairly weak in Venezuela. The weakness of programmatic representation in the Andes underscores that it is important to expand the scope of thinking about representation beyond what has been done for the advanced industrial democracies. An extensive literature on representation in the advanced industrial democracies has focused on programmatic representation. Yet an exclusive focus on programmatic representation

The Crisis of Democratic Representation in the Andes

29

does not take us very far in understanding the Andean region. The most important conclusion about programmatic representation in the Andes is its weakness. More important in the Andes is analyzing the widespread discrediting and turning away from conventional agents of democratic representation. As scholars expand the geographic scope of studies of representation beyond the advanced industrial democracies, it will be important to not only examine programmatic representation but also to look at alternative forms of representation, such as clientelism, and to consider not only how representation works but why it often fails to work in the perception of citizens. Building on Kitschelt (2000), with minor modications, we distinguish between programmatic, clientelistic, personalistic (mainly populist), and institutional-affective linkages between voters and parties. These linkages refer to the primary basis upon which a given voter supports a party or politician. First, with programmatic linkages, a voter chooses a party or candidate because of the congruence between her programmatic/ideological positions and the partys or candidates. Second, a voter may choose a party or candidate primarily on the basis of selective incentives that will personally benet the voter or some nonprogrammatically dened group (e.g., a neighborhood) of which the voter forms a part. In this case, a voter might cast a ballot for a politician or party even though a competitor is ideologically closer to her preferred position. By securing clientelistic goods, voters can advance their material interests in a way that would not be possible through public goods. When this occurs, clientelistic linkages are dominant. Third, a voter may choose a candidate on the basis of the candidates personality, without a strong link to ideological preferences or to sociological location. Finally, by institutional-affective linkages we mean that a voter supports a party based on a sense of loyalty to ita cultural/symbolic identication with the partyabove and beyond what can be explained on the basis of the voters programmatic and clientelistic interests. This kind of linkage has received no attention in the literatures on representation and on parties, yet it deserves some consideration. Examples where institutional-affective linkages probably help explain why voters remained attached to parties include the support of most poor voters for the Justicialist Party in Argentina even after it turned to market-oriented policies in the 1990s (Levitsky 2003); the Conservative and Liberal Parties in Colombia in the 1960s and 1970s, when the programmatic differences between them were narrow yet many citizens retained powerful traditional party loyalties that are probably not fully explained by clientelistic benets (Archer 1995); and the Blancos and Colorados in Uruguay, where strong party loyalties persisted for decades despite relatively small programmatic differences between them (Gonzlez 1991). Programmatic linkages are weak in the Andes, with the partial exception of Venezuela since 1998. Traditional affective linkages between citizens and parties

30

Mainwaring, Bejarano, Pizarro Leongmez

have profoundly eroded. Clientelistic linkages are alive and well, but in an era of state shrinking (especially in Bolivia and Peru among the Andean countries), the supply of public sector resources available to politicians for building clientelistic linkages has diminished. Finally, as several chapters in this volume underscore (see especially Tanaka and Mayorga), personalistic linkages have ourished given the deep discrediting of conventional institutional channels of democratic representation.

The Crisis of Institutionalized Democratic Representation


Our empirical indicators in this chapter have focused on the crisis of parties and assemblies at the national level. What is in crisis in the Andean region and in many struggling democracies throughout the world is these institutionalized channels of democratic representation. We have not focused on individual politicians as agents of democratic representation, an issue that is central to the chapters by Tanaka and Mayorga. The delegitimation and decay of party systems and the discrediting of assemblies has, as Mayorgas and Tanakas chapters show, paved the way for plebiscitarian forms of representation in which populist presidents displace parties as the primary vehicles of expressing the popular will (ODonnell 1994; Weyland 1999). Thus, it might be argued that personalistic, plebiscitarian representation is simply displacing more institutionalized democratic representation. If this is true, is it accurate to speak of a crisis of democratic representation? We believe that the answer to this question is afrmative. The institutional agents of democratic representationassemblies and partieshave long held a privileged position in democratic theory, and for good reason. In a democracy, representatives should programmatically advance the interests of voters. The expansion of personalistic representation subverts that central democratic principle. As Tanaka and Mayorga argue in their chapters, plebiscitarian representation easily erodes into less-than-democratic forms of governing, and indeed may pave the way to authoritarian or semi-democratic regimes, as occurred with President Alberto Fujimori in Peru in the 1990s and with Venezuelan President Hugo Chvez after 1998. Leaders who are elected on the basis of direct populist appeal, sometimes with demagogic claims and often with the express intention of weakening institutional forms of democratic representation, often undermine rather than strengthen democratic institutions (ODonnell 1994). What starts out as plebiscitary representation easily erodes into non-democratic or even antidemocratic representation. Parties are key agents of democratic representation for three reasons. First, they provide indispensable information shortcuts to voters (Downs 1957; Hinich and Munger 1994). If elections were organized exclusively around individual candidates rather than partly through parties, voters would face daunting difculties in

The Crisis of Democratic Representation in the Andes

31

obtaining the signicant information needed to make good electoral judgments. Second, for this reason, parties are important mechanisms of electoral accountability. If citizens want to vote the bums out, they need to be able to punish not only individual ofceholders, but also in most cases the political parties responsible for governing. Third, parties connect citizens to the state in a different way than other vehicles of interest-articulation because through elections they offer a means to state power. For good reason, Schattschneider (1942, 1) wrote that modern democracy is unthinkable save in terms of the parties. Thus, a profound erosion of parties is closely linked to a crisis of democratic representation. The rise of plebiscitarian representation does not compensate for and may evenas has happened frequently in the Andes exacerbate the crisis of democratic representation. Our focus in most of the book is democratic representation at the national level. As ONeill makes clear (Chapter 6), democratic representation can occur at the subnational level. Citizens might be satised with these subnational levels even if they are dissatised with their agents of representation at the national level. Many proponents of decentralization believe it builds closer linkages between voters and representatives, enhances electoral accountability and responsiveness, and hence can improve democratic representation. Subnational politics has become more important in addressing citizen needs as the process of decentralization has taken hold; hence the centrality of ONeills chapter in understanding democratic representation today in the Andes. Even if citizens were satised with their subnational agents, the overall system of democratic representation would be compromised if voters were disgruntled with their national agents. The national-level agents of democratic representation have the rst and foremost responsibility to resolve many pressing needs that citizens face. The national macroeconomic situation, for example, affects citizen well-being in a way that no subnational policies can adequately compensate. While some attention to subnational forms of democratic representation is important in the Andes, what transpires at the subnational level cannot adequately compensate for failures of democratic representation at the national level. Moreover, because major political parties function at both the national and subnational level, a crisis of democratic representation at the national level inevitably adversely affects the legitimacy of democratic representation at the subnational level. If decentralization weakens the national state, it might even exacerbate the perception of a deciency of democratic representation.

Political Reform and the Crisis of Democratic Representation


It would be inaccurate to picture these political systems as immobile, static, or reform averse. On the contrary, during the past two decades, the Andes have been a veritable laboratory for experimenting with institutional reforms. During the

32

Mainwaring, Bejarano, Pizarro Leongmez

1990s, all ve countries undertook major constitutional reforms.29 All have experimented widely with the laws governing elections and political parties, as the chapters by Pachano, Pizarro Leongmez, Tanaka, and Crisp make abundantly clear. As ONeill argues in her chapter, one of the most consequential reforms in the region has been the trend toward decentralization. Indeed, one of the fundamental themes of this volume is the seemingly endless effort to improve mechanisms of representation via political reform. Neither the political elites nor the electorate have remained paralyzed in the face of the erosion of the relations of representation. In an atmosphere of disenchantment and tension, the parties and the political elites, sometimes with the support of and other times under intense pressure from signicant actors in society (the indigenous movement, e.g.) have modied existing institutional arrangements, seeking to improve the representativeness of these political systems. The results of these reforms have been mixed. Despite signicant gains in terms of representation of previously excluded minorities, some problems can be traced in part to these reforms: the erosion of parties and, in some cases, the added difculties in achieving effective government. The mixed legacy of these reformist efforts stems in part from the fact that institutional reform produces some unintended, and at times undesirable, effects. The mixed results also stem from the inevitability of trade offs, something that has not been adequately addressed in the literature on institutional reform. The twin goals of representativeness and governability sometimes stand in tension. Efforts aimed at enhancing either one of these dimensions of democracy may have a deleterious impact on the other. Given the recent nature of the reforms and the limited number of cases, we do not undertake a rigorous assessment of their global long-term effects.30

The Chapters That Follow


The key agents of democratic representation are political parties, politicians, assemblies, and elected executives, and hence these agents and their relation to citizens (the represented) are at the core of our purview. Among the agents of democratic representation, in most countries political parties have historically occupied a particularly prominent position. In principle, democratic representation can occur with individual politicians rather than parties as the agents of representation.31 In practice, however, in most democracies parties are the primary means of representing and structuring interests in mass democratic politics (Sartori 1976). A deep discrediting and weakening of parties is therefore closely associated with a crisis of democratic representation. For this reason, Part I has three chapters on parties and their failures as agents of democratic representation in the Andes. The ways that parties represent, and the failures of parties as agents of representation, vary across the ve Andean countries. The failure of parties as agents

The Crisis of Democratic Representation in the Andes

33

of representation was most profound in the two cases of party-system collapse, Peru and Venezuela (see Tanakas chapter). This similarity, however, masks an equally important difference. In the aftermath of the collapse of the Peruvian party system, the rebuilding of institutionalized mechanisms of democratic representation has been painfully slow. Fujimoris disgraceful demise and his antiparty attitudes and practices left a greatly weakened party landscape. In contrast, Hugo Chvez has more successfully built a party in Venezuela. In Colombia, the traditional parties gradually eroded in the 1990s after a century of electoral dominance, but the party system did not collapse. The Conservative and Liberal parties continue to be the countrys most powerful electoral contenders. Among the traditional (pre-1990s) parties, they are electorally the strongest of all the parties in these ve countries. In Ecuador, notwithstanding a widespread perception that mechanisms of democratic representation have failed, the party system of the 1980s has so far remained in place (Pachanos chapter). In 2002, however, for the rst time, a political outsider won the presidency. In Bolivia, the three mainstays of the post-1982 party system remained competitive until 2002, when the ADN experienced a sharp demise. Since 2002, political outsiders have ourished, traditional mechanisms of representation have been in disarray, and direct popular mobilization has surged. The nal chapter in Part I, by Ren Antonio Mayorga, addresses a closely related theme: the emergence of outsider politicians in response to the crisis of parties and party systems in the Andes. As Mayorga demonstrates, in the Andes a climate of neopopulism prevails, in which multiple social sectors are demanding incorporation into the political system. This climate of neopopulism is not headed by the organized popular sectors, but instead by the urban unemployed, the indigenous communities, those selling their wares on the streets, other groups within the informal sectors, coca growers, and peasants. According to Mayorga, intense social and political pressures have confronted the party systems. In some cases, such as Peru and Venezuela, this pressure was more than parties could handle. Part II focuses on two important institutional issues related to democratic representation. Brian Crisps chapter analyzes National Congresses in the Andean region. Congresses are one of the most important bodies of representation, and they suffer from the same low credibility as parties. He addresses reforms intended to enhance democratic representation but concludes, as do Pachano (Chapter 4) and Mainwaring (Chapter 10), that such reforms have not countered a deepening popular sense that the mechanisms of democratic representation are not functioning adequately. Kathleen ONeill analyzes changes in the intergovernmental distribution of power, which affects democratic representation by allowing for (or not) the election of representatives at the local and state level. ONeills chapter raises an interesting question: Can innovations and apparent improvements in repre-

34

Mainwaring, Bejarano, Pizarro Leongmez

sentation at the local level offset serious deciencies at the national level? We are deeply skeptical, especially in cases where decision-making authority and resources are still centralized. In Part III, chapters by Daniel H. Levine and Catalina Romero and by Deborah Yashar address representation of popular groups under democracy. Popular organizations and movements do not t our denition of democratic representation because they can articulate interests under both democratic and authoritarian regimes. Moreover, some popular mobilizations involve direct participation in politics rather than representation. Nevertheless, these two chapters are essential for understanding the crisis of democratic representation. There is no clearer manifestation of this crisis than the repudiation some popular groups express for the agents of democratic representation. For this reason, these two chapters are clearly relevant to the subject of this book. An examination of the crisis of democratic representation requires attention to popular groups that epitomize the rejection of parties and legislatures. Levine and Romero focus on urban citizen demands and perceptions under democracy in the era of discredited formal mechanisms of democratic representation. Their chapter illuminates how poor citizens attempt to further their interests outside the formal channels of democratic representation when these channels are discredited. Some movements explicitly reject the traditional vehicles of democratic representation. Some prefer direct participation or selfrepresentation (Warren and Jackson 2002) to traditional forms of representation. Deborah Yashar examines an increasingly important and often mobilized group in the Andes, especially in Ecuador and Bolivia: the indigenous population. In Ecuador, Colonel Lucio Gutirrezs surprising electoral victory in the 2002 presidential election was due in part to support from CONAIE (Confederacin de Nacionalidades Indgenas del Ecuador, or Confederation of Indigenous Nationalities of Ecuador) and, above all, from its political branch, Pachakutik. In Bolivia, Evo Morales placed second in the 2002 presidential election and won in 2005 by mobilizing substantial support from indigenous peoples. This mobilization of indigenous and popular communities is simultaneously a means of widening the democratic spectrum and, as Mayorga argues in his chapter, a deep source of political tension. An antagonism has grown between the streets as an expression of social mobilization, and Congress as an expression of institutionalized democratic representation.32 Often, social mobilization is not seen as a complement to or reinforcement of institutionalized political activity, but instead as an alternative and, in many cases, an anti-systemic alternativean instrument with which to change elected leaders via unconstitutional means or even change the system by extra-institutional de facto means. Yashars chapter also calls attention to an aspect of representation that has surfaced anew in recent years: the desire of some indigenous groups that people of their shared ethnicity represent them.

The Crisis of Democratic Representation in the Andes

35

The concluding chapter by Scott Mainwaring addresses the causes of low condence in parties and assemblies (excepting Venezuela since Chvez took ofce). In an argument that shares similarities with Mayorgas, Mainwaring argues that state deciencies and the politicization of these shortcomings by competing parties are the primary causes of the crisis of democratic representation. Although this concluding chapter focuses on the Andes, we believe that state deciencies are central to understanding failures of democratic representation well beyond this region. State performance is key to understanding the vicissitudes of many struggling competitive regimes in the world.

Conclusion
Many authors have claimed that conventional mechanisms of democratic representation are undergoing questioning and face declining legitimacy in most contemporary democracies (Pharr and Putnam 2000; Pizzorno 1981). This chapter relativizes these claims. Parties may have faced some erosion in the advanced industrial democracies, butwith the exception of Italy in the early 1990sthey have faced nothing resembling the profound questioning that they now do in all ve Andean countries. The Andes show what a real crisis of democratic representation is. We hope in this chapter to have contributed to thinking about what a crisis of democratic representation is and how it can be measured. This subject is one of the fundamental issues in the Andean region today, and indeed well beyond the Andes. If citizens believe that they are not well represented for an extended period under democracy, democracy itself is easily imperiled, as has occurred in Peru, Venezuela, Bolivia, and Ecuador at different moments in the 1990s and the rst decade of the twenty-rst century. The growing dissatisfaction with democracy in most of Latin America (UNDP 2004) suggests that what has transpired already in the Andean region may be a harbinger of things to come in the rest of Latin America. Understanding why citizens believe that democratic representation is failing them and addressing these shortcomings is one of the huge intellectual and political challenges of our day. The rest of the volume explores the failings and successes of democratic representation in the Andes, the causes of the failings, and the consequences of this crisis. APPENDIX Coding Rules for Outsider Presidential Candidates
1. After a democratic transition, we did not count the rst presidential election. 2. A new party is one that didnt win more than 5 percent of lower-chamber votes in the previous election and that did not present any candidates for the National Congress or presidency prior to that.

36

Mainwaring, Bejarano, Pizarro Leongmez

3. If a party changed its name from Election t to Election t 1, we did not count it as a new party at t 1. 4. We did not count an alliance (coalition) of previously existing parties as a new party. A coalition whose basis is not preexisting parties, however, counts as a new party. For example, the Movimiento S Colombia (Colombia Yes Movement) led by Noem Sann, which obtained 27 percent of the votes in the rst round in Colombia in 1998, counts as a new party. Although her party was composed mainly of leaders and followers from the traditional parties, Sann created a new electoral vehicle for her campaign. 5. We did not count a merger of two previously existing parties as a new party. 6. In cases of a party schism, neither of the resulting parties is counted as new. 7. We count as independents candidates who do not have a party afliation. For example, we count Alvaro Uribe Vlez as an independent in Colombia in 2002. Although he was still formally a member of the Liberal Party, he did not run on that partys ticket in 2002 but rather as an independent. In a similar vein, Claudio Fermn is counted as an independent after his expulsion from AD in Venezuela.

Notes
We are grateful to Michael Coppedge, Brian Crisp, Paul Drake, Frances Hagopian, Eric Hershberg, Mala Htun, Wendy Hunter, Herbert Kitschelt, Soledad Loaeza, Gerry Mackie, Ren Antonio Mayorga, Carlos Melndez, Ken Roberts, Martn Tanaka, Matthew Shugart, Kurt Weyland, Deborah Yashar, and Edurne Zoco for comments. We also thank Edurne Zoco, Angel Alvarez, Bong-Jun Ko, and Kathleen Monticello for research assistance. 1. Manin et al. (1999a, 1999b) provide an important discussion of this problem at a general abstract level but without thinking about the great cross-national variance in the satisfaction with or repudiation of mechanisms of democratic representation. We reverse that focus: we provide cursory attention to the general reasons why democratic representation might fail and instead examine why it fails particularly in the Andes. 2. The literature on political disaffection has some relevance to our work; it, too, understands that representation sometimes fails. For a recent example, see Torcal and Montero (2006). 3. To this list one might add the Universidad Simn Bolvar, the Business Advisory Council, the Labor Advisory Council, and many other regional institutions. 4. In Colombia, which has had a persistent history of civil governments, a conservative restoration of civil institutions took place in 1958. In Venezuela, a process of establishing democratic institutions took place in 1958. Other than the short interlude from 1947 to 1948, Venezuela had no democratic tradition before 1958. Despite this difference, the agreement upon which Venezuelan democracy was founded (known as the Punto Fijo Agreement) was more inclusive than that of the Colombian Frente Nacional and gave rise to a more open and participatory democracy (see Bejarano 2000; Levine 1992). 5. In Venezuela and Colombia, Rafael Caldera (1993) and lvaro Uribe Vlez (2002), respectively, although not political outsiders, gained the presidency via dissident movements of their own political parties, the COPEI and the Partido Liberal (Liberal Party). 6. Venezuelan exceptionalism in some attitudinal questions, including the one in Table 1.1 gauging voter support for democracy, requires a brief comment. Before the rise

The Crisis of Democratic Representation in the Andes

37

of Hugo Chvez in 1998, Venezuelans support for democracy and democratic representation was low. After 1998, support for democracy enjoyed a notable surge (Table 1.1), probably reecting the very different ways in which the pro- and anti-Chvez poles interpret the question. Pro-Chvez individuals see the president as democratic, and hence respond that democracy is always the best form of government. Anti-Chvez individuals see him as authoritarian, and hence agreealbeit while understanding the question in a completely different waythat democracy is always the best form of government. 7. Although we have focused in this section on convergence among these ve countries, they also continue to have important differences. 8. Brennan and Hamlin (1999), Christiano (1996, 20724), Ferejohn (1999), Fearon (1999), and Maravall (1999) also explicitly view representation through the prism of principalagent relationships. 9. Although the notion of the best interest of the public is intuitively appealing, the social-choice tradition (e.g., Arrow 1954) presented implicit critiques thereof that rendered this concept problematic. 10. Moreover, the two denitions provided by Manin et al. are not the same. Acting in the best interest of the public is not the same as acting in the interest of the represented. Acting in the best interest of the public may entail curbing wage raises at a given moment, whereas workers parties and politicians who represent workers would act in the interests of the represented (the workers) by pressing for wage increases. 11. Weissbergs (1978) denition is particularly distant from ours because it completely severs the electoral linkage between specic voters and their representatives: Our analysis denes representation as agreement between legislative voting and citizen opinion (535n4). In his conception, an elected representative from one district can represent a voter from another even though there is no electoral connection between them; the only issue that matters is programmatic convergence between a voter and a member of an assembly. 12. Electoral accountability requires that citizens have the opportunity to vote politicians out of ofce. This possibility is diminished where reelection is prohibited, as is the case in presidential systems with no reelection or with presidents who cannot be further reelected (e.g., the U.S. president in his second term). If reelection is banned, it is still possible to punish or reward an incumbents party, but not a specic politician unless she runs for another ofce. 13. On the relationship between representation and responsiveness, see Eulau and Karps (1977) and Powell (2000). 14. This is not to claim that electoral accountability is completely ineffectual. Fiorina (1981), Key (1966), Manin (1997), and Popkin (1991), among others, have underscored the potential for electoral accountability through retrospective voting. Stimson et al. (1995) emphasize both retrospective voting and anticipatory shifts in policy to respond to changes in public mood as mechanisms that produce electoral accountability. See Mansbridge (2003) for a discussion of different kinds of representation and their relationship to electoral accountability. 15. We italicize seemingly because it is difcult to measure the impact of the large contingent of poor voters on the way parties, politicians, and assemblies formulate public policy. 16. The fact that a crisis of representation hinges proximately on citizen evaluations does not imply that a crisis of representation is driven by purely subjective processes. As Mainwaring argues in Chapter 10, citizens form their judgments of the agents of representation partly on the basis of state performance. Citizens have bounded rationality in their assessment of the agents of representation. They form reasonably rational

38

Mainwaring, Bejarano, Pizarro Leongmez

judgments. When states perform badly over a protracted period of time, citizens are unlikely to believe that the agents of representation are fostering some public good or are delivering goods to them, and hence they are more likely to repudiate these agents. 17. Compared to the countless analyses on trust in institutions in the advanced industrial democracies, there is a paucity of work on this subject in Latin America, including the Andes. For exceptions, see Cleary and Stokes (2006); Power and Jamison (2005); Turner and Martz (1997). 18. The question in the World Values Survey was the same as in the Latinobarmetro, but the coding was different. The options were a great deal of trust, quite a lot, not very much, and none at all. 19. See Mishler and Rose (2001, 42) for data on trust in institutions in ten postSoviet countries. 20. As occurred, for example, in Colombia during the Frente Nacional period (1958 74), when the predominance of the two traditional parties, the Liberals and Conservatives, coincided with majority support from the population for these two parties. 21. Appendix 1 gives details on how we coded whether candidates were outsiders or not. 22. There are two types of outsiders: individuals who have never held political ofce and run against the establishment, such as Alberto Fujimori, Hugo Chvez, and Lucio Gutirrez, and those outside the party system, such as Rafael Caldera after his defection from COPEI and lvaro Uribe Vlez after leaving the Liberal Party. The latter are dissidents from traditional parties but are well-known political gures. 23. Data from Carlos Melndez (personal communication). In their chapters, Mayorga and Tanaka analyze some consequences of the rise of political outsiders. 24. Much more remains to be done on programmatic representation in Latin America; see Luna and Zechmeister (2005) on this issue. 25. The 1996 Latinobarmetro used a 0 10 scale rather than the conventional 110 scale. 26. If the standardized coefcient for either party was less than 1, we inverted it; that is, we divided 1 by the standardized coefcient. If the coefcient was not signicant at p .10, we adjusted the coefcient to 1.00 because it is not statistically different from 1.00. 27. For data that show the weakness of ideological voting in Latin America (except Chile and Uruguay) compared to most of the advanced industrial democracies, see Mainwaring and Torcal (2006). 28. The distance for each voter from the elite position is measured as an absolute value. The elite survey question we used excluded deputies from placing their own party on the leftright scale. The Latinobarmetro surveys asked citizens to place themselves on a 0 10 scale, whereas the World Values Survey and the Proyecto de Elites Latinoamericanas use the more common 110 scale. To make the Latinobarmetro scale commensurable with the other two, we used the formula .9LB 1 WVS, where LB is the voters score on the 0 10 scale and WVS is the voters score adjusted to the 110 scale. 29. Colombia in 1991, Peru in 1992 93, Bolivia in 1994, Ecuador in 199798, and Venezuela in 1999. 30. For an evaluation of the consequences of institutional reform in Colombia and Venezuela, see Bejarano (2002). On the impossibility of predicting with certainty the consequences of major constitutional changes, see Elster (1988). 31. The extent to which parties rather than individual politicians are the more salient agents of democratic representation varies by country (Dalton 1985; Esaiasson and Holmberg 1996).

The Crisis of Democratic Representation in the Andes

39

32. The debate raging throughout the Andean region between representative democracy and participatory democracy (which is a mislabel) is an expression of this conict between Congress and the street.

References
Archer, Ronald. 1995. Party Strength and Weakness in Colombias Besieged Democracy. In Building Democratic Institutions: Party Systems in Latin America, ed. Scott Mainwaring and Timothy R. Scully, 164 99. Stanford, CA: Stanford University Press. Arrow, Kenneth Joseph. 1954. Social Choice and Individual Values. New York: Wiley. Barnes, Samuel. 1977. Representation in Italy: Institutionalized Tradition and Electoral Choice. Chicago: University of Chicago Press. Bartolini, Stefano, and Peter Mair. 1990. Identity, Competition and Electoral Availability: The Stabilisation of European Electorates, 1885 1985. Cambridge: Cambridge University Press. Beissinger, Mark R., and Crawford Young, eds. 2002. Beyond State Crisis: Postcolonial Africa and Post-Soviet Eurasia in Comparative Perspective. Washington, DC: Woodrow Wilson Center Press. Bejarano, Ana Mara. 2000. Perverse Democratization: Pacts, Institutions, and Problematic Consolidation in Colombia and Venezuela. Ph.D. diss., Columbia University. . 2002. Buenas intenciones y efectos perversos: Los lmites del reformismo institucional en Colombia y Venezuela. Comentario Internacional: Revista del Centro Andino de Estudios Internacionales 4, II semester: 177 86. Bejarano, Ana Mara, and Eduardo Pizarro Leongmez. 2005. From Restricted to Besieged: The Changing Nature of the Limits to Democracy in Colombia. In The Third Wave of Democratization in Latin America: Advances and Setbacks, ed. Frances Hagopian and Scott Mainwaring, 235 60. Cambridge: Cambridge University Press. Brennan, Geoffrey, and Alan Hamlin. 1999. On Political Representation. British Journal of Political Science 29 ( January): 109 28. Burt, Jo Marie, and Philip Mauceri, eds. 2004. Politics in the Andes: Identity, Conict, Reform. Pittsburgh: University of Pittsburgh Press. Chalmers, Douglas, Scott Martin, and Kerianne Piester. 1997. Associative Networks: New Structures of Representation for the Popular Sectors? In The New Politics of Inequality in Latin America: Rethinking Participation and Representation, ed. Douglas Chalmers et al., 543 82. Oxford: Oxford University Press. Cheresky, Isidoro. 2003. Las elecciones nacionales de 1999 y 2001: Fluctuacin del voto debilitamiento de la Cohesin Partidaria y crisis de representacin. In De la ilusin reformista al descontento ciudadano: Las elecciones en Argentina, 1999 2001, ed. Isidoro Cheresky and Jean-Michel Blanquer, 19 51. Rosario, Argentina: Homo Sapiens Ediciones. Christiano, Thomas. 1996. The Rule of the Many: Fundamental Issues in Democratic Theory. Boulder, CO: Westview Press. Cleary, Matthew, and Susan Stokes. 2006. Democracy and the Culture of Skepticism: Political Trust in Argentina and Mexico. New York: Russell Sage Foundation.

40

Mainwaring, Bejarano, Pizarro Leongmez

Conaghan, Catherine M. 2000. The Irrelevant Right: Alberto Fujimori and the New Politics of Pragmatic Peru. In Conservative Parties, the Right, and Democracy in Latin America, ed. Kevin J. Middlebrook, 255 84. Baltimore: Johns Hopkins University Press. Conaghan, Catherine M., and James M. Malloy. 1994. Unsettling Statecraft: Democracy and Neoliberalism in the Central Andes. Pittsburgh: University of Pittsburgh Press. Converse, Philip. 1969. Of Time and Stability. Comparative Political Studies 2, no. 2: 139 71. Converse, Philip, and Roy Pierce. 1986. Political Representation in France. Cambridge, MA: Harvard University Press. Crandall, Russell, Guadalupe Paz, and Riordan Roett, eds. 2005. The Andes in Focus: Security, Democracy, and Economic Reform. Boulder: Lynne Rienner. Crisp, Brian, Daniel H. Levine, and Jos Molina. 2003. The Rise and Decline of COPEI in Venezuela. In Christian Democracy in Latin America: Electoral Competition and Regime Conicts, ed. Scott Mainwaring and Timothy R. Scully, 275 300. Stanford, CA: Stanford University Press. Dahl, Robert A. 1970. After the Revolution? New Haven, CT: Yale University Press. . 1971. Polyarchy: Participation and Opposition. New Haven, CT: Yale University Press. Dalton, Russell J. 1985. Political Parties and Political Representation: Party Supporters and Party Elites in Nine Nations. Comparative Political Studies 18, no. 3 (October): 26799. . 1988. Citizen Politics in Western Democracies: Public Opinion and Political Parties in the United States, Great Britain, West Germany, and France. Chatham, NJ: Chatham House. . 1999. Political Support in Advanced Industrial Democracies. In Critical Citizens: Global Support for Democratic Governance, ed. Pippa Norris, 5777. Oxford: Oxford University Press. Diamond, Larry. 1999. Developing Democracy: Toward Consolidation. Baltimore: Johns Hopkins University Press. Dogan, Mattei. 1997. Erosion of Condence in Advanced Democracies. Studies in Comparative International Development 32, no. 3 (Fall): 3 29. Downs, Anthony. 1957. An Economic Theory of Democracy. New York: Harper and Row. Drake, Paul, and Eric Hershberg, eds. Forthcoming. Crises in the Andes: Comparative Perspectives. Pittsburgh: University of Pittsburgh Press. Dunn, John. 1999. Situating Political Accountability. In Democracy, Accountability, and Representation, ed. Adam Przeworski, Susan C. Stokes, and Bernard Manin, 329 44. Cambridge: Cambridge University Press. Eaton, Kent. 2004. Politics Beyond the Capital: The Design of Subnational Institutions in South America. Stanford, CA: Stanford University Press. Elster, Jon. 1988. Arguments for Constitutional Choice: Reections on the Transition to Socialism. In Constitutionalism and Democracy, ed. Jon Elster and Rune Slagstad, 303 26. Cambridge: Cambridge University Press and Universitetsforlaget. Esaiasson, Peter, and Sren Holmberg. 1996. Looking for Political Representation. In Representation from Above: Members of Parliament and Representative Democracy in Sweden. Aldershot, England: Dartmouth Publishing. Eulau, Heinz, and Paul D. Karps. 1977. The Puzzle of Representation: Specifying Components of Responsiveness. Legislative Studies Quarterly 2, no. 3 (August): 233 54. Fearon, James D. 1999. Electoral Accountability and the Control of Politicians: Selecting Good Types versus Sanctioning Poor Performance. In Democracy, Accountability,

The Crisis of Democratic Representation in the Andes

41

and Representation, ed. Adam Przeworski, Susan C. Stokes, and Bernard Manin, 55 97. Cambridge: Cambridge University Press. Ferejohn, John. 1999. Accountability and Authority: Toward a Theory of Political Accountability. In Democracy, Accountability, and Representation, ed. Adam Przeworski, Susan C. Stokes, and Bernard Manin, 13153. Cambridge: Cambridge University Press. Fiorina, Morris P. 1981. Retrospective Voting in American National Elections. New Haven, CT: Yale University Press. Gleditsch, Kristian Skrede. 2002. All International Politics Is Local: The Diffusion of Conict, Integration, and Democratization. Ann Arbor: University of Michigan Press. Gonzlez, Luis Eduardo. 1991. Political Structures and Democracy in Uruguay. Notre Dame, IN: University of Notre Dame Press. Grompone, Romeo. 1996. Representacin poltica, partidos y escenarios futuros. Revista Socialismo y Participacin 73 (March): 1118. Guevara Mann, Carlos. 2001. Forsaken Virtue: An Analysis of the Political Behavior of Panamanian Legislators, 1984 1999. Ph.D. diss., University of Notre Dame. Hagopian, Frances. 1998. Democracy and Political Representation in Latin America in the 1990s: Pause, Reorganization, or Decline? In Fault Lines of Democracy in PostTransition Latin America, ed. Felipe Agero and Jeffrey Stark, 99 143. Miami: NorthSouth Center Press at the University of Miami. Hetherington, Marc J. 1998. The Political Relevance of Political Trust. American Political Science Review 92, no. 4 (December): 791 808. Hill, Kim Quaile, and Patricia A. Hurley. 1999. Dyadic Representation Reappraised. American Journal of Political Science 43, no. 1 ( January): 109 37. Hinich, Melvin, and Michael Munger. 1994. Ideology and the Theory of Political Choice. Ann Arbor: University of Michigan Press. Huntington, Samuel P. 1991. The Third Wave: Democratization in the Late Twentieth Century. Norman: University of Oklahoma Press. Jackisch, Carlota. 1997. Representacin democrtica y sistemas electorales. In Sistemas electorales y sus consecuencias polticas, ed. Carlota Jackisch, 7134. Buenos Aires: Centro Interdisciplinario de Estudios sobre el Desarrollo Latinoamericano (CIEDLA). Key, V. O. 1966. The Responsible Electorate: Rationality in Presidential Voting, 1936 1960. Cambridge, MA: Harvard University Press. Kitschelt, Herbert. 1990. New Social Movements and the Decline of Party Organization. In Challenging the Political Order: New Social Movements in Western Democracies, ed. Russell J. Dalton and Manfred Kuechler, 179 208. New York: Oxford University Press. . 2000. Linkages between Citizens and Politicians in Democratic Politics. Comparative Political Studies 33, no. 6/7 (August/September): 845 79. Kitschelt, Herbert, Zdenka Mansfeldova, Radoslaw Markowski, and Gbor Tka. 1999. Post Communist Party Systems: Competition, Representation, and Inter-Party Competition. Cambridge: Cambridge University Press. Levine, Daniel H. 1992. Popular Voices in Latin American Catholicism. Princeton, NJ: Princeton University Press. Levitsky, Steven. 2003. Transforming Labor-based Parties in Latin America: Argentine Peronism in Comparative Perspective. Cambridge: Cambridge University Press. Lipset, Seymour Martin, and Stein Rokkan. 1967. Cleavage Structures, Party Systems, and Voter Alignments: An Introduction. In Party Systems and Voter Alignments: CrossNational Perspectives, ed. Seymour Martin Lipset and Stein Rokkan, 1 64. New York: Free Press.

42

Mainwaring, Bejarano, Pizarro Leongmez

Lipset, Seymour Martin, and William Schneider. 1983. The Condence Gap: Business, Labor, and Government in the Public Mind. Baltimore: Johns Hopkins University Press. Listhaug, Ola. 1995. Condence in Political and Private Institutions. In Citizens and the State, ed. Hans Dieter Klingemann and Dieter Fuchs, 298 322. Oxford: Oxford University Press. Luna, Juan P., and Elizabeth J. Zechmeister. 2005. Political Representation in Latin America: A Study of Elite-Mass Congruence in Nine Countries. Comparative Political Studies 38, no. 4: 388 416. Mainwaring, Scott, and Anbal Prez-Lin. 2005. Latin American Democratization since 1978: Democratic Transitions, Breakdowns, and Erosions. In The Third Wave of Democratization in Latin America: Advances and Setbacks, ed. Frances Hagopian and Scott Mainwaring, 14 59. Cambridge: Cambridge University Press. . Forthcoming. Regional Effects and Region-Wide Diffusion of Democracy: Why Regions of the World are Important in Comparative Politics. In Regimes and Democracy in Latin America, ed. Gerardo Munck. Oxford: Oxford University Press. Mainwaring, Scott, and Mariano Torcal. 2006. Party System Institutionalization and Party System Theory after the Third Wave of Democratization. In Handbook of Political Parties, ed. Richard S. Katz and William Crotty, 204 27. London: Sage. Manin, Bernard. 1997. Principles of Representative Government. Cambridge: Cambridge University Press. Manin, Bernard, Adam Przeworski, and Susan C. Stokes. 1999a. Introduction to Democracy, Accountability, and Representation, ed. Adam Przeworski, Susan C. Stokes, and Bernard Manin, 126. Cambridge: Cambridge University Press. . 1999b. Elections and Representation. In Democracy, Accountability, and Representation, ed. Adam Przeworski, Susan C. Stokes, and Bernard Manin, 29 54. Cambridge: Cambridge University Press. Mansbridge, Jane. 2003. Rethinking Representation. American Political Science Review 97, no. 4 (November): 515 28. Maravall, Jos Mara. 1999. Accountability and Manipulation. In Democracy, Accountability, and Representation, ed. Adam Przeworski, Susan C. Stokes, and Bernard Manin, 154 96. Cambridge: Cambridge University Press. Mayorga, Ren. 2005. Bolivias Democracy at the Crossroads. In The Third Wave of Democratization in Latin America: Advances and Setbacks, ed. Frances Hagopian and Scott Mainwaring, 149 78. Cambridge: Cambridge University Press. McCrone, Donald J., and James H. Kuklinski. 1979. The Delegate Theory of Representation. American Journal of Political Science 23, no. 2 (May): 278 300. Miller, Warren E., and Donald E. Stokes. 1963. Constituency Inuence in Congress. American Political Science Review 57, no. 1 (March): 45 56. Mishler, William, and Richard Rose. 2001. Where Are the Origins of Political Trust? Testing Institutional and Cultural Theories in Post-Communist Societies. Comparative Political Studies 34, no. 1 (February): 30 62. Newton, Kenneth, and Pippa Norris. 2000. Condence in Public Institutions: Faith, Culture, or Performance? In Disaffected Democracies: Whats Troubling the Trilateral Countries? ed. Susan J. Pharr and Robert D. Putnam, 52 73. Princeton, NJ: Princeton University Press. Nohlen, Dieter, ed. 1993. Enciclopedia electoral de Amrica Latina y el Cribe. San Jos, Costa Rica: Instituto Interamericano de Derechos Humanos. Novaro, Marcos. 1994. Pilotos de tormentas: Crisis de representacin y personalizacin de la poltica en Argentina (1989 1993). Buenos Aires: Ediciones Letra Buena.

The Crisis of Democratic Representation in the Andes

43

. 1995. Crisis de representacin, neopopulismo y consolidacin democrtica. Sociedad (April): 95 117. ODonnell, Guillermo. 1994. Delegative Democracy? The Journal of Democracy 5, no. 1 ( January): 55 69. . 2003. Democracia, desarrollo humano y derechos humanos. In Democracia, Desarrollo Humano y Ciudadana, ed. Guillermo ODonnell, Osvaldo Iazzetta, and Jorge Vargas Cullell, 25 147. Rosario, Argentina: Homo Sapiens Ediciones. ODonnell, Guillermo, and Philippe Schmitter. 1986. Tentative Conclusions about Uncertain Democracies. Part 4 of Transitions from Authoritarian Rule: Prospects for Democracy, ed. Guillermo ODonnell, Philippe Schmitter, and Laurence Whitehead. Baltimore: Johns Hopkins University Press. ONeill, Kathleen. 2005. Decentralizing the State: Elections, Parties, and Local Power in the Andes. Cambridge: Cambridge University Press. Pedersen, Mogens N. 1983. Changing Patterns of Electoral Volatility in European Party Systems: Explorations in Explanation. In Western European Party Systems: Continuity and Change, ed. Hans Daalder and Peter Mair, 29 66. Beverly Hills, CA: Sage. Peruzzotti, Enrique. 2004. Crisis de representacin poltica y sociedad civil en Argentina. Poltica y Gobierno 11, no. 1 (Spring): 43 72. Pharr, Susan J., and Robert D. Putnam, eds. 2000. Disaffected Democracies: Whats Troubling the Trilateral Countries. Princeton, NJ: Princeton University Press. Pierce, Roy. 1999. Conclusion: Mixed Signals. In Policy Representation in Western Democracies, ed. Warren E. Miller, Jacques Thomassen, Richard Herrera, Sren Holmberg, Peter Esaiasson, and Bernhard Wessels, 9 32. Oxford: Oxford University Press. Pitkin, Hannah Fenichel. 1967. The Concept of Representation. Berkeley: University of California Press. Pizzorno, Alessandro. 1981. Interests and Parties in Pluralism. In Organizing Interests in Western Europe: Pluralism, Corporatism, and the Transformation of Politics, ed. Suzanne Berger, 247 84. New York: Cambridge University Press. Popkin, Samuel. 1991. The Reasoning Voter: Communication and Persuasion in Presidential Campaigns. Chicago: University of Chicago Press. Powell, G. Bingham, Jr. 2000. Elections as Instruments of Democracy. New Haven, CT: Yale University. Power, Timothy J., and Giselle D. Jamison. 2005. Political Mistrust in Latin America. Comparative Sociology 4, no. 12: 4772. Przeworski, Adam. 1975. Institutionalization of Voting Patterns, or Is Mobilization the Source of Decay? American Political Science Review 69, no. 1: 49 67. Przeworski, Adam, Susan Stokes, and Bernard Manin, eds. 1999. Democracy, Accountability, and Representation. Cambridge: Cambridge University Press. Riker, William. 1982. Liberalism versus Populism. San Francisco: Freeman. Roberts, Kenneth M. Forthcoming. Changing Course: Parties, Populism, and Political Representation in Latin Americas Neoliberal Era. Cambridge: Cambridge University Press. Roberts, Kenneth M., and Erik Wibbels. 1999. Party Systems and Electoral Volatility in Latin America: A Test of Economic, Institutional, and Structural Explanations. American Political Science Review 93, no. 3 (September): 575 90. Rogowski, Ronald. 1981. Representation in Political Theory and in Law. Ethics 91 (April): 395 430. Samuels, David J., and Matthew Soberg Shugart. 2003. Presidentialism, Elections, and Representation. Journal of Theoretical Politics 15, no. 1: 33 60.

44

Mainwaring, Bejarano, Pizarro Leongmez

Sani, Giacomo, and Giovanni Sartori. 1983. Polarization, Fragmentation, and Competition in Western Democracies. In Western European Party Systems, ed. Hans Daalder and Peter Mair, 307 40. Beverly Hills: Sage. Sartori, Giovanni. 1976. Parties and Party Systems: A Framework for Analysis. New York: Cambridge University Press. Schattschneider, Elmer E. 1942. Party Government. New York: Farrar and Rinehart. Schumpeter, Joseph. 1946. Capitalism, Socialism, and Democracy. New York: Harper and Row. Schwartz, Nancy L. 1988. The Blue Guitar: Political Representation and Community. Chicago: University of Chicago Press. Silveira, Flvio. 1998. A deciso do voto no Brasil. Porto Alegre: Edipucrs. Stimson, James, Michael B. MacKuen, and Robert S. Erikson. 1995. Dynamic Representation. American Political Science Review 89, no. 3 (September): 543 65. Stokes, Susan C. 1999. What Do Policy Switches Tell Us about Democracy? In Democracy, Accountability, and Representation, ed. Adam Przeworski, Susan C. Stokes, and Bernard Manin, 98 130. Cambridge: Cambridge University Press. Thomassen, Jacques. 1999. Political Communication between Political Elites and Mass Publics: The Role of Belief Systems. In Policy Representation in Western Democracies, ed. Warren E. Miller, Roy Pierce, Richard Herrera, Sren Holmberg, Peter Esaiasson, and Bernhard Wessels, 33 58. Oxford: Oxford University Press. Torcal, Mariano, and Jos Ramn Montero, eds. 2006. Political Disaffection in Contemporary Democracies: Social Capital, Institutions, and Politics. New York: Routledge. Turner, Frederick C., and John D. Martz. 1997. Institutional Condence and Democratic Consolidation in Latin America. Studies in Comparative International Development 32, no. 3 (Fall): 65 84. United Nations Development Programme. 2004. La democracia en Amrica Latina: Hacia una democracia de ciudadanas y ciudadanos. Lima: United Nations Development Programme. Warren, Kay B., and Jean E. Jackson, eds. 2002. Indigenous Movements, Self-Representation, and the State in Latin America. Austin: University of Texas Press. Weissberg, Robert. 1978. Collective vs. Dyadic Representation in Congress. American Political Science Review 72, no. 2 ( June): 535 47. Weyland, Kurt. 1999. Neoliberal Populism in Latin America and Eastern Europe. Comparative Politics 31, no. 4 ( July): 379 401. Williams, Melissa S. 1998. Voice, Trust, and Memory: Marginalized Groups and the Failings of Liberal Representation. Princeton, NJ: Princeton University Press. Wlezien, Christopher. 1996. Dynamics of Representation: The Case of U.S. Spending on Defence. British Journal of Political Science 26: 81103. . 2004. Patterns of Representation: Dynamics of Public Preferences and Policy. Journal of Politics 66, no. 1 (February): 124. Wood, B. Dan, and Angela Hinton Andersson. 1998. The Dynamics of Senatorial Representation, 1952 1991. Journal of Politics 60, no. 3 (August): 705 36. Zoco, Edurne. Forthcoming. The Collapse of Party Systems in Italy, Peru, and Venezuela. Ph.D. diss., University of Notre Dame.

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

Martn Tanaka
eru and Venezuela, countries with very different historical trajectories, shared similar political outcomes in the 1990s. Both their party systems collapsed, one result of exhausted statist economic policies and the implementation of rstgeneration market reforms, resulting in the subsequent establishment of authoritarian, though formally democratic, regimes under presidents Fujimori and Chvez. Such an outcome was unusual. In the rest of Latin America, party systems evolved and declining traditional parties coexisted with new parties, although not easily and in the middle of a crisis of representation. This gradual evolution of party systems, despite its limits, helped to maintain the checks and balances inherent to democratic rule. The explanation for the party system collapses is to be found less in structural causes or in the poor economic performance of both countries than in the responses of political actors to challenges posed by crises of representation at critical junctures, when the actors were especially vulnerable. The organization of the political parties was crucial in determining these responses; it explains the appearance of sharp internal conicts and divisions that accentuated problems of representation and enabled leaders outside the system to come to power through the electoral route. Both Alberto Fujimori (Peru, 1990 2000) and Hugo Chvez (Venezuela, 1998 present) are personalistic leaders with neopopulist and antisystem discourses who expressed the popular dissatisfaction with traditional actors. Both presidents were relatively effective in dismantling the preexisting political order through institutional reforms that, although formally democratic, in practice created authoritarian governments. These leaders represented some previously excluded popular sectors, not under democratic, but under plebiscitary schemes. The collapse of the party systems ended the political balance that had existed in both countries, and new state institutions were created under the hegemony of a single political actor, which led to a concentration of power that

48

Tanaka

ultimately ended the checks and balances inherent to democratic rule. These kind of regimes leave an onerous heritage that confronts both countries with enormous challenges: how to rebuild state institutions when the political and social actors are greatly weakened; and how to construct a new system of democratic representation that is pluralistic and participatory, while at the same time ensuring governability in crisis contexts in which the expectations and demands of the population are high.

Peru and Venezuela: Different Trajectories, the Same Results, and Some Explanations
From very different trajectories, Peru and Venezuela confronted crises, marked by the exhaustion of statist economic policies and the implementation of the rst generation of market reforms, which devastated the entire region in the 1980s and 1990s. Venezuela, with a relatively long democratic history dating from the end of the 1950s, exhibited a stable party system that featured cooperative behaviors and centripetal political competition. The parties were representative, rmly established in society, with links to various spheres of civil society. Their rootedness had its source in a long period of economic growth, which fostered social integration and the civic involvement of excluded sectors. Peru, on the other hand, revived an always shaky democratic experiment in 1980, after twelve years of a military government that had carried out profound structural changes. The fragile party system that emerged from the transition faced great challenges. This was a highly ideologized system, interacting with social movements and organized interest groups strong enough to apply political pressure. To complicate matters further, the democratic experiment coincided with the beginning of the armed conict unleashed by two terrorist groups, the Sendero Luminoso (Shining Path) and the MRTA (Movimiento Revolucionario Tupac Amaru, or Tupac Amaru Revolutionary Movement), the rst of which was particularly dogmatic and bloodthirsty.1 Despite these differences, Peru and Venezuela ended up with the same result: the sudden collapse of their party systems and the subsequent establishment of authoritarian regimes. In Peru, the collapse of the party system occurred between 1989 and 1992; in Venezuela, between 1998 and 2000. In Peru, the vote share of the four political groups that had garnered more than 90 percent of the vote in most elections in the 1980s fell to 71.8 percent in the municipal elections of 1989, and to 68.2 percent in the 1990 presidential contest, when Alberto Fujimori was elected. It continued to fall until the parties became virtually extinct politically (see Table 2.1). In Venezuela, the breakdown was even more rapid and is more surprising, given the apparent consolidation of the party system. Democratic Action (Accin Democrtica, or AD) and the Christian Democratic Party (COPEI) were

From Crisis to Collapse: Peru and Venezuela


Table 2.1 Peru: Vote Percentages for the Major Political Parties, 1978 2000
AP PPC AP PPC APRA LEFT (IU) Total

49

1978 (C) 1980 (P) 1980 (M) 1983 (M) 1985 (P) 1986 (M) 1989 (M) 1990 (P) 1992 (C) 1993 (M) 1995 (P) 1998 (M) 2000 (P)
SOURCE :

45.40 35.80 17.50 7.30 NP NP 11.60 1.64 5.00 0.40

23.80 9.60 11.10 13.90 11.90 14.80 9.70 5.70 NP NP NP

23.80 55.00 46.90 31.40 19.20 14.80 31.20 32.60 9.70 17.30 1.64 5.00 0.40

35.30 27.40 22.50 33.10 53.10 47.60 20.40 22.60 NP 10.80 4.11 7.00 1.40

29.40 14.40 23.30 29.00 24.70 30.80 20.20 13.00 NP 3.90 0.57 NP NP

88.50 96.80 92.70 93.50 97.00 93.20 71.80 68.20 9.70 32.00 6.30 12.00 1.80

Tuesta (2001). notes: The 1978 and 1992 elections were for constitutional assemblies (C). The 1980, 1985, 1990, 1995, and 2000 elections were for the presidency (P). The elections of 1980, 1983, 1986, 1993, and 1998 were municipal elections (M). The Total column is the combined vote for the Peruvian Aprista Party (APRA), the Popular Christian Party (PPC), Popular Action (AP), and the United Left (IU). For 1978, we are treating as IU votes the total vote for the parties that formed the IU coalition in 1980. In 1989 and 1990, the AP and PPC votes are part of the Democratic Front (FREDEMO) vote. NP: No participation; did not participate in election.

the hegemonic actors in presidential contests until the presidential election of 1993, when the winner was Rafael Caldera, who headed the Convergencia Nacional (National Convergence) coalition. In 1993, the combined vote total for the AD and COPEI presidential candidates was 46 percent. The collapse in the 1998 presidential contest won by Hugo Chvez was spectacular: AD and COPEI did not even run their own candidates. Instead, they backed the independent candidacy of Henrique Salas Rmer; their contribution to his 40 percent total was only 11 percent. Nor did they run candidates in the 2000 presidential contest, when Chvez won again, this time under a new constitution (see Table 2.2). These are disconcerting outcomes. In Peru, it was generally expected that the conicts and problems of the 1980s would give rise to increasing ideological polarization, setting off another military intervention (a scenario like Chiles for the period 1970 73). Instead, an outsider brought an end to the existing political order, a result anticipated by no one. The expected outcome was polarization and ungovernability, but what actually occurred was a grave crisis of representation. In Venezuela, the crisis in bipartism could have led to a scenario like the one in Colombia, where the traditional parties have undergone a progressive decline, with increasing internal fragmentation. New parties have emerged and developed in this context, and the old two-party system has been replaced by a moderate multiparty system. In contrast, in Venezuela, an outsider quickly did away with an order that had seemed consolidated. Clearly, Peru and Venezuela

50

Tanaka
Table 2.2 Venezuela: Vote Percentages for Presidential Elections, 1973 1998

Party

1973

1978

1983

1988

1993

1998

2000

AD COPEI AD COPEI MAS La Causa R Convergenciaa Polo Patritico/MVRb Proyecto Venezuelac Others
SOURCE :
a

48.7 36.7 85.4 4.3 10.3

43.3 46.6 89.9 5.2 4.9

58.4 33.5 91.9 3.5 0.1 4.5

52.9 40.9 93.8 2.7 0.3 3.7

23.6 22.7 46.3 22.0 30.5 1.2

0.1 56.2 40.0 3.8

59.5 40.5

Roberts 2003. Electoral coalition for Rafael Calderas campaign; included MAS. b Electoral coalition for Hugo Chvezs campaign; included MAS. c Electoral movement organized by independent candidate Henrique Salas Rmer; AD, COPEI, and other minor parties contributed to its vote totals.

constitute exceptions to the regional scenario of the 1980s and 1990s. In all other cases, party systems survived, despite problems of legitimacy and representation, with high electoral volatility and institutional instability (in some cases involving the removal of presidents by non-constitutional means, as in Ecuador) leading to a slow mutation toward a new system in which new parties exist side by side with traditional ones.2 Why didnt Peru and Venezuela go down this road? Why and how did they pass from crisis to collapse of the party system and the prevailing institutional order? Why and how did two countries with such different political trajectories end up with the same result? On the surface, it might appear easy to explain what happened in these countries by pointing to structural and economic variables and to the obvious incapacity of the various political actors to deal successfully with the challenges they faced. In this perspective, the parties alternated in power, failed to solve the problems, and were discredited, and hence voters sought options outside the system. In Peru, the election of Alberto Fujimori was preceded by a severe economic recession, high rates of ination, and extremely high rates of political violence (see Table 2.3), which damaged the legitimacy of all the principal parties. In Venezuela, the 1980s were very bad years, especially the year 1983 (under the administration of COPEI president Herrera), the period from 1988 to 1990 (under the second administration of President Carlos Andrs Prez), and almost the entire second administration of President Rafael Caldera (1994 98), who won election as the head of an independent movement, having left COPEI when he failed to secure its presidential nomination (see Table 2.3). This difcult period is said to be the cause of a progressive crisis of representation leading to a steady drop in party identication and growing political disaffection

From Crisis to Collapse: Peru and Venezuela


Table 2.3 Peru and Venezuela: Ination, GNP, and Subversive Acts, 1980 1999
Venezuela Peru Year Annual rate of ination GNP growth rate Subversive acts: National police records GNP growth rate

51

1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999

60.8 72.2 72.9 125.1 111.5 158.3 62.9 114.5 1,722.3 2,775.3 7,649.7 139.2 56.7 39.5 15.4 10.2 11.8 6.5 6.0 3.2

4.4 4.3 0.3 11.8 4.7 2.3 8.7 8.0 8.4 12.9 5.4 2.8 0.6 6.0 13.6 8.6 2.5 6.8 0.4 1.4

219 715 891 1,123 1,760 2,050 2,549 2,489 2,415 3,149 2,779 2,785 3,002 1,918 1,195 1,232 883 681 474 168

3.8 1.0 1.6 5.5 1.5 0 6.6 3.8 5.9 8.8 5.5 9.7 7.1 0.5 3.7 5.9 0.4 7.4 0.7 5.8

SOURCES : Peru: INEI (Instituto Nacional de Estadstica e Informtica); Venezuela: Anuario estadstico de Amrica Latina y el Caribe (Santiago: CEPAL, 2001).

(see Table 2.4). Supposedly, these factors explain why someone like Chvez could come to power. Although the economic crises narrowed the margins of possibility and options available to the political actors in both countries, economic performance alone cannot explain the collapse of these two party systems. Other countries in the region passed through similar or worse economic experiences, which also created crises of representation, yet their party systems managed to survive. Crucial to an understanding of the collapse of the party systems are the political actors decisions, especially those made at junctures when the decision makers were at their most vulnerable (in Peru, the 1990 election; in Venezuela, the 1998 election). In a situation of a crisis of representation, change, and high vulnerability, internal conicts led the parties into processes of division, which enabled outsiders, anti-system and anti-party caudillos, to win power through the electoral route. The relative consolidation of these new leaders enabled them to overthrow the prevailing order and replace it with a new order with authoritarian tendencies (see Mayorga, this volume).

52

Tanaka
Table 2.4 Venezuela: Trends in Party Identication
Polling organization Member/ Sympathizer Independent/ Not interested

Baloyra 1973 BATOBA 1983 CIEPA 1993 REDPOL 1998


SOURCE :

48.6 38.1 29.4 14.2

51.2 61.3 66.4 61.0

Gonzlez (2002).

Collapse of the Party System in Peru: Polarization, Intraparty Conicts, and Crisis of Representation
In previous work on Peru (Tanaka 1998, 2005), I argue that the breakdown in the party system resulted not so much from the performance of the political actors throughout the 1980s as from their actions starting near the end of 1988, when ination had accelerated and the country had entered into a dynamic marked by the 1989 90 elections.3 Despite the complicated situation, nothing portended that in 1990 a grave crisis of representation would develop and that in the succeeding years the party system would collapse. On the contrary, both the analysts and the actors themselves perceived that the principal danger lay in the growing polarization of party members, the abandonment of the political center that accompanied the crisis of the ruling American Popular Revolutionary Alliance (Alianza Popular Revolucionaria Americana, or APRA), and the strengthening of extremist actors. These trends led to serious problems of governability. In the context of the threat posed by Sendero Luminoso, this situation could have led to a repressive military intervention. Until 1989, the parties seemed relatively strong, with possibilities for recovery in the not too distant future. At the extremes were the United Left (Izquierda Unida, or IU) on the left and the Democratic Front (FREDEMO) on the right. From its founding in 1980, the IU, a political front formed by seven leftist organizations, had steadily increased its electoral clout, winning more than 30 percent of the vote in the 1986 municipal elections. Several 1987 opinion polls indicated that Alfonso Barrantes, by that time the most likely IU candidate, would be the top choice among voters in the 1990 presidential race. In late 1987, the IU called its First National Convention for September 1988 (although, in the end, it was held in January 1989), to ne-tune the organization and its strategy for coming to power through the electoral route. At this convention, the Front would adopt rules, policy, and platform guidelines, formulate a plan for immediate political action, and choose a unied political leadership. As for the right, Popular Action (Accin Popular, or AP) and the Popular Christian Party (Partido Popular Cristiano, or PPC) suffered harsh

From Crisis to Collapse: Peru and Venezuela

53

setbacks in the April 1985 general elections, after the second administration of Fernando Belande (1980 85). But by August 1987, they had taken the political initiative once again, heading the opposition to President Alan Garcas proposal to nationalize the banking system. The rightist block then underwent a signicant revitalization. August 1987 saw the birth of the Liberty Movement (Movimiento de Libertad, or ML), led by the writer Mario Vargas Llosa and the economist Hernando de Soto, which promoted market-oriented ideas and state modernization. January 1988 saw the formation of a major alliance involving the ML, AP, and the PPCthe Democratic Front (FREDEMO). In the November 1989 municipal elections, FREDEMO emerged as the countrys main political group, and opinion polls showed that Mario Vargas Llosa was likely to be Perus next president.4 By 1989, the Peruvian political scene was highly polarized. On one pole was a left with revolution in mind, with a kind of electoral path to socialism similar to the one followed by the Popular Unity Front (Unidad Popular, or UP) in Allendes Chile (1970 73). On the other pole, the right advocated a liberal ideology and a profound modernization of the economy and the state within the framework of a market economy. Given the ideological polarization of these programs, the triumph of either the left or the right would have created problems of governability. What occurred, unexpectedly, was a crisis of representation: radicalized political groups abandoned the political center formerly occupied by APRA, and the empty space was lled by an outsider. Such an unusual and unexpected outcome is understood by analyzing the 1990 election campaign and the conicts within the parties. The campaign was marked by a deep recession, hyperination, and high levels of political violence. In 1989, the Sendero Luminoso announced that it had arrived at a strategic balance with the forces of orderthe stage prior to a strategic offensive that would lead to the seizure of powerand it began a siege of Lima. In this context, internal conicts within the major parties led to open struggles and divisions, leading a sector of the electorate to seek other options outside the system. How can we understand the actors behavior? The context of crisis and violence, coupled with the (correct) perception that here was an extreme situation involving the end of one political cycle and the chance to start another, led the actors to abandon risk-averse behaviors, to be audacious, and to make decisions marked by ideological reasoning rather than pragmatism. Such conduct intensied the contradictions and internal conicts among the principal actors and produced the vacuum of representation that the hitherto unknown Fujimori took advantage of. Around 1987 the IU was in need of a profound reorganization. Until then, despite its electoral gains, the IU had functioned mainly as a coalition of parties, which were represented by the general secretaries of the various participating parties in a National Executive Committee (CDN), where each party maintained

54

Tanaka

its own political line. The IUs internal problems grew more acute during the administration of Alan Garca, whose populist and revolutionary rhetoric created problems of identity and strategy. Barrantes, IU chairman until May 1987, had maintained a stand of critical collaboration with the Garca administration. In 1987 Barrantes resigned his post because he did not have the backing of the majority of the parties general secretaries, who espoused a much rmer opposition line toward the Garca government. A clear, unied course of conduct was urgently needed, and that is why the rst national convention was called. After an intense and interesting period of preparation, which saw the enrollment of more than 130,000 members, an extremely high gure by Peruvian standards, the convention was held. But far from fostering the consolidation of the IU, it initiated a tortuous process of division. On one side of the debate, aligned with Alfonso Barrantes, were those who believed that to win elections and fashion a minimally stable and successful government it was essential to exclude the IUs radical sector. The radical sector had not clearly rejected armed struggle and thus would make it impossible to surmount a veto by the armed forces and conservative sectors. On the other side of the debate were the parties of the Revolutionary Block (Bloque Revolucionario),5 which believed that the seeds of revolution were already present, making it appropriate to prepare for a large-scale political and possibly military confrontation. Accordingly, the real objective was not to arrive at a government through elections but to prepare for taking power through insurrection. So, on one side were those outlining a reformist program, broad in scope and appealing to the average voter; on the other were those propounding a strengthening of the parties bases, of strategic sectors, and a digging-in to prepare for the coming confrontation. In the middle of this controversy were the Peruvian Communist Party (PCP) and those independent IU activists who had no party allegiance. The breakup of the IU, amid mutual recriminations and accusations, unfolded between January and October 1989 (the month in which candidates for the 1990 elections had to formally declare their intentions) and ruined the lefts electoral chances.6 In the 1990 elections the left divided, presenting two presidential candidates. The IU candidate, Henry Pease, polled 8.2 percent of the vote, while Alfonso Barrantes, candidate for the newly created Socialist Left (Izquierda Socialista, or IS), won only 4.7 percent. The crisis of the left increased the electoral chances of the right. Throughout most of 1989, with the collapse of the left, presidential opinion polls indicated that Mario Vargas Llosa was the favorite. In the second half of 1989 and the early months of 1990, the question was whether or not Vargas Llosa would get the more than 50 percent of the vote needed to win in the rst round. FREDEMO, however, had its own internal problems. The leadership of Vargas Llosa and Movimiento Libertad within the alliance generated jealousies and rivalries in AP and PPC. This friction came to a head in June 1989 when the FREDEMO

From Crisis to Collapse: Peru and Venezuela

55

strategy for the November municipal elections became the subject of so much debate that Vargas Llosa tendered his resignation as a presidential candidate, a resignation he later withdrew.7 In spite of these problems, FREDEMO had a fairly good showing in the November 1989 municipal elections. Although those elections witnessed the appearance of the rst independent candidates, who made manifest a delegitimation of the major parties (see Table 2.1), most of these independents were aligned with the major parties.8 The polarization and sense of urgency in the country affected FREDEMO and its campaign strategy, which makes it easier to understand why Vargas Llosa did not come up with a more conclusive victory in the rst round of the 1990 election (he won only 32.6 percent of the vote). Vargas Llosa distanced himself from the median voter with a fairly ideological campaign, seeking a clear mandate to go ahead with profound neoliberal reform. This campaign did not inspire enthusiasm in the electorate, especially after the popular mobilization against neoliberal reforms in Caracas in February 1989, under the administration of Carlos Andrs Prez. The crisis and chaos into which the government plunged seriously damaged APRAs electoral chances, yet APRA could not be completely written off. In the 1989 municipal elections, APRA remained the second largest party at the national level, behind FREDEMO and slightly ahead of the left candidates (see Table 2.1). APRAs candidate, Luis Alva Castro, won 22.5 percent of the vote in the 1990 presidential election. But APRA, too, had internal problems that decreased its electoral chances. According to the 1979 Constitution, Alan Garca could not seek reelection, and his efforts between 1987 and 1988 to pass a constitutional reform allowing him to run ended in failure. As a result, the general secretary of the party, Luis Alva Castro, competed with Garca for control of APRA. Garca fought to maintain control, and he decided to maintain his distance from Alva Castro. Throughout most of the campaign, Garca gambled on leading the opposition to Mario Vargas Llosas candidacy (once again, for ideological reasons) and backed Alfonso Barrantes rather than the APRA candidate. The division of the left, FREDEMOs internal problems and the extreme ideologization of its campaign, and the weakness of an APRA candidate who had to assume the costs of the failures of Garcas administration without receiving the benets of support from the topall coming at an especially critical moment created a vacuum of representation. The political center, having been left more or less vacant, was subsequently occupied by a candidate who had not even been mentioned in the surveys until a few weeks before the election.9 From among the group of minor candidates, Alberto Fujimori, the former rector of the Universidad Agraria (Rural University), suddenly turned out to be an attractive option. Once Fujimori began to rise in the opinion polls, Garca began to support him through his connections in the press, and his support was decisive. With Garcias support, just a few weeks before the election, Fujimori

56

Tanaka

ceased to be a minor candidate and he went on to place second in the contest.10 In that rst round, Vargas Llosa won with 32.6 percent, and Fujimori, surprisingly, came in second with 29.1 percent. In round two, with the votes of APRA and the left, Fujimori won the presidency, with 62.4 percent, compared to Vargas Llosas 37.6 percent. Once in ofce, Fujimori found himself with a minority in Congress. In the elections for the Senate and the Chamber of Deputies, FREDEMO obtained 32.3 percent and 30.1 percent, respectively; Cambio 90 (Change 90), only 21.7 percent and 16.5 percent; APRA, 25.1 percent and 25 percent (that is, it topped Cambio 90 in both houses); IU, 9.8 percent and 10 percent; IS, 5.5 percent and 5.3 percent. Cambio 90 won only 32 of the 180 seats in the Chamber of Deputies, and in the Senate, only 14 out of 62. FREDEMO had 63 deputies and 21 senators; APRA, 53 deputies and 17 senators (again, more than Cambio 90); IU, 16 deputies and 6 senators; IS, 4 and 3. Fujimori had no possibility of aspiring to reelection in 1995 because reelection was prohibited by the 1979 Constitution. Therefore, his presidency was perceived as a singular episode, certainly ephemeral, and once it was over, the parties would again occupy center stage. Things turned out quite differently.

Collapse of the Party System in Venezuela: Crisis, Intraparty Struggles, and a Crisis of Representation
In the Venezuelan case, internal struggles and processes of division within the major parties, in a situation in which they were particularly vulnerable, again explain the unexpected rise to power of an outsider who went on to destroy the prevailing political and institutional order. The behaviors within the parties were a consequence of internal structures so highly disciplined that losing factions had no space to air their differences and had incentives to break away. Internal struggles and divisions within the Venezuelan parties are a tradition.11 Such conicts were a consequence of the way the Venezuelan parties were structured. They were extremely hierarchical and disciplined (Crisp 2001). This structure created obstacles to airing factional disputes openly, and thereby stimulated party division. The factions that lose internal disputes are weakened, without major access to top party posts or candidacies in popular elections. As a result, they are tempted to try their luck outside the party apparatus. Article 185 of the 1961 Constitution, by allowing for the reelection of a president after ten years, created the gure of the great caudillo waiting for the chance to make a comeback. Thus, former leaders do not disappear, but grow weak and then reappear. This phenomenon helps explain the return to the presidency of Carlos Andrs Prez in 1989 and of Rafael Caldera in 1993, both of whom were greatly weakened politically. Such comebacks are crucial in understanding Hugo Chvezs rise to power in 1998.

From Crisis to Collapse: Peru and Venezuela

57

An interesting comparison can be made with Peru, where the parties were traditionally organized around an unchallengeable caudillo, but where a certain space for factional disputes also existed, as long as they did not challenge the maximum leader. Vctor Ral Haya de la Torre and then Alan Garca in APRA, Fernando Belande in AP, Luis Bedoya in the PPC, and to some extent Alfonso Barrantes in IU between 1980 and 1986, were the indisputable leaders of their political groups. Under them, however, disputes could develop, and these would be internally arbitrated by their parceling out of posts and benets, which made for a measure of internal equilibrium. Problems cropped up when, in the context of the 1990 election campaign, all the main actors were simultaneously left without adequate mechanisms for handling internal difculties. Internal conicts ripped apart the IU when Barrantes lost his role as the arbiter of differences. The right, joining up with FREDEMO in an alliance among equals, had no mechanisms for internal arbitration, although each component party had ways of dealing with its own internal conicts. Lastly, APRA was also experiencing serious conicts. Garca could not stand for reelection, and presidential candidate Luis Alva competed against him for control of the party. In Venezuela, despite repeated internal conicts, the two-party system had functioned until 1993, with AD and COPEI alternating in power. In the second half of the 1980s, however, the parties had lost legitimacy and the capacity to represent the citizenry, as a result of the nations poor economic performance and the governments difculties in confronting problems associated with the debt crisis. Not that the parties did nothing to confront the situation. In 1984, the Presidential Commission for the Reform of the State (COPRE) was created. COPRE undertook substantial institutional changes. Among them were the introduction of direct popular elections of state governors (1989) and the Organic Law of Municipal Regimes, which instituted direct elections for mayors (1989). Previously, governors and mayors had been appointed. These changes were in line with the decentralization process that originated in the 1970s. Other changes included the establishment in 1993 of the mixed formula (proportional personalized vote) for electing deputies to the National Congress and to legislative assemblies. These reforms sought to open up a political system perceived as tightly closed (characterized as a partyocracy), renew the leadership, and loosen the control of party bosses. The reform efforts only partially fullled their objective of improved representation, and they increased internal party tensions and conicts, which in the long run were decisive factors in the collapse of the party system. The changes also coincided with an economic crisis and reform. The concomitant tensions produced by party reforms and economic crisis explain Chvezs coming to power.12 The implementation of market-oriented reforms really began during the second administration (1989 93) of Carlos Andrs Prez, who spoke eloquently of a great turnaround. Prez faced a difcult nancial situation that demanded

58

Tanaka

adjustment measures, which had long been postponed throughout the 1980s. The public-sector decit was 9.9 percent of GNP in 1988, with a currentaccount decit of 4.9 billion dollars, a situation without precedent in the 1970s and 1980s. In addition, the price of oil had fallen to around 13 dollars a barrel that year, after averaging 33 dollars in 1985, and the drop brought much uncertainty to a country whose income is largely based on oil (Nam 1993; Hidalgo 2000). Carlos Andrs Prez also faced problems in his relations with his own party. He had won the nomination, two terms after his rst administration, by prevailing in a tough struggle against Octavio Lepage, who was backed by the outgoing president, Jaime Lusinchi. Prez had always based his power within AD on his charisma and his ability to communicate with the citizenry at large, rather than on his role as a bureaucratic executive. This situation makes it easier to understand the rst decision of his second administration: the formation of a governing team made up of independent gures so as to secure a margin of maneuverability in the face of pressure from the party and interest groups. Engaged in rebuilding his political leadership with a new political base, he undertook neoliberal economic reforms, which entailed reorienting ADs traditional political identity as a party with a statist economic policy. Prezs gambit is not surprising: an interesting parallel exists with Mexicos president Carlos Salinas de Gortari (1988 94) and his policy of modernization, market reform, and relative distancing from the dinosaurs of the PRI (Partido Revolucionario Institucional, or Institutional Revolutionary Party). Prez also had the example of President Victor Paz Estensoros New Political Economy in Bolivia, launched in 1985. AD was a bureaucratized structure, and in many instances a corrupt one, accustomed to clientelistic and corporate relationships and to nancial favors. Prez had strong incentives to remake AD under a modernizing leadership more in tune with the need for structural reform that was becoming evident in the region. The policy of the great turnaround generated a massive wave of spontaneous protests in February 1989, especially in Caracas (Kornblith 1998; Lpez Maya 2000). There was a feeling of indignation over the new policies, launched by someone who had won the election by exploiting the image of a return to the prosperous days of the 1970s. As has been explained by several analysts of the adjustment process, the citizenry did not perceive that a severe adjustment was necessary and inevitable in Venezuela as it had been in countries with deeper recessions and hyperination (Corrales 2000; Roberts 2003; Weyland 2002). This popular resentment led to the increasing isolation of the Prez government, even within AD. This was expressed, for example, by the distancing of the Confederation of Venezuelan Workers (CTV), the main labor confederation, which in May 1989 called for a work stoppage to protest the governments economic policies, marking the rst time the CTV had taken action against an AD government. But that was just one expression of an even greater distancing. The traditional party apparatus, already dealt a blow when Prez defeated Lepage

From Crisis to Collapse: Peru and Venezuela

59

for the nomination, felt left out of government decisions. Prez intended to pursue neoliberal reform, a radical about-face from ADs traditional populist policies. Prez thus lost support within his party, including among AD legislators, and this ultimately explains his removal from ofce by Congress in 1993. Many analysts have argued that the social costs of the great turnaround destroyed the Prez governments opportunities. There is not much evidence, however, to support that position. As the data in Table 2.4 show, the economy fell into a deep recession in 1989, a result of the adjustment program, but growth recovered during the remainder of Prezs term. Why was Prez unable to make political capital out of a recovery, as did Fujimori and other leaders of successful stabilization efforts such as Paz Estensoro in Bolivia and Carlos Menem in Argentina? In the Venezuelan case, the interparty and intraparty conicts are just as much a key to understanding Carlos Andrs Prezs fall as is economic performance. The existing discontent and conict among and between Venezuelas political parties had devastating political effects. Failed coup attempts in February and November of 1992, nevertheless generated sympathy among the popular sectors. This sympathy grew and spread because of the stands taken by political leaders from AD and the opposition. The political space acquired by Hugo Chvez was in large measure handed to him by the parties and their caudillos with their opportunistic stances. After the coup attempts, eminent politicians, among them former COPEI president Rafael Caldera, far from condemning the participants and defending the constitutional order, declared support for the insurrectionists. Their support for the coups further weakened the party system and exacerbated the unrealistic expectations of the citizenry and the hope for a redeeming leadership that would nish off the old order and bring prosperity to the country.13 AD, too, kept its distance from the Prez government instead of defending it. Congress sought to hamper the administration and ultimately to remove it. In May 1993, barely three months before the end of his term, Carlos Andrs Prez was removed from ofce by Congress after impeachment proceedings based on a dubious accusation of improper use of public funds, and Ramn J. Velasquez became the interim president. These events discredited the political system, which fell captive to particularistic, narrow interests.14 The 1993 elections caught AD considerably weakened after the scandals connected with the dismissal of President Prez. Nevertheless, there was not the slightest foreshadowing of the party system collapse that occurred in 1998. As can be seen in Table 2.2, despite an enormous drop in ADs vote as compared to 1988, the partys candidate, Claudio Fermn, came in second behind Caldera. Fermns rise to leadership was a result of the renewal generated by the decentralization process and the COPRE reforms. Fermn built his reputation on his good management as mayor of Caracas. This new type of leadership had some complications, and it faced opposition from the traditional party apparatus. To secure the presidential nomination, Fermn competed against Luis Alfaro

60

Tanaka
Table 2.5 Venezuela: Percentage of Seats in the Chamber of Deputies, 1973 2000
Party 1973 1978 1983 1988 1993 1998 2000

AD COPEI AD COPEI MASa LCR Convergencia MVR Proyecto Venezuela Others


SOURCES :
a

51.0 31.8 82.8 4.5 12.5

44.2 42.2 86.4 5.5 8.1

56.5 30.0 86.5 5.0 8.5

48.3 33.3 81.6 9.0 1.5 7.9

27.6 27.1 54.7 20.1 25.1

29.1 14.3 43.4 10.6 3.2 1.6 25.9 12.7 2.6

18.8 4.2 23.0 12.7 2.4 .1 46.1 4.2 9.1

Roberts 2003, 253; data for 2000 elections taken from Payne et al. 2003. MAS deputies are regarded as part of Convergencia.

Ucero and Hctor Alonso Lpez (who was backed by Carlos Andrs Prez). In the 1993 congressional elections, AD, although dropping signicantly in comparison with previous elections, remained Venezuelas largest party, clearly topping Convergencia and Radical Cause (La Causa R, or LCR) (see Table 2.5). ADs internal conicts had helped COPEI come to power in 1968 and 1978, in a system in which these two parties alternated in power. Something different occurred in the 1993 election because COPEI too was seriously damaged by internal conict. Rafael Caldera failed to capture his partys presidential candidacy in 1988, and he distanced himself from the candidacy of Eduardo Fernndez. Then, in 1992, Caldera did not condemn the coups; on the contrary, he rode the wave of sympathy aroused by the perpetrators and joined in the criticism of the traditional order and the neoliberal economic policy, adopting a populist discourse. For the 1993 elections, Caldera formed a new political group, Convergencia Nacional, an alliance that elevated him to the presidency once again, although he had earlier encouraged, within COPEI, the candidacy of Oswaldo Alvarez against that of Eduardo Fernndez. The exit of Caldera, the party founder, was a harsh blow for COPEI, and it ultimately meant the end of the system of the two parties alternating in power.15 Despite all its problems, COPEIs showing in 1993, like ADs, did not herald as inevitable its steep demise in 1998. Although COPEI suffered a sharp drop in comparison with its 1988 electoral results, it remained the second largest party in Venezuelas Congress. The partys presidential candidate, Oswaldo Alvarez, came in third in the presidential election (see Tables 2.2 and 2.5). Oswaldo Alvarez had built his political leadership by serving as governor of the state of Zulia. Just as with Fermn in AD, decentralization resulted in a renewal in political leadership but also exacerbated internal conicts in a party with a hierarchical tradition.

From Crisis to Collapse: Peru and Venezuela

61

Tables 2.2 and 2.5 show that the vote lost by AD and COPEI made possible both the emergence of Convergencia and the growth of LCR, a leftist movement of union origin with bases in various regions of the country (Lpez Maya 1997, 2001). As of 1993, Venezuela seemed to be embarking on a path of evolution from a traditional two-party system toward a moderate multipartism in which AD and COPEI would coexist with new parties. That is why the collapse of the party system cannot be readily inferred from a crisis of representation or problems of legitimacy. Although party identication was clearly on the decline, paralleled by an increase in the number identifying themselves as independents or as disaffected with politics (see Table 2.4), citizen preferences could have followed a pattern similar to that of Colombia, Ecuador, or Bolivia. That did not happen. To understand why, it is crucial to analyze the 1998 election campaign and, again, to look at how intraparty conicts simultaneously ruined the chances of all the actors in the system. This is what allowed an outsider to come to power. The emerging parties, Convergencia and LCR, did not consolidate themselves between 1993 and 1998. Convergencia paid the price for bad governmental performance. As can be seen in Table 2.3, economic performance between 1993 and 1998 was poor.16 The administrations shortcomings also hurt Movement to Socialism (Movimiento al Socialismo, or MAS), which was part of the governing coalition. In the 1998 presidential election, Convergencia did not nominate its own candidate but backed Irene Sez, who polled barely 2.82 percent of the vote; in Congress, only three Convergencia deputies and two senators won seats. MAS suffered a schism in the 1998 election. On one side was the sector close to the Caldera government and the minister of planning, Teodoro Petkoff; on the other was the ministers critics (Leopoldo Puchi and Felipe Mujica), who ended up backing Hugo Chvez. MAS contributed nine points to Chavzs 56.2 percent total and elected twenty-two deputies and six senators. As a leftist movement that was not part of the traditional order, LCR might have been the most obvious contender to ll the space vacated by Convergencia and MAS. LCR, however, also split before the 1998 election. One sector challenged Andrs Velsquezs leadership and exited in April 1997 to found a new movement, Fatherland for All (Patria Para Todos, or PPT), under the direction of Pablo Medina. LCR nominated a presidential candidate, Alfredo Ramos, who polled barely 0.1 percent of the vote, and it elected one senator and six deputies. PPT backed Chvez, bringing him 2.19 percent of the vote, and it elected seven deputies and one senator. COPEI continued the decline that had begun with the departure of its founder, Rafael Caldera, in 1993. In 1998, it did not eld a presidential candidate, although Luis Herrera and Donald Ramrez sought the nomination. The party rst backed Irene Sez and then another independent, Henrique Salas Rmer. It contributed barely 2.15 percent of the 40 percent he polled. In Congress, COPEI

62

Tanaka
Table 2.6 Venezuela: Number of Governors Elected, by Party, 1989 2000
Party 1989 1992 1995 1998 2000

AD COPEI MAS La Causa R Convergencia Independents MVR PRZVL-PROCA MERI-COPEI-AD PPT Total
SOURCE :

11 6 2 1 20

8 9 4 1 22

12 3 4 1 1 1 22

7 3 3 1 7 1 1 23

2 1 3 1 1 12 1 2 23

Maingn (2002).

obtained less than half the votes it had in 1993. AD came out better in Congress, but in the presidential eld it suffered a disaster. For Congress, AD maintained roughly the same vote as in 1993, enough to remain the biggest party in the November 1998 elections, just as before (Table 2.5). Governors were also elected at that time, and AD again won the most gubernatorial races (Table 2.6). In the presidential election, perhaps the most logical approach would have been to rebuild the party image around the leadership of Fermn, who had fared relatively well in the 1993 elections and who gured as a favorite in presidential public opinion polls taken in the rst half of 1997. However, the visibility of Fermns leadership and the increasing presence of leaders coming out of mayorships and governorships generated a reaction from the traditional party apparatus led by Luis Alfaro Ucero. The end result was Claudio Fermns departure from AD, as Ucero won the presidential nomination. The latter gathered so little support that AD nally withdrew his candidacy and backed Salas Rmer, contributing 9.05 percent to his total. Ucero stayed in the race although AD no longer backed him, and he obtained only 0.42 percent of the vote. The discrepancy between the disaster of the presidential vote and the relatively good performance in the eld of governorships, along with acceptable results for Congress, suggests that ADs problem lay in its extremely poor handling of the presidential contest. In the end, the 1998 presidential election had two main protagonists: Hugo Chvez, who won 56.2 percent of the vote, and Henrique Salas Rmer, with 40.0 percent. Seemingly, Chvezs victory grew out of an inexorable need, given the context of crisis, the discrediting of the traditional system with its internal conicts, and the fragility of the forces that had emerged in 1993. Yet Chvez and MBR-200 (Movimiento Bolivariano Revolucionario 200, or Bolivarian Revolutionary Movement 200) always hesitated to stand for election, to enter into the game of the system. Only in April 1997 did the group decide to end

From Crisis to Collapse: Peru and Venezuela

63

its abstentionist stance (Lpez Maya 2003). Chvez began to lead in the presidential preference polls only at the beginning of 1998, after the disasters suffered by his competitors. Throughout 1997, rst Claudio Fermn and then Irene Sez were favored to win. The 1998 election campaign was extremely volatile; its outcome cannot be readily explained through macrovariables, such as the discrediting of the system or a crisis of representation, alone. The mistakes of the parties in the system, and their internal conictsprecisely at a moment when these parties were particularly vulnerable explain how Chvez came to power. The Venezuelan parties suffered splits because of their hierarchical structure, which left no space for dissidence or pluralistic competition among factions. As a result, conicts were very hard to handle and often ended in fractures.

Destruction of the Old Order and Transition to a Competitive Authoritarianism


Fujimori and Chvez are denitely opposites as political leaders: the former came to power based on an ambiguous campaign platform that promoted his independent character, beyond traditional politics. He opposed Vargas Llosas marketreform proposals, but, once in power, implemented them and came to personify neoliberal reform. Chvez always identied with a revolutionary project, a Bolivarian revolution based on nationalist and socialist features. However, both men are personalistic and authoritarian leaders with a populist and antisystem discourse. Both mens candidacies reected a crisis of representation, and both men exploited popular dissatisfaction with traditional politics. They represented some previously excluded popular sectors, under plebiscitary, not democratic, schemes. Despite their differences, both established authoritarian regimes with striking similarities. This resemblance illustrates that deep institutional reform, under the hegemony of a president who undermines political competition, eliminates the logic of checks and balances inherent to democratic rule and paves the road to authoritarianism. Once Fujimori was in power, his success in stabilizing the economy through market-oriented reforms and his later achievements in combating terrorism (see Table 2.3) allowed him to build a coalition that supported his leadership, which was authoritarian, anti-political, and anti-institutional.17 In Venezuela, Chvez used the 1998 election campaign to promote the idea that transforming the country should start with institutional change, with dismantling the order of the Punto Fijo agreements of 1958 set down in the 1961 Constitution. The upheaval involved in his coming to power, the traditional parties internal crises that reached unsuspected extremes, and the support he received from some members of the power elite18all help to explain how Chvez destroyed the preceding institutional order so rapidly. The relative consolidation achieved by both leaders kept them from ending up like other leaders who had come to power by

64

Tanaka

challenging an establishment that eventually drove them out (Collor in Brazil, 1992; Serrano Elas in Guatemala, 1993; Bucaram, 1997, and, recently, Gutirrez, 2005, in Ecuador). Through their anti-system discourse, both leaders embodied and represented traditionally excluded sectors, but did so within clientelistic and populist schema.19 Both Fujimori and Chvez constructed new institutional orders, under political hegemonies that ended those that had preceded through the adoption of new constitutions and the takeover of all public authorities. Fujimori proceeded by rst staging a self-coup (April 1992) and then calling for the election of the Democratic Constituent Congress (November 1992). Chvez acted by calling for a referendum to approve a Constituent Assembly (April 1999), which then ignored the authority of Congress and other institutions. Even though both presidents stayed roughly within the margins of legality and enjoyed broad electoral support, they substantially undermined pluralism, competition, and the balance of power (horizontal accountability), thereby creating what can be characterized as competitive authoritarianisms in which democratic representation was replaced by a plebiscitarian legitimacy.20 The disequilibrium was further accentuated by the inability of the opposition in both cases to consolidate viable alternatives and overcome obstacles to collective action. In Peru, after the April 1992 coup, Fujimori called for a new Congress that would also serve as a constituent assembly. In November 1992, the Constituent Democratic Congress (CCD) was installed, with a Fujimorist majority, unlike the 1990 92 Congress in which Fujimoris supporters had been a minority.21 The return to constitutional order was established by the October 1993 referendum, which approved the new constitution by a scant margin,22 and by the 1995 general election for president and Congress, in which Fujimori easily won reelection in the rst round. Fujimori won 64.4 percent of the vote, as well as a majority in the unicameral Congress; his movement won 52.1 percent of the congressional vote. His closest competitor, Javier Prez de Cullar, got 21.8 percent of the presidential vote, and his Union for Peru movement (Unin por el Per, or UPP) captured 14 percent of the congressional vote. In the Congress elected in 1995, there were two main blocks: one linked to the government, Cambio 90 Nueva Mayora (Change 90 New Majority) with 67 out of 120 seats, and the opposition block, led by UPP, with 17 seats. After these two, the party with the greatest number of seats was APRA, which had only 8 representatives. Problems appeared soon after the presidents reelection. Since Fujimorism was a highly personalized movement, it required that Fujimori the person stay in power more than Fujimorism itself. The movement had no signicant existence beyond its leader.23 The path followed by Fujimori after 1992 is an interesting illustration of how to construct an authoritarian order through democratic means when holding a majority in Congress, and of how a democratic invocation of the majority can

From Crisis to Collapse: Peru and Venezuela

65

be used to destroy republican balance and liberal principles.24 The path Fujimori followed to set up an authoritarian government may be summarized by giving an account of the reelection law and the maneuvers intended to impede any challenge to his 2000 presidential candidacy.25 In August 1996, Congress, with a solid pro-Fujimori majority, passed the law of Authentic Interpretation of the 1993 Constitution. According to this law, Fujimoris rst presidential term had not been from 1990 to 1995 but from 1995 to 2000, since his rst term had been governed by the 1979 Constitution, not that of 1993. This law allowed Fujimori to stand for his rst reelection in 2000. Shortly thereafter, in September 1996, several opposition leaders began collecting signatures to seek a referendum on the repeal of the Authentic Interpretation law. The response came in October 1996, when Congress passed legislation regulating the exercise of the referendum. A referendum would now require not only citizen signatures but also the approval of at least two-fths of the members of Congress (that is, 48 votes). The route to blocking the referendum subsequently involved a congressional confrontation with the judges of the Constitutional Court, which in January 1997 declared by a simple majority that the law of Authentic Interpretation was inapplicable. Congress responded in May of that year by dismissing the justices who had voted for that interpretation. In July 1998, the promoters of the referendum presented petitions with 1,441,535 citizen signatures to the National Ofce of Election Processes (ONPE). In August, the ONPE enforced the referendum law of October 1996. Instead of calling for the referendum, it sent the request to Congress, where the opposition did not have the forty-eight votes necessary to approve the referendum. Fujimori needed to do more than block the referendum. He also had to keep the National Elections Board ( JNE) from being able to declare that there was a basis for challenging his candidacy by invoking its unconstitutionality, so he had to control the JNE. According to Article 179 of the 1993 Constitution, the JNE is made up of ve members: one elected by the Supreme Court from among its retired and active justices; one elected by the Board of Supreme Prosecutors from among retired and active supreme prosecutors; one elected by the Bar Association of Lima from among its members; one elected by the deans of the law faculties of public universities from among their former deans; and one elected by the deans of the law faculties of private universities from among their former deans. Fujimoris strategy consisted of controlling the institutions with representatives on the JNE. Accordingly, in June 1996, the reorganization of the Judiciary and the Prosecutors Ofce was announced. The government took action to ensure that the two representatives of these institutions would not impede Fujimoris reelection plans. In November 1997, the government announced the takeover of the public universities, a move through which it assumed control over the deans of the law schools. That accomplished, Fujimori could now count on three out of ve votes on the JNE. As further insurance, in

66

Tanaka

May 1998 Congress passed a law changing the kind of vote required for the JNE to declare that there was a basis for challenging a candidacy; the vote went from a simple majority (three votes) to a qualied majority of four out of ve. After these machinations, in December 1999, Alberto Fujimoris candidacy was led. The opposition challenged his candidacy, but ultimately the challenge was rejected by the JNE. The route to reelection involved near-absolute control over all state institutions. This became even more evident during the 2000 election campaign, when public resources were mobilized to promote Fujimori. Even the armed forces got into the act.26 The Venezuelan case had some elements in common with the Peruvian one.27 Chvez rst did away with the Congress elected in November 1998, in which AD was the largest party. He also dismissed the governors, in whose ranks AD still gured prominently (see Tables 2.5 and 2.6). To accomplish these objectives, on the day he took ofce, February 2, 1999, Chvez called for a referendum on convoking a National Constituent Assembly (ANC). The referendum took place on April 25, 1999. More than 80 percent of the valid vote among those who cast a ballot was in favor of holding an ANC, but absenteeism was over 60 percent. On July 25, 1999, the election for the members of the ANC was held, also with a high rate of absenteeism (over 53 percent). Chvez organized this election with a majoritarian electoral system (a personalized one) that enabled him, with 65.5 percent of the vote, to control 94.5 percent of the seats (121 out of 128). The opposition won 34.5 percent of the vote but just 5.5 percent of the seats.28 With this comfortable majority, Chavism could draw up a constitution without having to make any major concessions to the opposition.29 On December 15, 1999, a referendum was held to approve the new constitution, which would replace the 1961 Constitution. Although yes received 72 percent of the valid votes, 56 percent of eligible voters did not cast ballots. What really destroyed the balance of power in Venezuela was the ANCs action on December 23, 1999. Invoking its role as the incarnation of the sovereign will and the expression of a new institutional order, the assembly dissolved the other public authorities, including the Congress of the Republic, the Council of the Judiciary, the Supreme Court of Justice, and the Legislative Assemblies.30 Most of the new authorities, whose mandates had already been established under the new constitution, were elected in 2000. On July 30, elections were held for president of the republic, deputies to the National Assembly (formerly the National Congress), governors of the twenty-three states, deputies to the State Legislative Councils (formerly Legislative Assemblies), metropolitan mayors, mayors, councilors to the city councils, and delegates to the Latin American Parliament and the Andean Parliament. These elections were carried out in the context of Chvezs hegemony, allowing Chvez to use constitutional and majoritarian means to build an almost absolute authority with no institutional counterweights. The result was an

From Crisis to Collapse: Peru and Venezuela

67

authoritarian wielding of power backed by a legality largely devoid of content. Hugo Chvez was reelected with 59.75 percent of the vote (against 37.5 percent for Francisco Arias), and the Movimiento Quinta Repblica (Fifth Republic Movement, or MVR) obtained more than 46 percent of the congressional seats.31 All these maneuvers against the rule of law in both countries generated antagonism, and during long periods of time the majority of the public disapproved of the performance of Fujimori and Chvez. Why, then, was the political opposition not able to put effective limits on these authoritarian presidents? The defeat of the opposition can in large measure be attributed to its internal weaknesses and its fragmented nature, the absence of a common strategy, and the lack of a sufciently supported clear alternative offering more than a simple return to a past the citizenry rejected. The opposition raised institutional banners and respect for the rule of law and had little to say in social and economic terms; that was precisely the strong point for both Fujimori and Chvez. These leaders expressed the interests of previously excluded popular sectors and mobilized them in clientelistic schemes. They drew their legitimacy from plebiscitary schemes and signicant popular support, not from democratic legitimacy and respect for the rule of law. In this sense, the problems of democratic representation remain in both Peru and Venezuela. In Peru, Fujimorism always enjoyed the support of the winners of the economic reform process: that sector of the business community which is linked to large-scale mining interests, nance, and commerce, who beneted from trade liberalization, privatization, and foreign investment incentives (Tanaka 2003; Gonzales de Olarte 1998). Yet this sector, while strategic, is extremely small and thus was not able to deliver the votes needed by Fujimori to win the 2000 elections. Electorally, the regimes legitimacy depended on the support of the poor. Thanks to privatizations, increased tax revenues, and greater access to loans from abroad, and despite its neoliberal character, the Peruvian state renewed its economic presence under Fujimori. The second Fujimori administration saw the highest social expenditure levels in more than two decades, and this helps explain the regimes greater support among the poorest of the poor. This support was built through effective clientelistic schemes which targeted social expenditure under a centralized structure, controlled by the presidency (Tanaka and Trivelli 2002). In Venezuela, President Chvez has always enjoyed considerable popular support, but his support has uctuated and his authoritarian actions have also generated signicant opposition. Chvez began his administration with very high approval ratings, and they allowed him to destroy the existing institutional order and establish a new one under his political hegemony. However, his popular support declined from 2001 until at least 2004, and social and political polarization increased. Chvezs policies generated rejection by the upper and middle classes and, at the same time, created high expectations among the poor, which identied with his populist, nationalistic, and revolutionary rhetoric. This

68

Tanaka

polarization reached high levels after the Decree of 49 Laws in November 2001, under extraordinary legislative measures granted by the Congress, which included a land reform law and a new oil law, among others. The polarization culminated in a coup dtat in April 2002. After the coup failed, the opposition led a general strike, between December 2002 and the rst months of 2003, calling for the removal of the president (Medina and Lpez Maya 2003). After the defeat of the general strike, a dialogue was established, in May 2003, between the government and the opposition, with the mediation of the Organization of American States (OAS). An institutional mechanism to resolve the political conict was invoked, and it turned into the call for a referendum to remove President Chvez. This tortuous process began by August 2003 with the rst campaign to collect the signatures needed for a recall referendum. After many conicts, the recall vote nally took place in August 2004. Chvez obtained 58.25 percent of the vote, and the opposition 41.74 percent. What happened? Despite all the allegations raised by the opposition, which denounced the governments dirty maneuvers and alleged electoral fraud, the Venezuelan case can be analyzed along the same lines as the Peruvian one. As in Fujimoris Peru, in Venezuela there is an authoritarian government and an active opposition, but one that lacks a common strategy and is unable to present a clear alternative. While the opposition emphasized an institutional discourse, Chvez underscored redistributive measures and expanded social expenditures, beneting from rising oil prices in 2003 and 2004. Despite the considerable differences between Fujimori and Chvez, the Peruvian and Venezuelan political regimes have remarkable similarities, as does the relationship between the government and the opposition. These presidents headed formally democratic governments. Based on their political hegemony and their control of the Congress, they rebuilt the institutional order and destroyed the checks and balances inherent to democratic rule. Their measures provoked an important opposition, which focused on institutional banners that appealed to the middle classes but not to the popular sectors, which were seduced by populist rhetoric and increasing social expenditures. Both countries are suffering the consequences of the non-existence of a party system, along with the consequences of the fragility of the new leaderships and movements that have sprung up in recent years. Because of the precariousness of the groups in power and those in opposition, the problems of democratic representation are not going away. In the vacuum left by the collapse of the party systems, authoritarian governments emerged, but they failed to establish new hegemonic parties. The new movements appearing in recent years are characterized by personalism, precariousness, improvisation, and volatility. This helps explain why Fujimorism collapsed unexpectedly after the irregular reelection of 2000. Fujimorism fell apart because of its internal contradictions and its personalistic nature; more specically, the main explanation for the fall lies in the conicts between the president and his intelligence advisor, Vladimiro Montesinos (Tanaka 2005).

From Crisis to Collapse: Peru and Venezuela

69

The Challenges: How to Build Democratic Representation without Parties


Peru and Venezuela are exceptions in Latin America and among the Andean countries, where crises of representation have brought about a gradual evolution of party systems. In these other countries, the traditional parties are competing with new parties that represent new social sectors. In Peru and Venezuela, a sudden collapse of the party systems occurred, opening the way for political outsiders. The key to understanding the different outcome here lies in the internal dynamics of the political parties at junctures when they were particularly vulnerablethe 1990 election in Peru and the 1998 election in Venezuela. The parties were unable to handle their internal conicts, which resulted in schisms. As a result, the traditional parties were discredited in the eyes of the voters. Once in power, Fujimori and Chvez destroyed the preceding institutional order and established competitive authoritarian regimes, in which democratic representation was replaced by plebiscitarian mechanisms of legitimation. For both countries, major challenges lie ahead. In Peru, the challenge now is to surmount the legacy of a decade of authoritarianism that destroyed state institutions, which were run so as to keep Fujimori in power. Another negative legacy is the fragility of social and political organizations. The party system of the 1980s was destroyed; the gap was lled by Fujimorism and independent movements, but none of these have consolidated themselves. Fujimorism also damaged Peruvian societys capacity for collective action, undermining the representativeness of the social actors and isolating them from society. Contrary to some views, the fall of Fujimorism was not the result of the growth of the opposition or of social protests. That false notion has led to overestimating the capabilities of the political groups now at the center of the political scene and in control of the Congress, and to underestimating the continuity of patterns characteristic of Fujimorism, which continue within the political culture, the media, the judiciary, and other institutions. The campaign for the new presidential and congressional elections in 2001 showed the great weakness of the participating actors, which is also demonstrated by the precariousness of the current government led by President Alejandro Toledo (20012006). In Venezuela, the challenge is to see that the Chvez government does not follow a path similar to that of Fujimorismin other words, to prevent Chvezs control of institutions and the absence of checks and balances from leading to a critical weakening of political competition and pluralism that generates increasing arbitrariness, authoritarianism, and corruption. Such a path may foster a polarization greater than already exists and may open the way for a cycle of greater instability and violence. The opposition to Chvez uctuates between strategies of violence, with calls for a coup (such as the April 2002 coup), and a negotiated institutional solution (a referendum, for example, leading to a recall and new elections). The opposition to Chvez is fractured and offers no clear alter-

70

Tanaka

native, making it hard to visualize a way out of the current situation of profound polarization. In both countries we are witnessing the difculties of conducting politics without parties (Levitsky and Cameron 2001). Precarious and volatile movements and parties generate two perverse logics. First, these groups have limited time horizons; they give priority to short-term logics and are unable to devise long-term policies or strategies. Second, since the movements are new and precarious, their expectations regarding political rewards are not great. It is enough for them to obtain a few mayorships and governorships or congressional seats, and from these positions strive for future growth and consolidation. This logic hinders the formation of coalitions, generates fragmentation, and impedes resolving problems of collective action. For these reasons, in Peru and Venezuela both the actors in power and the opposition are weak. In the Peruvian case, this is the story of movements such as Unin por el Per (Unity for Peru), Somos Per (We Are Peru), Solidaridad Nacional (National Solidarity), and Per Posible (Possible Peru).32 In Venezuela, the same is true of movements such as Convergencia, Proyecto Venezuela (Project Venezuela), and new groups such as Primero Justicia (First Justice) and others. Amid the climate of questioning of politics and parties present throughout the Andean region, the cases of Peru and Venezuela furnish valuable lessons about the importance of parties. Despite their shortcomings, their absence makes problems worse rather than solving them. The cases of Peru and Venezuela offer a remarkable contrast, compared to the other Andean countries, with their party systems in crisis but not in a state of collapse. Citizen discontent and crises of representation have encouraged the traditional parties to promote institutional reforms in order to refurbish their images and to compete successfully with emerging political forces. As limited as these initiatives may be, they open up the political system and create spaces that can be occupied by new social and political actors. In recent years, Colombia, Bolivia, and Ecuador have undertaken decentralization processes, instituted the popular election of local ofcials, introduced new constitutions or constitutional changes that broadened recognition of social rights, and effected political reforms to enhance the accountability of elected ofcialsall with results that are far from insignicant. In Bolivia and Ecuador, the political systems have made great progress toward recognizing the rights of indigenous groups, achieving substantive advances in greater representation for ethnic groups that were traditionally ignored; in Colombia, the heritage of the National Front and its exclusionary nature is denitely gone. In short, despite all the problems, the existence of a party system constitutes an advantage for democratization. Between 1998 and 2002, the region faced a new economic recession, a consequence of the exhaustion of rst-generation market reforms. The Economic Commission for Latin America (ECLA) labeled this period the lost half a decade, and its effects are still impacting the political arena. We currently face

From Crisis to Collapse: Peru and Venezuela

71

a new critical juncture that will surely have medium and long-term consequences. The institutional changes mentioned above have been exhausted, in the sense that the political systems have been opened and new sectors have emerged, but in the crisis context, this emergence has created renewed governability problems. Bolivia may also face the collapse of its party system in the next general elections; Ecuador and Colombia face increasing instability. Democratic representation is still a central issue in the Andes. Parties do not represent adequately. New anti-system leaders emerge, but they further weaken democratic institutions. How to achieve an equilibrium between representation and democratic governance is part of the pending agenda.

Notes
My thanks to Mara Jess Osorio for compiling some of the data presented here, and to Thais Maingn, Scott Mainwaring, and Eduardo Pizarro Leongmez for their intelligent comments on the rst version of the text. Responsibility for its limitations is of course entirely mine. Part of the research for this paper was made possible through assistance from the Consejo Latinoamericano de Ciencias Sociales (CLACSO), under its regional grant program (Programa Regional de Becas CLACSOAsdi) for senior researchers on Latin America and the Caribbean 2001. 1. For an overview of Venezuela, see Caballero (2000) and Levine and Crisp (1999); for Peru, see McClintock (1999b). 2. For the Colombian case, see Pizarro Leongmez; for Ecuador, see Pachano; for Bolivia, see Mayorga, all in this volume. 3. In November 1989 there were municipal elections, and in April 1990 elections for the president of the republic and all congressional seats. 4. According to an October 1989 APOYO poll, 47 percent of voters intended to vote for Vargas Llosa in the April 1990 presidential election. 5. Made up of the Unied Mariateguista Party (PUM), the National Union of the Revolutionary Left (UNIR), and the Popular Front of Workers, Peasants, and Students (FOCEP). 6. Candidates for the municipal elections led in August 1989; presidential hopefuls led in October 1989; congressional candidates in January 1990. During that entire period, the internal struggles of all the parties were daily topics in the news media. 7. Vargas Llosa wanted FREDEMO to run Front candidates in the municipal elections, whereas AP and PPC (Partido Popular Cristiano, or the Popular Christian Party) wanted to run candidates from their own ranks, with Front candidates being nominated only for the presidential and congressional contests. This disagreement led to Vargas Llosas decision to temporarily withdraw his presidential bid. 8. Ricardo Belmont, for example, elected mayor of Lima as an independent, campaigned openly for Vargas Llosa, even making a speech at the latters end-of-campaign rally in Lima. 9. Some writers maintain that the Fujimori phenomenon was the expression of a grave crisis of political representation in Peruvian society, and that it expressed ethnic,

72

Tanaka

cultural, class, and other problems of representation. In my view, such positions illustrate the fallacy of retrospective determinism. Once an event has taken place, an argument is constructed presenting that event as inevitable. Yet less than a month before the election, it was almost impossible to imagine such an outcome. 10. According to a survey rm, APOYO, Fujimori no longer appeared under the heading Others (for very minor candidates) in its poll taken between March 8 and 11, when he had 3 percent of popular preferences. In the March 16 18 poll, he registered 6 percent; in the March 24 26 survey, 9 percent. According to IMASENs March 5 7 poll, Fujimori had 2.5 percent; in the March 9 12 survey, 6.1 percent; and in the March 14 16 survey, 9.5 percent. These gures began to increase at a faster rate, and Fujimori reached 29.1 percent on April 8. 11. See Coppedge (1994); Benton (1997); Corrales (2000); Crisp et al. (2003); Molina and Alvarez (2004), among others. 12. On COPRE, see Jcome (1999). On the relationship between decentralization and the breakdown of the party system, see Lalander (2004) and Penfold (2001). On the effects of the proportional personalized vote, see Crisp and Rey (2001) and Kulisheck and Crisp (2001). 13. The former COPEI candidate, Eduardo Fernndez, was more consistent and supported the constitutional president. The reason for Calderas stance is clear. Unable to win the presidential nomination through COPEI, he needed to create a political space outside the party. Once president, Caldera granted amnesty to Chvez, a decision that allowed the latter to run in the 1998 election. On the imaginary redeemer who would nally take bodily form as Chavism, see Arenas and Gmez Calcao (2000). 14. For an extended discussion of attempts to remove Latin American presidents in recent years, see Prez-Lin (2001). 15. It is revealing to compare the stance of Rmulo Betancourt, long-time leader of AD, with that of Caldera. The former, after serving as president, did not run again. Conversely, Caldera participated in every election he could. The effects of these two behaviors on the destinies of their respective parties are quite different. Betancourts behavior strengthened his party; Calderas damaged his party and the party system. On Caldera, see Crisp et al. (2003); and lvarez (2004). 16. For an overview of the economy in the period from 1989 to 1998, see Hidalgo (2000) and Kelly (2001). 17. On Fujimorism, see Cotler and Grompone (2000); Degregori (2000); Rospigliosi (2000); Marcus and Tanaka (2001), among others. 18. The media were relatively favorable to Chvez until the 1999 Constituent Assembly, when a distancing that later turned into open confrontation began. An illustrative example is Alfredo Pea, former editor of the daily El Nacional. Pea was elected by MVR to the Constituent Assembly and then to the mayorship of Caracas. He is now one of the leaders of the opposition to Chvez (see Petkoff 2002). 19. On this point, see Tanaka (2002). For a different perspective on the Venezuelan case, see Lpez Maya (2001, 2003) and Medina and Lpez Maya (2003). 20. I take the concept of competitive authoritarianism from Levitsky and Way (2002); see also Schedler (2002) and Diamond (2002). For the Peruvian case, see also McClintock (1999a), Conaghan (2001), and Tanaka (1999, 2002). On Venezuela, see Coppedge (2002). 21. In the April 1990 elections, Cambio 90 obtained 21.7 percent of the votes for the Senate and 16.5 percent of the votes for deputies. In November 1992, Cambio 90 Nueva Mayora won 49.2 percent of the votes and won 44 out of a total of 80 congressional seats.

From Crisis to Collapse: Peru and Venezuela

73

22. Yes votes prevailed over No votes, 52 to 48 percent, amid accusations of fraud. 23. In the 2000 election, with Fujimori as the candidate, the Peru 2000 movement got 42 percent of the votes for Congress. Just one year later, the movements identied with Fujimorism, Cambio 90 Nueva Mayora and Solucin Popular, obtained a meager 4.8 percent and 3.6 percent, respectively. 24. See ODonnell (1998). This phenomenon should lead to an in-depth discussion of how to conceptualize democracy and how the international community ought to deal with these matters. The OAS was lax with Fujimori, despite his movements authoritarianism, because Fujimori had won a popular mandate. Now the OAS is in a similar situation with Chvez. 25. On these points, see Ames et al. (2001); Bernales (2000, 2001); Sanborn et al. (2000), among others. 26. On the 2000 election, see the many election-observation reports produced by the OAS mission, the Carter Center, the National Democratic Institute, the U.S. State Department, the International Federation for Human Rights, the Electoral Reform International Service, and the Washington Ofce on Latin America. Also see the reports by Peruvian groups such as Transparencia, Foro Democrtico, Consejo por la Paz, and the Defensora del Pueblo (the government ombudsmans ofce). 27. On the subjects discussed in this section, see Gmez and Patruyo (2000) and Maingn et al. (2000), among others. 28. It is interesting to compare Chvezs and Fujimoris electoral systems. Given that Fujimori had the majority of the votes, it might have been better for Fujimori had he, too, elected his unicameral Congress by a majority system in 1993, 1995, and 2000. On all three occasions, however, Congress was elected from a single countrywide district, a highly proportional system with preferences for minorities. Why did he do this? Fujimori wanted to have total control over his party and did not want to negotiate with local and regional political bosses, something unavoidable with a system of single-member districts, even if he could have controlled the nomination process. Fujimori proceeded in this way because he never established a solid political movement. Chvez, conversely, by 1998 had built a minimal political organization, which he put together between his release from prison and the 1998 election. 29. Neither the Venezuela Constitution of 1999 nor the Peruvian Constitution of 1993 is explicitly designed to build an authoritarian regime. Both constitutions focused on promoting direct democracy and maintained the basic formalities of a democratic constitution. These constitutions later became signicant obstacles for Fujimori and Chvez, and served as useful tools for the opposition in both Peru and Venezuela. Both constitutions allowed for referendums as a way to limit presidential power. Both presidents repeatedly violated the constitutional order created under their political hegemonies. 30. In January 2000 the ANC dissolved itself and was replaced by a mini-Congress made up of some ANC members and others appointed by the ANC. This body operated until the new Congress began its duties. 31. On the 2000 elections, see Carrasquero et al. (2001); and Maingn (2002). Although AD did not nominate a presidential candidate, the party is still a signicant presence in Congress as the prime opposition force. La Causa Radical, despite being the main pillar of the vote for Arias Crdenas (contributing some 19 percent of his 37.5 percent), got a minuscule vote for Congress. 32. The only exception is APRA, but it is partial, given the accentuation of its caudillismo. APRAs presidential candidate, Abel Salinas, obtained a meager 1.38 percent of the

74

Tanaka

vote in the 2000 election, when its congressional list won 5.5 percent. Alan Garca won 25.8 percent in the rst presidential round of 2001 and 46 percent in the second, and the partys congressional list polled 19.7 percent of the vote.

References
lvarez, ngel. 2004. COPEI: La triste historia de un partido sin vocacin de poder. In Los partidos polticos venezolanos en el siglo XXI, ed. Jos Molina and ngel lvarez, 159 94. Caracas: Vadell Hermanos. Ames, Rolando, Enrique Bernales, Sinesio Lpez, and Rafael Roncagliolo. 2001. Situacin de la democracia en el Per (2000 2001). Lima: PUCP. Arenas, Nelly, and Luis Gmez Calcao. 2000. El imaginario redentor: De la revolucin de octubre a la Quinta Repblica Bolivariana. Temas para la discusin, Serie Arbitrada, no. 6. Caracas: CENDES. Benton, Allyson L. 1997. Patronage Games: The Effects of Economic Reform on IntraParty Politics in Venezuela. Paper delivered at the 1997 meeting of the Latin American Studies Association, Guadalajara, Mxico, April. Bernales, Enrique. 2000. La ilegitimidad constitucional del tercer gobierno de Alberto Fujimori. In Per 2000: Un triunfo sin democracia, ed. Cecilia Anicama et al., 57108. Lima: Comisin Andina de Juristas. . 2001. Aspectos constitucionales de la transicin democrtica. In Las tareas de la transicin democrtica, ed. Cecilia Anicama et al., 33 58. Lima: Comisin Andina de Juristas. Caballero, Manuel. 2000. La gestacin de Hugo Chvez. 40 aos de luces y sombras en la democracia venezolana. Madrid: Catarata. Carrasquero, Jos Vicente, Thais Maingn, and Friedrich Welsch, eds. 2001. Venezuela en transicin: Elecciones y democracia, 1998 2000. Caracas: REDPOL. Conaghan, Catherine. 2001. Making and Unmaking Authoritarian Peru: Re-Election, Resistance, and Regime Transition. The North-South Agenda, Paper no. 47, University of Miami ( June). Coppedge, Michael. 1994. Strong Parties and Lame Ducks. Presidential Partyarchy and Factionalism in Venezuela. Stanford, CA: Stanford University Press. . 2002. Venezuela: Popular Sovereignty versus Liberal Democracy. Working Paper, no. 294. Helen Kellogg Institute for International Studies, University of Notre Dame, Notre Dame, IN. Corrales, Javier. 2000. Presidents, Ruling Parties, and Party Rules: A Theory on the Politics of Economic Reform in Latin America. Comparative Politics 32, no. 2 ( January): 127 49. Cotler, Julio, and Romeo Grompone. 2000. El fujimorismo: Ascenso y cada de un rgimen autoritario. Lima: Instituto de Estudios Peruanos. Crisp, Brian. 2001. Candidate Selection in Venezuela (and Its Impact on Legislator Behavior). Paper presented at the Latin American Studies Association 13th International Congress, Washington, D.C., September. Crisp, Brian, and Juan Carlos Rey. 2001. The Sources of Electoral Reform in Venezuela. In Mixed-Member Electoral Systems: The Best of Both Worlds? ed. Matthew S. Shugart and Martin Wattenberg, 173 93. Oxford: Oxford University Press.

From Crisis to Collapse: Peru and Venezuela

75

Crisp, Brian, Daniel Levine, and Jos E. Molina. 2003. The Rise and Decline of COPEI in Venezuela. In Christian Democracy in Latin America, ed. Scott Mainwaring and Timothy R. Scully, 275 300. Stanford, CA: Stanford University Press. Degregori, Carlos Ivn. 2000. La dcada de la antipoltica: Auge y huida de Alberto Fujimori y Vladimiro Montesinos. Lima: IEP. Diamond, Larry. 2002. Thinking about Hybrid Regimes. Journal of Democracy 13, no. 2: 2135. Gmez Calcao, Luis, and Thanal Patruyo. 2000. Entre la esperanza popular y la crisis econmica: Transicin poltica en Venezuela. Cuadernos del CENDES (Caracas) 17, no. 43, segunda poca ( JanuaryApril). Gonzlez, Sonia. 2002. La desconanza en los partidos en Venezuela. Unpublished manuscript. Gonzales de Olarte, Efran. 1998. El neoliberalismo a la peruana: Economa poltica del ajuste estructural, 1990 1997. Lima: Instituto de Estudios Peruanos. Hidalgo, Manuel. 2000. Liderazgo poltico y reforma econmica: El caso de Venezuela, 1989 1998. Zona Abierta (Madrid: Ed. Pablo Iglesias), no. 90/91: 91160. Jcome, Francine. 1999. Reformas polticas en Venezuela: Una evaluacin preliminar. Ciencias de Gobierno (Instituto Zuliano de Estudios Polticos, Econmicos y Sociales), no. 6 ( JulyDecember). Kelly, Janet. 2001. The Syndrome of Economic Decline and the Quest for Change. Unpublished manuscript. Kornblith, Miriam. 1998. Venezuela en los noventa: La crisis de la democracia. Caracas: IESA. Kulisheck, Michael, and Brian Crisp. 2001. The Legislative Consequences of MMP Electoral Rules in Venezuela. In Mixed-Member Electoral Systems: The Best of Both Worlds? ed. Matthew S. Shugart and Martin Wattenberg, 404 31. Oxford: Oxford University Press. Lalander, Rickard. 2004. Suicide of the Elephants? Venezuelan Decentralization between Partyarchy and Chavismo. Helsinki: University of Helsinki. Levine, Daniel, and Brian Crisp. 1999. Venezuela: The Character, Crisis, and Possible Future of Democracy. In Democracy in Developing Countries: Latin America, ed. Larry Diamond et al., 367 428. 2nd. ed. Boulder: Lynne Rienner. Levitsky, Steven, and Maxwell Cameron. 2001. Democracy without Parties? Political Parties and Regime Change in Fujimoris Peru. Paper delivered at the 13th Congress of the Latin American Studies Association, Washington, DC, September. Levitsky, Steven, and Lucan Way. 2002. The Rise of Competitive Authoritarianism. Journal of Democracy 13, no. 2: 51 65. Lpez Maya, Margarita. 1997. The Rise of Causa R in Venezuela. In The New Politics of Inequality in Latin America: Rethinking Participation and Representation, ed. Douglas Chalmers et al., 117 43. Oxford: Oxford University Press. . 2000. La protesta popular venezolana entre 1989 y 1993 (en el umbral del neoliberalismo). In Lucha popular, democracia, neoliberalismo: Protesta popular en Amrica Latina en los aos de ajuste, ed. Margarita Lpez Maya, 21135. Caracas: Nueva Sociedad. . 2001. Partidos de vocacin popular en la recomposicin del sistema poltico venezolano: Fortalezas y debilidades. Paper delivered at the 13th Congress of the Latin American Studies Association, Washington, DC, September. . 2003. Hugo Chvez Fras: Su movimiento y presidencia. In La poltica venezolana en la poca de Chvez: Clases, polarizacin y conicto, ed. Steve Hellner and Daniel Hellinger, 97120. Caracas: Nueva Sociedad. Maingn, Thais. 2002. Comportamiento poltico-electoral del venezolano y construccin de tendencias: 1998 y 2000. Cuadernos del Cendes 49: 49 78.

76

Tanaka

Maingn, Thais, Carmen Prez, and Heinz Sonntag. 2000. La batalla por una nueva constitucin para Venezuela. Revista Mexicana de Sociologa 62, no. 4 (October December): 91124. Marcus, Jane, and Martn Tanaka. 2001. Lecciones del nal del fujimorismo: La legitimidad presidencial y la accin poltica. Lima: Instituto de Estudios Peruanos. McClintock, Cynthia. 1999a. Es autoritario el gobierno de Fujimori? In El juego poltico: Fujimori, la oposicin y las reglas, ed. Fernando Tuesta, 65 96. Lima: Fundacin Friedrich Ebert. . 1999b. Peru: Precarious Regimes, Authoritarian and Democratic. In Democracy in Developing Countries: Latin America, ed. Larry Diamond et al., 309 65. 2nd ed. Boulder, CO: Lynne Rienner. Medina, Medlo, and Margarita Lpez Maya. 2003. Venezuela: Confrontacin social y polarizacin poltica. Bogot: Ed. Aurora. Molina, Jos, and ngel lvarez, eds. 2004. Los partidos polticos venezolanos en el siglo XXI. Caracas: Vadell Hermanos Editores. Nam, Moiss. 1993. Paper Tigers and Minotaurs: The Politics of Venezuelas Economic Reforms. Washington, DC: Carnegie Endowment for International Peace. ODonnell, Guillermo. 1998. Accountability Horizontal. Agora (Buenos Aires), no. 8: 91160. Payne, Mark, et al. 2003. La poltica importa: Democracia y desarrollo en Amrica Latina. Washington, DC: BIDIDEA. Penfold, Michael. 2001. El colapso del sistema de partidos en Venezuela: Explicacin de una muerte anunciada. In Venezuela en transicin: Elecciones y democracia, 1998 2000, ed. Jos Vicente Carrasquero, Thais Maingn, and Friedrich Welsch, 36 51. Caracas: REDPOL. Prez-Lin, Anbal. 2001. Public Opinion and Executive Accountability: The Political Economy of Impeachment Crisis. Paper delivered at the 2001 meeting of the Latin American Studies Association, Washington, DC. Petkoff, Teodoro. 2002. El presidente acusa a los medios: La luna de miel ha terminado. Etctera (Mexico City) (March). Roberts, Kenneth M. 2003. Party System Collapse Amid Market Restructuring in Venezuela. In Post-Stabilization Politics in Latin America: Competition, Transition, Collapse, ed. Carol Wise, Riordan Roett, and Guadalupe Paz, 249 72. Washington, DC: Brookings Institution Press, 2003. Rospigliosi, Fernando. 2000. Montesinos y las fuerzas armadas: Cmo control durante una dcada las instituciones militares. Lima: Instituto de Estudios Peruanos. Sanborn, Cynthia, Francisco Eguiguren, and Bruce Kay. 2000. Democracy and Governance in Peru: An Assessment. Lima: Management Systems International (MSI), under contract to the U.S. Agency for International Development. Schedler, Andreas. 2002. The Menu of Manipulation. Journal of Democracy 13, no. 2: 36 50. Tanaka, Martn. 1998. Los espejismos de la democracia: El colapso del sistema de partidos en el Per, 1980 1995, en perspectiva comparada. Lima: Instituto de Estudios Peruanos. . 1999. La consolidacin democrtica en Amrica Latina y la importancia de la competencia poltica: Lecciones desde la experiencia peruana. In El juego politico: Fujimori, la oposicin y las reglas, ed. Fernando Tuesta, 43 63. Lima: Fundacin Friedrich Ebert. . 2002. La situacin de la democracia en Colombia, Per y Venezuela a inicios de siglo. Lima: Comisin Andina de Juristas.

From Crisis to Collapse: Peru and Venezuela

77

. 2003. The Political Constraints on Market Reform in Peru. In PostStabilization Politics in Latin America: Competition, Transition, Collapse, ed. Carol Wise, Riordan Roett, and Guadalupe Paz, 221 48. Washington, DC: Brookings Institution Press. . 2005. Peru, 1980 2000: Chronicle of a Death Foretold? Determinism, Will, Actors, and De Facto Powers. In The Third Wave of Democratization in Latin America: Advances and Setbacks, ed. Frances Hagopian and Scott Mainwaring, 261 88. Cambridge: Cambridge University Press. Tanaka, Martn, and Carolina Trivelli. 2002. Las trampas de la focalizacin y la participacin: Pobreza y polticas sociales en el Per durante la dcada de Fujimori. Documento de Trabajo, no. 121. Instituto de Estudios Peruanos, Lima. Tuesta, Fernando. 2001. Per poltico en cifras, 18212001. 3rd ed., revised and expanded. Lima: Fundacin Ebert. Weyland, Kurt. 2002. The Politics of Market Reform in Fragile Democracies: Argentina, Brazil, Peru, and Venezuela. Princeton, NJ: Princeton University Press.

3 Giants with Feet of Clay: Political Parties in Colombia

Eduardo Pizarro Leongmez

ntil the late 1990s, Colombias two-party system was one of the oldest and most stable in the world. In 1849, Don Ezequiel Rojas created the programmatic foundations for the Liberal Party (PL). The same year, Mariano Ospina Rodrguez and Jos Eusebio Caro wrote the founding doctrines of the Conservative Party (PC). This was over 150 years ago. Neither England nor the United Statesthe two classic models for the two-party system can claim such a long tradition. The two parties that dominated the nineteenth century in England (Conservative and Liberal) gave way to two others (Conservative and Labour) that prevailed throughout the twentieth century. In the United States, the two current parties (Republican and Democrat) emerged shortly before the Civil War (1861 65). Unlike the party systems in England and the United States, the two-party system in Colombia actually functioned as a multiparty system, owing to a long tradition of factional struggle. In Colombia, as in Uruguay, party factions were political entities with a higher degree of discipline and cohesion than the parties proper; the parties were, in fact, little more than two political subcultures behind which the real political machinery moved. For all practical purposes, factions separate from the ofcial Liberal or Conservative Partiessuch as the Liberal Revolutionary Movement (Movimiento Revolucionario Liberal, or MRL) or the Conservative Unionwere parties in themselves. Each had its national directorate, its departmental and municipal directorates, its parliamentary caucus, and its government platform. This system has undergone a deep transformation in the last decade. The internal fragmentation of the Liberal and Conservative Parties has intensied, with both moving from an internal structure based on institutionalized factions (Morgenstern 2001, 236)1 to one based on personalistic factions or, to use the term popularized in Colombia, electoral micro-enterprises (Pizarro

Giants with Feet of Clay: Colombia

79

Leongmez 2002). At the same time, there has been a progressive decline in bipartisanshipfrom a system in which the two parties controlled over 90 percent of congressional seats toward a system with other political movements that are gaining greater access to representation in local, departmental, and national legislatures with each passing day. The Colombian party system is undergoing a profound transformation, the results of which are still uncertain. Along with Uruguay and Honduras, Colombia until the late 1990s had one of the three oldest and most stable two-party systems in Latin America. A reconstruction of the two-party system is unlikely, especially given the apparently irreversible crisis in the PC, whose decline has been steady over the past three decades.2 Is Colombia headed for a collapse of the party system, as in Venezuela, Peru, and Italy? Or toward a reorganization of the party system based on political coalitions that have blossomed in recent elections? No one dares make any predictions. A climate of enormous uncertainty reigns. In this context of party atomization, the forms of political representation are also undergoing profound change. Two or three decades ago, the two traditional parties, typical multiclass, catch-all parties, monopolized the immense majority of partisan support. For generations, Colombians of all social classes identied with Liberals or Conservatives (Losada and Vlez 1981). Today, party identity survives only in a few rural or semi-urban zones and among elderly urban voters. The electoral panorama is dominated by electoral micro-enterprises. This form of personalistic atomization is the manifestation of a phenomenon unparalleled elsewhere in the world, except probably in Israel: the political parties inability to select candidates for legislative assemblies. In Colombia, candidates from all the political or social movements (with only a few exceptions) increasingly nominate themselves, design their own campaigns, and organize their own nances. The political parties simply give their labels so that the candidacies can run for ofce. The political panorama is therefore dominated by hundreds of micro-representations of all sorts (political, corporative, regional, ethnic, religious) and on all levels (national, regional, and local) in which each parliamentarian, deputy, or councilman covers a certain nucleus of the population.3 The following analysis of the personalistic and particularistic representation that prevails in Colombia draws upon studies by Carey and Shugart (1995), Shugart (1999), and, more recently, the Inter-American Development Bank (Seddon et al. 2002) and Panizza (2001), which focus on the impact of electoral systems on the level of particularism in a given political regime.4 In these studies, particularism is dened as a political leaders ability to further his or her career by supporting specic social groups rather than national platforms. As we will see, increasingly there are few incentives in Colombia to build a political career based on a party platform. Several institutional factors motivate politicians to cultivate careers based above all on their personal reputations, and therefore

80

Pizarro Leongmez

on the constitution of electoral micro-enterprises composed of specic voters (the Colombian expression for a clearly identiable and narrow electoral constituency). This phenomenon has given rise to a model of political representation that I refer to as dual representation. 5 On the one hand, legislative power is generally an expression of segmented and particularistic representation; on the other, executive power is the expression of more general interests. The same phenomenon is at the root of the increasing hypertrophy of presidential power, upon which the states conduct and the very governability of the political system rest, especially in relation to the gure of the president.6

Institutionalization and De-institutionalization


In 1995 the well-known book Building Democratic Institutions: Party Systems in Latin America (1995), edited by Scott Mainwaring and Timothy Scully, focused attention on the degree of institutionalization or de-institutionalization of the continents party systems. According to this conceptual framework, in a system of institutionalized parties, party organization is generally solid, electoral volatility is low, parties are deeply rooted in society, and their operations are a result of institutional routines rather than charismatic personalities or leaders. And viceversa: in a weakly institutionalized system, party organization is weak, electoral volatility is high, parties are not deeply rooted in society, and personalities dominate party life. In this comparative analysis, the Colombian party system appeared to be one of Latin Americas most institutionalized. How, then, can one explain the rapid de-institutionalization process in the Colombian party system? Randall and Svasand (2002) suggest that analyses of party systems are based on a supposition that may be true in certain contexts and false in others. This supposition assumes that the institutionalization of a party system is equivalent to the institutionalization of the parties comprising the system. This is not always the case. The four criteria that Mainwaring and Scully (1995, 4 5) used to measure the institutionalization of the party system placed Colombia on a high level but, in turn, did not mention the internal erosion of the Liberal and Conservative Parties.7 The decline of the two historical parties generates many questions since both parties had, for decades, been able to survive through even dramatic changes.8 Their ability to adapt was considered one of their most specic traits. What happened? The literature on parties suggests many hypotheses for the weakening of political parties on a global scale. Pennings and Hazan (2001, 267) maintain that in the majority of modern democracies, the relationship between parties and voters has weakened. The reasons for this are often related to growing levels of education and social well-being that render citizens less dependent on parties, unions and other collective representational bodies. This is indeed probable in

Giants with Feet of Clay: Colombia

81

countries with a high level of social well-being.9 But in nations where poverty and income inequality have increased, the explanation is most likely different. In Latin America, and particularly in Colombia, different explanations exist. Institutionalist explanations have shown that different political and electoral reforms, beginning in the 1980s, undermined the traditional forms of political action in Colombia and created a climate that favored the current party atomization (Cox and Shugart 1995). Authors, including Kenneth Roberts (2001, 183 84), have emphasized political economy factors, afrming that the crisis of importsubstitution industrialization and the neo-liberalism that took its place in the 1980s and 1990s meant a new critical juncture for Latin Americas political development. The retreat of the state, the scal crisis, and austere economic policies reduced the ow of resources needed to sustain both parties founded on client-based networks (Colombia) and those founded on corporatist networks (Venezuela). According to Roberts, the new economic model brought with it the need to construct a new matrix of political representation. Some parties in the region adapted (e.g., in Chile, Uruguay, and Costa Rica); others did not. Finally, there are cultural explanations for rapidly declining party loyalty, a consequence of the weakening of the old channels of political socialization such as family and region (Pinzn 1998). An analysis of the steady decay of the two-party system in Colombia can and should be based on a number of explanations, including institutional as well as cultural and economic factors. This brief discussion, however, will be limited to the institutional dimensionwith an emphasis on the impact of electoral laws and rules regulating party operations on the growing trend toward personalistic atomization.

Personalism and Particularism


The growing atomization of the traditional parties (the Liberals and Conservatives) and the profound fragmentation of the so-called third forces (the minor parties) in Colombia have caused two phenomena that feed off one another: on the one hand, an extreme personalization of political life and, on the other, an increasingly particularistic political representation, oriented more toward specic sectors of the population. In a pioneering work, Carey and Shugart (1995) proposed a method of estimating the relative value of personal reputation versus party reputation for elected parliamentarians or congressional candidates seeking to further their political careers. Recently, a team from the Inter-American Development Bank revived this proposal to measure the impact of electoral systems on the high or low level of particularism in the world (Seddon et al. 2002; Panizza 2001).10 The central thesis of these studies is that levels of particularism are closely related to politicians predominant reputation. If the party label is central to politicians

82

Pizarro Leongmez

electoral prospects, politicians tend to value projects of collective interest that benet the parties as channels for aggregation and articulation of social interests. If, on the contrary, a personalistic reputation is central to politicians election, politicians tend to favor their personal reputation and, therefore, the interests of specic social segments in which their political inuence is rooted. The most important conclusion of the IDB study is that both excessively party-dominant systems (in which parties develop corporative interests and pay little attention to the demands of voters) and excessively personalistic systems (in which politicians subordination to local or circumscribed interests prevents them from generating national agendas) have the most negative effect on good legislative governance that can lead to economic development. Conversely, intermediate systems in which stable party systems existwhether strong (Great Britain) or weak (the United States), but coupled with channels for integration and accountability among the voters and their representativesare ideal. This might be achieved in the one-member districts found in some countries.11 To determine the impact of electoral institutions on the type of reputation and, consequently, the model of representation (national or particularist), the authors use three variables: the nomination of candidates (or ballot), whether votes are pooled so that votes for one candidate of a party can benet other candidates of the same party, and the vote. Each receives a rating of 0, 1, or 2, where the lowest number indicates that the system favors parties, while the highest number indicates that the system tends to favor personalistic attitudes among candidates and legislators.12 The Index of Political Particularism measures the existing incentives for politicians to build political support based on particular electoral districts rather than adherence to party platforms. A high score on this index indicates that the system is candidate-centric with strong incentives for politicians to cultivate, above all, circumscribed interests. A low score is associated with party-centric electoral systems. In Colombia, the electoral and nomination systems motivate politicians to cultivate personal reputation, therefore favoring particularist representation. Initially, this would seem a contradiction, given the fact that the system of proportional representation with closed lists tends to favoras proven in other international experiencesthe role of the party elite in selecting candidates and ranking them on the ballot. Proportional representation (PR) with closed lists should in theory encourage a highly party-centric system. However, due to a number of factors, PR in Colombia has transformed into what Carey and Shugart (1995, 429) refer to as a Personal-List Formula. The origins of this situation, or its institutional dimension, lie in several factors: rst, the total freedom of legally established parties to present a limitless number of lists of candidates for public ofces; second, the ridiculous nature of the current CNE (National Electoral Council) standards for forming a political party or movement; 13 and third, the almost total absence of entrance barriers to the electoral arena.

Giants with Feet of Clay: Colombia

83

According to Article 108 of the 1991 Constitution, political actors are those parties, political movements, social movements, or signicant groups of citizens able to participate in elections through endorsement, the creation of a party or movement recognized by the CNE, or payment of a registration fee known as a caucin mnima.14 Let us analyze each of the three variables (ballot, pool, and vote) separately.

Nomination of Candidates
The selection of political candidates by parties is fundamental to a democratic system. As Gallagher and Marsh (1988, 1) afrm, The quality of the selection of candidates determines the quality of representatives elected to parliament, often of the members of government and, by extension, national politics. A change in the selection of party procedures in any country has profound implications on the way politics function. A 1999 issue of the journal Party Politics reected on the current function of political parties, recognizing that party functions are not what they once were but that, nonetheless, the nomination of candidates continues to be an essential function. The modern party is a voluntary association whose declared aim is to be represented, and to lead the institutions of government, in a given state or political community. To this end, the party regularly engages in fullling three critical functions: (1) Nominating candidates for public ofce; (2) Adopting statements of public policy, primarily in an election platform; (3) Mobilizing support for each of the above candidates (public ofcials) and policies (Yanai 1999, 7). The selection of political candidates is undergoing changes on an international level. In Western Europe, for example, parties are more open to having a greater number of their members involved in choosing parliamentary candidates. What is the reason for this change? Why, in spite of the fear aroused by the weakening of parties in the United States due to the role of the primaries, are European parties opening, however timidly, the oodgates for popular participation? Many analysts relate this change to the weakening of linkages between parties and their voters. Many parties are experiencing a reduction in the number of members or activists (Mair and van Biezen 2001) and are adopting different strategies for recovering their ties to voters. One such strategy is the democratization of the selection of candidates (Pennings and Hazan 2001, 268). By increasing the number of people involved in selecting the candidates, parties seek to strengthen the feeling of belonging to the party. In Colombia, the process by which candidates are selected for legislative bodies was not democratized: it was disjointed. There is no selection of candidates from above (i.e., from the party leaders). In fact, there is no selection of candidates at all. Candidates are self-chosen and the parties simply grant them their

84

Pizarro Leongmez
Table 3.1 Number of Lists that Registered for Senate and Lower-Chamber Elections, 1958 2002
Year Senate House

1958 1960 1962 1964 1966 1968 1970 1974a 1978 1982 1986 1990 1994 1998 2002
SOURCE :
a

67 97 147 206 176 210 225 202 213 251 319 321

83 113 143 192 215 221 316 253 308 343 330 351 628 692 883

Registradura Nacional del Estado Civil. Members of the House were elected, from 1958 until 1970, for a period of two years. Beginning in 1970, terms were changed to correspond with those of the Senate (four years).

respective endorsements to participate in the elections on their behalfin the so-called Carnival of Endorsements.15 Add to this indiscriminate concession of endorsements the near absence of CNE requirements for recognition of a political party or movement, or the simple chance to participate by paying a small fee (caucin mnima), or signing a responsibility policy (pliza de seriedad) (Art. 9, Law 130 of 1994), and, for all practical purposes, there are no entrance barriers to the electoral system.16 Table 3.1 illustrates the degree to which the electoral system has been fragmentedin other words, its degree of personalism. This is the main conclusion of a report by a commission of international consultants comprised of Arturo Valenzuela, Joseph Colomer, Arend Lijphart, and Matthew Shugartunder contract to Andrs Pastranas government: Colombias current electoral system is the most personalistic in the world (Lijphart et al. 1999, 237). Although political parties currently control the party labels, which can be used only with the partys endorsement,17 in practice, given the lack of entrance barriers and the fact any candidate who wishes to run can do so through the mechanisms described earlier (endorsement, fee, or simply a new label), the traditional parties, as well as many new parties, make indiscriminate endorsements. In the literature on Brazil these are called rental partiesthat is, parties that are nothing more than exible umbrellas designed to shelter all kinds of personalist factions.

Giants with Feet of Clay: Colombia

85

The system that allows a legislator to resign and turn his or her seat over to someone else who ran for ofce under the same party label has reinforced factionalism. The Constitutional Assembly of 1991 put an end to the old, much criticized system of suplentes that had existed until that time, whereby the principal and the suplente rotated in the seat.18 When Article 134 of the 1991 Constitution (Vacancies due to complete absences of members of Congress will be lled by unelected candidates, according to the order of inscription on the corresponding list) became law, it opened the way for a completely disjointed system known as the carrousel in Colombian parliamentary jargon. Legislative Act 03 of 1993 perversely replaced the gure of the personal suplente: now all members on the list are potential substitutes who can assume the seat of the legislator. According to this legislative act, a member of Congress can request an unpaid leave of absence for three months, renewable. Leaves are granted in cases of temporary absence, suspension of investiture by virtue of a judicial decision, incapacity certied by an ofcial measure, domestic calamity duly proven, and cases of force majeure. Owing to this lax regulation, members of Congress plan in advance within their electoral micro-enterprise the amount of time each member on the list will spend occupying a particular congressional seat. This perverse mechanism is completely functional for the purposes of electoral microenterprises. These are constructed from an agglomeration of votes from different political leaders in a region, with the goal of building a viable electoral enterprisethat is, one capable of capturing a seat based on electoral remainders (residuos) (i.e., winning a seat after the rst seats in an electoral district are allocated by the electoral quotient for that district) in the House or Senate. A study by the Visible Congress Project (Congreso Visible) of the University of the Andes revealed that, by the end of 2001, 247 substitute legislators had served in the Senate and the Chamber of Deputies between 1998 and 2001 without being directly elected. These individuals had occupied the second, third, fourth, and even lower positions on the list that had originally won the congressional seat.19 This is one of the greatest aberrations of the Colombian political system, allowing persons not elected by popular vote to assume legislative ofce. Substitutes who failed to win election routinely become members of Congress when the elected legislator takes a leave. The phenomenon not only reinforces the large-scale atomization of the current Colombian Congresssince the political aims of the substitute member of Congress may differ from those of the elected parliamentarianbut also inhibits the formation of a professional political class. With regard to the subject at handpolitical representationits impact is to reinforce particularism in Congress. Whether their stay in Congress is long or short, legislators are interested only in successfully promoting those laws that directly affect their electoral efdom.

86

Pizarro Leongmez

War of the Leftovers


The disjointing of the Colombian electoral system, in which only occasionally do single-party lists or candidates exist, generates a formidable intraparty war among multiple lists and candidates in each electoral district. This system has no parallel elsewhere in Latin America. As Rodrguez maintains (2003, 6), The electoral formulai.e., the specic way votes are translated into seatsreinforces this tendency. The Colombian electoral system allocates votes using the Hare quota, without vote pooling at the party level. This formula contains no structural incentives to enhance internal party cohesion. It disproportionally rewards small lists, increasing intraparty competition. With this organizational weakening of the parties and their institutionalized factions, it is impossible to speak of vote pooling at the sub-party level. The votes that one list wins cannot be transferred to another list, even within the same party in the same electoral district. Therefore, there is zero-sum competition among the many lists of the same party in the same district. If vote pooling occurs at all, it is within each individual list. Seats are allocated rst according to how many electoral quotients a party reaches in a district. The electoral quotient equals the number of seats divided by the number of votes in a district. After this, lists win seats according to the order determined by the Hare quota of electoral remainders. As Table 3.2 shows and Cox and Shugart (1995) argued, lists that win more than one seat are disappearing. The Colombian electoral system has come to resemble, in spirit if not in letter, the pre-1993 Japanese system of the single nontransferable vote. As seen in Table 3.2, in the 1991 elections, nine lists elected more than one senator, obtaining a total of 34 of the 100 elected senatorial seats. In contrast, in subsequent elections only three lists have elected more than one senator. In 1991, sixty-six lists were able to elect only the head of their list, ninety-four lists did so in 1994 and 1998, and ninety-three elected only the head of their list in 2002.
Table 3.2 Electoral Performance of Lists That Elected More Than One Senator, 19912002
Seats won by Seats won by Number electoral quotients Hare remainders of lists that (lists that elected (lists that elected elected more than more than more than one senator one senator) one senator) Total seats elected by lists that elected more than one senator

Year

1991 1994 1998 2002


SOURCE :

9 3 3 3

26 3 3 4

8 3 3 3

34 6 6 7

Registradura Nacional del Estado Civil.

Giants with Feet of Clay: Colombia

87

At all levels, from the municipal to national, the overwhelming majority of Colombian electoral lists are unable to elect more than one candidate. This phenomenon is even more pronounced at the local (municipal council members) and regional (departmental assemblies) levels, or in lower-house elections (with departmental constituencies), than in the Senate, due to the high costs of campaigning for a list with a national constituency.

The Ballot
Due to the large number of electoral lists and candidates that represent the same political party or movement in each district, except in special situations a voter cannot support a political partys or movements ofcial candidate or list. Instead, she must support the candidate or list of one of the factions competing in that district. For example, during the 2002 senatorial elections, the National Liberal Directorate (Directorio Nacional Liberal) did not present just a single list, but instead endorsed 148 lists, each representing the Colombian Liberal Party. The CNE recognized not only an assortment of lists representing the different liberal factions, but also 26 lists of autonomous parties or movements. The National Conservative Directorate presented 25 ofcial lists and 31 dissident lists from among its different internal factions. The result has been brutal intraparty competition in which personal reputation, above all, is used to distinguish oneself from the dozens of ofcial candidates within the same party. In this respect, the Colombian electoral system, according to Carey and Shugarts criteria, receives the highest possible score of two (each voter casts one vote, either for a candidate or a party faction). The electoral ballot reinforces this phenomenon. Distributed by the state through the National Registry of Civil Status (Registradura Nacional del Estado Civil), the ballot contains a photograph of the person heading the list. In the 2002 congressional elections, the ballot was an enormous poster with 322 photographs. Voters were thus able to see the head of each respective list, but none of the other candidates appeared on the ballot.

National Constituency and Particularism


One of the principal political-electoral reforms of 1991 was the creation of a nationwide electoral district for the Senate. Since then, all 100 senators have been elected in a single nationwide district. Members of the Constituent Assembly sustained two motivations to justify this reform: rst, they argued that political minorities would have easier access to Congress through a larger district given that, by way of a disperse strategy (Botero 1998), these groups would be able to win votes throughout the entire country and gain one or more seats; and second, they felt that the creation of a nationwide district could serve as an incentive for new national leadership and projects.

88

Pizarro Leongmez
Table 3.3 Senate Seats Won by Political Parties and Movements, 19912002
Party or Movement 1991 1994 1998 2002

Partido Liberal Partido Conservadora Unin PatriticaPartido Comunista Colombiano Accin Democrtica M-19 Movimiento Unitario Metapoltico Partido Nacional Cristiano Unin Cristiana Laicos por Colombia Movimiento Compromiso Cvico Cristiano Alianza Nacional Popular Movimiento Autoridades Indgenas de Colombia Movimiento Obrero Independiente y Revolucionario Educacin, Trabajo y Cambio Social Movimiento Cvico Independiente Movimiento Convergencia Popular Cvica Movimiento Ciudadano Alianza Social IndgenaMCIConar Antanas Movimiento Bolivariano Vamos Colombia Movimiento de Defensa Ciudadana Movimiento Independiente Frente de Esperanza Alianza Social Indgena Movimiento Fuerza Independiente Movimiento Popular Unido Movimiento Ciudadanos por Boyac Movimiento Convergencia Ciudadana Movimiento Dejen Jugar al Moreno Movimiento Huella Ciudadana Movimiento Independiente de Renovacin Absoluta Movimiento Progresismo Democrtico Movimiento Somos Colombia Movimiento Unionista Partido Social Demcrata Colombiano Partido Unidad Democrtica Movimiento Vanguardia Social y Moral Total
SOURCE :
a

59 26 1 9 1 1 1 1 1

59 32 1 1

61 27

53 26

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 1 1 1 1 1 1 1

100

100

100

1 2 1 1 1 1 1 1 1 1 1 1 1 1 100

Registradura Nacional del Estado Civil. The gures for the two traditional parties include all the factions that belong to those parties.

The rst objective was fullled. The scope of political and social representation has widened, as seen in Table 3.3. Many parties have gained representation in the Senate. The second objective, to the contrary, has generally speaking been an enormous failure, due to the spreading phenomenon of personalist factions. What happened? In general, small districts tend to favor personalism and particularism. For example, Thomas Lancaster (1986, 70) has stated that due to the fact client-based

Giants with Feet of Clay: Colombia

89

(pork barrel) projects are distribution policies directed at geographically dened voters, a geographically unlimited electoral system must be related to some degree of client-based activity. Conversely, large districts are better suited to the creation of national agendas supported by nationwide parties. Carey and Shugart share this point of view, but only in cases where intraparty competition does not exist. According to them, with closed-list proportional representation, as the size of an electoral district grows, and therefore as the number of candidates on a list grows, party reputation will be the dominant factor. With other electoral formulae, on the contrary, increasing district size can heighten the importance of individual reputation since each candidate must compete with candidates from his or her own party. As the number of co-party members from which a candidate must distinguish himself increases, so does the importance of establishing a personal reputation which is distinct from that of the party (Carey and Shugart 1995, 430). As we have shown, intraparty competition in Colombian elections at all levels is erce, owing to the dozens of lists (or candidates) that represent all the political parties or movements. This intraparty ghting is even ercer in the Senate due to the number of seats (100) and candidates involved in the countrywide district. In the Senate, candidates have a great need to differentiate themselves from others within the party by means of a dened personal image. This need to emphasize ones personal reputation in turn accentuates the need to build ones own electoral niche. As demonstrated by Felipe Botero (1998), the weight of the local vote continues to be a determining factor in electing senators. On average, 66.72 percent of senators were elected by votes coming from just one department in elections between 1991 and 1998. In sum, not only do the three variables (ballot, pool, and vote) tend to generate an extremely personalistic system, but national constituencies, far from favoring the formation of national leadership and reinforcing national parties, have instead deepened the extreme atomization of the political parties in Colombia. In summary, Colombia scores a 2 2 2 with regard to Carey and Shugarts three variables. This is the highest possible level of personalism, comparable only to the single non-transferable vote (SNTV) system that existed in Japan and the SNTV open-endorsement system used in the Philippines. Carey and Shugart (1995) gave Colombia a score of 2 12, presupposing the existence of pooling at the individual list level. As we have seen, although electoral norms foresee this possibility if a list wins more than one seat, in fact, few lists win more than one seat in the Senate, and almost no lists win more than one seat in other legislative bodies (the Chamber of Deputies, departmental assemblies, and municipal councils). Furthermore, the national constituency for the Senate reinforces the impact of the electoral system, owing to the high degree of intraparty competition generated by elections to this body.

90

Pizarro Leongmez

Electoral Niches and Lost Votes


The transition from institutionalized factions to personalistic factions in the traditional parties, as well as the severe fragmentation among the minor parties known in Colombia as third forces, has driven each parliamentary group, and even each individual member of Congress, to seek a determined electoral nichethat is, a specic social sector as a foundation for electoral support. This specic voter, whether through client-based, corporatist, regional, ethnic, or religious networks, is the object of countless diverse micro-representations. 20 There are many examples, among them the Liberal Partys Social Security Movement (Movimiento por la Seguridad Social). Led by Liberal senator Alfonso Angarita Baracaldo, the movement provides congressional representation to retired people who receive state pensions, and it monopolizes parliamentary initiatives that favor this social sector. Another example is the Education, Work, and Social Change Movement (Movimiento Educacin, Trabajo y Cambio Social) led by a senator from the independent left, Jaime Dussn, who is the spokesman for public school teachers. The Christian churches are divided among those who follow the Christian Union (Unin Cristiana) and supporters of the National Christian Party (Partido Nacional Cristiano). Indigenous communities are split between the Social Indigenous Alliance (Alianza Social Indgena) and the Indigenous Authority Movement of Colombia (Movimiento Autoridades Indgenas de Colombia). These personalist factions of traditional parties, micro-parties, or micromovements founded to exploit the wishes of specic voters do not make up the bulk of the members of Parliament. In numerical terms, they are a small parliamentary minority, but they have enormous political signicance. They are typically elected by independent urban voters, generally well-educated individuals with higher than average incomes (Pinzn 1998, 407). These members of Congress, whose electoral support is based on programmatic issues (in Colombia referred to as voto de opinin, or opinion vote), do not require huge sums of money for their political campaigns because they are all well known and highly regarded. These legislators therefore tend to be more critical of political corruption. The absence of a specic electorate allows them greater autonomy than most members of Congress have, and in general they are more apt to propose and defend national agendas. This means that the existence of national and particularist agendas as a dividing line between the executive and legislative branches contains one notable exception, in that these members tend to be the main leaders in Congress for both the government and the opposition. To a large extent, if the Colombian Congress occasionally produces some worthwhile legislation, it is because of presidential leadership and the leadership of these more programmatically oriented members of Congress.21

Giants with Feet of Clay: Colombia

91

One of the most negative consequences of this extreme atomization of political representationbeyond the disproportionate growth of lists and candidates under the rubric of dozens of political parties, micro-parties, and movements is the increase in wasted votes. As political labels proliferate along with the number of lists and candidates, the number of unrepresented sectors has also increased. Cumulatively, a large number of electoral lists fail to win any seats in Congress but still win a substantial share of the vote; the citizens who vote for these lists are effectively unrepresented. In 1998, 222 lists for the Senate 70 percent of all the listswon a total of 2,540,000 votes without winning any seats. In the Chamber of Deputies, the situation is even more worrisome: of the 696 lists that competed in elections that same year, more than 500 were left without representation. At the level of departmental and municipal legislatures, the situation is even more dramatic. Particularist representation can contribute to improving political representation for some social sectors (e.g., indigenous communities),22 but on the whole, there are more losers than winners. Faced with the parties inability to aggregate and represent collective interests, the corporatist, regional, or other types of micro-representations come together in a game with few winners and many losers.

Parliamentary Administration
How does this atomization and personalization of party life affect Congress? The central hypothesis of many studies on this subject is that the more personalistic the vote, the more individualistic the legislators conduct. And conversely, the more partisan the vote, the more partisan parliamentary conduct will be (Amorim Neto and Santos 2001, 214). In Colombia, the predominance of the personalistic vote results in severe internal fragmentation of all popularly elected bodies, especially Congress, which generates party indiscipline.23 This correlation between personalization of the vote, fragmented representation, and congressional indiscipline should surprise no one since, as recent studies demonstrate, the more fragmented the Congress the greater the likelihood of strategic conduct and manipulation of the legislative agenda by individual legislators.24 Such constant negotiation between the government and individual members of Congress in order to sway the vote drastically increases the transaction cost of moving legislation forward.25 A study by the Center for Socio-Legal Investigations (Centro de Investigaciones Sociojurdicas, or CIJUS) at the Universidad de los Andes (Bejarano et al. 2001) illustrates this relationship. This study analyzed the legislative period from July 20, 1998, through July 20, 1999. Immediately striking is the number of proposed laws (354) and legislative acts (35) processed. According to Brian Crisp (1999), legislative ination in Colombia contrasts sharply with the moderate

92

Pizarro Leongmez

level of legislative activity in Venezuela, where in forty years only some 750 laws were proposed in the Chamber of Deputies, among which almost none were local or regional. Various factors explain this legislative ination. First is the lack of congressional parties or caucuses as a source of coherent parliamentary agendas, which generates a lack of lters to control the quality, nancial sustainability, and legal rigor of proposed legislation.26 Second, as noted earlier, is the phenomenon of personalist factions as political and electoral units. The authors of the CIJUS study examined whether proposed bills were local, regional, or national. They discovered that 78 percent of the laws proposed by legislators had a strictly local or regional impact. Only 22 percent of the bills addressed national issues. In addition, the authors of the study also noted a proliferation of local or regional proposals presented by the suplentes on the list. The suplente phenomenon . . . allows persons who were not popularly elected to use their time in Congress to improve their political pull in certain provinces by proposing legislative projects (Bejarano et al. 2001, 232). The suplentes propose, above all, projects that honor regional personalities or commemorate dates of local interest, or laws that authorize the issuance of postal stamps as a way to raise departmental or municipal revenue. The proposed honors, commemorations, and stamps are always accompanied by local investment projects. While a signicant amount of proposed legislation in Congress is targeted at a specic beneciary, proposals originated by the government are usually oriented toward addressing national problems. This is what I call dual representation. The aforementioned CIJUS study of the origins of proposed laws or legislative action made an important discovery: In no other case were regional or local proposals presented by the rest of the state. The government and other branches of public power are in charge of looking out for national interests, while in Congress, these national interests are frequently mixed in with regional or local proposals (Bejarano et al. 2001, 233). As is the case for Brazil (see Amorim Neto and Santos 2001), in Colombia it is because of the executive connection that the governability of the political system does not disintegrate into extreme particularism. The undisciplined vote is not prevalent in parliamentary voting on proposals of governmental interest.27 In Colombia, the executive makes use of a wealth of resources (constitutional and informal) for gaining support for proposed laws. In the constitutional arena, the president enjoys the privileges of parliamentary initiative (Art. 154), design of the national budget, total or partial veto of laws approved by Congress (Art. 167), exclusive legal initiative on certain subjects (Art. 154), and legislative emergency (Art. 163), among other prerogatives. In addition, as is true in many countries, the president can also use informal resources, especially pork barrel legislation, to inuence the conduct of members of Congress. Needless to say, most members of Congress are anxious to obtain these resources. Members of Congress with

Giants with Feet of Clay: Colombia

93

specic voters are, in practice, a kind of lobby working for their efs or corporative interests. As such, they simply negotiate their vote on behalf of certain government proposals in exchange for pork barrel resources. But with few exceptions, they never commit to unconditional and lasting support to the government. Support must be negotiated on each occasion. In this way, in Colombia today, exchanging pork barrel resources for parliamentary support is one of the components of the legislative/executive relationship and congressional conduct.

Conclusion
Colombia has not escaped the party crisis affecting the Andean region. There has been no collapse of the party system such as happened in Peru and Venezuela. Yet the level of disintegration of the traditional Liberal and Conservative Parties and the severe splintering of the third forces seriously affects the governability of the state and the very legitimacy of democratic institutions. If we rely on the results of the 2002 Latinobarmetro, the level of support for and trust in democracy, political parties, and the Congress is very worrisome. Only 39 percent of Colombians support democracy, as opposed to an average of 53.2 percent throughout the Andean region and 56 percent in Latin America generally. Satisfaction with democracy is even lower: 11 percent compared with 21.8 percent among the Andean nations and 32 percent for the entire subcontinent. Only 10 percent of Colombians trust the parties, and 14 percent the Congress. The roots of this deterioration of citizen support for and trust in democratic institutions are complex. However, the level of organizational deterioration in the parties and the party system as a whole has had a pronounced inuence. In a system splintered into hundreds of particularist micro-representations, the number of losers in the electoral game increases exponentially. The electoral and party systems in Colombia since 1991 have not only aggravated the work of governing bodies on all levels, but have helped to heighten the overall crisis of representation in the political system.

Notes
I am extremely grateful for the generous critical comments of Michael Coppedge, Scott Mainwaring, and Ana Mara Bejarano. 1. Morgenstern (2001) differentiates between institutionalized factions ( Japan, Italy, Uruguay) and the ephemeral factionsin English politics, for example centering on a specic topic or election. In Uruguay, the principal role of party factions, in both elections and political decision making, has generated debate as to whether parties or factions

94

Pizarro Leongmez

should be counted to characterize the system (see, e.g., Solari 1964; Gonzlez 1991). The same debate is present in Colombia (Posada-Carb 1997). 2. Conservative Party votes fell from 40 percent in the 1982 parliamentary elections to 23 percent in 1998, which represents a permanent decline. Another manifestation of this decline is the absence, for the rst time since 1942, of an ofcial Conservative Party candidate in the 2002 presidential elections. Unlike the PC, the PL has made an effort to renew itself ideologicallythrough adherence to the Socialist Internationaland organizationally. Time will tell. 3. I later refer to a small parliamentary nucleus with enormous political inuence, whose election by the urban populace grants them greater autonomy and, more importantly, a real incentive to build less particularistic parliamentary agendas. 4. These studies are based on an ever-increasing literature founded on comparative studies, which shows that there is a clear relationship between electoral institutions (rules for nominations, electoral formulae, and scope of districts), electoral strategies, and the conduct of members of Congress (Lancaster 1986; Cain et al. 1987; Mainwaring 1991; Shugart and Carey 1992). 5. The phenomenon of dual representation is present at every level of the state. It is also visible at departmental and municipal levels, where the public policy of governors and mayors contrasts with the more markedly particularistic vision of the deputies and councilmen. 6. Interestingly enough, the Constitution of 1991 sought to strengthen the role of parliament and to improve the balance between the branches of public power. In practice, the atomization of the party system and, therefore, the absence of parliamentary caucuses, rendered the constitutional text on this issue ineffectual (Pizarro Leongmez 1996, 2001). 7. Mainwaring and Scully were, however, well aware of the dark cloud hanging over the two traditional parties: the party system may be entering a dissolution phase after decades of considerable stability. . . . Pronounced factionalism is simply a manifestation of the erosion of party organization in recent decades. Factions can introduce their own group of candidates . . . ; the loss of organizational control over the selection of candidates is extreme (Mainwaring and Scully 1995, 18). 8. For example, military governments (1953 58) and, above all, the attempt by President Gustavo Rojas Pinilla to create a political/military alternative to the parties, the PeopleArmed Forces duo. 9. Or even in certain urban sectors of Latin America. In Colombia, without a doubt, a primary source of the volatile vote in favor of independent candidates comes, above all, from the well-off and well educated (Pinzn 1998). 10. The IDB database covers 155 countries between 1978 and 1997. 11. In the words of Shugart (1999, 319): Personalistic systems fail to provide elections that turn primarily on collective policy, because of the incentives individual members have to collect personal votes, which are better captured through small-scale service provision (clientelism). Camarillian systems fail to connect parties with collective policy preferences because individual members have no incentive to nd out what voters want. Efcient systems, on the other hand, place members in a position where they must balance the interests of both voters and party leaders. 12. This is an ordinal scoring system. The nomination (ballot) measures the degree of control party leaders have over access to party nominations: 0: Party leaders present only one closed list that cannot be altered by voters; 1: Leaders present a list that can be altered by voters; 2: Leaders have no control over the lists.

Giants with Feet of Clay: Colombia

95

Vote pooling measures whether or not the votes received by a party candidate contributed to the number of seats won by the party altogether in any given district: 0: A general pooling in favor of the party as a whole takes place; 1: Pooling takes place at the party faction level; 2: Pooling does not take place. The vote variable indicates: 0: Voters can cast only one vote in favor of the party. 1: They can vote for multiple candidates. 2: They can cast only one vote in favor of a candidate or party faction. 13. According to Article 108 of the current constitution, The National Electoral Council will grant legal status to those political parties or movements . . . proving their existence with at least 50,000 signatures, or that obtained at least this same number of votes in the last elections, or that attained representation in the National Congress. This constitutional norm is developed in Article 3 of Law 130 of 1994, Basic Statutes of the Political Parties and Movements. 14. The caucin mnima allowed a candidate or a list that did not have the support of a party recognized by the National Electoral Council to pay a small fee to register for elections. In the Constitutional Assembly of 1991, a logic of incorporation was imposed as a supposedly suitable mechanism for overcoming the two-party system that was perceived as one of the sources of the national crisis (Pizarro Leongmez and Bejarano, forthcoming). Its effects have been contradictory: on the one hand, it opened the oodgates for new political forces; on the other, it accelerated the tendency toward the disintegration of the party system that began with the National Front of 1958 74 (Pizarro Leongmez 2001). 15. La feria de los avales (The carnival of endorsements), El Tiempo, January 29, 2002. In this editorial, Colombias most inuential daily newspaper denounced the way in which uncontrolled endorsements from some 75 political parties or movements legally recognized by the National Registry of Civil Status was opening the door to candidacies from armed groups (guerrillas and paramilitaries) and drug trafckers. The Liberal Party alone endorsed 148 lists for the national Senate. 16. As a way of explaining the erosion of the two largest parties from the 1968 88 period in Venezuela, some authors have used approaches based on rational choice or game theory models to explain political behavior, arguing that the decentralizing reforms of 1989 weakened the AD and the COPEI by lowering entrance barriers for new parties and encouraging enterprising politicians to abandon or declare their autonomy from political parties (Penfold 2001; Benton 1997). In Colombia, both the 1991 Constitution and the 1994 Law of Parties created perverse incentives that ended up fueling the current party atomization (Pizarro Leongmez and Bejarano, forthcoming). 17. Carey and Shugart (1995, 429) erroneously state that candidates do not require party endorsement to use the party label. According to Article 5 of Law 130 of 1994, Political parties and movements are owners of their names and of the symbol they have registered with the National Electoral Council. Article 9 of the same law states that the party label may be used only with the endorsement of a partys legal representative. 18. Suplente can be roughly translated as substitute or replacement. Until the 1991 Constitution, both senators and representatives were elected along with a suplente, who could replace the elected member of Congress. Article 261 of the 1991 Constitution abolished the old system of suplentes.

96

Pizarro Leongmez

19. Some examples are remarkable: the seat held by the well-known legislator Sergio Cabrera was nally occupied by the seventh candidate on his list, a complete unknown, Edgar Antonio Ruiz. In the Senate, Gabriel Ignacio Zapata temporarily ceded his seat to the eleventh candidate on his list, Julio Acosta. The system is anarchic. The cure proved worse than the disease, afrmed Senator Luis Guillermo Vlez (El Tiempo, November 25, 2001). 20. Japans electoral and party systems have been the subject of interesting comparative studies with the Colombian case. Bouissou (1994, 385) speaks of mini-parties with a single cause (mini seito). In Colombia, the analysis of the political campaigns of clientelistic senators and deputies is enlightening. They state their intentions to work on behalf of either certain social sectors (actors, athletes, retirees, teachers), specic social movements (community movements, cooperatives, unions), particular ethnic groups (black communities from the Pacic region, Paez indigenous people), or particular regions (northern Valle or southern Bolvar provinces). Generally, in Colombia the only caucuses that operate relatively efciently are the regional caucuses, especially the Atlantic Coast caucus that is, members of Congress of all political persuasions who were elected from the Atlantic coastal region. 21. This is the case, for example, with senators such as Antonio Navarro, Germn Vargas, Luis Alberto Ramos, Ingrid Betancourt, Carlos Gaviria, Claudia Blum, Rafael Pardo, and a few others. 22. The indigenous communities are one of the social sectors that has most beneted from the new electoral institutions. This is not merely because of the special electoral district for the Senate (two senators), but also because of the national scope of the electoral district, which has allowed indigenous candidates to amass scattered votes and in this way double their chances of widespread political representation (Pearanda 2002). This has also been the case for the black communities of the Pacic coast, the Christian churches, and other sectors. In other words, not everyone is a loser. 23. Tsebelis (1995) denes party discipline as the degree of party unity in congressional voting. The Rice Index is an operational indicator: RI percentage of votes in favor minus percentage of votes against. 24. This type of conduct from highly fragmented and undisciplined congressional representatives in turn leads to high rates of party switching (or transfugismo, as it is known in Brazil), as well as to the formation of unstable coalitions. Coalitions form with each new issue, and every coalition has a different conguration (Amorim Neto and Santos 2001). 25. In the United States, it is said, no doubt exaggeratedly, that 535 members of Congress equals 535 parties, given the extensive pork barrel legislation that is required to ensure efcient legislative operations. In Colombia, the 1991 Constitution expressly prohibits legislators slush funds (called auxilios parlamentarios) for clientelistic purposes. (These slush funds originated with the constitutional reform of 1968, which institutionalized them as a way to grease congressional approval of constitutional reforms.) Nevertheless, they are constantly revived under different forms: co-nancing funds, co-nancing funds for rural investment (DRI), and special interministerial or interparty funds for the provinces, as occurred with the latest National Development Plan (Plan Nacional de Desarrollo, 2002 2004). The reason is simple: in an atomized parliament, pork barrel spending is the fuel that feeds the legislative dynamic given the absence of caucuses founded on a certain party or ideology. The 1968 auxilios parlamentarios allowed individual members of Congress to select their favorite recipients for pork barrel projects. 26. In Colombia, the media have created a perverse form of evaluating congressional conduct: the number of proposed laws presented by members of Congress, regardless of

Giants with Feet of Clay: Colombia

97

their quality, coherence, and signicance. With this kind of media pressure, and given the lack of party or other quality lters, hundreds of laws are proposed, generating enormous legislative gridlock. 27. One noteworthythough not surprising exception is the attempts by recent governments at political reforms aimed at changing the current rules of the game regarding political parties and legislation. These reform proposals have met with open resistance.

References
Amorim Neto, Octavio, and Fabiano Santos. 2001. The Executive Connection: Presidentially Dened Factions and Party Discipline in Brazil. Party Politics 7, no. 2: 213 34. Bejarano, Ana Mara, Laura Zambrano Robledo, Felipe Botero Jaramillo, Laura Wills Otero, and Francisco Jos Quiroz. 2001. Qu hace funcionar al Congreso? Una aproximacin inicial a las fallas y los aciertos de la institucin legislativa. Estudios Ocasionales (CIJUSUniversidad de los Andes, Bogot). Benton, Allyson. 1997. Patronage Games: The Effects of Economic Reform on Internal Party Politics and Party System Stability in Latin America. Paper presented at the 1997 meeting of the American Political Science Association, Washington, DC, August 28 31. Botero, Felipe. 1998. El Senado que nunca fue: La circunscripcin nacional despus de tres elecciones. In Elecciones y democracia en Colombia 19971998, ed. Ana Mara Bejarano and Andrs Dvila, 285 335. Bogot: Universidad de los Andes. Boussiou, Jean-Marie. 1994. Les lections lgislatives japonaises du 18 de juillet 1993. Revue Franaise de Science Politique 44, no. 3: 379 423. Cain, Bruce, John Ferejohn, and Morris Fiorina. 1987. The Personal Vote: Constituency Service and Electoral Independence. Cambridge, MA: Harvard University Press. Carey, John, and Matthew Shugart. 1995. Incentives to Cultivate a Personal Vote: A Rank Ordering of Electoral Formulas. Electoral Studies 14, no. 4: 417 40. Cox, Gary, and Matthew Shugart. 1995. In the Absence of Vote Pooling: Nomination and Vote Allocation Errors in Colombia. Electoral Studies 14, no. 4: 441 60. Crisp, Brian. 1999. El comportamiento de los congresistas en Amrica Latina. Paper presented at the Universidad de los Andes, July 13. Gallagher, Michael, and Michael Marsh. 1988. Candidate Selection in Comparative Perspective: The Secret Garden of Politics. London: Sage. Gonzlez, Luis Eduardo. 1991. Political Structures and Democracy in Uruguay. Notre Dame, IN: Notre Dame University Press. Lancaster, Thomas. 1986. Electoral Structures and Pork Barrel Politics. International Political Science Review 7, no. 1: 67 81. Lijphart, Arend, et al. 1999. Sobre la reforma poltica en Colombia. Informe de la Consultora Internacional. In Reforma Poltica: Un propsito de nacin. Memorias. Serie Documentos, no. 17. Bogot: Ministerio del Interior. Losada, Rodrigo, and Eduardo Vlez. 1981. Identicacin y participacin poltica en Colombia. Bogot: Fedesarrollo.

98

Pizarro Leongmez

Mainwaring, Scott. 1991. Politicians, Parties, and Electoral Systems: Brazil in Comparative Perspective. Comparative Politics 24: 21 43. Mainwaring, Scott, and Timothy Scully. 1995. Building Democratic Institutions: Party Systems in Latin America. Stanford, CA: Stanford University Press. Mair, Peter, and Ingrid van Biezen. 2001. Party Membership in Twenty European Democracies, 1980 2000. Party Politics 7, no. 1: 5 21. Morgenstern, Scott. 2001. Organized Factions and Disorganized Parties: Electoral Incentives in Uruguay. Party Politics 7, no. 2: 235 56. Panizza, Ugo. 2001. Electoral Rules, Political Systems, and Institutional Quality. Economics and Politics 13, no. 3: 311 42. Pearanda, Ricardo. 2002. Los Nuevos Ciudadanos: Las Organizaciones Indgenas en el Sistema Poltico. In Degradacin o cambio: Evolucin del sistema poltico colombiano, ed. Francisco Gutirrez Sann, 131 81. Bogot: Grupo Editorial Norma. Penfold, Michael. 2001. El colapso del sistema de partidos en Venezuela: Explicacin de una muerte anunciada. In Venezuela en transicin: elecciones y democracia 1998 2000, ed. Jos Vicente Carrasquero et al., 36 51. Caracas: RedPolCDB Publicaciones. Pennings, Paul, and Reuven Hazan. 2001. Democratizing Candidate Selection: Causes and Consequences. Party Politics 7, no. 3: 26775. Pinzn, Patricia. 1998. Una aproximacin al voto urbano: El voto en las ciudades colombianas. In Ana Elecciones y democracia en Colombia 19971998, Mara Bejarano and Andrs Dvila, eds., 40132. Bogot: Universidad de los Andes. Pizarro Leongmez, Eduardo. 1996. La crisis de los partidos y los partidos en la crisis. In Tras las huellas de la crisis poltica, ed. Francisco Leal, 205 34. Bogot: Tercer Mundo Editores/IEPRI. . 2001. Colombia: Renovacin o colapso del sistema de partidos? In Colombia ante los restos del siglo XXI: Desarrollo, democracia y paz, ed. Juan Ibeas and Manuel Alcntara. Salamanca: Ediciones de la Universidad de Salamanca. . 2002. La atomizacin partidista en Colombia: El fenmeno de las microempresas electorales. In Degradacin o cambio: Evolucin del sistema poltico colombiano, ed. Francisco Gutirrez, 357 401. Bogot: Editorial Norma-IEPRI. Pizarro Leongmez, Eduardo, and Ana Mara Bejarano. Forthcoming. Political Reform in Colombia after 1991: Is There Anything Left to Reform? In Democracy, Peace, and Human Rights in Colombia, ed. Christopher Welna and Gustavo Galln, Notre Dame, IN: Notre Dame University Press. Posada-Carb, Eduardo. 1997. Limits of Power: Elections under the Conservative Hegemony in Colombia, 1886 1930. Hispanic American Historical Review 77, no. 2: 245 79. Randall, Vicky, and Lars Svasand. 2002. Party Institutionalization in New Democracies. Party Politics 8, no. 1: 5 29. Roberts, Kenneth. 2001. La descomposicin del sistema de partidos en Venezuela vista desde un anlisis comparativo. Revista Venezolana de Economa y Ciencias Sociales 2: 183 200. Rodrguez, Juan Carlos. 2003. The Regional Diversity of Electoral Competition in Colombia. Paper presented at the 24th LASA Congress, Dallas, March 2729. Seddon, Jessica, Alejandro Gaviria, Ugo Panizza, and Ernesto Stein. 2002. Political Particularism around the World. Working Paper, no. 463, Research Department, InterAmerican Development Bank, Washington, DC. Solari, Aldo. 1964. Uruguay: Partidos polticos y sistema electoral. Montevideo: Fundacin de la Cultura. Shugart, Matthew. 1999. Efciency and Reform: A New Index of Government Responsiveness and the Conjunction of Electoral and Economic Reform. University of California, Irvine (mimeo).

Giants with Feet of Clay: Colombia

99

Shugart, Matthew, and John Carey. 1992. Presidents and Assemblies: Constitutional Design and Electoral Dynamics. Cambridge: Cambridge University Press. Tsebelis, George. 1995. Decision Making in Political Systems: Veto Players in Presidentialism, Parliamentarism, Multicameralism, and Multipartism. British Journal of Political Science 25, no. 3: 289 325. Yanai, Nathan. 1999. Why Do Political Parties Survive? Party Politics 5, no. 1: 5 17.

4 Ecuador: The Provincialization of Representation

Simn Pachano

mong the many causes alluded to when explaining the problems of the Andean countries, and especially those of Ecuador, the crisis of representation has grown in importance in recent years. Scholarly analyses as well as politicians refer to a crisis of representation as an unquestionable fact that obstructs policy making and implementation (Barrera 2001; F. Bustamante 2000). Allegedly, deciencies in representativeness result in problems of governability and conditions unfavorable to the consolidation of democracy. This perspective suggests that those deciencies derive from the voters dissatisfaction with the meager results of politicians actions in their role as authorities of popular representation, and that, at the same time, this dissatisfaction leads to mistrust not only of the people involved but of the institutions and the system as a whole. As a result, following a period of exploration as voters experiment with different options, they nally reject representative democracy and focus on alternative options, ranging from seemingly innovative proposals to the election of anti-system leaders. The validity of this analysis hinges on the relationship between the expectations and the results of political representation. How representation is evaluated depends on the returns derived from the representatives actions, which also supposes that voters expect those actions to bring about specic results. This analysis is therefore based on voters expectations on the one hand, and the results of authorities actions on the other. The crux of the analysis of representation lies in the relationship between voters and their representatives, not in an isolated analysis of each. What needs to be addressed is the bond between voters and representatives, or, in other words, the mandate emerging from voter expectations. An investigation of this relationship has been one of the weak points in Ecuadorian studies of representation. Most such studies have focused on either voters or representatives but not on both at the same time, and even less so on the relationship between the two.

Ecuador: Provincialization of Representation

101

The role of the political parties as fundamental actors in this relationship has garnered much attention in recent years and provided important clues with which to ascertain the nature of representation in Ecuador. Most studies have emphasized the conditions under which representation is carried out (Len 1994; F. Bustamante 2000), especially representatives role in the clientelistic and corporatist practices that characterize politics in Ecuador. Recent studies have concentrated on parties ideological orientation (Freidenberg 2000) and political culture (Burbano 1998), as well as internal organization (Freidenberg and Alcntara 2001). Party dispersion, fragmentation, atomization, uncertainty, and volatility are highlighted in these studies. Most allude to negative or problematic aspects of parties that impede their capacity to carry out their responsibilities (Conaghan 1994; Arias 1995; Meja 1998; F. Bustamante 2000). The limited capacity for representationregardless of how one understands itparticularly stands out, generally with respect to the predominance of clientelistic and corporatist practices, as well as personalism. These analyses focus on the parties problems or inability to carry out their responsibilities; few point to the parties ability to survive in a hostile environment. This is a good starting point, but it is necessary to go further and explain this capacity to survive. There is no doubt that a crisis of representation exists, but it is important to know what that means. Despite negative public opinion and even contradicting actions taken to undermine their weight and inuence, the parties have secured a role as vehicles for political representation.1 In the post-1979 period, four partiesPartido Social Cristiano (Social Christian Party, or PSC), Partido Roldosista Ecuatoriano (Ecuadorian Roldosist Party, or PRE), Izquierda Democrtica (Democratic Left, or ID), and Democracia Popular (Popular Democracy, or DP)have consolidated and together have won about three-fourths of the vote. This has occurred within the framework of a highly fragmented and atomized system. Just as important, however, is the increasing share of the vote that these parties have managed to accumulate over time. One of the prominent characteristics of the Ecuadorian party system is this apparently contradictory combination of fragmentation and concentration. The large number of parties that win seats in Congress and gain access to representational positions in provincial and local assemblies is offset by the predominance of a relatively small number of parties. Generally speaking, the parties have demonstrated a greater ability than independents to secure voters support. This chapter uses as a starting point this ability of the Ecuadorian parties to survive in a hostile environment in order to propose an alternative understanding of the problems concerning representation. I argue that there are other forms and mechanisms of representation and that all the political actors, especially the parties, can adapt to them. The survival of the parties is due to their ability to adapt to conditions that are not necessarily part of the institutional design of the

102

Pachano

political system. This adaptation clears the way for complex situations affecting party consolidation since it requires renunciation of many of the classic functions allegedly performed by parties in a democratic regime. Because of the type of relationship parties have with their electorates, they are forced to produce results that do not satisfy the expectations of the population as a whole. The parties must be rooted in territorially and socially dened groups of voters in order to survive. This situation transforms parties into expressions of partiality and not of a public good, and leads them to develop a great ability to represent specic interests and local arenas, but also leaves them with an enormous deciency in representing national interests. The main argument of this chapter is that the provincialization of partiesthat is, their predominant focus on provincial issueshas facilitated their survival but has also caused their main deciencies. I do not deny that there is a crisis of representation, but I try to identify the nature of this crisis in Ecuador, where the term seems too broad and may cause confusion. As it has been applied to Ecuador, the expression a crisis of representation confuses several different levels, and it treats different kinds of problems indiscriminately. Most analyses of a crisis of representation focus on three central themes: the political systems outputs, the structure or formation of representative institutions, and the concept of political representation. When analysts refer to the outputs yielded by the political system, they emphasize the poor social and economic performance throughout the post-1979 period.2 From this perspective, problems of representation derive from the inability to satisfy the demands of society. However, it is not clear to what extent this failure to produce better results is a consequence of the system of representation rather than of non-political factors. The governmentsand, in general, democratic institutionsproblems of efciency cannot be attributed entirely to the forms, mechanisms, and procedures of political representation. Certainly, representation has an effect on government because it contributes to the formation of governments and establishes limits for governments and assemblies. In this sense, the forms of representation are one of the means of attaining the goals of formation of both decision-making instances and operative institutions, but the degree to which they are successful or the importance of their role in these instances is precisely what ought to constitute the focus of our analysis. This will be the object of the rst part of this chapter, which analyzes the main characteristics of the electoral system and especially the structure and formation of the National Congress and its relationship to the executive.3 The representativeness of the popularly elected bodies and ofcialsthe second theme in this chapterfocuses on one of the political systems main problems, not only in Ecuador, but also in conceptual terms. Institutional architecture, design, and procedures are fundamental factors in the study of representation. A lesser or greater capacity for inclusion of the different social actors, interests, and conicts depends signicantly on the design of representative institutions. The degree of satisfaction with representation itselfnot

Ecuador: Provincialization of Representation

103

necessarily with the results of the political system, which is another matter depends to a great extent on institutional design. With a few exceptions (Meja 2002; Freidenberg and Alcntara 2001), this subject has received little attention in the Ecuadorian case. Few analyses have been concerned with institutional issues; most have been oriented more toward sociological or anthropological explanations of representation. My analysis emphasizes the cleavages in national politics, focusing on the institutional structures ability to reect and process these cleavages. I argue that there is a divide between these two (national cleavages and institutional structure) that clearly causes problems in representation. On the other hand, the generalization of certain political practices has created alternative forms of representation and of satisfying the demands of particularistic actors (through clientelism and corporatism), which have allowed parties to survive as mechanisms of representation. By substituting the formally established channels and mechanisms of representation, these forms of particularism have eroded them; however, they have also, simultaneously, been able to respondhowever partiallyto demands and also to constitute an alternative arena for representation. The problems arising from this situation are related to the temporal dimension of this coexistence. The main question for political science and for parties and politicians is: For how long and in what conditions can this balance between institutions and everyday practices be maintained if these practices erode the institutions? Particularistic practices such as clientelism and corporatism ensure immediate results, but they corrode the institutions by draining them of content. Therefore, a basic question in this chapter is the relationship between formal institutions and political practice. This is the primary focus of the second part of the chapter. Finally, it is important to consider the conception of political representation underlying the claim that there is a crisis of representation. In most Ecuadorian studies on this subject, representation is seen as an expression of a binding mandate or at least as a direct channeling of interests. Most of these studies highlight the limited ability of institutions to process conict, clearly one of the basic functions of representative mechanisms. Other observations focus on the limited capacity of the political system to adequately represent diverse social interests. Apparently, each social sector is expected to get a quota in representative bodies in order to ensure not only the processing of their respective demands but also participation in decision making. This contradicts three basic principles of a representative regime: majority rule, autonomy of the representatives, and, derived from this last one, the non-binding mandate. To a great extent in Ecuador, the argument that there is a crisis of representation is based on this erroneous perception of representation, leading to a demand for results that cannot be produced. I discuss this perception in an attempt to demonstrate that in order to tackle the problems of representativeness, we need an adequate concept of representation.

104

Pachano

Endless Reform and a Contradictory Institutional Framework


The Ecuadorian normative and institutional framework has been continually altered by both Congress and the executive and by a Constitutional Assembly that issued a new constitution in 1998. In 1983, before the end of the rst posttransition presidential and congressional terms, the rst constitutional reforms were introduced. This was the beginning of a litany of institutional reforms that apparently will continue to be an integral part of Ecuadorian political practice. Political reformlegal, constitutional, proceduralhas been used as a tool for solving political conict. Even small problems, those constituting the customary practices of political actors that must be processed politically, lead to questions concerning the institutional and normative framework, leaving this framework constantly up for grabs. As a result, the possibilities of consolidating valid reference points for the actors involved are very limited. Basic aspects of the electoral system have been constantly altered: the representational formula, the electoral calendar, district size, and the way in which votes are cast (see Table 4.1). This has been one of the greatest obstacles to the institutionalization of the party system.4 Constant change has made it impossible for two consecutive elections to take place under the same set of rules, and neither the voters nor the political parties have enjoyed certainty concerning the rules of the game. These problems are due not only to the frequency and number of reforms, but also to the contradictions from one reform to the next. Competing particularistic interests and pressure from diverse social groups that function with short-term logic has produced a complex institutional system rife with contradictions (Conaghan 1995; Meja 2002). Many of the components of the electoral system contradict one another. For example, some aspects of the system were intended to strengthen parties. In contrast, the personal vote system introduced in 1998 worked against parties. Even if the 1978 Constitution and the party and electoral laws of 1979 had some birth defects, they have only become worse over time, mainly due to successive changes brought about by particularistic interests and the need to respond to specic situations. Although the three main objectives in the return to democracy were to strengthen parties, attenuate the personalistic character of Ecuadorian politics, and prevent party system fragmentation, the new institutional rules had the opposite effect. Parties have had serious problems with consolidation, and in the 2002 elections the solid electoral performance displayed by the four largest parties was reversed.5 The parties without exception have been electorally successful in limited geographic spaces. In national politics, personalism continues to be a salient characteristic. Finally, the dispersion of representation has increased noticeably in Congress; more parties win seats with a small number of votes.

Ecuador: Provincialization of Representation


Table 4.1 Ecuador: Main Political-Electoral Reforms, 1983 2003
Year Main reforms

105

1983

1985

1986 1994

1996 1997

1998

2000 2003

Reduction of the presidential and legislative terms (from 5 to 4 years) Introduction of intermediate election (every 2 years) for provincial deputies General election for deputies coincides with the rst runoff presidential election (instead of the second) Name of legislature is changed from Cmara Nacional de Representantes to Congreso Nacional Mechanism for budget approval is simplied Executive is given special powers to propose laws in situations of economic emergency Majority system replaces proportional representation system Elimination of the electoral threshold as a requisite for the permanent registration of parties Return to proportional representation (with Hare formula) Immediate reelection is approved for all elected ofces, with the exception of the president Deputies are prohibited from managing or lobbying for budget appropriations The prohibition on alliances is eliminated Introduction of the system of personalized voting with open lists Seats are allocated according to individual votes by simple majority, regardless of list total votes; proportional formula is eliminated National deputies increase from 12 to 20 National deputies are eliminated Number of provincial deputies is increased (with a base of two per province instead of one as was formerly the case) Presidential and legislative elections are separated from local and provincial elections (electoral calendar is diversied) Return to proportional representation (DHondt formula) Congress loses the ability to promote a vote of no-condence against cabinet ministers Mechanisms for the appointment of legislative authorities introduced (president and two vice-presidents are appointed according to size of party benches but with a vote of the entire legislative body) Change in executive-legislative relations (powers of Congress are restricted in aspects such as budget approval, appointment of accountability authorities, among others) New conditions for the runoff presidential election: absolute majority or 40 percent threshold plus 10 percentage points above next candidate Return to the allocation of seats by lists (DHondt formula), keeping the personalized voting system Elimination of DHondt formula ( January) Introduction of Imperiali formula (September)

106

Pachano

This complex, contradictory, and constantly changing institutional design has operated in an environment that is socially, economically, and culturally hostile to the consolidation of parties and to representative institutions in general. The political practices and behaviors of the social and political actors have obstructed the achievement of the objectives proposed at the beginning of the transition. Because of its importance, this subject has been frequently discussed from many perspectives (Menndez 1986; F. Bustamante 1997; Burbano 1998; De la Torre 1996; CORDES, n.d.). Most authors have concentrated on practices and behavior, without paying sufcient attention to the institutional aspects. Most analysts have posited a cause-effect relationship whereby institutions are determined by the social structure and political culture. Such analyses express the sociological and cultural bias of Ecuadorian political studies.

Diffuse Multipartism: Rules and Their Implications


One of the outstanding characteristics of Ecuadorian democracy has been the dispersion and fragmentation of the party system. Since the return to a democratic regime in 1979, at least ten parties have secured congressional representation. All of them even the largest ones that have maintained the most stable share of voteshave experienced erratic electoral fortunes (see Table 4.2, which includes only the four largest parties). Many parties have failed to survive more than two elections and have been replaced by new parties that are generally as small as those that disappeared.6 Several components of the electoral system have fostered party proliferation: the use of the province as an electoral district, proportional representation, the prohibition of local or subnational parties, the no immediate reelection rule (in effect from 1979 until 1994), and the implementation of two-round presidential elections.
Table 4.2 Share of Congressional Vote Won by Four Major Parties, 1979 2002
(% of valid votes, provincial deputies) Parties 1979 1984 1988 1992 1996 1998 2002

PSC PRE ID DP Others Total


SOURCE :
a

6.4
a

14.8
b

78.8 100.0

11.5 5.1 20.0 7.3 56.1 100.0

12.4 16.3 22.6 10.9 37.8 100.0

23.2 16.0 9.5 7.2 44.1 100.0

27.9 21.3 7.1 11.9 31.8 100.0

20.3 17.5 11.9 24.1 26.2 100.0

26.4 11.9 11.9 3.1 46.7 100.0

Supreme Electoral Court. Formed in 1982. b Not ofcially recognized; participated under the auspices of the CFP.

Ecuador: Provincialization of Representation

107

The Province as an Electoral District


The use of provinces, the countrys administrative-political divisions, as electoral districts causes ve problems in representation. First, the effect of their size range is translated, at the electoral level, in the coexistence of districts of different magnitudes, with results characteristic of this situation (Taagepera and Shugart 1989; Snyder 2001). Parties can win seats with very few votes, especially if they are concentrated in small provinces. This has been the strategy followed by parties that obtain a very limited percentage of the vote on a national level but that win seats in the National Congress with votes obtained in provinces with small populations.7 The most common size for electoral districts is two seats, with seven provinces electing that number. In 2002, four provinces elected three deputies each, six provinces elected four, two provinces elected ve, one province elected fourteen, and one district elected eighteen deputies (see Table 4.3). If small districts are dened as those that elect less than 4.0 percent of the members of Congress (the median is 3.5 percent), then half of the districts t into this category.
Table 4.3 Share of National Electorate and Number of Deputies per Province, 2002
Province Share of national electorate Number of deputies Percent of total deputies

Galpagos Zamora Orellana Pastaza Morona Sucumbos Napo Bolvar Carchi Caar Esmeraldas Imbabura Cotopaxi Loja Chimborazo Tungurahua El Oro Azuay Los Ros Manab Pichincha Guayas Total
SOURCE :

0.1 0.3 0.4 0.4 0.6 0.6 0.9 1.5 1.5 1.6 2.6 3.0 3.1 3.3 3.8 4.5 4.5 4.8 5.0 10.1 20.5 27.0 100.0

2 2 2 2 2 2 2 3 3 3 4 3 4 4 4 4 4 5 5 8 14 18 100

2.0 2.0 2.0 2.0 2.0 2.0 2.0 3.0 3.0 3.0 4.0 3.0 4.0 4.0 4.0 4.0 4.0 5.0 5.0 8.0 14.0 18.0 100.0

Supreme Electoral Court.

108

Pachano

Parties with little national presence can concentrate their efforts in one or several of these provinces and win seats in Congress. If we add to district size the effects of the use of proportional representation or the system of personalized, open-list voting (introduced in 1997), it is clear that the doors have been wide open to the dispersion of voting and the fragmentation of the party system. Secondly, as it stands, the system creates imbalances and distortions among provinces in terms of the relationship between representatives and represented. The proportion of votes needed to elect a deputy varies signicantly from district to district; voters from different districts do not have the same weight. As pointed out by Taagepera and Shugart (1989, 14) and Snyder (2001, 146ff.), this violates the one person, one vote rule since the weight of each individual vote is not the same in all districts.8 The representativeness of the deputies is affected by malapportionment. Each deputy represents a very unequal proportion of the population, and the deputies possibilities of establishing a relationship with voters varies substantially, depending on district size. In the smaller districts, the possibility of establishing direct, practically face-to-face relationships is greater, which may create a fertile ground for binding mandates (mandatos vinculantes), which in turn may form the basis for clientelistic and corporatist forms of representation. The number of members each province has in Congress depends on one of two rules: a minimum of two seats per province, or one seat for every 300,000 inhabitants. These rules create some malapportionment. The smallest provinces benet and the largest are adversely affected. The rule that a province gets one seat for every 300,000 citizens clashes with the idea of not increasing the number of deputies and restricting parliament to a reasonable size, and it meets the resistance of the small and mid-sized provinces that view any increase in the number of seats for the large provinces as a threat to their interests. Malapportionment also has regional effects.9 As the countrys most populated region, with 50.5 percent of the national population, the Coastal region, comprised of only ve provinces, elects only 39 percent of the members of Congress. At the other end of the spectrum, the Amazon and Galpagos regionswith a total of seven provinces that benet from the minimum of two seats per province, and with only 3 percent of the countrys population elect 14 percent of the seats. Comprised of ten provinces, the Sierra region is the only one to achieve representation that is proportional between its population (46.6 percent) and its share of seats (47 percent) in the Congress. This has been one of the few subjects related to the electoral system that has been on the political agenda and debated in terms of its repercussions on the representativeness of the various provinces and regions. Third, with the use of the province as an electoral district the myth of territorial representation is created, whereby the deputy is the representative of provincial interests rather than of a political movement. This is an alien and even contradictory concept given the unitary character of the Ecuadorian state, yet it

Ecuador: Provincialization of Representation

109

is widely generalized in national politics and shapes the behavior of Ecuadors political parties. Parties have to favor representation of provincial interests, even if it means sacricing their own positions and a vision for the country as a whole. The vindication of regional, provincial, and local interests has become almost an obligation since it is one of the ways to win electoral support. This logic in turn has fueled the conguration of captive voters and electoral bastions, as part of the logic of a narrower and more particularistic political arena. Fourth, the use of the province as an electoral district has contributed to party indiscipline (Meja 2002). Most deputies who have abandoned their parties allude to the parties nonexistent or meager concern with their province of origin, which constitutes a tacit vindication of a binding mandate. Most deputies who switch parties have been rewarded by resources or payoffs for the province, either through negotiations with the government or by a relatively powerful party boss. These agreements between deputies and governmentthe presidential connection referred to by Amorin Neto and Santos (2001, 221)are obvious from the time of the elections and not only in the deputies performance. The deputy thus fullls his/her commitment to his/her province. Fifth, spurred by their quest to obtain the greatest possible number of seats, the parties (especially the largest ones, with electoral bastions in the most populated provinces) must seek votes in the small provinces, which leads them to seek out local candidates who can win votes. Generally, to do this they must sacrice their own principles and become catch-all parties, adapting their discourse and proposals to particularistic local realities. Deputies who win election have enormous negotiating power and enjoy considerable autonomy with respect to the parties. Making provinces the electoral districts has fostered the provincialization of politics. The electoral rules do not favor the national distribution of party voting (or nationalization, as referred to by Mainwaring and Jones 2003), but instead tend to force parties into subnational arenas. This also contributes to the overburdening of the national level by channeling demands to the upper levels (government and Congress), a trend that is also spurred by the countrys administrative centralization, which leaves little space for decision making at the lower levels of municipalities and provincial councils. Lastly, the electoral system is an incentive for the corporatist and clientelistic practices that characterize Ecuadorian politics. The elimination of the national deputies in 199810 heightened the negative effects of the electoral system given that they were a push factor for the conguration of a national political arena.

Proportional Representation
The proportional representation (PR) electoral system fostered the fragmentation of the party system by allowing minor parties to win seats in Congress. The allocation of seats by means of a double quotient mechanism (using the Hare

110

Pachano

formula), useful for maintaining proportionality between votes and seats, became an incentive for the proliferation of small parties that could gain representation with few votes. This was particularly evident in provinces with the greatest number of voters (Guayas and Pichincha, both electoral bastions of the large parties) and in intermediate ones (Manab, Los Ros, Azuay, and El Oro), where votes are more dispersed. Parties gained representation with an insignicant number of votes as a result of PR with the Hare formula. Small parties have used two strategies to gain representation in Congress: rst, as seen in the preceding section, they can concentrate their efforts on provinces with the fewest voters; or, they can compete in the large and intermediate provinces where the proportional formula favors them. Either way, parties can win seats with a minimal proportion of the national vote. This system results in the consistent presence of legislative parties with only a few seats. The Ecuadorian Congress has consistently had a signicant number of small parties, operationalized here as those with less than 5 percent of the members of the national assembly. (This 5 percent maximum was equal to three deputies in the legislature from 1979 to 1984, four from 1984 to 1996, and six from 1998 to 2000.)11 The dispersion in the National Congress makes it difcult to assemble majorities in support of or in opposition to the government. These small parties are important because no party has ever obtained the majority of deputies in the National Congress. Small parties have consistently been necessary to pass laws and form opposition blocks. The small provinces (especially the Amazonia provinces) have tended to bring together parliamentary coalitions outside party lines, especially in situations where their votes can be negotiated (Rowland 1998; Meja 1998). They have acquired an importance disproportionate to the number of their legislators, giving them considerable negotiating power in important congressional votes and in electing congressional leaders. Also contributing to the power of the small provinces is the relatively small size of the Ecuadorian Congress; a few votes can make the difference in crucial decisions.12 The effects of the proportional system are heightened by the lack of an electoral threshold that prevents parties from obtaining seats in Congress with less than a certain percentage of the vote. The threshold established by law (which has uctuated between 4 percent and 5 percent for elections of deputies, provincial councilors, and municipal councilors and has not been in effect during the entire period under discussion) is for registration purposes only. Parties that fail to meet the minimum share of votes in two consecutive elections lose the Supreme Electoral Courts ofcial recognition. However, parties that fail to meet the 4 5 percent threshold may still win seats and function as parties during their term in ofce. Furthermore, because registration is forfeited only after the given percentage has not been achieved in two consecutive electionsthey cannot run in the third electionthose who win ofce with below-threshold percentages

Ecuador: Provincialization of Representation

111

may still keep their seats for up to two consecutive terms (which might mean as many as eight years). In addition, proportional representation has been an incentive for personalism. Many analysts have argued that PR with closed and blocked lists should strengthen parties (Nohlen 1993). However, in conjunction with the use of provinces as electoral districts, the parties obligation to participate nationally, and the prohibition of alliances in the proportional elections, as well as the establishment of PR within the context of reduced institutionalization and predominating caudillismo, it has produced the opposite effect. The parties have had to incorporate candidates who can bring in votes. In 1997, in response to a referendum, the Ecuadorian electoral system, including PR, underwent the greatest reform in its history.13 A majoritarian system based on personalized voting with open lists was introduced. However, the electoral system introduced in 1997 was soon replaced, and for all practical purposes has reverted to the proportional system.

National and Subnational Parties


One of the main objectives of the 1979 Constitution was to strengthen political parties. The history of instability during the preceding half century was associated with the absence of parties of national scope capable of aggregating interests and forming governments founded on popular legitimacy. For the rst time in the countrys history, and together with the new constitution, electoral and party laws were approved, both with considerable regulatory content. The new provisions were intended to promote the formation of strong parties, whose stability would be assured in time by ample organizational support and their presence throughout the entire nation. The goal was the elimination or at least reduction of the formation of electoral machines that might be capable of winning votes but that would have no real long-term life of their own, no roots in society, and would be limited to certain regions or provinces. The electoral and party laws had meticulous provisions that forced parties to carry out a series of activities in order to obtain and maintain their registration with the Supreme Electoral Court. Parties were required to maintain organizational structures in at least ten of the twenty provinces that existed at the time. Once they obtained legal recognition, they were required to present candidates in at least ten provinces. Failure to comply with these two provisions resulted rst in the cancellation of registration, and after a second election, in the loss of legal recognition. These provisions have acted as more effective barriers to the fragmentation of the party system than the electoral threshold.14 These regulations did not achieve their main objective of promoting the formation of national political parties. Parties have concentrated their votes in certain provinces or at most in one region. Except for a brief period in which ID

112

Pachano

and PSC maintained a national presencein terms of their votesthe predominance of provincial or regional parties has been the main characteristic of the Ecuadorian system. Electoral bastions, in which each party concentrates its efforts and to which other parties nd it difcult to gain access, have grown steadily in strength. Even the dominant party nds it difcult to move beyond these boundaries and compete in other provinces. In addition to structural determinationsEcuador is characterized by very distinctive regional societiessome legal regulations, including the very ones designed to promote the formation of parties of national scope, have fostered the provincial focus of parties. The legal regulations force the parties to act on a national level and compete for seats in the National Congress. Otherwise, the aforementioned provisions would apply and parties would lose their registration and be unable to run candidates. In this way, minor local or provincial politicians and parties have been shifted to the national level, and the particularistic concerns of these politicians have found their way into the National Congress. This has a double effect. On the one hand, it lls the national scene with small parties, generally with local orientations that represent the interests of very narrow sectors of society. Consequently, the overload of subnational concerns and demands that might under different circumstances be resolved at the local level becomes more pronounced on a national level. On the other hand, the larger partieswhich in Ecuador tend to be more ideological and to have a more national orientationare forced to compete in elections with locally or provincially rooted parties. They sacrice principles in order to win votes in these localities. In this manner, they contribute to the overload of subnational topics in national politics, thus reinforcing the regional cleavages that characterize Ecuadorian politics. Both large and small parties, whether rooted in a certain sector or a certain ideology, must adapt to the provincial or local orientation of politics. The inexibility of the provisions aimed at helping parties achieve a more national scope has had a harmful effect on this same objective. Some degree of exibilityallowing, for example, local or provincial parties to compete in municipal and provincial council electionswould have brought about positive results and helped to strengthen national parties. Better results would have been achieved if effective barriers to participation on a national level had accompanied this exibility at the municipal and provincial level. These problems have worsened since the 1998 Constitution eliminated the national deputies. Until 1998, a minority of deputies was elected in a single nationwide district that attenuated the provincialism of political life, whereas the majority was elected using provinces as the electoral district. The national deputies were seenboth by the voters and by themselvesas guardians of a national mandate that the provincial deputies lacked.15 Their elimination strengthened the perception of Congress as a forum of territorial representation that focuses on local problems.

Ecuador: Provincialization of Representation

113

The provincialization of the parties is also fueled by the parties selection of candidates. In the process of candidate selection, parties constantly negotiate with local leaders, who can usually impose their own conditions because they have captive voters. The local leaders are usually linked to local interest groups, so parties are forced to be responsive to those interests. For this reason, deputies tend to feel a greater connection and commitment to the local interests than to the parties, which further promotes the idea that the deputy is a territorial representative with a binding mandate. Despite the problems derived from these negotiations over candidate selection, the debate over the degree of democratization in selecting candidates is important. A more open candidate selection could open up the space for the participation of sectors that might not otherwise participate in the process;, however, it could also be deemed as a way of including local oligarchic groups that in turn fuel the corporatist tendencies of Ecuadorian politics. In any case, candidates selected in this manner are the least likely to become disciplined party members on their legislative benches (Meja 1998).

Alliances and Their Limitations


The prohibition of interparty electoral alliances that existed until the reforms of 1996 created an obstacle to building coalitions in Congress (see Table 4.1). The electoral law established that for municipal council members, provincial councilors, and both national and provincial deputies, each party needed to present its own list. This provision sought to strengthen parties, assuming that participation at all these levels of elections would require stable organizations and solid structures. However, the regulations brought about unintended consequences. The prohibition on electoral coalitions fostered party system fragmentation since each party had to compete on its own. Pressured by the obligation to secure a minimum number of votes and present candidates in at least a minimum number of provinces, parties were forced to participate at each and every level. The inevitable result was the fragmentation of the system due to the enormous number of parties, most of them small, which under different circumstances might have formed alliances and thereby contributed to the formation of large ideological and electoral trends. Instead, parties competed with one another in a battle for access to government resources, and there were more incentives for interparty confrontation than for reaching agreements. The confrontational tendencies of Ecuadorian politics are largely due to these provisions rather than to the political culture. Many local or provincial parties found that the regulations supported their strategy wherein seats are obtained via the proportional formula and the awarding of seats by remainder. Forced to participate on their own on all electoral levels and in the greatest possible number of provinces, parties used this opportunity to their own advantage. Many local caudillos employed this provision to negotiate successfully with small parties that, forced to compete on their own,

114

Pachano

needed a certain number of votes to guarantee their presence in Congress or, at the very least, to comply with the required minimum number of votes. Although the ban on coalitions was revoked in 1996, some barriers to coalitions remain. Electoral alliances are now allowed, but the label of only one of the coalition partners is used to identify the coalition. The other party or parties are forced to give up their identity. Because of this, parties have incentives to form alliances only when their chances of obtaining seats on their own are limited or nonexistent. The formation of coalitions depends mainly on whether a party believes it would fare better by running on its own or as part of a coalition, and not on the political and programmatic orientation of the alliance or the ideological principles guiding it. Coalitions are created for instrumental electoral purposes and not for the formation of large fronts identied by their principles, objectives, or platforms. Since 1997, national coalitions have been uncommon. Generally, coalitions have been formed in provinces and municipalities, for elections held for provincial and municipal councils and for provincial deputies.16 The elimination of national deputies and the exibility afforded by establishing coalitions in specic provinces without compromising the respective parties in the rest of the country have motivated this pragmatic behavior. They have also heightened dispersion, since a cost-benet calculation by a party can lead to innite combinations, most of them inexplicable in terms of the coalition partners programmatic positions. The prohibition of alliances from 1979 to 1996 and later their liberalization and increased exibility have transformed parties into umbrellas that shelter a wide range of factions that enjoy great autonomy in selecting candidates. Although parties are formally national organizations, in electoral practice they operate like provincial organizations with relative autonomy in selecting candidates.17 A game is set up, revolving around parties with more or less ability to represent local interests, which is what really matters to the groups with which the parties have to negotiate. An additional ingredient surfaces in the provincialization of the parties and their increased exibility or loss of ideological-programmatic positions (in other words, in their transformation to catch-all parties).

Immediate Reelection
From 1979 until 1994, the immediate reelection of all authorities chosen in popular elections, including deputies, was prohibited. Reeleccin cruzada, or crossover reelection, was established, whereby a deputy could move from one type of post to another, either from national to provincial deputy or vice versa. However, since there were only twelve national deputies, the possibility of returning to Congress via this path was slim. At most, only twenty-four deputies (34.8 percent of the total members of Congress at that time) would be able to win reelection, and only if all the national deputies were reelected as provincial

Ecuador: Provincialization of Representation

115

deputies and, at the same time, their seats were taken by provincial deputies who were elected as national deputies. This outcome was practically impossible, and it never occurred. In 1983, when the Constitution underwent initial reforms, terms for all provincial deputies were set at two years, while the term of a national deputy remained four years.18 Consequently, more than four-fths of Congress had to be replaced every two years, with no possibility of immediate reelection and minimal hope of crossover reelection. The ban on immediate reelection brought instability to parliamentary activity. This instability in Congress was reinforced by the annual election of congressional leaders and annual renewal of legislative committees. It became a substantial burden for parties to nd candidates for all of these positions, given the ban on immediate reelection. The negative effects of constant congressional turnover became apparent not only in the instability in Congresswhich assumed a short-term logic that affected legislative outcomes as well as its relationship with the executive branch but also because political parties were forced to improvise to keep up with the situation. None of the parties, not even those with the most solid structures, could respond to this challenge. Their reserves of leaders and militants capable of carrying out legislative functions were exhausted. Parties had to call on individuals outside the party, generally local caudillos with popular electoral appeal but with no guarantee of loyalty or discipline to the party. This is one of the explanations for the emergence of oating politicians (Conaghan 1994) with limited loyalty to their parties. Once again, the legal provisions resulted in outcomes radically contrary to those desired. Instead of supporting the renewal of political leaders, encouraging greater participation in popular elections, and helping to reduce personalism, the prohibition of immediate reelection fostered improvisation, bred instability, and accelerated the deterioration of the parties. It was an additional incentive for the presence of local caudillos in national politics and for the growing tendency toward the representation of local and corporatist interests.

Runoff Elections
In an attempt to strengthen the presidential mandate, the Constitution of 1979 established runoff elections if no candidate wins more than 50 percent of the valid votes in the rst round. The runoff system was intended to guarantee that a presidents legitimacy would be greater than that of presidents elected in the 1950s and 1960s, who were elected with a low percentage of the vote and only a small margin over their competitors. Allegedly, this lack of a clear popular mandate was one of the reasons for governmental instability.19 The runoff system requires a number of conditions not present in the country at the time it was established. As well, other components of the institutional

116

Pachano

arrangement stood in the way of achieving the necessary conditions. A basic requirement for the runoff system to operate adequately is the existence of strong parties, with stable electoral support and, above all, the ability to inuence the way their constituencies vote so that the second round reects organic political decisions and not just the isolated electoral inclinations of each voter. In the absence of parties that fulll this requirement, the second round of presidential voting represents the joint aggregation of separate wills, which does not generate stable and organic support for the government. These disparate wills have been, for the most part, channeled into negative votes against one of the nal candidates rather than into votes in favor of another (Seligson and Crdova 2002). For several reasons, including their inability to inuence the way their followers vote, Ecuadorian parties have consistently avoided publicly supporting presidential candidates (except of course their own) in the second round (Conaghan 1995). As a result of the failure to forge electoral coalitions for the presidency, governing coalitions have not formed and sustained collaboration between the executive and the legislature. The entire post-1979 period has been characterized by confrontation between these two branches of power. This so-called pugna de podereslegislative/executive conicthas on occasion placed regime stability at risk and has generally hampered governments (Snchez-Parga 1998). This destructive behavior by parties is due to several factors, among them formal institutional design, and in particular the lack of incentives for parties to develop collaborative practices. The cost of participating in governing coalitions, especially when parties hope to see governments rapidly erode, is much higher than the cost of avoiding any electoral commitment in the second round. The use of the two-round voting format in a system characterized by high fragmentation and volatility serves as an incentive for many parties to participate in presidential elections.20 Because of the dispersion of votes, small parties have an opportunityunavailable under other circumstancesto reach the runoff round and even win presidential elections. Parties can go on to the second round with relatively few votes, as has occurred on several occasions (see Table 4.4). Since 1984, congressional elections have taken place at the same time as the rst round of the presidential election, creating an additional incentive for parties to present presidential candidates. With a presidential candidate, a partys deputies enjoy better prospects of getting elected. Without a presidential candidate, parties have no way to offer future governmental benets, so they are deprived of one of the main attractions of congressional elections in an environment where clientelism dominates. Therefore, parties generally present their own presidential candidate even when their chances of winning might be greater as part of an interparty coalition. The benets obtained by parties in legislative elections come at the expense of presidential elections. Party strategy is shaped by this context of great fragmentation. Parties know they can obtain inuence disproportionate to their

Ecuador: Provincialization of Representation


Table 4.4 Share of Presidential Vote in First Round, 1978 2002
(% of valid votes) Candidates 1979 1984 1988 1992 1996 1998

117

2002

First place Second place Third place Fourth place Fifth place Sixth place Seventh place Eighth place Ninth place Tenth place Eleventh place Twelfth place
SOURCE :

27.7 23.9 22.7 12.0 8.0 4.7

28.7 27.2 13.5 7.3 6.8 6.6 4.7 4.3 0.8

24.7 17.7 14.7 12.5 11.5 7.8 5.0 3.3 1.6 1.2

31.9 25.0 21.9 8.5 3.2 2.6 1.9 1.9 1.4 0.9 0.5 0.3

27.2 26.3 20.6 13.5 4.9 3.0 2.4 1.2 0.9

34.9 26.6 16.1 14.7 5.1 2.6

20.6 17.4 15.4 13.9 12.1 11.9 3.7 1.7 1.2 1.1 0.9

Supreme Electoral Court.

size, and the two-round system for electing the president creates this possibility. This strategy consists not only of gaining seats in Congress, but also of laying the foundation for future relationships with the executive, regardless of who wins. As pointed out in the case of Brazil quite similar to Ecuador in some ways this strategy is generally linked to the pursuit and procurement of participation in the distribution of the national budget (patronage) (Amorin Neto and Santos 2001). This subject cannot, therefore, be considered merely a question of the electoral timetable, or in other words, the election of deputies during the rst round of the presidential election. The main problem lies in the incentives generated by the runoff system. This system creates an incentive for most parties to participate in presidential elections and lays the foundation for clientelistic relationships between the president and the members of Congress. Congressional elections have been held concurrently with the second round of presidential voting only once, in 1979, and afforded insufcient experience with which to judge whether this might reduce the dispersion of presidential votes and the number of parties represented in Congress (see Table 4.5).21

Personalized Voting with Open Lists


In 1997, based on the results of a popular referendum, the Ecuadorian electoral system underwent a major reform that eliminated the system of proportional representation and replaced it with personalized voting with open lists. Under this system, the parties lists of candidates become nothing more than a means of presentation since voters cast their ballots for as many individual candidates as there were seats to be lled, regardless of their party afliation. The voter could

Table 4.5 Size of Legislative Delegations, 1979 2002


(percentages of parties with different-sized delegations) 1984 1986 1988 1990 1992 1994 1996 1998 2002

No. of deputies elected

1979

1 2 3 4 5 6 7 to 10 11 to 20 21 and over

30.0% 20.0 10.0 10.0 0.0 0.0 10.0 10.0 10.0

26.7% 13.3 20.0 6.7 6.7 6.7 13.3 0.0 6.7

28.6% 0.0 21.4 14.3 7.1 7.1 7.1 14.3 0.0

25.0% 25.0 0.0 8.3 0.0 8.3 25.0 0.0 8.3

9.1% 18.2 27.3 0.0 0.0 0.0 18.2 27.3 0.0

30.8% 15.4 7.7 0.0 7.7 7.7 7.7 15.4 7.7

28.6% 28.6 0.0 0.0 0.0 14.3 14.3 7.1 7.1

9.1% 36.4 9.1 9.1 0.0 0.0 9.1 18.2 9.1

0.0% 22.2 11.1 11.1 0.0 0.0 11.1 11.1 33.3

0.0% 16.6 25.0 8.3 8.3 0.0 0.0 25.0 16.8

SOURCE :

Supreme Electoral Court; Andrs Mejas database.

Ecuador: Provincialization of Representation

119

vote a straight party line, but this option did not eliminate the personalized character of voting because it was but one of the multiple ways of accruing votes. Contrary to a proportional representation system with a personal vote (also called a preference vote) in which the voter chooses one candidate from a given party or coalition, in Ecuadors 1997 system, each voter could chose as many candidates as there were seats in each province, from various lists. Dispersion could occur in the very act of voting, since the voter had several votes or fractions of votes, something that does not happen under most systems. The vote itself carried the seeds of dispersion. Therefore, a single persons vote could produce the same effect that would take the votes of several persons to accomplish in other electoral systems. This system provided maximum exibility in choosing parties or, if one prefers, ideologies. In a large district, voters could cast votes for candidates from all over the political spectrum, causing the spatial model for voting (Downs 1957) to lose its power and the relationship between voters and parties to be almost completely annihilated. The systems most notorious effects were seen in the large districts where the possibilities of selecting from different parties were greatest. The largest partiesthose that underwent a process of consolidation throughout the post1978 period and that helped support the stability of the party systemwere the most affected, mostly because their electoral strongholds are in the largest districts. Because it adversely affected the main parties, this electoral system dealt a blow to the institutionalization of the party system. The open-list system also produced a dispersion of votes in the smaller districts. In most small districts,22 the majority of voters distributed their votes widely. The open-list system weakened parties and furthered the extreme personalism of politics (Pachano 1998). It is difcult to nd a system that does a better job of fostering personalism and fragmentation. This electoral system fostered the oating character of both voters and politicians (Conaghan 1994, 1995). The displacement of votes from one party to anotherthe very foundation of dispersion and fragmentationneed not be put off until a later election since it could be accomplished in a single act of voting. And with it also came reduced possibilities of interpreting electoral results as a sanction or reward for different parties since no unied party preference was expressed when a voter chose candidates from several parties. In this way, the role of elections as a mechanism for assessing party performance (accountability) was signicantly reduced. Although at the national level general tendencies could be discerned, they did not necessarily reect the voters positions since multiple positions were expressed in a single act of voting. In conjunction with the ample opening awarded independents established in 1994 as a result of another referendumthis electoral system left the party system vulnerable to deepening problems. It contributed to personalism, already a clear tendency in previous elections and one of the main factors contributing

120

Pachano

to the weakening of parties. It was introduced in a context of animosity toward parties, arising fundamentally from the poor performance of governments invariably identied with certain partiessince the mid-1990s.

Diffuse Multipartism: Interests and Practices


The institutional framework described in the previous pages unfolds within a social context characterized by diversity. Ecuadorian society is plural in social, ethnic, linguistic, cultural, and regional terms (Almeida 1999; T. Bustamante 1992; Ibarra 1992; Deler 1987; Handelsman, n.d.; Rivera 1998; Pachano 1985). Social scientists have identied the ethnic aspects of this diversity and regional differences as important political factors, cleavages that dene behavior and identities. Considerable scholarly attention has centered on ethnicity, understandably, given the impact of the indigenous mobilizations beginning in 1990, as well as on the formation of Pachakutik, the rst party of ethnic origin, and its participation in national politics. A constant and active presence has made the indigenous a visible actor on the national scene, although Pachakutik is a small party conned to a few provinces.23 The presence of an ethnic party has generated widespread interest in ethnicity and politics in the social sciences (Van Cott 2003, 2004). Ecuadors regional differences have been studied at length (Quintero 1991; Pachano 1985; Len 1994; Deler 1987). Diversity is expressed in the form of regional societies differentiated along economic, social, cultural, and political lines. The sources of this differentiation are structural, by which I mean that it derives from those factors that constitute a society and therefore greatly impact its formation and behavior. Each of these regional societies takes the shape of relatively differentiated spaces in which specic social relationships are established and build their own power structures, giving rise to strong regional identities as well as unique behavior and habits. Social and political actors play the national political game more with their own regional demands than as actors on a national level. Political parties and national instances of representation are always heavily loaded with subnational demands and interests. This problem is aggravated by long-standing administrative centralization. The existence of regional societies means that politics takes place on two levels. First, a political game in the regions or in subnational arenasrevolves around controlling provincial and cantonal institutions. In this arena, local issues are salient, politicians proceed through an important stage in their careers, and collective actors are formed and battle for representation of regional interests. These political actors must also operate as expressions of national forces or at least establish a close relationship with them. The prohibition of the formation of subnational parties creates a mandatory relationship between local issues and the parties that, because of constitutional and legal provisions, must be national. Although this relationship has grown more exible with the introduction of

Ecuador: Provincialization of Representation

121

independent political candidatessince independents are not required to maintain a national organization and can limit themselves to local levelsit is still a burdensome imposition for parties. Subnational issues are very present in national politics. Regional demands and the social groups that represent them have an enormous effect on national issues. The power of subnational identities and regional issues in Ecuadorian politics is clearly visible in the constant presence of these regional problems, needs, and demands at the national level. National political actors are forced to take a stand on subnational issues, thus completing a circle that inhibits the identication, processing, and solution of national problems. This interaction between the national and the subnational is at the heart of political representation in Ecuador. What is represented, who represents, and how they are represented are the fundamental questions. In this game, powerful subnational actors are forced to act as emissaries of a binding mandate issued from their regions in order to ensure their own survival, while weak national actors, attempting to ensure their own survival, are forced to embrace subnational demands. The subnational actors do not prioritize the interests of the country as a whole, even though they act in national fora such as the National Congress. To abandon this provincially oriented behavior would be political suicide for politicians from the provinces since they would be giving up their reason for existing, as well as for those coming out of the national arena since they would no longer have access to the subnational levels. If a political movement emerges at the subnational level, it must make the transition to the national level, not only because of legal determinations but also because that is where decisions are made and resources are distributed. Conversely, if a political movement emerges at the national level, it must penetrate the subnational levels because that is where the interests that motivate the voters lie.24 The decisive factors in this two-sided game lie for the most part in the institutional/legal framework, especially in elements of the electoral system outlined above. That is the problem that confronts the political parties. Their dilemma lies in the necessity of either consolidating into national parties capable of working for the general interest and structuring government proposals, or remaining subnational parties with loyal constituencies but continually dependent upon socially and spatially limited interests. In the light of the last twenty years, the latter is clearly the stronger tendency. To ensure their permanency, parties have strengthened their links to regional or provincial interests and secured positions in electoral strongholds, even at the risk of sacricing proposals of national scope and giving up the possibility of producing positive results during their terms in government. A result of this dynamic has been the provincialization of parties and of politics in general. Provincialization can be understood in two ways. First, it refers to the electoral reclusion of parties in the strictly dened arenas in which they obtain their votes. Second, it also refers to the predominance of subnational

122

Pachano
Table 4.6 Electoral Strongholds of the Main Political Parties, 1979 2002
(share of national party vote won in First and Second provinces, provincial deputy elections) Party PSC PRE % Province
a

ID % Province %

DPa Province %

Year

Province

1979 1984 1988 1992 1996 1998 2002

Guayas Pichincha Pichincha Guayas Guayas Pichincha Guayas Pichincha Guayas Pichincha Guayas Pichincha Guayas Pichincha

30.72 30.05 30.55 27.60 38.05 19.15 51.25 6.35 44.78 13.53 44.80 15.55 78.51 6.98

Guayas Pichincha Guayas Pichincha Guayas Manab Guayas Manab Guayas Manab Guayas Manab

76.37 10.92 65.82 5.37 38.97 9.24 34.34 15.24 40.03 14.47 59.96 12.88

Pichincha Guayas Pichincha Guayas Pichincha Guayas Pichincha Guayas Pichincha El Oro Pichincha Azuay Pichincha Guayas

34.11 16.23 26.51 11.02 27.03 14.59 26.06 10.92 29.55 15.74 47.38 9.59 66.02 14.14

Manab Pichincha Pichincha Guayas Pichincha Azuay Pichincha Guayas Pichincha Guayas Pichincha Manab

16.39 13.45 33.22 11.04 30.37 12.77 42.39 9.25 27.86 26.34 39.45 29.22

SOURCE :
a

Supreme Electoral Court. Did not compete in 1979.

issues in national politics, which in turn has a negative effect on policies and governability. Provincialization is one of the main characteristics of the party system, and in Ecuador it contributes to others such as fragmentation and atomization. To appreciate the magnitude of provincialization in Ecuadorian parties, consider the parties electoral behavior in terms of territorial origin and respective number of votes. As Table 4.6 shows, the parties with the most seats in Congress (PSC, ID, PRE, and DP) have won a high percentage of their votes in only one province, clearly out of proportion with that provinces importance within the national electorate. Guayas, Pichincha, Manab, and Azuay are the provinces with the greatest population. But the gures that the parties win in their strongholds greatly exceed the proportion of voters that these provinces represent countrywide. While during the post-1978 period Guayas has uctuated between 24.0 percent and 27.5 percent of the national electorate, and Pichincha between 18.0 percent and 20.0 percent, the parties with electoral strongholds in these provinces exceed these gures by amounts that have grown in recent years.25 Some parties fare well in the Coastal provinces (Guayas, Manab) and others fare well in the Sierra provinces (Pichincha, Azuay). The pronounced regional electoral differences have been a constant in Ecuadors political and electoral history. Electoral strength in one region automatically equals weakness in another, which explains the formation of impenetrable electoral strongholds. The parties are severely limited in achieving a proportional distribution of votes throughout

Ecuador: Provincialization of Representation

123

the nation. Given the relatively balanced distribution of population between the Coast and the Sierra, and given the absence of a third region capable of offsetting this balance (due to the small population of the Amazon and Galpagos provinces), no party is likely to win a majority at the national level, something that in fact has not occurred during the entire post-1979 period. In this sense, the provincialization of the parties is one of the main reasons for party weakness and the fragmentation of the system as a whole. Regional discord is one of the most visible characteristics of Ecuadors political system. Region tends to overshadow other political cleavages, so that the Ecuadorian political game is dened more by the conict between territorially constructed identities than by economic or ideological cleavages. Its inuence is obvious in the actors behavior and in the content of the national political agenda, and it forces political parties to act accordingly. The parties must represent spatially dened interests. The possibility of obtaining an even distribution of the vote for the different parties across the whole of the national territory is minimal, as is the space in which to build a national agenda. This regionalization is clearly seen in the work by Mainwaring and Jones (2003), who document that Ecuador had the least nationalized distribution of the vote among seventeen countries in the western hemisphere. The authors created an index of party system nationalization. Between 1979 and 1996, Ecuador attained an average coefcient of 0.57 on a scale of 0 to 1.26 Only Brazil (0.58) approximated Ecuadors low level of nationalization. Bolivia scored 0.77, Chile and Uruguay 0.87, Costa Rica 0.90, and Honduras 0.92. Another indicator of nationalization, the territorial distribution index (TDI) measures the distance between the number of votes a party wins in each province and the proportion of the national electorate in that province. A party is national in character if its votes are distributed by province in approximately the same proportion as the provinces share of the national electorate. This indicator is constructed by adding together the differences between the proportion represented by each province in the national electorate and the proportion of votes each province has in the partys total number of party votes. It compares the weight of each province in the nationwide electorate with that of the provincial votes in the total number of party votes. Each party is assigned a gure for each election (Table 4.7).27 A lower number indicates less distance from the nationwide distribution of the electorate and consequently a partys greater national presence. Regardless of the indicator used, the distribution of each partys votes differs markedly across different provinces. Based on the TDI, national distribution has deteriorated throughout the post-1979 period. Smaller parties (below 10 percent of the vote) show the most uneven electoral performance across different provinces. This means there is a relationship between the fragmentation of the party system and the uneven distribution of party votes across provinces. This is an expression of the relationship between small parties and local interests, and

124

Pachano
Table 4.7 Territorial Distribution Index (TDI) of Main Parties, 1979 2002
1979 1984 1988 1990 1992 1996 1998 2002 Average

PSC ID MPDa DP FRA UDPFADIa PRE CFPa PCEa APREa PSEa PLREa MUPPNPa Average
SOURCE : NOTE :

22.1 22.8 18.0 22.6 17.9 45.3 46.3 41.2 28.6 29.4

18.1 20.1 24.6 24.1 23.7 24.3 51.4 32.3 41.6 43.6 50.4 26.5 31.7

14.0 18.5 30.9 26.1 27.9 57.8 39.0 22.9 23.5 22.9 35.9 38.5 29.8

29.1 19.2 26.7 36.9 39.0 26.6 27.4 36.6 24.1 31.9 38.4 48.4 32.02

26.5 25.6 31.5 32.2 29.1 33.6 22.4 35.8 48.6 30.7 34.4 53.3 33.6

20.4 33.9 25.4 28.6 27.1 23.1 37.8 44.9 36.4 45.5 27.1 54.2 33.7

21.1 23.9 24.1 21.2 44.4 26.4 27.1 46.3 33.4 60.8 39.5 52.0 87.7 39.1

52.4 48.1 34.2 50.0

25.45 26.50 26.93 31.30 31.86 31.87 32.53 33.20 37.32 38.93 40.99 43.31 74.22

37.3 35.9 42.5 72.1 80.8 50.4

Supreme Electoral Court. Empty cells indicate party did not compete that year. a Parties with an average number of votes below 10 percent in that period.

of the fact that their presence in the national arena is due to legal imperatives and that the national arena is the only real space in which important decision making occurs. Remaining on the fringes of the national institutions, specically Congress, would cost the local parties dearly. The provincialization of the parties is directly expressed in parliamentary representation. As seen in Table 4.8, the conguration of largest parties from the Coastal and Sierra provinces is clearer at this level.28 In sum, the regulations designed to foster the formation of parties with national scope have turned out to be useless. Probably, the explanation is that these measures were not meant to regulate already existing behavior, but instead to generate new behavior designed to consolidate a modern political system. Therefore, instead of being measures aimed at channeling the demands and the representation of regional or local interests, they were a way of denying or hiding these interests. These regulations were intended to impose certain behaviors, and they ignored the concrete conditions of the provinces and of the regional arenas in general. For this reason, from the outset there was a risk that actors would use other channels to articulate their provincial or local demands. And when these other ways failed to materializewhich could have been prevented through a process of decentralization of and increased exibility in party and electoral lawsregional and local demands quickly found their way into the mechanisms designed specically to evade them. Due to the legal impossibility of forming parties with strictly local or regional scope, the national partiesrather, those forced to be nationalhad to take on

Ecuador: Provincialization of Representation


table 4.8 Regional Distribution of Origin of Deputies, by Party, 1979 2002
Regional origin Party Costal (Coastal) Sierra AmazoniaGalpagos Total

125

PSC PRE ID DP
SOURCES :

66.7% 75.3% 30.2% 28.3%

31.8% 23.4% 65.5% 66.4%

1.6% 1.4% 4.3% 5.3%

100% 100% 100% 100%

Supreme Electoral Court; Freidenberg 2000.

the demands and the representation that would have been the province of the former. Conceivably, this might not have occurred within a exible framework that allowed regional or local parties to coexist with national parties, provided that the functions and scope of action for each of these had previously been clearly dened. However, by applying general laws to diverse situations and, above all, by giving these laws the power to transform practices and to generate behavior that turned out to be articial, the local parties were checked but their functions were transferred to parties expected to be national in character. Thus, national parties were forced to adapt to this distortion or run the risk of isolating themselves from voters and losing their ability to represent them. This forced them into a situation of dependency with regard to local or regional interests, and the effort to respond to local interests overshadowed ideological and programmatic considerations. Thus was completed a circle comprised of (a) the presence of regional identities; (b) inexible laws that sought to deny or minimize them; (c) the absence of adequate mechanisms to express these local and provincial interests; and (d) parties forced to meet the interests of regional electorates. The main consequence was the shifting of local and regional issues to the national arena, especially Congress, where debate can no longer be separated from territorial determinations and the game described earlier between the national and the subnational must be played. Forced to act as representatives of particularistic local interests, parties act as voices for narrow social and economic groups. The corporatist nature of politics in Ecuador can be explained to a great extent by this relationship between regionally dened interests and political representation since pressure groups achieve a dominant presence in local arenas and dominate representation. Political operations become tremendously complex, especially with regard to the pursuit of agreements and the fostering of national politics, which takes place in an arena where particularistic and directly represented interests battle one another. The indigenous peoples presence in Ecuadorian politics is emblematic of this link between localized interests and the provincialization of the parties. These

126

Pachano

indigenous parties have stronger regional roots than other parties because the geographic location of the indigenous populations creates a regional bias. The indigenous population is located almost exclusively in the Sierra and Amazonia provinces. Therefore Pachakutik, the principle partisan voice of these groups, wins votes almost exclusively in these regions. It is an important player in the Sierra and Amazon, but faces enormous difculties in winning votes in the Coastal provinces. Its electoral shortcomings in the Coastal provinces have prevented it from developing a broader base, not only in electoral terms, but also with regard to the possible structuring of proposals of national scope that go beyond the particularistic interests of the indigenous peoples. Pachakutik has adopted the same logic as the system as a whole, forced to take refuge in local bastions in order to build up its electoral strength at the cost of not having a presence in other regions. One can extend what has been said about Pachakutik to all Ecuadorian political parties. Even the largest parties have adopted this strategy of representing group interests in order to win a large number of votes in some provinces. This is the dilemma facing the parties and giving rise to the problems of representation that, paradoxically, are not the ones most analysts emphasize when they refer to the crisis of representation.

Crisis of Representation or Crisis of Regulation?


The problems facing the party system originated basically in the rules that regulate them. Their inorganic characterthe fact they do not all point in the same directionthe contradictions between their separate components, and the reforms constantly introduced in response to short-term interests prevent the system from attenuating the structural conditions surrounding it. These structural conditions give rise to actors, orientations, and behaviors that are ill suited to the construction and consolidation of a political forum of national scope or politics built around an arena wherein the general interest can take shape. These structural conditions would have had a less negative impact if another institutional designspecically, a different electoral systemwere in place. Structural heterogeneity is not necessarily an obstacle to the elaboration of national proposals and, consequently, to the consolidation of national parties. The experiences of countries as diverse or more diverse than Ecuador (Spain, United States, Germany, Switzerland) have proven the power of institutional design to forge national parties and interests. When they speak of the Ecuadorian crisis of representation, most analysts allude to aspects other than institutional design and refer instead to social, economic, and cultural factors (Arias 1995; Burbano 1998; Dvalos 2001). These analyses emphasize the results produced by the system, and they question the capacity of parties and democratic institutions more generally to represent interests.

Ecuador: Provincialization of Representation

127

They usually claim that links between the represented and the representatives are weak. In turn, such weak linkages are considered a threat to the smooth operation of democratic institutions and even to the systems stability. In this manner these analysts nally arrive at problems of governability, through a forced identication with the problems of representation or representativeness. A connection does exist between problems of representation and governability, but not the kind of connection that has usually been suggested in Ecuador. The political system fails to yield satisfactory results not because of a rupture between the represented and the representativessuch a rupture does not exist or does not exist as acutely as the analysts claim. Nor is the main problem a limited ability to represent intereststhis ability is actually excessive given the particularistic nature of representation in Ecuador. Rather, the problems of governability that stem from the system of representation arise because of the game that emerges out of a defective institutional design. The impossibility of fostering policies with national scope, the short-term focus of political action, and the predominance of local and group interests impose a logic that leads to the immobilization of governments and Congress. The ongoing game between powerful local actors and weak national actors, driven and fostered by the institutional design, goes a long way in explaining the political systems low capacity. The provincialization of the parties, a result of the electoral system described above, largely explains problems that have not been treated frequently enough by Ecuadorian social scientists and, on the contrary, have remained hidden behind ideas such as the crisis of representation.

Notes
1. Until the 2002 elections, the four biggest parties of the post-1979 period (PSC, PRE, ID, and DP) displayed a tendency toward an increased share of the vote notwithstanding cyclical oscillations. Ecuador is halfway between the collapse of the political parties of Peru and Venezuela and the stability of Colombia and Bolivia. 2. Since the transition to democracy, a decline is visible in indicators such as the gross domestic product, poverty indexes, distribution of income, the proportion of the budget assigned to social expenditures, and the populations buying power. From 1980 to 2000, there was zero growth in the gross domestic product; per capita income fell by 0.3 percent between 1981 and 1991 and by 0.6 percent between 1991 and 2001; poverty increased from 34 percent of the population in 1990 to 56 percent in 2002. 3. An analysis of the substantive outputs of the Ecuadorian political system is beyond the scope of this chapter. I only make general references to this subject without ignoring its importance in any analysis of the problems of representation. 4. According to Mainwaring and Scully (1995, 4), one of the criteria for the institutionalization of party systems is the permanency of electoral rules, together with the

128

Pachano

solidity of the organizations, reduced electoral volatility, the existence of roots in the society, and operations dependent upon bureaucratic routines more than on personalities or charismatic leadership. 5. The Social Cristiano, Roldosista Ecuatoriano, Izquierda Democrtica, and Democracia Popular parties have won as much as 80 percent of the valid vote. In the 2002 elections this vote share dropped noticeably, although this is not an indication of a party collapse of the magnitude experienced in Peru and Venezuela. 6. Small parties have disappeared as a result of a legal provision that requires that they win a minimum share of the vote in two consecutive elections in order to maintain legal recognition. This legal barrier has uctuated between 4 percent and 5 percent, and has not remained continuously in effect during this period. This is a barrier only to registration and not to representation since parties that do not meet the minimum maintain their seats in parliament and in other elective ofces even after the second consecutive failure to meet the threshold. 7. The Partido Socialista Ecuatoriano, the Frente Radical Alfarista, the Partido Liberal Radical, and the Movimiento Pachakutik have repeatedly done just this. Supported by the absence of any true barrier to representation, they have survived several elections. 8. Snyder (2001, 149) considers the problems of malapportionment between voters and seats to be one of the causes of unjust elections, on a par with the buying of votes, the altering of outcomes, and electoral fraud. 9. Although regions do not constitute an ofcial administrative-political division and are not a part of the electoral design, in the countrys political and social life they carry considerable weight. 10. This was one of the reforms introduced by the National Constituent Assembly during the constitution-making episode of 1998. 11. The number of members of the National Congress has uctuated constantly. The number of deputies increased from 69 in 1979 to 123 in 2000, with 71 between 1984 and 1988, 72 in 1990, 77 in 1992, 72 in 1994, 82 in 1996, 123 in 1998, and 100 since 2000. 12. The most notorious example of the inuence of small parties was the Frente Radical Alfarista (Radical Alfarista Front, or FRA). Although it never had more than three deputies, it gained the presidency of the Congress on two occasions. When Congress unseated President Abdal Bucaram in 1997 and appointed an interim presidentin clear violation of the Constitutionit elected the supreme leader of the FRA, Fabin Alarcn. 13. The Ecuadorian political system has been constantly reformed since 1979. This has become a source of instability since the country lacks a stable normative framework to guide the behavior of political actors. A summary of the many reforms introduced since 1979 can be found in Table 4.1. 14. Other provisions regulate various aspects of internal party life and express the orientation of the new regulations and the parties role. The obligation to participate in a minimum number of provinces refers to multimember elections: elections for municipal councils, provincial councils, and the National Congress. 15. There was always a differentiation between national and provincial deputies, with regard not only to their electoral districts but also to their functions. When in 1983 the provincial deputys term was reduced to two years, the national deputys term remained at four. The minimum age requirement for provincial deputies is 25, while it was 30 for national deputies. And although not in the end adopted, a proposal was made that would require candidates for the presidency of the Congress to be limited to national deputies. 16. Only twice, in 1996 and 2002, have national alliances been formed for presidential and legislative elections. But even so, in 2002, the parties that formed this national alliance entered into different coalitions in the provinces.

Ecuador: Provincialization of Representation

129

17. The newspaper El Comercio drew attention to the importance of the parties provincial politics in a series of reports published between February and August 2003. Each partys selection of candidates responds to the specicities of a certain province. 18. The change in the electoral calendar was more profound. The presidential and legislative terms were cut from ve to four years and the term of a provincial deputy to two years. The goal of increased stability and continuity through longer terms was therefore subordinated. 19. This perception was wrong. There is no correlation between presidents elected with a low percentage of the votes and instability of their governments. 20. The Ecuadorian party system is one of the most fragmented and volatile in Latin America (Conaghan 1995; Mainwaring and Scully 1995; Meja 2002). There are few studies on this subject. For example, there has been little exploration of the relationship between party system fragmentation or electoral volatility and particularistic practices such as clientelism and corporatism, or between the provincialization of politics and parties. 21. The scheduling of parliamentary elections to coincide with the second round of the presidential election may affect the percentage of votes won by the party of the winning candidate; in 1979 this candidates party achieved the highest percentage of votes for Congress during the entire period. But this too can be questioned, since it applied only to the winner and not to the other second-round presidential candidate, whose party did not fare well in the congressional election. 22. In the 1997 election, in seven of the nine districts that elected two deputies, candidates from two different parties won. In ve of the seven districts that elected three deputies, three different parties elected one candidate each (Pachano 1998). 23. Pachakutik has taken part in elections since 1996. Its share of the vote (based on the average number of deputies and provincial and municipal councilors) peaked in 1998 at less than 5 percent of valid votes. In 2002, although it backed the winning presidential candidate, Pachakutik barely surpassed that percentage. The party has achieved signicant results in local elections, especially mayoral elections in cantons with a large indigenous population, but it has been unable to penetrate several provinces, especially the Coastal ones. Certain actions, such as Pachakutiks support of the January 2000 coup that ousted President Mahuad, have led to greater renown but have at the same time limited the partys electoral growth. 24. Political parties have pushed this tendency to the limit by granting privileges to the municipalities and provincial councils, where they have strengthened themselves electorally and where at the same time they have been able to develop successful administrations. The cases of the Partido Social Cristiano (PSC) in the mayors ofce in Guayaquil and the Izquierda Democrtica (ID) in Quito are examples. 25. The only exceptionsthe Partido Social Cristiano and the Izquierda Democrtica between 1979 and 1986 illustrate the provincialization of parties that had a national scope during the rst elections in the post-1979 period. 26. The indicator uses the Gini coefcient to measure inequality of distribution, in this case the votes obtained by each party in electoral districts or subnational units. In this application it has been inverted (1/Gini): a higher score equals a more nationalized distribution of votes (Mainwaring and Jones 2003, 142). 27. The indicator is the product of the sum total of absolute values taken from the difference between the weight of the province in the census (padrn) and the partys provincial votes, multiplied by the weight the province carries. The following formula can be used to express this: TDI ( |Pn VPn|P)/2, where Pn is the weight carried by each province in the electoral census and VPn is the weight of provincial votes over the

130

Pachano

partys national voting. This is similar to the procedure used by Taagepera and Shugart (1989, 104ff.) to measure deviation from proportionality. Thanks to Andrs Meja for help in arriving at this indicator. 28. The Coastal provinces are underrepresented as a result of using provinces as electoral districts and because only parties with the greatest number of votes during the period are included. The small parties are local or provincial groups and including them would mean working with a constant and not a variable.

References
Almeida, Jos. 1999. Identidades en el Ecuador: Un balance antropolgico. Antropologa (Universidad Catlica, Quito), no. 4: 14 32. Amorin Neto, Octavio, and Fabiano Santos. 2001. The Executive Connection. Party Politics 7, no. 2: 213 34. Arias, Natalia. 1995. Partidos polticos: Hroes o villanos? Ecuador Debate 36: 49 61. Barrera, Augusto. 2001. Accin colectiva y crisis poltica: El levantamiento indgena ecuatoriano en la dcada de los noventa. Quito: Abya-Yala. Burbano, Felipe. 1998. Cultura poltica y democracia en Ecuador: Una aproximacin a nuestros vacos. Quito: CORDES. Bustamante, Fernando. 1997. La cultura poltica y ciudadana en el Ecuador. In Ecuador: un problema de gobernabilidad, ed. Esteban Vega, 103 57. Quito: CORDES. . 2000. Los partidos como orientaciones culturales. Iconos 9: 88 97. Bustamante, Teodoro. 1992. Identidad, democracia y ciudadana. In Identidades y sociedad, by various authors, 43 76. Quito: CELA. Conaghan, Catherine. 1994. Loose Parties, Floating Politicians, and Institutional Stress: Presidentialism in Ecuador, 1979 1988. In The Failure of Presidential Democracy, ed. Juan Linz and Arturo Valenzuela, 254 85. Baltimore: Johns Hopkins University Press. . 1995. Politician against Parties: Discord and Disconnection in Ecuadors Party System. In Building Democratic Institutions, ed. Scott Mainwaring and Timothy Scully, 434 58. Stanford, CA: Stanford University Press. CORDES. n.d. La ruta de la gobernabilidad. Quito: CORDES. Dvalos, Pablo. 2001. Movimiento indgena ecuatoriano: La construccin de un actor poltico. Revista Ciencias Sociales 20: 11138. De la Torre, Carlos. 1996. Un solo toque. Quito: CAAP. Deler, Jean-Paul. 1987. Ecuador: Del espacio al estado nacional. Quito: Banco Central del Ecuador. Downs, Anthony. 1957. An Economic Theory of Democracy. New York: Harper Collins. Freidenberg, Flavia. 2000. Las posiciones ideolgicas programticas en los partidos ecuatorianos. Paper presented at the seminar Political Parties in Latin America, Institute of Latin American Studies (ILAS), London. Freidenberg, Flavia, and Manuel Alcntara. 2001. Los dueos del poder: Partidos polticos en Ecuador, 1978 2000. Quito: FLACSO. Handelsman, Michael. n.d. La globalizacin y la construccin de nuevas expresiones de identidad: El caso plurinacional del Ecuador. Unpublished paper.

Ecuador: Provincialization of Representation

131

Ibarra, Hernn. 1992. El laberinto del mestizaje. In Identidades y sociedad, by various authors, 95 123. Quito: CELA. Len, Jorge. 1994. El n de un ciclo poltico electoral: El regreso de las elites tradicionales, apata y cambio. Ecuador Debate 32: 76 89. Mainwaring, Scott, and Mark P. Jones. 2003. The Nationalization of Parties and Party Systems: An Empirical Measure and Application to the Americas. Party Politics 9, no. 2: 139 66. Mainwaring, Scott, and Timothy Scully, eds. 1995. Building Democratic Institutions: Party Systems in Latin America. Stanford, CA: Stanford University Press. Meja, Andrs. 1998. Partidos polticos: El eslabn perdido de la representacin. Quito: CORDES. . 2002. Gobernabilidad democrtica. Quito: Honrad Adenauer. Menndez, Amparo. 1986. La conquista del voto: De Velasco a Rolds. Quito: CEN. Nohlen, Dieter. 1993. Sistemas electorales de Amrica Latina. Lima: Fundacin Friedrich Ebert. Pachano, Simn. 1985. Movimientos sociales regionales. In Movimientos sociales en el Ecuador, ed. Luis Verdesoto, 151 81. Quito: ILDIS. . 1998. La representacin catica. Quito: FLACSO. Quintero, Rafael. 1991. La cuestin regional y el poder. Quito: CENFlacsoYork University. Rivera, Fredy. 1998. Los indigenismos en Ecuador: De paternalismos y otras representaciones. Amrica Latina Hoy, no. 19: 57 63. Rowland, Michel. 1998. La crisis de representatividad del rgimen poltico ecuatoriano: Una aproximacin institucional. In Representacin poltica y democracia, ed. Carlota Jacksich, 9 43. Buenos Aires: Konrad Adenauer. Snchez-Parga, Jos. 1998. La pugna de poderes: Anlisis crtico del sistema poltico ecuatoriano. Quito: Abya-Yala. Seligson, Mitchell, and Polibio Crdova. 2002. Auditora de la democracia. Quito: CEDATOS. Snyder, Richard. 2001. Devaluing the Vote in Latin America. Journal of Democracy 12, no. 1: 146 59. Taagepera, Rein, and Matthew Shugart. 1989. Seats and Votes: The Effects and Determinants of Electoral Systems. New Haven, CT: Yale University Press. Van Cott, Donna Lee. 2003. Institutional Change and Ethnic Parties in South America. Latin American Politics and Society 45, no. 2: 139. . 2004. Los movimientos indgenas y sus logros: La representacin y el reconocimiento jurdico en Los Andes. America Latina Hoy, no. 36: 14159.

5 Outsiders and Neopopulism: The Road to Plebiscitary Democracy

Ren Antonio Mayorga

emocracy in the Andean countries is in a dismal situation. A process of decline and even reversal has been under way for some time and reached a critical stage by the end of the 1990s. Representative democracy has followed contradictory and regressive paths, leading in some countries to authoritarian regimes, delegative democracies, or semi-democracies. In the last decade, most countries in the region have undergone devastating political turmoil, with major implications for the regions stability, political foundations, and future prospects. Throughout the Andes, the signs of strain are manifest. Between 1997 and 2005, Ecuador underwent a chaotic period of instability, witnessing the election and overthrow of ve presidents. Since the coup dtat of paratrooper Hugo Chvez in February 1992, Venezuela has suffered a mounting crisis of the state and the economy, compounded by a dramatic collapse of its party system. The rise to power of an autocratic outsider has pushed the country into a political stalemate and an even deeper crisis of ingovernability. Chvezs government has triggered the polarization of society, in which a widespread opposition has launched four general strikes but failed to topple Chvez in either the farcical coup dtat of April 2002 or the protracted general strike between December 2002 and January 2003. A decade earlier, Peru became the rst Latin American country to see its party system collapse, giving rise to a decade-long dictatorship that dismantled state institutions and degenerated into a maa-type regime. Colombias party system is undergoing a process of dangerous atomization, and the multifaceted threat of guerrilla violence, drug trafcking, and terrorism grips the state. Bolivia now faces serious strains after more than a decade of institution building (1985 97) that was not backed by sufcient economic growth and poverty reduction. The whole regions democratic system is at stake and faces a twofold crisis of political representation and governability.

Outsiders and Neopopulism

133

One of the most troubling aspects of the crisis of democracy has been the emergence of outsidersthat is, neopopulist and anti-political actorsin almost the whole region. The rise of outsiders is relevant to the subject of this book for two main reasons. First, the rise has been the dramatic outcome of the crisis of democratic representation, and particularly of the collapse of parties, as argued in Chapter 1. Second, it has had disruptive consequences for representative democracy. I argue that the crisis of party systems stemmed from the failure of predominant parties as governing parties. Ultimately, problems of governability were the underlying cause of the crisis of democratic representation, that is, of the increasing inability of parties to reect and articulate electoral preferences that became apparent in the deep distrust of citizens and the sharp decline of electoral support for parties. In Peru and Venezuela, outsiders sprang onto the scene with overwhelming success. Both Chvez and Alberto Fujimori in Peru seized power democratically and established political regimes that do not t the category of liberal democracies.1 In Bolivia, two outsiders, Carlos Palenque and Max Fernndez, created neopopulist parties with relative success and took an ambiguous stance toward representative democracy. Yet they could not prevail completely, given their integration into the party system in which they played a signicant role. Bolivias democratic regime has faced, however, a different threat, from indigenous movements that seek to destroy democratic institutions and replace them with utopian, ethnic-based, direct democracy and nationalist populism. Highlighting common patterns and differences among outsiders politics in Peru, Venezuela, and Bolivia from a comparative perspective, this chapter will address three main issues. First, it examines the causes for the emergence of outsiders and their rise to power. Second, it looks at the sequences and patterns of party system collapse. Third, it analyzes the far-reaching destructive consequences of outsiders politics on liberal democracy and democratic institutions. The main purpose is both to nd common ground explaining neopopulism and the emergence of outsiders in the Andean countries and to dwell on politically rooted differences between them. Why did successful outsiderssuccessful in the sense that they rose to power emerge in Peru and Venezuela? Why were outsiders in Bolivia only partially successful? What accounts for the rapid rise of radical indigenous political movements?

Theoretical Approach
The concept of neopopulism is useful for addressing the nature of the politics carried out by outsiders. This contention rst calls for a denition of the conceptual differences between neopopulism and the classical term populism, widely used in Latin American social sciences since the 1960s. Such diverse nationalist, anti-imperialist political movements as the PRI (Partido Revolucionario Institucional, or Institutional Revolutionary Party) in Mexico, Peronism in

134

Mayorga

Argentina, the APRA (Alianza Popular Revolucionaria Americana, or American Revolutionary Popular Alliance) in Peru, and the MNR (Movimiento Nacionalista Revolucionario, or Nationalist Revolutionary Movement) in Bolivia have all been labeled as populist. As a widespread and crucial historical phenomenon in Latin America, populism has spawned competing theoretical perspectives that reveal strong disagreement about the meaning of populism as a concept (Weyland 2001, 13). The concept of populism has been anchored in four theoretical perspectives: (1) a historical-sociological perspective, which stresses social mobilizations and sociopolitical coalitions arising in the context of the crisis of oligarchic domination, the early stages of industrialization, and the transition from a traditional to a modern society; (2) an economic perspective, which draws attention to populism as a type of redistributive policy and state interventionism responding to economic elites weakness and inability to develop class hegemony; (3) an ideological perspective, which identies populism with a specic discourse articulating the constitution of a popular actor and the contradiction between this actor and the dominant classes, and (4) a political perspective, which explains populism as a pattern of mobilization of subaltern and/or excluded masses by personalistic leaders that is not based on institutional structures of political mediation (Pcaut 1987, 245 54; Roberts 1995, 84 85). Mostly embedded in modernization and dependency theories, theoretical work on populism since the 1960s led to diverse meanings and a wide dispersion into social, economic, ideological, and political domains. As Weyland (2001) asserts, this theoretical work produced divergent cumulative and radial concepts. The result was broad dissemination and fuzzy meanings of populism as a concept. In contrast, theoretical efforts to understand and explain the paradoxical resilience of populism in the last decadeincluding the cases I am dealing with in this chapteraim to reshape the concept as a theoretical tool by delimiting it in terms of a key domain and a predominant meaning, thus making it useful for empirical research. The recent attempts to build a theoretical approach based on a political-institutional perspective provide a synthetic construction of the concept of populism, integrating phenomena associated with classical populism that diverge from contemporary, neopopulist features (Weyland 2001; Mayorga 1995; Novaro 1996; de la Torre 2000; Martucelli and Svampa 1992; Pcaut 1987). Given the exibility and diversity of contemporary populism and the emergence of populist leaders in contexts that are a far cry from past nationalist and statist populism, a mainstream in the current theoretical work restricts the concepts key domain to the political realm and denes it as a predominantly political phenomenon. From this perspective, present forms of populism are no longer linked to specic economic policies or social constituencies. Against the backdrop of paradigmatic cases such as Peru and Venezuela, I argue that contemporary

Outsiders and Neopopulism

135

populism should be conceptualized fundamentally as a pattern of personalistic and anti-institutionalist politics, rooted mainly in the appeal to and/or mobilization of marginalized masses. In this specic pattern of politics, the charismatic leader exploits an ideological discourse of defending the poor and excluded, through which he garners electoral support and democratically legitimizes the quest for and exercise of power. In this regard, unlike historical populism, neopopulism is involved in the democratic game. It accepts the rules of political competition, but at the same time resorts to the higher quality and legitimacy of the leader, who presents himself as redeemer and embodiment of the people and the nation. As an ideology, neopopulism is therefore a pattern of ideological legitimation that is not at odds with representative democracy. In fact, it takes advantage of the resources and incentives that representative democracy and its electoral mechanisms provide. Yet the Caesaristic conception of politics, the leaders central role, and the lack of an institutionalized party and support inevitably combine to undermine democratic institutions and to concentrate state power in the hands of the leader once he comes to power. As anti-institutionalist practice, neopopulism therefore is a pervasive form of anti-politics, that is, of politics carried out against parties, democratic institutions, and established political and economic elites (Schedler 1994, 4). Neopopulist discourse is basically anti-political insofar as it questions the established political parties as corrupt institutions and blames the political class and the economic elites for the problems facing the country. Thus, the outsiders discourse assumes fundamentally not only a radical rejection of the existing party systems as such, but also the idea that parties are useless and pernicious organizations. This discourse is equivocal insofar as it is mostly an appeal to excluded people and simultaneously a commitment to neoliberal economic policies (Mayorga 1995). Neopopulist discourse has fostered an extreme neoliberal economic model of structural adjustment characterized by deregulation of markets, privatization of state enterprises, foreign-trade liberalization, and the absence of social policies (Gonzlez de Olarte 1998; Roberts 1995, 101 8). As the Peruvian experience under Fujimori showed, nonetheless, these policies were not incompatible with economic populism. Fujimori managed social policies through an extraordinary concentration of power in the executive and by relying upon direct paternalistic relationships that were conducive to the micro-level exchange of material benets for political support, even in a context of macroeconomic austerity (Roberts 1995, 106).2 This is paradoxical, given the link of classical populism with nationalism, antiimperialism, and state intervention. For precisely both of these reasonsthe use of populist rhetoric supporting neoliberal economic and political strategies, and the type of ideological legitimationI use the term neopopulism in order to differentiate classical populism from its contemporary forms. Weyland denes populism in a similar way, as a political strategy through which a personalistic

136

Mayorga

leader seeks or exercises government power based on direct, unmediated, uninstitutionalized support from large numbers of mostly unorganized followers (Weyland 2001, 20 21). This denition better ts neopopulism. It does not explain, however, past forms of populism, nor, more importantly, why and how populist leaders like Juan Pern, Ral Haya de la Torre, and Victor Paz Estenssoro, who were not democrats in the liberal sense, had a well-organized and structured mass support and engaged in institution building. Most importantly, neopopulism differs from historical populism in that its fundamental characteristics include not only Caesaristic politics but the phenomenon of outsiders springing up from outside the established party system. At rst glance, the emergence of outsiders seems akin to thunder in a clear blue sky. But outsiders become key players essentially because of an auspicious context: a crisis of governability and a profound decay and breakdown of party systems. Thus, to understand the rise of outsiders and its disastrous consequences for democratic development, it is critical to draw attention to the fundamental fact that when parties as government agents fail to perform reasonably in tackling the fundamental problems and needs of citizens, they lose their capacity for political representation. Both phenomena engender a power vacuum that outsiders can exploit for their benet. Neopopulism in Peru has turned out primarily to be somewhat of an odd marriage between anti-politics and neoliberalism, aimed at reducing the state and establishing a market-centered economy. In contrast, in Venezuela, neopopulism has been linked to statist economic policies that are more compatible with classical populism. Despite the predominant economic policy of orthodox control of scal decits, Chvez has made a contradictory attempt to return to state-led capitalism with the 49 decrees of December 2001, which quickly triggered the widespread opposition of powerful business groups.3 Neopopulism is an ambiguous and exible phenomenon that has gained ground by assuming two ideologically different stances thatusing traditional categories could be labeled as left-wing and right-wing. Contrary to past populism, however, the rightwing tendency has linked neopopulism and anti-politics with neoliberal adjustment policies, as the cases of Fujimori in Peru and Bucaram in Ecuador demonstrate (Mayorga 1995; Weyland 1996, 2001; Novaro 1996; Knight 1998).4 Drawing upon recent research and my previous work (Mayorga 1995; Weyland 1996, 2001; Roberts 1995), this chapter posits two key theses. I contend, rst, that the emergence of neopopulism and anti-system actors has been the outcome of two main processes: the decomposition of party systems and a deep crisis of the statein fact, a crisis of governability. Second, I argue that in order to esh out these reections on common patterns and qualitative differences in the cases of Peru, Venezuela, and Bolivia, it is critical to pay attention to politicalinstitutional contexts and processes. An institutional approach provides satisfactory theoretical tools for addressing these issues, focusing on the relevance of

Outsiders and Neopopulism

137

institutional variables such as party systems, electoral systems, state structures, and governability problems. As Linz has stressed, structural characteristics like class and economic structures constitute a series of opportunities and constraints for both social and political actors and political institutions. Yet structural variables are not laws that causally determine historical and political development. Within a given institutional context, actors adopt choices and make decisions that substantially affect political outcomes. Since these decisions are not structurally predetermined (or structured contingencies), outcomes are probabilistic and inuenced by contingencies and chancethat is, more than one outcome is possible. For this reason, the variable of leadership can be decisive and cannot be predicted by any model (Linz 1978, 4 5). Therefore, my historical-institutional approach focuses on the interaction between political institutions and contexts, on the one hand, and processes and decisions of political actors, on the other.

Causes of Emergence: Crises of Governability and Political Representation


Several key questions underlie the emergence of neopopulist outsiders. Why did political parties fall prey to a structural crisis? Why did a collapse of the party systems take place in Peru and Venezuela? Why and how did outsiders seize power in these countries, and why did this not happen in Bolivia? Turning to the rst question, political parties and party-based governments in Peru and Venezuela put the sustainability of democracy in jeopardy. Parties and their leaders were unable to respond with effective policies to aggravating social and political problems caused by socioeconomic decline and state crisis in a period of collapse of the state-centered economy, requiring a shift to a market-centered economy. These problems were not only structural but also a result of political decisions and bad performance of governing parties, thereby creating a providential scenario for the rise of outsiders. In this sense, the key problems provoking the decline of parties were fundamentally problems of governability and not of political representation in terms of the reection of societal interests and demands.5 The crisis of political representation unfolded in several stages as an outcome of an underlying and deepening crisis of governability that caused a growing gap between society and the parties, and consequently a crisis of political representation. Over time, the dominant political parties in Peru and Venezuela suffered a signicant loss of votes and seats because a majority of voters no longer trusted them due to their failure as governing parties. These parties were still able to win a signicant share of votes and seats, in the 1990 elections in Peru and the 1998 elections in Venezuela, yet they fell short of winning presidential elections. In a presidential system, losing the presidential contest to outsiders amounts to a dramatic loss of political representation and power.

138

Mayorga

Analyzing the crumbling of political representation as a backdrop for the outsiders rise, it is necessary to dwell on the double dimension of political representation. In a democratic system, political representation consists fundamentally of a transfer of power to party representatives by citizens through fair and transparent electoral processes (Manin 1997; Sartori 1999). The fundamental assumption of political representation is that representatives must articulate the interests of society by acting in the best interest of the public, both in the executive and in parliament.6 Furthermore, as Sartori contends, two souls and demands coexist in representative government: to govern and to represent (Sartori 1999, 269). Since modern democracy is a system of representative government based on parties, and elections have the paramount aim of leading to the formation of a legitimate government with the responsibility to govern, it is a mistake to address problems of political representation solely in terms of parties ability to reect electoral preferences. Although the main political parties in Peru and Venezuela did not lose political representation all at once, they nevertheless failed as representative actors after performing poorly in government. As a consequence, they lost presidential power and subsequently were unable to survive as opposition parties.7 A context of worsening socioeconomic crisis and the decline of political parties as governmental actors brought about a crisis of governability, providing auspicious conditions for the rise of outsiders and the anti-political logic of neopopulist discourse. Prompted by the traditional ruling parties failure and loss of credibility, outsiders could present themselves as a radical alternative to the party system and political elites, and as charismatic leaders claiming to carry out a mission of national redemption. Fujimori and Chvez played the political game by established electoral rules; however, they claimed their authority not from democratic principles and rules, but from a higher legitimacy as charismatic leaders.8 For this purpose, they used a radical, anti-political discourse as an effective tool for identifying themselves with the needs of excluded people, playing the role of paternalistic leaders who embodiedmore effectively than democratic institutionsthe unity of the state and the people. My second thesis is that the neopopulist politics of outsiders are not only a political strategy and an anti-institutional style of politicsas Weyland holds but a strategy leading to the weakening and breakdown of liberal representative democracy and, particularly, to its transformation into a plebiscitary democracy. The so-called return of the leader has meant the destruction of democratic institutions and the rise of authoritarian political regimes. When Fujimori was sworn in for his second term in July 1995, the question arose as to whether he was an exceptional case or the spearhead of a new type of dictatorship that could spread to other countries in Latin America (Rospigliosi 1995, 314). As events in Venezuela later demonstrated, the tendency toward authoritarian neopopulism has not been an exception.

Outsiders and Neopopulism

139

The breeding ground for outsiders and neopopulism has been complex and multifaceted. Political-institutional and leadership-mediated causes are the crucial explanatory factors, while socioeconomic problems (which were both causes and effects) have constituted a critical context. Social cleavages, inequalities, and fragmentation deepened greatly during the 1980s. The rise of the informal economy and the atomization of social groups impaired social organizationsand especially labor unions. At the same time, economic decline generated unemployment, hyperination, dissatisfaction with political parties, and, eventually, the demise of populist policies (Gonzlez de Olarte 1998; Naim 2001). These socioeconomic factors put a strain on the whole political system, and particularly on the party system and the states ability to cope with socioeconomic crisis. Partiesboth in a polarized party system such as Perus and a moderate party system such as Venezuelas could not adequately respond to this crisis. Due to a progressive weakening of the basic functions of representation and governance that they exercise in a democratic system, parties lost their linkages to social organizations. Thus, the key factor explaining the emergence of outsiders is the dramatic crisis of party systems resulting from a failure of democratic governability. From the onset of the transition to democracy in Peru, political parties failed as governmental actors to carry out policies that could solve the populations grim socioeconomic problems (See Table 10.1 in Mainwarings concluding chapter in this volume). Furthermore, they did not modernize their patterns of action, persisting in a zero-sum game of confrontation. Democracy implies not only dissent and confrontation but also consent and agreement on fundamental issues. But the populist governments of AP (Accin Popular, or Popular Action) (1980 85) and APRA that preceded Fujimoris government used their electoral victories as carte blanche for a vertical style of leadership, refusing negotiation and agreements with the opposition (Lynch 1999, 260). The presidential system and the tradition of personalistic politics embodied in caudillismo nurtured this tendency of presidential power, with the executive excluding the opposition and exerting control over all state institutions. In Venezuela, the populist system of elite conciliation and the dominant party system of AD (Accin Democrtica, or Democratic Action) and COPEI (Comit de Organizacin Poltica Electoral Independiente, or the Committee for an Independent Electoral Political Organization) were for decades the lynchpin of democratic stability and a strong presidentialist system (Rey 1991; Kornblith 1998). Venezuelan democracy was mainly based on negotiation and consent. However, partyarchy had perverse effects, leading to a pathological kind of political control (Coppedge 1994, 2) and to pragmatism as the predominant political style: Ironically, the same characteristics of parties that had promoted democratic governance in the rst two decades of the regime worked to undermine it in the last two decades (Coppedge 2002, 10 11).

140

Mayorga

Because political parties in government failed to ameliorate economic and social problems, they lost their capacity to represent and channel social interests. At the end of Garcas government, between 1987 and 1989, Perus GDP had decreased by 20 percent, poverty had grown, and salaries had lost almost 60 percent of their purchasing power. By 1988, the ination rate had reached 1,722 percent, peaking at 7,649 percent in 1990. In this context, the war waged by Shining Path had produced almost seventy thousand victims and economic losses of about US $20 billion, equivalent to Perus external debt at that time (Ferrero 1993; McClintock 1989; Comisin de la Verdad y Reconciliacin 2003). In Venezuela, poverty had increased dramatically in the last two decades. Beginning in 1983, the successive governments of AD and COPEI could not stem the economic decline, although they had the opportunity to take advantage of two oil booms in the 1970s. At the time of Chvezs rise to power, 68 percent of the population lived below the poverty line. Moreover, per capita income in real terms was back at the level of 1962, unemployment had reached 15 percent, and 45 percent of the workforce was employed in the informal economy (Naim 2001, 21). To summarize, at the end of the 1980s parties became targets of discredit and distrust. In both Peru and Venezuela, a crisis of party legitimacy erupted, mostly as a consequence of bad performance, inefciency, and, last but not least, corruption. As a consequence, people affected by the socioeconomic crisis turned to neopopulist outsiders who promised to overcome poverty, corruption, and social inequalities. Meanwhile, parties became hermetic organizations that were increasingly alien to society. In Peru, parties detached themselves from underlying developments in society, thereby losing their constituencies by sticking to the traditional game of movement-like (movimientista) politics while unions were losing their grip. Political representation and electoral politics were further eroded by a mediadominated logic of political competition that had begun to hold sway over politics (Tanaka 1998, 180 82). While crucial transformations of the social structure and the political arena were taking place, at the end of Garcas government, selfsufcient parties were not able to change their practices and strategies. Thus, the thesis that political parties in Peru committed suicide and were not victims of murder sounds quite adequate (Lynch 1999, 257). In Venezuela, although the presidential system was often prone to stalemate following the 1958 pact of Punto Fijo that restored democracy, partyarchy was instrumental and successful in moderating political parties conict and guaranteeing democratic stability and governance. Nonetheless, the overwhelming control over the political system and civil society generated a dangerous lack of horizontal accountability, abuses of power, widespread corruption, and impunity. Thus, as the economy went downhill and poverty spread in the 1980s, the parties could not switch entrenched political practices. They instead became accomplices to their own destruction and accomplices also in the sense that they stubbornly and tragically resisted pressure to reform themselves (Coppedge 2002, 10 11).

Outsiders and Neopopulism

141

The Process of Decline and Collapse of Party Systems


A useful theoretical approach is to analyze the breakdown of predominant party systems and political leadership in a context of economic, societal, and state crisis as a sequential and perhaps also patterned process that will allow generalizations in the Andean region.9 In Peru and Venezuela, the process of the decline and breakdown of representative democracy and party systems evolved in three distinct phases: (1) a phase of detachment and estrangement of parties vis--vis society, and their weakening due to internal struggles and inability to change; (2) a phase of electoral defeat, party breakdown, and democratic takeover of power by outsiders; and (3) a phase of nal destruction of parties, wrought by Fujimoris autogolpe and an ensuing constituent assembly in Peru, and in Venezuela by a constituent assembly, after which Chvez achieved full control of power, dominating the executive, Congress, and the judiciary. In Peru, the rst stage was a protracted process spanning a decade, from 1980 to 1990, in which socioeconomic and political problems worsened (Cotler 2000). Democratization converged with acute problems of governability, including economic crisis and hyperination, a wave of strikes, the erce offensive by Shining Path, and the crisis and inability of parties to respond to these problems. By the end of the decade, most parties found themselves in a deep internal crisis, characterized by internal struggles within APRA and FREDEMO (Frente Democrtico, or Democratic Front) and a fragmentation of the left (Tanaka 1998, 170 73; Lynch 1999, 254 57). At the time of the November 1989 municipal elections and the April 1990 presidential election, The political class as a whole was alien to citizens worries and locked in intra-party struggles (Tanaka 1998, 173). The terrorist war that Shining Path had waged against the state and society had reached Lima, sparking uncertainty and fear. Thus, in the process of democratization beginning in 1979, Peru had not addressed major problems in its society and political system. Unlike Bolivia, the democratization process led to the decomposition of the party system and traditional leadership. Two problems emerged together. On the one hand, parties could solve neither socioeconomic problems nor the problems caused by guerrilla violence and terrorism. On the other, party organizations were unable to adapt to structurally rooted changes in political representation and to the new logic of interparty competition. Party strength was no longer determined by the mobilization of interest groups and corporate social actors, but by a logic of media-structured linkages to a diffuse, fragmented public (Tanaka 1998, 92 93, 168). Parties could not and were not willing to reshufe their structures and change both their political styles and policy orientations. The second stage in the process of decline was the electoral defeat of traditional parties and Fujimoris democratic election in April 1990. At the end of 1989, popular perceptions about parties began to undergo a dramatic change.

142

Mayorga

A rst sign that the scenario for the emergence of outsiders had been set was the November 1989 election of Ricardo Belmont, an outsider, as mayor of Lima. The trend was reinforced when the famous writer Mario Vargas Llosa, also an outsider, became the leading candidate of a new political front, FREDEMO. During the 1980s, high electoral volatility and a high degree of polarization and confrontation characterized the Peruvian party system. The electoral campaign of 1990 was no exception. Vargas Llosa polarized the electoral contest mainly because of his proposed economic shock program, prompting the APRA government and leftist parties to make him the main target of their attacks. President Garca turned his back on his own partys candidate, Alva Castro, deciding two months before the election to support the unknown Fujimori instead. Thus, Fujimori obtained strong government backing. After March 11, 1990, when polls gave Fujimori only 3 percent of voter preference, government-linked media gave him a decisive boost. Regional development corporations also offered the tractors Fujimori used so effectively in his campaign across the country.10 By mid-March, his support had reached 9.5 percent, and on the day of the rst-round election, April 8, 1990, he obtained 29.1 percent of the vote. Finally, in the May 1990 runoff election, Fujimori achieved a resounding victory with 56.7 percent of the vote. Were both results really a surprise, then? Did Fujimori rise to power by accident and chance, as Tanaka (1998, 164) suggests? The collapse of the party systemthe nal stage occurred between 1992 and 1995. Why did this collapse take place? According to Tanaka, who takes issue with retrospective determinism, the collapse of the party system and the ensuing breakdown of democracy were not inevitable (Tanaka 1998, 200). Events could have transpired differently. The opposition parties could have pressed Fujimori to step down, since the electoral results in 1990 had not led to a catastrophic defeat of FREDEMO and the traditional parties. FREDEMO obtained 32.3 percent in the rst round of the presidential election, and 30.1 percent of the seats in the Senate and the Chamber of Deputies. APRA obtained 25.1 (presidency) and 21.5 percent (Congress), respectively, while the United Left (Izquierda Unida, or IU) obtained 9.8 and 10 percent, and IS (Izquierda Socialista, or Socialist Left) 5.5 and 5.3 percent, respectively. Cambio 90 (Change 90), Fujimoris movement, became a minority faction in Congress, obtaining only 21.7 percent of Senate seats and 16.5 percent of the seats in the Chamber of Deputies. The outcome was a Congress in which opposition parties had an overwhelming majority and, consequently, a minority government, leaving Fujimori in a very weak position. Moreover, Fujimori had neither organized party support nor a government team with which to govern. Fujimoris politics and the choices made by the opposition parties brought about a showdown, eventually causing the breakdown of the party system. First, with the crucial aid of Vladimir Montesinos, Fujimori built a coalition with the

Outsiders and Neopopulism

143

military. Their pact was based on the Green Plan that the armed forces had developed with the aim of establishing a regime under military control. Fujimori resorted to an authoritarian project anchored in the Green Plan to establish a market-oriented economy within the context of a controlled democracy (Rospigliosi 1995, 31112; Reyna 2000, 141). Second, he forged a coalition with de facto powersbusiness groups, foreign investors, international organizationsin the framework of his neoliberal economic policies. Third, by 1991, successful economic policies provided Fujimori with great legitimacy. Finally, based on the authoritarian project, he decided to override congressional opposition and subordinate other democratic institutions. Because he was the head of a minority government, Fujimori asked Congress for emergency powers in order to cope with the economic crisis. In November 1991, he deliberately presented 124 bills to Congress at once, but Congress demanded a partial revision. Fujimoris underlying motive was to provoke a showdown with Congress by accusing it of incompetence and obstructive opposition to government policies. A stalemate between the legislative and executive branches ensued, and Fujimori reacted by threatening to close the Congress. By that time, Fujimori and the parties in Congress were adversaries, engaged in a war of attrition (Tanaka 1998, 213) that culminated with the April 5, 1992, autogolpe that shut down Congress, the judiciary, and other state institutions. Due to the success of economic policies to reduce ination and create stability, Fujimori won a decisive battle against the parties and other institutions with the backing of the armed forces, big business and, most notably, overwhelming popular support. Politically, Fujimori justied his move by arguing that the decomposition of prevailing institutions, widespread chaos and corruption, and the obstacles posed by Congress and the judiciary made effective governance impossible. In addition, Fujimori contended that prevailing democratic institutions were deceptive and false, and that it was necessary to take an exceptional approach to advance the process of national reconstruction (Maniesto a la Nacin, April 5, 1992). Thus, the autogolpe was one part of the strategy to concentrate power. The Constituent Assembly was the other part. In response to pressure from the Organization of American States (OAS), Fujimori decided to redemocratize the political regime by calling for a Constituent Assembly in November 1992. The Constituent Assembly introduced a new constitution that entrenched executive powers and established a provision that allowed for the reelection of the president. The main parties of the traditional party systemAPRA, AP, and IUrefused to participate in the assembly. Only the PPC (Partido Popular Cristiano, or the Popular Christian Party) and a new leftist front, MID (Movimiento de la Izquierda Democrtica, or Democratic Left Movement), took part, but they suffered a grave defeat in the election for the Constituent Assembly, whereas Fujimoris candidates (Cambio 90 Nueva Mayora, or Change 90 New Majority) pulled in 49.2 percent of the vote.

144

Mayorga

A contested referendum in October 1993 approved the new constitution with a slight majority of 52.3 percent of the vote. A historic defeat of the traditional parties occurred in the April 1995 presidential election, in which Fujimori won 62.3 percent of the vote while all the traditional parties combined did not obtain even 5 percent. It was the rst time ever in the contemporary history of democracy in Latin America that a general election demonstrated the complete breakdown of the traditional party system. In Venezuela, the process of party system decay and decomposition was even more protracted than in Peru. The process spanned fteen years, from 1983 to 1998. It began slowly, with an economic crisis in 1983 stemming from a severe fall in oil prices and a massive ight of capital. That led to the rst devaluation of the bolivar (the Venezuelan currency) in twenty-ve years. The 1980s witnessed the increase of the external debt, the instability of oil markets, capital ight, scal crisis, administrative corruption, and, most notably, ination (Kornblith 1998, 132). An attempt by the government of Carlos Andrs Prez (1989 92) to address the crisis with a neoliberal program failed, provoking a violent upheaval in the nations capital, the Caracazo, in February 1989. In hindsight, the Caracazo turned out to be the turning point in the crisis of Venezuelan democracy. An end had come to the economic stability and prosperity generated by the rent economy based on oil wealth. From 1978 to 1989, per capita GDP shrank 29 percent; from 1985 to 1998, declining oil prices hit the country hard. The democratic stability stemming from the partyarchy established by AD and COPEI since 1958 also came to an end, as the two coups dtat in 1992 and the impeachment of President Prez in the same year made evident. Electoral abstention grew from 12 percent in 1983 to 39.8 percent in 1993, an indication of increasing disaffection with the party system (Coppedge 2002, 11). The electoral victory of Rafael Caldera in the December 1993 presidential election signaled a second turning point in the process of party system decline. Caldera won 30.46 percent of the vote as the candidate of an alliance of small partiesMAS (Movimiento al Socialismo, or Movement toward Socialism), Convergencia Nacional (National Convergence), the Communist Party, and others. His thoroughly anti-political campaign triggered a serious delegitimization and breakup of the two-party system and its transformation into an unstable and weak multipartism (Molina and Prez 1994, 74 75). The AD candidate, Claudio Fermin, and the COPEI candidate, Oswaldo Alvarez, obtained 23.6 and 22.73 percent, respectively; in the case of AD, this was half the percentage won in 1988. This weakened, fragile party system broke down completely in the December 1998 election. Another key factor in the process of decomposition was the failure of constitutional reforms and attempts to change political parties (Kornblith 1998, 165 82). Paradoxically, the dominant parties undertook a process of constitutional reforms starting with the proposals made by the Presidential Commission for the Reform of the State (COPRE) at the end of the 1980s. Their goal was to

Outsiders and Neopopulism

145

stave off the weakening of the political system and to restore the party systems legitimacy. COPRE proposed sweeping reforms including decentralization, direct election of governors and mayors, the mixed-member electoral system, referendums at the state and national level, and internal democratization of parties. With the exception of the last two reforms, Congress enacted decentralization and the reform of the electoral system in 1989. The two coups dtat in February and November 1992 produced a critical situation, however, that further crippled constitutional reforms (Kornblith 1998, 61114). The suspension of the constitutional reform process damaged Congresss reputation and weakened the parties and political leadership. Attempts to curb the political systems deterioration through state reforms such as decentralization and electoral reforms were unsuccessful. The question, then, is why the reforms were not enough to restore Venezuelans faith in their political system and why they did not prevent the progressive delegitimization of democracy (Coppedge 1994, 164). Arguably, institutional reforms were carried out too late and only partially. They did not have an impact on solving crucial problems. According to Kornblith, the leading parties, AD and COPEI, did not share common criteria and were not really committed to political reforms (1998, 11114). They lacked both the necessary ability to innovate and the political will to carry out reforms. In a similar vein, Coppedge (1994, 164) contends that the basic problem has been that while political parties were the only actors in a position to adopt reforms, they were unwilling to make such reforms. This turned out to be one of the key problems. By the mid-1990s, after the failed coups dtat and the impeachment of Carlos Andrs Prez, doubts about the stability of the democratic regime had become evident. The nal stage of breakdown of the Venezuelan party system took place over a two-year period that included the 1998 presidential election, the Constituent Assembly, the 1999 referendum, and the July 2000 general election. The November 1998 presidential election was the turning point in a fteen-year process of decline of the dominant parties. Accordingly, Chvez did not destroy the old parties; he rather lled a political vacuum (Coppedge 2002, 14). Chvez won this presidential election in a landslide, with 56.2 percent of the vote, while the candidate of the governing party, Convergencia Nacional (Salas Rmer), those of the AD and COPEI, and the independent, Irene Salas, obtained low percentages of the vote. AD fared relatively well in the congressional election that took place one month before the presidential election, winning 24.1 percent of the vote, which was similar to the percentage the party had obtained in 1993. COPEI was the big loser, garnering only 12 percent of the vote. In comparison, Chvezs party won 19.9 percent of the vote, and his ally, MAS, obtained 8.9 percent. Thus, separate congressional and gubernatorial elections in 1998 brought about a minority government and the danger of a deadlock between the executive and Congress. The prevailing constitution ensured that Chvezs adversaries

146

Mayorga

would control Congress and other institutions. Chvez controlled only one-third of the seats in the two chambers. This fact, and not the alleged shortcomings of the 1961 Constitution, was what prompted Chvez to convene the Constituent Assembly. His primary motivation was not to tinker with the constitution, but to use the Constituent Assembly strategically as a mechanism to concentrate power and to neutralize Congress, the courts, and all other guarantors of horizontal accountability (Coppedge 2002, 1718). After another landslide in the Constituent Assembly election, in which his alliance won 122 out of 131 seats, Chvez reinforced his strategy of forging a strong, personalized power base. He substantiated this in the 2000 presidential election, winning 59.75 percent of the vote, while the AD candidate obtained only 2.72 percent (Molina 2000, 34). In conclusion, from a comparative perspective the rise to power of Fujimori and Chvez was characterized by similarities and differences in both origins and processes. The key similarities lie in the decomposition of the party system, the failure of political elites, and the ensuing governability problems that set the stage for the emergence of outsiders. Important similarities also relate to the incentives of presidentialism for outsiders and institutional arrangements such as the electoral runoff system. As Linz argues, there are structural reasons for the candidacy of outsiders. While an institutionalized party system makes it difcult for outsiders to enter a presidential competition, the personalized character of a presidential election makes possible, especially in the absence of a strong party system, the access to power of outsiders (Linz 1994, 26 27). The differences are rooted mainly in contingent political causes and certain contextual factors, such as guerrilla terrorism in Peru, the coups dtat in Venezuela, and self-destructive decisions of political actors. Contingent political causesassociated with the rationality of actors, and specic choices and decisions that cannot be explained as a predetermined, logical result of the political crisisalso had a major impact on the triumph of outsiders. Key political actors such as Alan Garca in Peru (president from 1985 to 1990) and Rafael Caldera in Venezuela (president from 1993 to 1998, and earlier from 1968 to 1973) helped anti-system actors like Fujimori and Chvez to seize power. Caldera himself came to power in 1994 with an anti-establishment electoral campaign. Once he took ofce, he pardoned Chvez, who was in jail after the failed 1992 coup dtat. Nonetheless, these contingencies do not justify Tanakas (1998) conclusion that the emergence of outsiders was thoroughly contingent in Peru. Tanakas (1998, 197) claim that Fujimoris rise to power is very much a random product understates structural causes, favorable strategic contexts, and the rationality of actors prompting Fujimoris success. As Tanaka himself asserts, the crisis of the party system and the electoral runoff system paved the way for and beneted Fujimori. In a plurality system, Fujimori never would have come to power (Tanaka 1998, 197; Mayorga 1995, 57). In an extremely polarized campaign, he took advantage of the strategic vote of antiVargas Llosa voters.

Outsiders and Neopopulism

147

From the standpoint of rational choice, it might have been rational for the Garca government, APRA, and leftist voters to vote for Fujimori in the runoff. Although in the end this decision proved to be self-destructive for them and other parties, their disenchantment with the traditional parties, as well as the polarization of the campaign between a candidate linked to the traditional elites and an unknown candidate associated with the Cholo and Indian population, made it rational for antiVargas Llosa voters to prefer the unknown demagogue. Besides, given the alleged weakness of a future Fujimori government, APRA was counting on the possibility of controlling Fujimori.

Consequences: Neopopulism as a Political Regime


Both in Peru and Venezuela, the politics of outsiders have led to the same farreaching consequences: full control of the state with the support of a majority of the population andas a key outcomea plebiscitary democracy leading to the weakening and demise of liberal, constitutional democracy.11 Neopopulist regimes in Peru and Venezuela emerged as electoral democracies, but they have mainly been regimes based on plebiscitary mechanisms, restricted pluralism, concentration of power in the head of government, the elimination of mechanisms of horizontal accountability, and popular demobilization.12 These regimes derive their legitimacy not only from democratic elections but also from the higher or deeper legitimacy of the leader himself, owing to the plebiscitary character of leadership. As Weber argues, the leader attains the condence and trust of the people through mass-demagogic means. This kind of leadership is Caesaristic, and its main tool is the plebiscite (Weber 1964, 1094). Thus, the correlation between neopopulism, deinstitutionalization, and autocratic rule has become manifest (Roberts 1995, 116; Weyland 2001, 25). From the outset, neopopulist regimes under the sway of outsiders have been marked by a tension between democracy based on popular sovereignty (i.e., electoral majorities) and liberal democracy (based on constraints on presidential action, horizontal accountability, and checks and balances). Chvezs regime is best dened as illiberal because his political movement controlled the executive, the courts, and the legislature, and handpicked all the members of supposedly independent agencies (Coppedge 2002, 15 16, 36). In the context of the decline and collapse of the party systems and of the frailty of state institutions, Fujimori and Chvez carried out a politics of tabula rasa, or a deliberate politics of deinstitutionalization. Although democratic-liberal institutions were not suppressed (except for the period from April to December 1992 in Peru), they were subverted and eroded, while loose political movements and the growing political intervention of the armed forces lled the vacuum left by parties. Furthermore, Fujimoris regimeand, increasingly, Chvezshas clearly shown that, in the absence of parties, neopopulist regimes can turn into outright dictatorial regimes,

148

Mayorga

not only by eroding civil liberties and concentrating power, but also by substituting the armed forces for parties. The breakdown of parties as mediation structures prompted a tendency toward autocratic regimes, proving that no working democracy is possible without parties.13 Is the concept of delegative democracy useful for dening and explaining neopopulist regimes? As ODonnell depicts it, delegative democracy rests on the premise that a candidate winning the presidency is entitled to govern as he or she sees t. Elected presidents present themselves as above political parties and organized interests (ODonnell 1994, 59 60). This concept stresses both the crucial democratic element of electoral legitimacy and the absolute predominance of the presidency. Yet the concentration of power and the destruction of independent democratic institutions have been so extreme in Peru and Venezuela that the concept of delegative democracy does not sufciently encompass the extent to which representative democracy has been undermined. The category of semi-democratic regime is more pertinent because it implies a signicant or total removal of checks and balances, an absence of horizontal accountability, human rights abuses, and wide autonomy for the armed forces (Levitzky 1999, 80; Mainwaring 1999, 102). Since it has led to pervasive authoritarianism, neopopulism ought to be assessed not only as a political strategy of outsiders but, when successful, also and foremost as a strategy breeding an authoritarian, dictatorial regime, or at least an illiberal regime. Fujimoris regime, in particular, turned out to be an extreme case of an authoritarian regime with a varnish of legitimacy that degenerated into a government of corrupt cliques that made political decisions as if they were state secrets (Grompone 2000, 109). After the autogolpe in April 1992, Fujimoris government depended upon an extended spoils system and not on a pact of domination or political hegemony. Although he lacked his own political organization and power structure, Fujimori built a broad power coalition consisting of himself and the de facto powersthe military, business, and international organizations (Lynch 1999, 244 52). An inner circle made up of Fujimori, Vladimir Montesinos, and the military eventually dominated this alliance. It quickly became a criminal maa and a corrupt gang of cronies engaged in embezzlement of public funds, blackmail and corruption of media and business groups, inuence peddling, illicit enrichment, and arms dealing. The most disturbing aspect of Fujimoris government was that Montesinos, who directed the SIN (National Intelligence Service), became the all-powerful executor of the pact with the military that grew out of the strategic Green Plan (Rospigliosi 1995, 329 31; Grompone 2000, 95 97; Reyna 2000, 138 41). Through the underground leadership of Montesinos, Fujimori carried out a strategy of toma de casilleros14 within the armed forces and business groups (Grompone 2000, 95). The pact with the military was not made with the armed forces as an institution but with the highest-ranking ofcers within a system of personal mutual loyalties and

Outsiders and Neopopulism

149

favors, which Montesinos rmly controlled. From the ofces of the SIN, Montesinos forged the pact by dismantling the armed forces institutional structures and hierarchies and by arbitrarily planning promotions and passing over ofcers who were due for promotion (Rospigliosi 1995; Obando 2001). The SIN extended its responsibilities into the armed forces and thoroughly penetrated them through a wide range of activities. The service became a watchdog over the military, inltrated the ministries and the state administration, gathered information about members of the opposition by tapping telephones and bribing members of Congress, exerted control over media information, and manipulated electoral campaigns (Grompone 2000, 101; Reyna 2000, 138 43). Due to the absence of institutionalized power structures, Fujimoris autocratic regime depended heavily on the secret service led by Montesinos. As the SIN became the core of Fujimoris power, his political fate was closely linked with Montesinoss scheming. In the end, Montesinos became not only the guarantor of Fujimoris power but also the source of his ruin and downfall, which took place shortly after the start of his third presidential term in July 2000. The collapse of Fujimoris regime was not the result of the resurgence of a vigorous political opposition but of the scandal over a video that led to Fujimoris resignation and revealed how dependent he had become on the network of corruption that Montesinos managed. Similarly, because he lacked his own political organization and power base, Chvez created a political movement in Venezuela, the MVR (Movimiento V. Repblica, or Fifth Republic Movement), for the 1998 election. The MVR was a loose organization that joined civilian and military members of diverse leftist origins who had brought together the popular sectors supporting AD and COPEI in the past. Nevertheless, the MVR was more than an electoral device and faade; it could have transformed itself into a single, hegemonic party, perhaps, and into an effective instrument of power. For the election, the MVR built an alliance with small leftist parties such as the MAS and the PPT (Patria para Todos, or Fatherland for All). Fourteen months after Chvez seized power, however, the MVR split up when Chvez himself decided to get rid of it by reorganizing his original movement, the golpista MBR-200 (Movimiento Bolivariano Revolucionario, or Bolivarian Revolutionary Movement). Apparently, Chvez disbanded the MVR because he saw it as an obstacle to his strategy of establishing a personalistic dictatorship. Neither movement has been a political organization on which Chvez has based his power. He does not rely nor does he intend to rely upon a broad national political structure, that is, on a party capable of organizing the masses as well as mediating and channeling interests and conicts. As a popular leader, Chvez has attempted instead to make up for this lack of political structure in two ways: by creating direct, plebiscitary links with his constituency through the so-called Bolivarian circles, and by turning the armed forces into his political instrument

150

Mayorga

for state administration (Gmez Calcao and Arenas 2001; Gmez Calcao 2002). Chvez has engaged both active and retired military ofcers in the executive at the national and regional levels, in state enterprises (especially PDVSA, the huge state-owned oil company),15 and particularly in the Plan Bolvar, which has put the armed forces in charge of repairing schools, building infrastructure, providing health care for the poor, and selling basic goods (Trinkunas 2002, 68 69). In short, Chvezs strategy has been to transform the armed forces into his own political instrument, engaging them directly in state administration and in an array of activities that go beyond the responsibilities dened for them in the constitution (Trinkunas 2002, 65 66; Manrique 2001, 325 26). According to some analyses, a new model of military intervention in politics through the leader has emerged. The model is characterized by the politicization of the armed forces and, according to Manrique (2001, 327), even by their transformation into a military party, since they have taken on functions inherent to a political party and hold key posts in the system of decision making. This thesis is debatable, however. The armed forces are not organized as a party; they have no need to legitimize their existence or to participate in elections, which is the fundamental function of a party. Instead, the armed forces play a political role reluctantly, subject to the political will and imposition of the charismatic leader on whom they depend. They are internally split into three factions: pro-Chvez ofcers (so-called revolutionaries by the government), institutionalists, and opponents. The latter two constitute the biggest factions, according to a classication of the Military Intelligence Unit (DIM, Divisin de Inteligencia Militar) (Manrique 2001, 330). The military has become an ersatz or surrogate party in Chvezs strategy.16 Unlike Fujimori, Chvezs nationalist, statist orientation has prevented him from forging an alliance with business groups; and he has not sought the support of international organizations. From the outset, he opted for confrontational politics against private business, the media, unions, and even the Catholic Church, triggering a dangerous polarization between them and his own followers.17 By keeping political control in his own hands, he has attempted to militarize the state apparatus and rely on military support to create an inner circle of followers. But apparently, he has neither an operator like Montesinos nor an SIN to assist him. The DIM does not seem to play a similar role. Chvezs policies have caused a deep internal rupture, the defection of several ofcers, and growing opposition within the armed forces to the expansion of the militarys role, to the militarization of the state, and to Chvezs authoritarian populist policies (Coppedge 2002, 27). The failed coup dtat of April 11, 2002, revealed that Chvez could not rally the armed forces behind him, although he has subsequently achieved signicant control over the military. Chvez feels committed neither to his own constitution nor to representative democracy.18 Despite the introduction of the new constitution in 1999, which strengthened presidential powers, Chvez claimed that Venezuela lives under a

Outsiders and Neopopulism

151

regime of constitutional transition and that he seeks extraordinary powers beyond those already conferred to him by the ley habilitante (empowering law). He also stressed that he was making a superhuman effort to carry out a peaceful revolution, but that should this fail an armed revolution would be the only alternative (El Nacional, May 5, 2001). This statement was a clear indication that Chvez does not respect the constitutional order he helped to establish, and would rather pursue the path of strengthening his own personal power. Hence, a deep contradiction between the constitutional order and Chvezs personal power project has become apparenta contradiction that does not create an adequate foundation for a long-term, institutionalized, neo-populist regime but rather renders it an oxymoron. From a historical perspective, their reliance on personalistic power structures makes neopopulist regimes inherently less stable than institutionalized democratic regimes, and they tend, in fact, to be short-lived. Since his rise to power, Chvezs political projectgrounded in a militaristic vision of politicshas left little room for doubt about his aim of dismantling the previous political regime and imposing a peaceful revolution against middle-class, business, and labor interests.19 To achieve this, he resorted to constitutional means introduced by a constitutional reform, expanding presidential powers under the guise of participatory democracy and allowing himself absolute legislative and decree powers in any matter.20 The project gained momentum at the end of 2001, when the government emitted 49 presidential decrees that signaled its course toward stronger state intervention in the economy, particularly in the agrarian sector. At the same time, Chvezs legislative majority 61.2 percent of the seats after the 2000 electionhad dwindled to little more than 50 percent due to defections in his coalition. Chvez thus sparked a widespread and radical confrontation with his political and social opponents, bringing about a historical rupture with the procedures of negotiation that had dominated party politics from 1958 until 1998. Both the refusal to allow political bargaining and the instrumental use of participatory democracy have shed light on the authoritarian character of Chvezs regime (Molina 2003). The 49 presidential decrees were a turning point in the relationship between government and the opposition since they provoked the rebellion of the middle classes, organized labor, and business associations, and as a consequence an even deeper rupture within the armed forces.21 In April 2002, the rst general strike against Chvezs regime set the stage for an uprising by high-ranking ofcers and a coup dtat led by conservative business groups, which intended to establish absolute powers by abolishing the constitution. The failure of the coup dtat aggravated the conict between the government and the opposition, however, given the fact that after his restoration to power, Chvez maintained his strategy of imposing his peaceful revolution and thereby increasing the high-handed, autocratic concentration of power. Chvezs rst reaction to the coup was to start a dialogue of reconciliation with the opposition

152

Mayorga

and to attain a modicum of governability. Yet he also took advantage of a weakened and leaderless opposition and the split in the armed forces to strengthen his political power by purging the armed forces and by putting loyal ofcers into high commands. Simultaneously, the opposition forces came together in the socalled Democratic Coordinatora loose coalition of eighteen parties and forty non-governmental organizations, business associations, and labor unionsand switched their strategy in November 2002 by calling a consultative, non-binding referendum that the National Electoral Council declared and supported in order to cut Chvezs mandate. When the government refused this demand, the Democratic Coordinator hardened its stance by calling a general strike in December 2002. Private business, trade unions, the media, the Church, and PVDSA workers and management all supported the strike as a means to force either a recall referendum on Chvezs government, his immediate resignation, or early elections. Thanks mainly to the loyalty of the armed forces, the government was able to withstand this assault. As a result, the clash between government and opposition turned into protracted trench warfare, which weakened the opposition even further. After two months, the general strike failed to achieve its political aims. Once again, Chvez succeeded in clinging to power, while business interests and the economy as a whole bore the brunt. Between 2000 and 2002 the countrys economic decline was impressive. While the GDP declined 8.9 percent in 2002, in 2003 it slumped 10 percent, with the ination rate soaring to 27 percent and the unemployment rate to 15 percent (A Tale of Two Years, Latin American Regional Report: Andean Group, January 6, 2004, 3). Exchange controls, price controls of production costs, the massive import of consumer goods for subsidized sale, and an absence of investment marked the economy. Factories that stopped producing were taken over by the military. Unexpectedly, the government survived a pervasive and seemingly permanent political crisis, but the price was mounting polarization and ungovernability. In hindsight, the political crisis led to the defeat of the oppositions strategy, putting at least a temporary end to the stalemate in Venezuelan politics in Chvezs favor. Chvez decided to burn bridges by escalating the pace of his revolution. After gaining control over the state oil company in February 2003, he embarked upon a revolutionary offensive against the media, private business, and the legal system. He has shown his determination to carry out the policies envisaged in the 49 decrees of December 2001 by detaining opposition political leaders, harassing the media, manipulating the judiciary, and dismissing PDVSA staff. Emulating Juan Velasco Alvarados military regime in Peru (1968 75), Chvezs government is apparently heading toward establishing a mixed economy, fostering the development of a powerful state sector and a sector of agrarian cooperatives under state control (Molina 2003).

Outsiders and Neopopulism

153

Following its bungled attempt at a general strike, the opposition found itself without a coherent strategy, in a state of confusion and internal struggle. The Democratic Coordinator eventually achieved consensus for a much-needed change of political strategy by accepting the mediation of the OAS and Jimmy Carters proposal to seek an electoral, democratic solution based on the constitution, which seemed to be the only way out. In the end, the opposition yielded to Chvezs proposal for solving the political crisis: the recall referendum allowed for in Article 72 of the Bolivarian Constitution. The government played a double game by simultaneously participating in the OAS-backed negotiations with the opposition and embarking on the radicalization of the Bolivarian Revolution. Its strategy has been to control potential sources of destabilization politically, to muzzle the media, to put state institutions and enterprises under tight political control, and to boost state planning in the agrarian sector through cooperatives based on the model of endogenous development launched in March 2003. After seven months of thwarting any electoral solution whatsoever, the government nally accepted a recall referendum in an OAS-sponsored agreementsigned with the opposition on May 29, 2003 which seemed to integrate both sides demands. The agreement reected Chvezs conviction that he had strengthened his power enough to win a recall referendum. Chvez achieved, in fact, an astonishing victory, with 59 percent of the votes. Four factors help to explain this outcome: the successful strategy of delaying the recall referendum, assuming that economic recovery and massive public spending for social programsthe misiones targeting the poorwould bolster Chvezs prospects for triumph; the tight control over state institutions, particularly over the National Electoral Council; the enduring support of Chvezs constituency to his government; and the oppositions inability to put forward a credible political alternative.22 Chvezs overwhelming victory in the August 2004 recall referendum engendered vast political consequences. First, by shifting the balance of power to his favor, his victory apparently put an end to the regimes instability and tempered the high degree of polarization that beset the country. Second, the Democratic Coordinator, which claimed that Chvez had committed fraud, broke up immediately after its defeat.23 Several parties like AD, Justice First, and Radical Cause withdrew from the oppositions umbrella. The business federation did the same and acknowledged the new situation by signaling their willingness to seek agreements with the government. Third, to further consolidate his grip on power, Chvez decided to step up the Bolivarian Revolution through various initiatives. A law passed in April 2004 permitted Chvez to gain total control over the Supreme Court by expanding its members from twenty to thirty-two justices. Moreover, a so-called social responsibility law, approved in January 2005, gave him the tools to regulate the media and to restrict freedom of

154

Mayorga

expression. Finally, the collapse of the opposition in the recall referendum set the stage for another Chvez tour de force in the state and municipal elections of October 2004, in which he won 20 of the 23 statesand 193 out of 332 municipalities, among them the most importantwhile the opposition could only retain the oil-rich state of Zulia. Chvez accomplished his objective to secure an almost total control over the state with great success, mainly by establishing through electoral and massdemagogic means a populist, plebiscitary democracy that subverted the underpinnings of constitutional democracy. But he was also successful in taking full advantage of the oppositions shortcomings and mistakes. Arguably the most powerful president in Venezuelas history, Chvez is determined to hold on to power until 2030 in order to create an endogenous socialist model, which is now his proclaimed goal. Consequently, he stepped up the Bolivarian Revolution by expropriating several agro-industrial farms and by establishing social production companies and workers co-management schemes (Cowing the Private Sector 2005). Moreover, he intends a new change of the constitution that would allow him an unlimited reelection as president. Given his farreaching control of the state and backed by windfall prots from oil exports, the long-term viability of his regime therefore looks brighter than ever. Yet it seems contingent upon three key factors: maintaining the support of his constituency by delivering on his promises, that is, by effectively reducing poverty and improving the livelihood of the poor; sustaining the welfare programs, which, in turn, hinge on a high and steady oil revenue; and last but not least, diversifying the economy to alleviate the dependence on oil exports.

The Waning of Neopopulist Parties and the Politicization of Indigenous Movements in Bolivia
Neopopulist outsiders in Bolivia constituted a different case because they did not become a threat to the democratic system.24 At the end of the 1980s, two neopopulist parties emerged. One was CONDEPA (Conciencia de Patria, or Conscience of the Fatherland), founded by Carlos Palenque, the owner of TV and radio networks in La Paz, and the other was UCS (Unidad Cvica Solidaridad, or Civic Solidarity Unity), a political movement built by Max Fernndez, then the most powerful stockholder of the countrys most important brewery. Both largely represented and channeled the demands of informal and marginalized sectors of the population. CONDEPAs constituency was the rural and migrant population of the department of La Paz, a population that was affected by adjustment policies and unrepresented by the established parties. Both CONDEPA and UCS, which had support from urban popular sectors, were neopopulist parties with authoritarian leaders who developed an anti-political discourse against traditional elites and parties. They blended a personalistic,

Outsiders and Neopopulism

155

clientelistic style of politics with especially in the case of the UCSa plebiscitary appeal to the masses and a commitment to market-oriented policies. Thus, the upsurge of neopopulist parties occurred in the contradictory context of reinforcement of democratic institutions and a relative delegitimization of governing parties arising from the negative social impact of adjustment policies. Despite their initial strong anti-systemic bias, these parties did not undermine the legitimacy of the democratic system. On the contrary, they became predominantly systemic parties that played a signicant institutional role by integrating their constituencies, participating in the management of a few important municipalities, and forging interparty agreements aimed at institutional reforms. Most importantly, they became coalition partners in the governments led by MNR and ADN (Accin Democrtica Nacionalista). They did not achieve presidential power, but their impact on the party system was important for a decade. UCS won a notable percentage of seats in the 1993 and 1997 elections (15.4 and 14.6 percent, respectively). It became a minor coalition partner in the governments of Snchez de Lozada (1993 97) and Banzer (19972001), holding the Ministries of Sustainable Development and Labor. CONDEPA also won a signicant share of seats in the 1989, 1993, and 1997 elections, and was a coalition partner in Banzers government for one year. Thus, although these parties emerged outside of and against the established party system, they soon became incorporated through the integrative capacities of the moderate multiparty system. Nevertheless, both parties have also been dual parties: on the one hand, given the neopopulist, plebiscitary bias of their democratic ideology, they have not been fully committed to democratic institutions; on the other, they have participated in electoral processes and became relevant political actors. Neopopulist elements affected their discourse, yet they aimed at political integration. The reasons for this political dualism were threefold. First, after 1985 a process of institutionalization strengthened the party system and transformed it into a moderate one. Party fragmentation and polarization were reduced, and a pattern of consensual politics superseded traditional confrontational politics. Second, in the framework of the prevailing constitution, consensual politics became the driving force for crafting coalition governments, which became the bedrock of Bolivias democratic system. Such a system provided strong incentives for cooperation among parties, so that even small parties could participate in building coalition governments. Third, the crisis of populism and of the state-led economy leading to the failure of the rst democratic government between 1982 and 1985 was overcome through successful structural adjustment policies, which for more than a decade and a half legitimated the democratic system (Mayorga 1995). The death of their leaders and, more importantly, their inefcient and corrupt participation in state administration weakened these personalistic political movements greatly. CONDEPA suffered a catastrophic defeat in the 1999 municipal elections and lost its stronghold in La Paz. This failure was the initial step toward

156

Mayorga
Table 5.1 Bolivia: Party Votes and Seats, Lower-Chamber Elections, June 30, 2002
Number of lower-chamber seats 47 35 27 31 6 5 5 1

Party MNR MAS NFR MIR MIP UCS AND PS LJ MCC CONDEPA
SOURCE :

% of votes 22.46 20.94 20.91 16.32 6.09 5.51 3.40 0.65 2.72 0.63 0.37

% of seats 29.93 22.29 17.19 19.74 3.82 3.18 3.18 0.63

Corte Nacional Electoral (National Electoral Court).

the destruction of the party in the June 2002 general election, when CONDEPA obtained only 0.36 percent of the vote. UCS also experienced a substantial loss of political power, gaining just 5.5 percent of the vote and ve seats in parliament in the same election (Table 5.1). This party also seems doomed to disappear. In Bolivia, neopopulist parties were powerful actors, but as political organizations they ended up being an ephemeral phenomenon that failed to alter the main features of a surprisingly stable party and government system. Nevertheless, due to the persistent problems of poverty and social exclusion, the potential source for neopopulist and anti-systemic actors has remained. The antisystemic pressures coming from neopopulist parties that emerged in the 1980s have withered away. At the same time, in recent years the widening of democracy in a context of economic depression and deepening social conict has resulted, paradoxically, in a new polarization of the political system, stemming mainly from the politicization of indigenous social movements. The taming and ultimately the demise of neopopulist parties left a vacuum that peasant and indigenous movements have lled, while at the same time constituting a different political trend and cleavage.25 This new paradox calls for explanation: while Bolivian democracy developed the capacity to include previously excluded social groups politically, the dynamics of the inclusion of indigenous movements spawned contradictions and tendencies that have put representative democracy in jeopardy. New ethnic-political cleavages and short-term problems have led to the politicization of indigenous social movements that evince some neopopulist tendencies. What are the factors explaining the politicization of the Chapare and Northern Altiplano indigenous movements, the most important to emerge since the democratic transition in 1982? Structural factors provide important background.

Outsiders and Neopopulism

157

Bolivia is going through an intense period of politicization of economic and social conicts, mainly due to economic decline, to the incapacity of the state to tackle problems, and to the party systems lingering inability to mediate and channel social interests and demands. This latter aspect is critical because it is linked, paradoxically, with an institutional reformthe 1994 Law of Popular Participation. The law created new political opportunities for social movements, helping to strengthen participation in local politics and thereby differentiate patterns of action and interest intermediation.26 The laws implementation fostered the erosion of traditional linkages between parties and social organizations, and particularly linkages with indigenous movements. Moreover, the law failed to relieve economic and social conicts or to alleviate the stagnation of economic development in rural areas and the worsening of living conditions. It also exacerbated ethnic cleavages instead of strengthening a national identity and consciousness. Short-term factors also lie at the root of the politicization of indigenous movements. The poor and marginalized sectors growing unrest and disenchantment with the performance of the economy and with the social consequences of limited government policies have put serious strains on the democratic system, after more than a decade of institution building with insufcient economic growth and poverty reduction. During the fragile Banzer government, political mismanagement, insufcient economic growth, and social conicts intensied the struggle of contentious social movements. Like other Latin American countries, since 1999 Bolivia has experienced deep economic stagnation and deterioration of living conditions. Real GDP grew only 0.4 percent in 1999 and only about 2 percent between 2000 and 2002, while informal self-employment amounted to about 65 percent of the labor force. Under pressure from entrenched U.S. interests, Banzer and Quirogas government (19972002) aggravated economic and social conicts by sticking to a radical policy of coca eradication and the prohibition of coca trade in the Chapare without taking into account socioeconomic and political costs. This policywhich was not backed by a comprehensive alternative agrarian development programadversely affected the economy, provoking a GDP decrease of about 5 percent. It had the concomitant political effect of sparking the mobilization and radicalization of the coca growers unions, led by Evo Morales. The conict spread to include other contentious groups such as labor and teachers unions and, in particular, informal-sector organizations, which expressed their discontent not only with specic policies but with the post-1985 economic model. The immediate background for the social and political crisis in April and September 2000 was the massive rejection of arbitrary increases in the water tariff in Cochabamba, where an international company had signed a contract with the municipality and the central government to provide water and energy to the city. Mass mobilizations and road blockages erupted in the so-called water war organized by peasant unions, teachers, labor unions, and informal-sector

158

Mayorga

organizations. Paralyzing key regions of the country, they issued a vast array of over a hundred demands that were basically economic in character, including the cancellation of the water contract in Cochabamba, salary increases, an end to the coca eradication policy, repeal of the agrarian reform law, and abolition of the privatization of state enterprises. This social crisis was a watershed for the political system. It revealed both a great accumulation of social conicts and a multiplicity of old and new actors that overwhelmed the countrys established political institutions. The crisis also demonstrated a great potential for mobilization against a weak national government and fragile regional state structures. Above all, it reected an increasing gap between political parties and social movements, due not mainly to a dramatic loss of political representation but to the governing parties inability to channel conicts and to carry out effective public policies (Mayorga 2005). Social and economic tensions began to erode political stability as never before by turning into political conicts. The most far-reaching result of the 2000 social and political crisis was to extend and enhance the politicization of indigenous movementsa process that the Law of Popular Participation and the 1995 and 1999 municipal elections had already fostered. The Law of Popular Participation provided incentives and opportunities for the political inclusion of locally based social organizations and leaders into municipal governments. The mixedmember electoral system established in 1994 helped bolster a locality-centered, constituency-serving political representation (Mayorga 2001). In the 1995 and 1999 municipal elections, the Chapare peasant unions running on the IU and MAS tickets had already won a signicant number of rural municipalities. In 1995, their candidates obtained 3.7 percent of all seats (60 seats total) in municipal councils, although not all belonged to the Chapare social movement. In the 1999 municipal elections, the MAS was the only indigenous party participating, and it obtained 4.7 percent of all seats (80 seats total) in municipal councils.27 From 1995 onward, the Chapare peasant unions became the driving force in the local municipal arena. This political advance at the local level extended to the national level with the 1997 presidential election. Morales and three leaders of the coca growers unions successfully ran on the IU ticket, gaining four seats in the Chamber of Deputies. While the government expected to cripple peasant coca unions by cracking down on the coca economy, the peasant unions had already changed the thrust of their resistance into a political struggle and institutional presence. Relying upon the coca unions as key networks and organizational resources, they achieved a high degree of politicization by linking their mobilizations against the U.S.-backed eradication policy with the struggle for national autonomy. They also interwove their defense of coca growing with the principles and values of Indian cultural identity (Alb 2003).28 The indigenous movements became identity-based political movements. They framed their political action in a double strategy favoring

Outsiders and Neopopulism

159

the preservation of the coca economy and the restoration of community traditions and rights, claiming jurisdictional control over indigenous territories. In this way, persistent historical ethnic cleavages were successfully transformed into political issues. They focused not only on ethnic political representation but on a radical restructuring of the state, a task that small, short-lived Indian parties had undertaken in vain at the outset of the democratic transition. Thus, the Indian movements developed a political struggle whose logic of action can be dened as the ethnic demarcation of limits (Eder 2001, 202). According to this classication, social and political conicts are constructed as ethnic-cultural conicts of identity, so that existing economic and social interests turn out to be mediated by identity conicts. In this way they are less negotiable, at least for the leaders of these movements. The extent of the hitherto unprecedented politicization of indigenous movements became apparent in the June 2002 general election, which triggered the most far-reaching political consequences since the emergence of a moderate multiparty system and the collapse of the workers movement. The two indigenous movements strengthened in the aftermath of the 2000 crisis and established themselves as political movements or parties by participating in these elections. Serious political mistakes by Congress and Quirogas government bolstered the MAS, which obtained a stunning triumph with 20.94 percent of the vote and thirty-ve seats in the lower chamber.29 According to the constitutional provision that established that Congress elects the president from among the two candidates with the largest number of votes if no candidate wins an outright majority of the electorate, Morales, as the leader of MAS, the party receiving the second-highest number of votes, was entitled to participate as a presidential candidate in the decisive congressional arena. He lost to Snchez de Lozada, the MNR candidate, who built a majority coalition in order to be elected president. Since Morales already had political experience and had served as a deputy in the legislature between 1997 and 2002, he cannot be considered an outsider. Moreover, he did not run in the 2002 election as an independent or with a new party label. The other indigenous movement participating was the MIP (Movimiento Indgena Pachakuti, or Pachakuti Indigenous Movement), a party founded in November 2001 by Aymara peasant leader Felipe Quispe. As executive secretary of the national peasant confederation, Quispe was a political outsider who also became a powerful leader by building a stronghold in the Aymara communities of some provinces in La Paz. His party obtained 6.09 percent of the vote and six lower-chamber seats (Table 5.1). The noteworthy electoral performance of both parties reected the dissatisfaction and distrust of rural and poor urban voters in the countrys Andean region vis--vis the established parties. The second consequence of major political signicance was the widening of political inclusion and participation. For the rst time in Bolivias democratic history, the June 2002 elections permitted the political inclusion and autonomous

160

Mayorga

political representation of indigenous social movements, thereby demonstrating the adaptability of the political system (Mayorga 2001, 2002). Because both Indian parties combined obtained 27.03 percent of the vote and 26 percent of the seats in Congress, the elections brought about a historical shift in political representation. They qualitatively advanced ethnic representation and conrmed the mutation of these indigenous movements into strong political actors. But do they have the capacity to transform themselves into stable political movements and, above all, into political parties? Given their meteoric rise and growth, the MAS and the MIP are more social and political movements than structured parties. Indeed, they are mainly social protest movements that articulate both the utopian and substantive demands of peasant sectors, which now face the challenge of building political parties in order to extend their social and political base. In this regard, the key question is whether these movements will be able to consolidate themselves as strictly ethnic-based parties or whether they will manage to develop as parties with a broad national constituency and political program. Since Bolivia is currently undergoing economic and urban changes that erode the structural, social, and economic foundations for the development of ethnicindigenous political parties, my hypothesis is that indigenous movements cannot become national political parties if they remain identity-based and bound by their Indian social constituencies.30 First, identity-based indigenous movements do not appeal to broad social groups in the increasingly mestizo and culturally diverse Bolivian society; and secondly, these movements are constrained by particularistic, ethnic-corporatist issues and have been unable hitherto to develop political programs and strategies involving relevant national issues. The MAS has taken on this challenge by championing the nationalization of natural resources, mainly of the huge natural gas resources, and an ethnic-based constituent assembly aimed at a radical restructuring of the state. This strategy, based predominantly on direct forms of contention, has apparently strengthened electoral support for the MAS in rural and urban indigenous sectors in the countrys western region. It has not, however, signicantly broadened the partys constituency to either middle-class sectors or to the eastern lowlands that claim regional autonomy and are hostile to the indigenous movement.31 The third consequence of indigenous politicization has been renewed polarization of the party system, demonstrating the extent to which ethnic-cultural cleavages were politicized. Since 1985, Bolivia has had a moderate multiparty system with three relevant parties at its axis: MNR, MIR (Movimiento de Izquierda Revolucionario, or Revolutionary Left Movement), and ADN. Yet the indigenous movements conquest of parliamentary power has polarized the party system. They are hostile to representative democracy and to the market economy, advocating instead a utopian model of ethnic identity-based, participatory democracy and the return to a state-led economy. Both the MAS and the MIP put forward a political program with a strong anti-systemic bias drawing on an

Outsiders and Neopopulism

161

ethnicist and fundamentalist ideology and upholding the utopian idea that the aylluthe traditional indigenous communityshould serve as the mainstay of a profound reorganization of the state and society. These political movements and the established parties have had deep normative and political disagreements on democratic principles and rules of the game. The MAS rejected outright the basic tenets of representative democracy and the market economy on the grounds that they are alien to Indian cultures. Accordingly, it attempted a radical, strongly anti-institutional strategy, dubbed siege strategy, aimed at blocking and destabilizing the government and the state by using both the tactics of mobilization and its veto power against government initiatives in Congress, which require a two-thirds majority. After the overthrow of Snchez de Lozadas government by an urban indigenous uprising in El Alto that claimed the right to nationalize gas resources, the MAS veered from confrontational politics to an electoral strategy and to supporting Mesas government, which assumed the radical goals of the MAS and the indigenous movement. The main reason for this change seemed to be that Morales aims at transforming his political movement into a national party capable of contending for state power. Because of the polarization of the party system and the rise of a radical opposition with veto power, three crucial tensions have arisen that affect the stability and future prospects of the democratic system. First, since the MAS and the MIP are basically social movements, their political practices respond to the logic of social protest and contention, applying anti-institutional tools of pressure on political institutions. Both parties are, in fact, extra-parliamentary movements that do not differentiate social and political styles of action and instead subordinate the logic of politics to the logic of social protest movements. Consequently, to the extent that indigenous movements entered the political system conceiving of themselves as anti-systemic social movements, the MAS and the MIP are dual political movements.32 As a political movement, the MAS faces the dilemma of continuing a politics of confrontational opposition to the democratic system or shifting to the role of a responsible opposition. To do the latter, it must turn into an institutionalized party that abides by the rules of the game and acts as an institutional catalyst of social and political change. With its weak leadership and locally restricted social base, the MIP does not appear to have the capacity to transform itself into an organized party. As a strong movement, the MAS, on the other hand, could probably rise to the challenge of developing national political structures, thereby overcoming its origins as a network of peasant unions. Second, the fundamentalist ideology of the peasant movements gives rise to contradictions with their own pragmatic and concrete demands. The identityoriented framing of the coca peasants struggles, for example, distorts the interests of broad sectors of the rural population who are more interested in economic integration as a means to improve their living conditions than in a utopian revival of the pre-Columbian past. Conicting ideological principles

162

Mayorga

and strategic guidelines that try to blend issue mobilizations with identity mobilizations lie at the root of the MASs political action. Third, the paradoxical conuence of political inclusion of indigenous movements and polarization of the party system has stirred up contradictory perspectives. The political inclusion coming out of the June 2002 election turned out to be a serious threat to the democratic system, as polarization and existing political strains between the traditional parties and the indigenous movements undermined democratic stability, governmental capacity, and state unity. Yet the current situation can also be seen as a historic opportunity to enhance the quality of democracy by developing an integrated, pluralistic, multicultural, and multiethnic democracy. The key challenges facing Bolivian democracy, therefore, are overcoming polarization, catastrophic stalemate, and state crisis, on the one hand, and achieving full political inclusion of indigenous political movements, on the othera task that will be more complex and demanding than the successful integration of neopopulist parties at the beginning of the 1990s.

Conclusion
Stressing common patterns and qualitative differences in Peru, Venezuela, and Bolivia, this chapter set out to explain from a comparative perspective the emergence of neopopulism and anti-system actors as an outcome of two main processes: the decomposition of party systems, and a deep crisis of the statein fact, a crisis of governability. The key causal factor for the breakdown of political parties was this crisis and not a crisis of political representation, which in any case was a result of the former. This line of reasoning agrees with the conclusion that Mainwaring draws in the nal chapter of this book: namely, that at the core of the contemporary crisis of representation in the Andes there is a crisis of democratic governability, associated with grave deciencies in state capacity. Political parties and leadership in Peru and Venezuela put the sustainability of democracy in jeopardy, becoming obstacles and problems as a consequence of their failure as governmental actors. A context of poor state performance, socioeconomic crisis, and political party decline brought about favorable conditions for the rise of outsiders. By virtue of the failure of democratic governability and the traditional parties loss of credibility, outsiders sprang up, claiming to be the only way out of the crisis. Outsiders politics in Peru and Venezuela had destructive outcomes. The most negative consequences for the democratic system were the concentration of power in the hands of high-handed leaders and the erosion of democratic institutions, leading to the breakdown of liberal-representative democracy and to its transformation into plebiscitary, semi-democratic regimes. By contrast, in Bolivia a moderate centripetal party system and a coalition-based government system absorbed neopopulist outsiders. Due to this political integration, they did not become a threat to democracy. Yet anti-systemic threats have sprung more recently from ethnic-fundamentalist and populist indigenous movements that

Outsiders and Neopopulism

163

became powerful veto players and achievedin the midst of an economic recession and mounting social conicta vital political space in Congress. This fact demonstrated the political systems capacity to respond positively to the sweeping politicization of these movements. Nevertheless, the conuence of the strengthening of ethnic-political representation, the crumbling of the moderate party system, and the emergence of an ethnicist populist alternative has prompted a dangerous political destabilization, raising another historical challenge for Bolivian democracy. Again at stake is its capacity to reestablish democratic governability by enhancing the political integration of indigenous movements and at the same time fostering an efcient state, capable of meeting the urgent needs of the poor and of advancing integral citizenship. Whether Bolivian democracy will be able to tackle this weighty task will to a great extent depend on overcoming polarization and stalemate, which in turn hinges on rebuilding political leadership and the party system as well as on restoring consensus-based politics.

Notes
I am very grateful to Scott Mainwaring, Eduardo Pizarro Leongmez, Richard Snyder, and James Mahoney for their insightful comments on previous versions of this chapter. 1. Ecuador also experienced an outsiders rise to power with the short-lived government of Colonel Lucio Gutirrez from January 2003 until his downfall in April 2005. 2. Stokes (2001, 142 48) also refers to reasons for social spending in support of the poor. 3. The political programs of Palenque and Fernandez in Bolivia, which proposed to restore state capitalism, were similar cases. 4. Chvezs difculties in carrying out his statist program seem to indicate that classical populism encounters structural limits and is therefore not viable. In the end, regardless of his ideological objectives, Chvezs populism seems doomed to resemble Fujimoris petty cash populism. I owe this observation to Eduardo Pizarro Leongmez. 5. By governability I mean the capacity of government to tackle fundamental problems of society through effective decision making and public policies. 6. I draw upon the idea that dening representation as acting in the interest of the represented, provides a minimal core conception (Przeworski et al. 1999, 2). 7. Tanaka highlights the critical dimension of governability, but he explains the crisis of Peruvian political parties mainly as a result of awed short-term political decisions (Tanaka 1998, 71 85). 8. As Weber stressed, charismatic leaders do not feel bound by institutional rules and constraints and instead demand faith and strict adherence from their followers (Weber 1964, 834). 9. These reections are based on Linzs (1978) ideas about the breakdown of democracy. 10. Based on this support and the contacts Fujimori had within the APRA, Planas (2000, 295 301) contends that Fujimori was neither an independent candidate nor an outsider.

164

Mayorga

11. Some analysts characterize Fujimoris regime as a delegative democracy. Other terms used to dene these regimes have been soft dictatorship, Caesarist democracy, and neopopulist democracy. 12. This closely ts Linzs denition of an authoritarian regime: authoritarianism with limited, not responsible, political pluralism, without elaborate and guiding ideology, without extensive nor intensive structured social support and political mobilization in which a leader or occasionally a small group exercise power within formally ill-dened limits but actually quite predictable ones (Linz 2000, 159). 13. The Peruvian experience offers stark evidence of the indispensability of parties as mechanisms of representation (Levitsky and Cameron 2003, 27). 14. Translated as takeover of pigeonholes, this phrase describes the gradual control achieved by Fujimori over the hard cores of military and economic power. 15. See Venezuelas Crisis: Towards the Endgame, The Economist, April 13 19, 2002. 16. According to a survey conducted by Arturo Keller in May 2001, only 17 percent of respondents supported Chvezs strategy of civilian-military government; 9 percent supported an exclusively military government, and 68 percent supported civilian government (El Universal, May 6, 2001). 17. This profound polarization was the backdrop for the coup dtat of April 11, 2002. In May 2001, Chvezs politics had already led to saber rattling within the armed forces and to the possibility that military adversaries might be inuenced by the radical pro-coup discourse of some political actors and leaders of the old Punto Fijo democracy. 18. In the OAS meeting in Quebec in May 2001, Chvez disagreed with the Democratic Charter and did not endorse the basic principle of representative democracy. He ratied this on his visit to Russia in May 2001, asserting that he believed in democracy, but not in the forms of democracy imposed on us. He also declared on his trip to various Asian countries: I am the second Latin American Castro. 19. Before his rise to power, Chvez declared in an interview that he nurtured the idea of a new-style militarism, which in his view had almost been established in Peru in 1968 75 (Blanco Muoz 1998, 73). 20. As General Medina stated: The vast majority realize that Chvez has taken his mandate for social change and used it for a revolution that takes the country down a road it doesnt want to go (A Tragic and Dangerous Stalemate, The Economist, October 12, 2002). 21. At the beginning of his mandate, polls indicated that Chvez had the support of 89 percent of the country. Three years later this support decreased to about one third (survey of Data Anlisis 2002). 22. Moises Naim stresses the oppositions miscalculation that Chvez could not last and its lack of a long-term strategy (Latin American Regional Report: Andean Group, September 7, 2004, 2). 23. Fraud accusations could not be substantiated. (see, e.g., Jennifer McCoy, What Really Happened in Venezuela? The Economist, September 2, 2004). 24. Linz (1994, 29) assessed Fernndez as a potential threat to democracy. 25. NFR (New Republican Force), a neopopulist party with regional roots in Cochabamba harking back to traditional populism, emerged in 1995 and participated in the June 2002 election, obtaining 20.91 percent of the vote and twenty-seven seats in Congress. 26. Contentious politics is produced when political opportunities broaden (Tarrow 2002, 23).

Outsiders and Neopopulism

165

27. Strangely enough, the MAS was the result of a leftist split from FSB, a rightist party, in 1988. Since his own party organization was not recognized by the National Electoral Court, Morales rented this label for the 1999 election. Thus, the MAS became the political umbrella for the six Chapare unions. 28. Morales became a political leader in the context of this struggle. He can be considered the offspring of U.S. policies and Banzers government. 29. Moraless electoral stance got a decisive push from the U.S. ambassador, Manuel Rocha, when he declared shortly before election day that Bolivia would risk the suspension of U.S. aid if its citizens voted for Morales. 30. Levitsky and Cameron (2003, 17) are skeptical about the chances for partybuilding in the Andean region, pointing out the persistence of exclusion from citizen rights and the enduring legacy of colonialism. 31. The MAS obtained the highest percentage of votes (18.4 percent) in the municipal elections in December 2004. According to a survey done in October 2005 on vote intention in the presidential elections scheduled for December 2005, Morales would get 28 percent and ex-president Quiroga 22 percent (La Razn, October 7, 2005). 32. There are analytical differences between a social movement, a political movement, and a political party. As Kitschelt contends, Collective organizations have distinctive proles of political practices (or action repertoires according to Tilly) which allow us in a given moment to associate them more with one of these ideal typessocial movements, interest groups, parties (Kitschelt 2001, 356).

References
Alb, Xavier. 2003. Pueblos Indios en la Poltica. La Paz: Plural. Ayo, Diego. 2000. Los Desafos de la Participacin Popular. La Paz: CEBEM. Blanco Muoz, Agustn. 1998. Habla el Comandante. Caracas: Fundacin Po Tamayo, Universidad Central de Venezuela. Comisin de la Verdad y Reconciliacin. 2003. Informe Final. Coppedge, Michael. 1994. Strong Parties and Lame Ducks: Presidential Partyarchy and Factionalism in Venezuela. Stanford, CA: Stanford University Press. . 2002. Venezuela: Popular Sovereignty versus Liberal Democracy. Working Paper, no. 294, Helen Kellogg Institute for International Studies, University of Notre Dame, Notre Dame, IN. Cotler, Julio. 2000. La gobernabilidad en el Per: Entre el autoritarismo y la democracia. In El fujimorismo: Ascenso y cada de un rgimen autoritario, ed. Julio Cotler and Romeo Grompone, 31136. Lima: IEP. Cowing the Private Sector. 2005. The Economist, September 15. Eder, Klaus. 2001. De los intereses a la identidad y de sta de vuelta a los intereses: Los lmites de la movimientizacin moderna. In Construccin de Europa: Democracia y globalizacin, vol. 1, ed. Ramn Mainz, 197218. Santiago de Compostela: Universidad de Santiago de Compostela. Ferrero Costa, Eduardo. 1993. Perus Presidential Coup. Journal of Democracy 1 ( January). Gmez Calcao, Luis. 2002. Los crculos bolivarianos: El mito de la unidad del pueblo. Unpublished paper, CENDES, Caracas.

166

Mayorga

Gmez Calcao, Luis, and Nelly Arenas. 2001. Modernizacin: Autoritaria o actualizacin del populismo? La transicin poltica en Venezuela. In Venezuela: Rupturas y Continuidades del Sistema Poltico (1999 2001), ed. Marisa Ramos, 37 68. Salamanca: Ediciones Universidad de Salamanca. Gonzlez de Olarte, Efrain. 1998. El neoliberalismo a la peruana: Economa poltica del ajuste estructural. Lima: IEP-CIES. Grompone, Romero. 2000. Al Da Siguiente: El fujimorismo como proyecto inconcluso de transformacin poltica y social. In El fujimorismo: Ascenso y cada de un rgimen Autoritario, ed. Julio Cotler and Romeo Grompone, 77178. Lima: IEP. Kitschelt, Herbert. 2001. Panoramas de intermediacin de intereses polticos: Movimientos sociales, grupos de inters y partidos polticos a comienzos del siglo XXI. In Construccin de Europa: Democracia y globalizacin, vol. 1, ed. Ramn Maiz, 361 86. Santiago de Compostela: Universidad de Santiago de Compostela. Knight, Alan. 1998. Populism and Neopopulism in Latin America. Journal of Latin American Studies 30, no. 2 (May): 223 48. Kornblith, Miriam. 1998. Venezuela en los 90: La crisis de la democracia. Caracas: IESA. Levitsky, Steven. 1999. Fujimori and Post-Party Politics in Peru. Journal of Democracy 10, no. 3 ( July): 78 92. Levistky, Steven, and Maxwell Cameron. 2003. Democracy without Parties? Political Parties and Regime Change in Fujimoris Peru. Latin American Politics and Society 45, no. 3 (Fall): 133. Linz, Juan J. 1978. Crisis, Breakdown, and Reequilibration. Baltimore: Johns Hopkins University Press. . 1994. Presidential or Parliamentary Democracy: Does It Make a Difference? In The Failure of Presidential Democracy, ed. Juan Linz and Arturo Valenzuela, 3 87. Baltimore: Johns Hopkins University Press. . 2000. Totalitarian and Authoritarian Regimes. Boulder, CO: Lynne Rienner. Lynch, Nicols. 1999. Una tragedia sin hroes: La derrota de los partidos y el origen de los independientes. Per, 1980 1992. Lima: Universidad Nacional Mayor de San Marcos. Mainwaring, Scott. 1999. The Surprising Resilience of Elected Governments. Journal of Democracy 10, no. 3: 10114. Manin, Bernard. 1997. The Principles of Representative Government. Cambridge: Cambridge University Press. Manrique, Miguel. 2001. La participacin poltica de las fuerzas armadas venezolanas en el sistema poltico, 1998 2001. In Las fuerzas armadas en la regin andina: No deliberantes o actores andinos? ed. Martn Tanaka, 305 36. Lima: Comisin Andina de Juristas. Martucelli, Danilo, and Maristella Svampa. 1992. La doble legitimidad del populismo. Pretextos (OctoberDecember): 63 79. Mayorga, Ren Antonio. 1995. Antipoltica y neopopulismo. La Paz: CEBEM. . 2001. The Mixed-Member Proportional System and Its Consequences in Bolivia. In Mixed-Member Electoral Systems: The Best of Both Worlds? ed. Matthew Shugart and Martin Wattenberg, 194 208. Oxford: Oxford University Press. . 2002. La metamorfosis del sistema de partidos. Opiniones y Anlisis: El Escenario Post-electoral en Bolivia 60 (OctoberDecember): 67112. . 2005. Bolivias Democracy at the Crossroads. In The Third Wave of Democratization in Latin America, ed. Scott Mainwaring and Frances Hagopian. Cambridge: Cambridge University Press. McClintock, Cynthia. 1989. Peru: Precarious Regimes, Authoritarian and Democratic. In Democracy in Developing Countries: Latin America, ed. Larry Diamond, Juan Linz, and Seymour Lipset, 335 86. Boulder, CO: Lynne Rienner.

Outsiders and Neopopulism

167

Molina, Jos Enrique. 2000. Comportamiento electoral en Venezuela, 1998 2000: Cambio y continuidad. Cuestiones Polticas 25: 27 66. . 2003. La revolucin bolivariana en Venezuela: Socialismo autoritario en un mar de contradicciones? Paper presented at the Latin American Studies Association Congress, Dallas, March 2729. Molina, Jos Enrique, and Carmen Prez. 1994. Venezuela: Un nuevo sistema de partidos. Cuestiones Polticas 13: 63 90. Naim, Moises. 2001. The Real Story Behind Venezuelas Woes. Journal of Democracy 12, no. 3 (April): 1731. Novaro, Marcos. 1996. Los populismos latinoamericanos transgurados. Nueva Sociedad 144: 90 103. Obando, Enrique. 2001. Las relaciones cvico-militares en el Per en la dcada del 90: Lecciones para el futuro. In Las fuerzas armadas en la regin andina: No deliberantes o actores polticos? ed. Martin Tanaka, 247304. Lima: Comisin Andina de Juristas. ODonnell, Guillermo. 1994. Delegative Democracy? Journal of Democracy 5, no. 1: 55 69. Pcaut, Daniel. 1987. Lordre et la violence. Paris: EHESS. Planas, Pedro. 2000. La democracia voltil. Lima: Fundacin Ebert. Przeworski, Adam, Susan Stokes, and Bernard Manin, eds. 1999. Democracy, Accountability, and Representation. Cambridge: Cambridge University Press. Rey, Juan Carlos. 1991. La democracia venezolana y la crisis del sistema populista de conciliacin. Revista de Estudios Polticos 74: 533 78. Reyna, Carlos. 2000. Un rgimen de espritu militar. In Per 2000: Un triunfo sin democracia, 133 59. Lima: Comisin Andina de Juristas. Roberts, Kenneth M. 1995. Neo-liberalism and the Transformation of Populism in Latin America. World Politics 48, no. 1 (October): 82 116. Rospigliosi, Fernando. 1995. La amenaza de la fujimorizacin: Gobernabilidad en condiciones adversas, Per y los pases andinos. In Partidos y clase poltica en amrica latina en los 90, ed. Sonia Picado and Daniel Zovatto, 31136. San Jos: IIDH-CAPEL. Sartori, Giovanni. 1999. Elementos de teora poltica. Madrid: Alianza Editorial. Schedler, Andreas. 1994. Antipolitical Opposition: A Framework for Comparative Analysis. Paper presented at the Vienna Dialogue for Democracy on The Politics of Antipolitics, Institute for Advanced Studies, Vienna, July 710, 1994. Stokes, Susan. 2001. Mandates and Democracy: Neoliberalism by Surprise in Latin America. Cambridge: Cambridge University Press. Tanaka, Martn. 1998. Los espejismos de la democracia: El colapso del sistema de partidos en el Per. Lima: IEP. Tarrow, Sidney. 2002. Power in Movement: Social Movements and Contentious Politics. Cambridge: Cambridge University Press. Torre, Carlos de la. 2000. Populist Seduction in Latin America. Athens: Ohio State University. Trinkunas, Harold. 2002. The Crisis in Venezuelan Civil-Military Relations: From Punto Fijo to the Fifth Republic. Latin American Research Review 37, no.1 ( January): 4176. Weber, Max. 1964. Wirtschaft und Gesellschaft. Vol. 2. Kln: Kiepenhauer and Witsch. Weyland, Kurt. 1996. Neopopulism and Neo-Liberalism in Latin America: Unexpected Afnities. Studies in Comparative International Development 31 (Fall): 3 31. . 2001. Clarifying a Contested Concept: Populism in the Study of Latin American Politics. Comparative Politics 34, no. 4 (October): 122.

Part II DECENTRALIZATION, LEGISLATURES, AND DEMOCRATIC REPRESENTATION

6 Decentralized Politics and Political Outcomes in the Andes

Kathleen ONeill

n the 1980s and 1990s, the Andean countries, along with most of the rest of their Latin American neighborsand, indeed, a great deal of the developed and developing world experienced a series of decentralizing reforms. These reforms generated a great deal of optimism, based on accumulated scholarship suggesting that decentralization would yield signicant scal and political benets. Economic theories indicated that decentralization (or scal federalism)1 would increase efciency (Tiebout 1956; Musgrave 1959; Rubinfeld 1987; Bird 1990; Oates 1998).2 On the political side, decentralization was linked to democratic consolidation and improved democratic practice (Diamond 1999; Huther and Shah 1998; Fox 1994; Dahl 1971).3 Nearly ten years after the most recent of the major reforms in the region, assessments of decentralizations effects have delivered a mixed verdict on these predictions. This is particularly true on the economic side, where some studies have found that decentralization increases the size of government (Stein 1998), impedes scal restraint (Alesina, Carrasquilla, and Echavarra 2002; Rodden 2002; Alesina et al. 1999), or increases corruption (Treisman 1999; Tanzi 1994; but see also Fisman and Gatti 2000). Assessments of decentralizations effects on democracy and other political outcomes have lagged well behind economic assessments.4 Instead, most political analyses have focused on the causes of decentralization (ONeill 2005; ONeill 2003; Garman, Haggard, and Willis 2001; Willis, Garman, and Haggard 1999; Barr 2001; Grindle 2000). Those studies that do explore the democratic dividends of decentralization tend to discuss the relationship between decentralization and political outcomes as one of many parts of their work. In addition, they tend to focus on particular cases that exhibit exceptionally good (Campbell 2003; Tendler 1997) or bad (Eisenstadt 1999) results, painting an incomplete picture of decentralizations effects. In short, there is little focused, systematic analysis of decentralizations effects on political outcomes.

172

ONeill

Exploring decentralization in the Andes is particularly interesting, as the same period in which this supposedly democracy-enhancing reform has blossomed has coincided with a crisis of democratic representation. This suggests two alternative interpretations: either decentralization has contributed to this crisis (through the mechanisms Scott Mainwaring labels the paradox of democratic representation in this books Conclusioni.e., by opening up new avenues of political contestation through which competing parties criticize each other in order to attract votes and, as a byproduct, fostering a pervasive sense of state crisis); or decentralization, though contributing positively to democratic representation, has served only to dampen what would have been a far worse crisis of democratic representation had decentralization not been adopted. This chapter investigates the relationship between decentralizing reforms, political parties, and political representation in Bolivia, Colombia, Ecuador, Peru, and Venezuela. A great deal of scholarship paints a rosy picture of decentralizations likely contributions to improved democracy. Diamond and Tsalik (1999) summarize ve mechanisms through which decentralization should improve democratic practice:
First, it helps to develop democratic values and skills among citizens. Second, it increases accountability and responsiveness to local interests and concerns. Third, it provides additional channels of access to power for historically marginalized groups and thus improves the representativeness of democracy. Fourth, it enhances checks and balances vis-vis power at the center. Fifth, it provides opportunities . . . for parties and factions in opposition at the center to exercise some measure of political power. (12122)

This chapter looks at the ways in which decentralization has affected political representation either through, or in spite of, political parties, examining many of the mechanisms identied by Diamond and Tsalik. In particular, I explore the extent to which decentralization has affected participation and public opinion (both related to the rst mechanism outlined above). Next, I look at the ways political parties have responded to opportunities at subnational levels of government and the different extent to which they have been successful (this touches on mechanisms three and ve outlined above). Finally, I explore a consequence of decentralization not predicted above: the way in which decentralization has changed the career paths of politicians at the national level and has, as a result, affected political party organizations.5 These same three mechanismsvoter turnout, party volatility across elections, and the growth of outsider candidates for president are three of the dening characteristics of a crisis of democratic representation described in this books Introduction. To the extent that decentralization increases turnout, dampens party volatility across elections, or contributes to fewer independent presidential candidacies, it might be said to enhance representation. As this analysis will show, decentralizations relationship to these categories is complex. I nd that the relationship between decentralization and political

Decentralized Politics and Political Outcomes

173

representation varies signicantly across countries and over time within countries. What is more, it varies quite signicantly within countries at any particular point in time. Instead of uniformly improving political representation, in some areas decentralization has empowered capable, enthusiastic reformers,6 while in others it has increased the power of landed oligarchs or, worse still, armed enemies of the state. In every Andean country, one can point to several subnational units in which political representation has been unambiguously improved by decentralization, and to at least a handful of units that have been made unambiguously worse off. The effects of decentralization on political representation are experienced locally. Given this wide range of variation (along the three axes of country, subnational unit, and time), and the multiple ways that decentralization is hypothesized to affect political outcomes, this assessment can only scratch the surface. In this chapter, I explore variation at the level of countries and over time, leaving the majority of subnational variation unexamined. The chapter is organized into four sections. The rst summarizes the regions decentralizing reforms, highlighting the unique origins and features of each countrys particular experience. The next three sections analyze the ways in which decentralization has affected political representation across the region, examining decentralizations effects on political representation from three vantage points. The rst of these looks at political participation through the eyes of citizens. Looking at rates of electoral participation in subnational contests and at scattered public opinion data, this section nds that enthusiasm for decentralization is high in all countries. The second section explores the link between decentralization and political representation from the perspective of political parties. How have traditional and emerging parties taken advantage of the new local and regional arenas of power that have been created and expanded by decentralizing reforms? Here, variation exists not just at the national level but at the subnational level within countries and party by party. In most cases, decentralization has led to a decline in the ability of traditional parties to control subnational positions. However, in some cases, the decline in traditional party support at the subnational level has lagged behind the decline in their support at the national level. In other cases, parties seem to have adapted or used institutional rules to maintain their strong showing at subnational levels. A third section explores the consequences of decentralization from the politicians point of view. Here I examine the impact of decentralization on the career paths of presidential candidates. As subnational governments have grown more powerful, ambitious politicians have increasingly sought out elected subnational positions. Using their record of subnational rule as a signicant credential, many former mayors and governors have launched campaigns for national ofce, including the presidency. This represents a major departure from more traditional career paths and may represent an unintended consequence of decentralization for political parties: a threat to their ability to

174

ONeill

maintain a monopoly on political advancement within the system. However, this development may also provide an opportunity for traditional parties, if this increased competition within parties revitalizes their connection to their constituents.

Decentralization in the Andes


Unfortunately, a clear and universally accepted denition of decentralization does not exist.7 It is thus critically important to dene decentralization as I will be using it in this chapter: decentralization is a reform (or series of reforms) that increases political power through the election of subnational ofcials where they have been previously appointed and that also accords some level of autonomous scal power to those elected ofcials.8 For a very rough sense of comparative levels of scal decentralization across the Andes, see Table 6.1. To speak of decentralization in the Andes requires careful attention to the variety of experiences even within this subregion of Latin America. Each country in the region decentralized at a different time and through a different process. As a preface to my exploration of decentralizations effects on democratic representation in Bolivia, Colombia, Ecuador, Peru, and Venezuela, this chapter begins by briey surveying these experiences. For ease of exposition, this section proceeds chronologically through the cases (see Table 6.2 for dates of key reforms in each country).
Table 6.1 Expenditure Decentralization in Latin America
Country Subnational spending as a % of total spending

Argentina Brazil Colombia Bolivia Mexico Venezuela Uruguay Chile Peru Ecuador Paraguay Latin American averagea OECD average

49.3 45.6 39.0 26.7 25.4 19.6 14.2 13.6 10.5 7.5 6.2 14.6 34.9

source: These gures are based on 1995 data and are taken from IDB (1997, 157). note: Countries in italics are Andean countries. a Not all Latin American countries included in the average are listed in the table.

Decentralized Politics and Political Outcomes

175

Unlike neighbors such as Brazil, Argentina, or Mexico, the Andean countries do not have a long history of federalism (Venezuelas nominal federal structure notwithstanding). Decentralization in these ve cases thus represented a major change from a history of rather centralized rule. Peru began to decentralize before its most recent period of democracy, during Fernando Belande Terrys rst presidential administration (1963 68). The election of municipal governments was reintroduced in the constitutional assembly (1978), and the election of mayors began once again, in 1980, when Belande was re-elected to the presidency. Despite the fact that subnational ofcials were elected in Peru, these ofcials suffered from a severe shortfall in resources. During Belandes second term (1980 85), he strengthened their base of scal resources somewhat, creating a formulaic system that transferred slightly more (but still largely insufcient) resources to local governments. Decentralization to local governments changed little during Alan Garcas turbulent presidency (1985 90). As it became increasingly clear that Garcas Alianza Popular Revolucionaria Americana (American Popular Revolutionary Party, or APRA) would not be a strong competitor in 1990 national elections, his government began to push (successfully) for the election of regional governors (Thedieck and Buller 1995), and the rst election of this type took place in 1990. While APRA did abysmally in national polls in 1990, its candidates won regional executive elections in all but one province (Buller 1993, 151). When Fujimori dismissed the legislature and other elected ofcials of the government in his 1992 autogolpe, these regional ofcials were also dismissed and the election of regional ofcials ceased. Fujimori did, however, allow new local elections in 1993. Instead of increasing funding to municipal governments, Fujimori introduced Decree Law 776 in 1993; this law drastically cut back transfers to local
Table 6.2 Popular Election of Subnational Executives
Country Year Summary

Bolivia Colombia Ecuador Peru

1995 1988 1991 1980 1980 1990 1992 1992

Venezuela
a b c

First countrywide popular election of mayors First popular election of mayors First popular election of governors First popular election of mayors and prefects in new democratic periodb First popular election of mayors in new democratic periodc First popular election of governors Postponement of mayoral elections (1 year); end of the popular election of governorsd First popular election of mayors and governors

Mayors had been elected in large municipalities before 1995. Mayors and prefects had been elected prior to the authoritarian period (1972 79) Mayors were rst elected in 1963, but elections were later interrupted by authoritarianism (1968 79). d President Toledo promised gubernatorial elections in 2002.

176

ONeill

governments. In fact, municipal budgets shrank by more than 75 percent in the year following this reform (Kay 1995). In place of these funds, Fujimori created a local development program called FONCODES that provided direct central government funding for particular local projects; he maintained a great deal of discretion in determining the recipients of this money. In addition, he eased the rules for seeking local election, allowing a ood of independent candidates that made it hard for either the remnants of old political parties or nascent political parties to build a base through these contests. In this sense, many scholars point to Fujimoris Peru as an example of recentralization (Dammert Ego Aguire 1999; Delgado Silva 1995). Currently, local government in Peru continues to receive very little funding; however, support for democratic election of regional ofcials appears to be growing (Tanaka 2002, this volume). One of the earliest decentralizers in the region, and one of those that has taken the greatest strides toward extending scal and political devolution, is Colombia. Colombia decentralized power to subnational governments in two stages: during the presidency of Belisario Betancur (1982 86) and during its constitutional assembly (1990 91). Betancurs government initiated laws that introduced the direct election of local mayors (the rst elections were held in 1988) and that provided local governments with a signicant percentage of the central governments scal resources, to be allocated through a complex formula that included such items as population, percentage of the population whose basic needs went unsatised, and scal effort in raising local resources. A second round of decentralization during Colombias constitutional assembly in 1990 91 introduced the popular election of regional governors, created automatic transfers to regional governments, and included provisions strengthening local governments still further. There is no question that, for better or worse, Colombia is the most decentralized government in the Andean region. After Colombias decentralizing moves in the early 1980s, the next attempt at decentralization occurred in Venezuela, beginning with legislative acts passed in 1989. These acts allowed for the popular election of mayors and governors and guaranteed these governments both nancial resources and new responsibilities. The outline of the decentralization process in Venezuela differed signicantly from that in Colombia andas we shall seein Bolivia; instead of a major change toward increased funding, Venezuelas decentralization reforms followed a more gradualist approach. Each regional government seeking more autonomy has to petition the legislature for its approval of increased responsibilities and the increased funding to go with them (de la Cruz 1992). This purportedly allows the legislature to ensure that the subnational unit has the proper capacity for carrying out such important responsibilities as local educational policy; in practice, it also gives the legislature enormous power over the pace and extent of decentralization (Kraemer 1999).

Decentralized Politics and Political Outcomes

177

In 1994, Bolivia began an ambitious experiment in decentralizing reform. Bolivias Law of Popular Participation (LPP) created 311 municipalities, largely in communities where no ofcial local government had ever existed. It allowed for the election of local councils. If any party in the municipal race achieved greater than 50 percent of the vote, its mayoral candidate was elected directly to become the mayor; otherwise, municipal councils chose one of their members to be the mayor. In addition, this law guaranteed that 20 percent of the national scal intake would be divided among the municipal governments, on a purely per capita basis. One of the innovative features of Bolivias decentralization is its ofcial recognition of grassroots civil organizations (neighborhood associations, peasant unions, and indigenous groups). The LPP encourages members of these groups to form civic oversight committees (Comits de Vigilancia) to oversee the allocation of funds within the municipality. These organizations can petition the central government to freeze scal transfers to the municipality pending a central government inquiry into municipal management. Another feature calculated to increase the accountability of mayors to their constituents allows the municipal council to remove mayors from their position at one-year intervals and to replace the mayor with another of their number.9 Regional governors remain appointed by the president. Finally, the laggard in the region on decentralizing reform has been Ecuador. Despite the election of local and provincial government ofcials, these ofcials receive very little money, and much of the money that does arrive is subject to the discretion of the central government. Extensive debates over decentralization have raged in Ecuador in recent years, peaking in 1997 and 1998 with the Ley de 15%, which promised a transfer of 15 percent of the central governments funds to the local and provincial governments. Implementation of this law languished until after it was reformed in Ecuadors 1998 Constitution. The 1998 Constitution changed the basis on which central government transfers (including those transferred under the Law of 15%) would be distributed. The Law of 15% had called for 10 percent of the funds to be distributed in equal parts to each canton: 50 percent according to an index of basic needs unsatised among the population; and 40 percent based on the population of each canton. The new constitution dropped the rst criterion and added three others: nancial capacity, improvement in level of life betterment, and administrative efciency. Finally, the Law of 15% was criticized for transferring funds without responsibilities, which led to a change in the wording of the constitution. Article 226 now reads, There cannot be a transfer of responsibilities without a transfer of equivalent resources, nor a transfer of resources without (corresponding) responsibilities. The decentralization law also allows for the decentralization of funds and responsibilities to occur through a bargaining process between the central and subnational governments (similar to Venezuelas system). In addition, the

178

ONeill

constitution also allows for the direct election of mayors in all municipalities (previously this had occurred only in provincial capitals and in cities with populations greater than 50,000). The political innovations called for in the new constitution have been implemented, but the nancial aspects have lagged behind as the country grapples with the worst nancial crisis in its history. In 1999, although scal decentralization had only reached about 9 percent 10 (roughly half of the 15 percent required by law), critics argued that the government simply could not afford to transfer these funds to subnational levels during a period of such national crisis. The economic problems contributed not only to the removal of Ecuadors president, but to the rst ever default on Brady bonds and to the adoption of the U.S. dollar as the nations currency. In the governments attempts to deal with these problems, decentralization has been set aside as a major issue.11 In addition, the focus of the decentralization debate has shifted away from the decentralization of political and economic resources to municipal governments. Instead, the debate has been refocused on whether or not the government should grant autonomy to ethnolinguistic groups in specically dened indigenous territories and to increasing the power of the provinces (Cameron 2000). Clearly, the Andean reforms reect a wide range of experience with decentralization (see Table 6.1, above, for a sense of how the Andean countries compare to other Latin American countriesand to each otherin terms of how much subnational governments spend relative to what the national governments spend). Colombia has gone the furthest toward giving both mayors and governors a democratic basis and a signicant ow of scal resources, allocated according to transparent criteria. In addition, each level of government has its own sources for raising revenue and the criteria for intergovernmental transfers give subnational governments incentives to raise their own resources in addition to receiving transfers from above. Bolivia has also taken major steps toward decentralizing its government, but has focused its efforts on mayors and local councils, stopping short of empowering regional governments. Its innovative features include several provisions for incorporating civil society into local government decision making. Both Venezuela and Ecuador have recently taken signicant steps toward decentralization, but have adopted a more gradualist approach than either Colombia or Bolivia, allowing individual subnational governments the opportunity to petition for increased resources and responsibilities. On the political side, Venezuela (especially) and Ecuador (to a lesser extent) have made some signicant strides toward increasing the electoral accountability of subnational ofcials to their communities, even where they do not control signicant scal resources. Perus experience is an important counterpoint to the stories of its neighbors: here decentralization proceeded tentatively during its early years of democracy, but many decentralizing reforms were either nullied or signicantly rolled

Decentralized Politics and Political Outcomes

179

back during Fujimoris presidency, leaving elected local governments with few scal resources. With such a wide range of experiences, the Andean region provides an exciting set of cases for exploring the effects of decentralization on democratic representation. The next section begins this exploration by investigating patterns of participation in local elections.

Participation
One of the clearest ways that decentralization contributes to democratic practice is by creating a wider variety of opportunities for citizens to participate in the democratic process. Voter turnout is one way to judge the involvement of citizens in decentralized contests. Rather than compare the turnout of citizens across the countries in the region, this section compares local turnout to national turnout within each country, in an attempt to control for various national laws and cultural factors that might affect an individuals interest in casting a ballot. For example, in Venezuela, where voting is mandatory, one would expect higher turnout than in Colombia, where it is not. The point of comparing national with subnational turnout within a country is to control for these kinds of factors. The evidence that follows suggests that Andean citizens are quite excited about the opportunity to participate in a greater range of democratic contests. Throughout the region, turnout in local contests relative to national contests compares favorably with more established democracies, suggesting that individuals are interested in their local governments and eager to participate in choosing local ofcials. Of course, there is important variation within the region; in some cases, after an initial increase, turnout in subnational contests declines over time relative to turnout in national contests. The gap between turnout in subnational and national contests is also affected by the timing of elections and the actual power of subnational ofcials across countries. One of the major political development arguments for decentralization is that the opportunity for citizens to participate in elections and government at the local level gives them an education in democracy more generally. Perhaps this will have the effect of conditioning citizens to expect the delays and compromises inherent in democratic debate and decision making at all levels of government, making democracy more robust in the face of economic downturns and more resistant to populism and demagoguery. While not fully exploring this claim, looking at electoral turnout gures does show a strong tendency for citizens who are given the chance to participate in local elections to take that opportunity. Before delving into the Andean electoral turnout results, I want to include some data on turnout in local versus national elections in the United States. Although the United States is widely noted as a case with low electoral turnout in national elections within the developed world, its turnout rates in local contests are even more abysmal. Most scholars of elections and voting behavior have re-

180

ONeill

marked on a trend in which voters turn out in much smaller numbers for elections at more local levels of government. Looking at the turnout rates in the United States, there appears to be a linear decline in participation rates as the power invested in the ofce for which elections are being held declines. In the 1996 presidential elections, national voter turnout in the United States reached 48.91 percent, while for congressional elections, the number reached 45.6 percent. Looking at one county,12 gubernatorial turnout in 1996 reached just shy of 45 percent, while voting in local elections reached only 25 percent in a year when the mayor was being elected (1997), and did not even reach 15 percent in 1999, when the mayors position was not open to election. This gap between voter turnout in national and subnational elections in the United States is illustrative of a trend throughout the developed democracies. In the Andes, local electoral turnout also lags behind turnout in national elections, but the divergence is not nearly as high as it is in the United States. Looking at Figure 6.1, which plots national and subnational turnout in millions of valid votes in each of the ve Andean countries, the Colombian case immediately catches the eye. Here local elections stretch back to 1988, with national elections continuing uninterrupted throughout this period.13 As Figure 6.1 demonstrates, local turnout in 1994, a year in which both national and local contests occurred, was signicantly higher in the national polling: local turnout reached only 63 percent of national turnout. What is more striking, however, are the sharp increases in local turnout in 1997 and 2000, when turnout in local contests was more than double the turnout in national contests two years earlier. In the 2000 election, a ballot for peace was held concurrently with the municipal elections; a great deal of voter turnout must therefore be seen as a response to this peace initiative rather than as an extraordinary increase in voter enthusiasm for local contests. Still, the increased turnout in 1997a year when local elections occurred independently of national elections (which should depress turnout) does signal a strong level of interest in local politics by the electorate. Bolivias local turnout follows a very different path than Colombias: here local turnout skyrockets and then declines signicantly. Prior to the 1994 decentralizing reforms, Bolivia held local contests only in its largest cities and their suburban neighborhoods. As a consequence, turnout in these contests was restricted. In 1995, the rst nationwide round of local contests took place in 311 municipalities. Turnout was extraordinarily high, with the total number of voters in the 1995 local elections exceeding the total number of voters who turned out in 1993 for national legislative and presidential elections. After this peak in the rst local elections, turnout declined in Bolivias 1999 elections. Most troubling, the turnout in 1999 amounted to less than the number of voters who took part in the 1991 local contests that were restricted to the largest cities and towns in the country. What is more, the 1991 elections took place in the context of a system that devolved very few resources to local governments, in

Figure 6.1

Turnout in millions of voters, by level


Colombia 10 20 8 6 4 2 0 1980 Peru 20 15 10 5 0 2000 National 1980 1985 Municipal 1990 Year Local 1995 2000 President 20 15 10 5 0 1980 1985 Governor 1990 Year 1995 2000 1985 1990 Year 1995 2000 1980 1985 1990 Year Venezuela 1995 2000 15 10 5 0 2000 Chile

Bolivia

20

15

10

1980

1985

1990 Year

1995

Ecuador

20

15

10

1980

1985

1990 Year

1995

182

ONeill

contrast to the much better funded system in place in 1999. Looking for an answer, one might concentrate on the institutional mechanisms governing local politics in Bolivia. Declining voter turnout may reect voter frustration with the indirect nature of municipal elections for mayors or the furious pace at which mayors were overturned and replaced by municipal councils between 1995 and 1999 due to the censura procedures embedded in the earlier reforms. A third hypothesis might link declines in turnout to the restrictions that have kept independent candidatesthose not afliated with a recognized political partyfrom running in local contests.14 In short, the result is overdetermined; certainly greater investigation would be necessary to understand this drastic change in turnout. A third case to explore is the Venezuelan system. Since the greatest degree of decentralization conferred resources on regional rather than local governments, I compare turnout in national elections for president with turnout for gubernatorial contests. The striking feature of the Venezuelan case is the huge increase in turnout between 1995 and 2000. While the 1995 gubernatorial elections met with extremely low turnout (only 32 percent of 1993s national turnout), turnout increased dramatically in the 2000 elections, when subnational turnout nearly equaled national turnout. Given the gradualist nature of Venezuelas decentralization reform, this increase in turnout could correspond to increases in real resources to regional governments. To investigate this hypothesis, one could look at the degree of decentralization granted to each regional government and the increase in electoral turnout within that region between 1995 and 2000. Of course, as we will see in both the Ecuadorian and Peruvian cases, concurrent elections generally lead to much higher turnout. In Venezuela, the 2000 elections were particularly exciting and elicited a high degree of turnout, making our task of parsing out the differential contribution of decentralization to turnout impossible. To determine what is driving turnout in Venezuela, we would have to observe at least one more non-concurrent set of regional elections; in addition, recent constitutional changes have changed electoral rules in new ways that may also be contributing to changes in voter turnout at both national and regional levels. It seems too soon to draw any conclusions in the Venezuelan case. Ecuador presents yet another pattern of national and municipal voter mobilization. When Ecuador returned to democratic rule in 1980, turnout in both national and local contests barely differed (municipal turnout was 95 percent of national turnout); however, as time went by, turnout increased at a higher rate for national elections than for municipal elections. While I do not have the data to determine whether municipal turnout increased after 1998, when electoral rules changed to increase the number of local ofcials being popularly elected, the period 1980 95 suggests a trajectory of growing divergence between national and municipal electoral turnout. National turnout climbed at a faster rate than municipal participation. Given the patterns we have seen in Perus neighbors, we should perhaps expect to see very low turnout in local elections in Peru, since these local governments

Decentralized Politics and Political Outcomes

183

have little scal power. Instead, Figure 6.1 shows strong turnout at the local level. When national and municipal elections are concurrent, local turnout is signicantly higher than in other years; still, in 1998, local turnout is considerably higher than national turnout in 1995 and it is 85 percent of national turnout in 2000. Tanaka (2002) persuasively argues that local elections became an outlet for venting opposition to Fujimori in the later years of his presidency. Comparing across these ve countries, there does not appear to be a strong and direct relationship between the strength of municipal governments and the rate at which citizens turn out to vote for subnational ofcials. Based on the most recent rounds of elections, turnout in municipal elections reached the following percent of turnout in national elections in each country: 95 percent in Colombia,15 88 percent in Venezuela,16 85 percent in Peru,17 74 percent in Ecuador,18 and a dismal 54 percent in Bolivia.19 Extrapolating from the U.S. data noted above, the comparable local/national gure for the United States would be just over 51 percent, lower than the gures in any of the ve Andean countries. It is fascinating, then, that subnational voter turnout as a percentage of turnout in the most proximate national elections in each of the Andean countries is much higher than it is in the United States, even though subnational governments in the United States control far greater resources. Within the Andean cases, however, there does appear to be some rough correlation between high voter turnout and the level of power controlled by subnational ofcials. A rough classication of the relative strength of decentralization reforms across the ve cases would probably rank the cases as follows: Colombia, Bolivia, Venezuela, Peru, and Ecuador. The real surprises here, then, are Bolivia (for its low turnout) and Peru (for its high turnout). A static view fails to capture the whole story, however. In Bolivia, local turnout skyrocketed and then declined; in Colombia, local turnout has been steadily and sharply rising; in Ecuador, local turnout is slowly climbing; in Venezuela, it is rising sharply but this may reect a simple difference between concurrent and non-concurrent elections; and in Peru, local turnout is declining relative to national turnout. This pattern, except for the Bolivian anomaly, seems to t more clearly with the scope and pace of decentralizing reforms across the ve countries. This suggests that citizens are quite savvy in responding to devolution of power; where subnational governments are not very strong, citizens do not exert themselves to vote at the same rate as do voters electing more powerful subnational ofcials.

Political Parties, Local Elections, and Public Opinion


One of the hypothesized benets of decentralization is that it might make government more responsive to its citizens. In particular, decentralization should give political parties and leaders incentives to develop a closer relationship with their constituencies, thus reversing the crisis of representation. Alternatively, decentralization may create the conditions for new parties to form at subnational

184

ONeill

levels, gain credibility through good government performance, and then launch national electoral campaigns. This section will examine these hypotheses using public opinion data and election results by party. The two points investigated here are: First, how do citizens view their local elected ofcials and how does this compare with their views of national politicians? Second, how well have traditional parties fared in local contests over time? While turnout in local contests may be considered one measure of how citizens feel about their local politicians, there are many other factors that determine whether or not an individual might vote. One might vote to get rid of corrupt ofcials; because it is a civic duty; in order to secure a position in the local bureaucracy; or because one happens to be walking by the polling place and nds that all of ones friends are there voting. To get a more direct feel for how citizens view their local governments and the political parties who contest elections, I turn to some scattered public opinion data. The rst thing to note about public opinion toward political parties in Latin America is that it is exceptionally low. If parties pushed through decentralizing reforms in order to increase the popularity of political parties, then it has been a resounding failure. A 1998 poll conducted by the Wall Street Journal of the Americas reports that the percentage of respondents who said they had some or much trust in political parties was: 10 percent in Bolivia, 15 percent in Colombia, 6 percent in Ecuador, 19 percent in Peru, and 6 percent in Venezuela. Compare this with 37 percent in the United States and 22 percent in Costa Rica. Political parties ranked well below the press, the armed forces, large companies, the police, unions, and even the legislature, in every single Andean country. In Colombia, where the traditional Liberal and Conservative Parties have fared better than most traditional parties in the region, trust in political parties has consistently ranked below trust in other institutions. In 1989, 15.9 percent expressed trust in parties (Semana 1990, 89); by 1994 that number was 22 percent (Semana 1996); by 1995, it had returned to 16 percent (Semana 1996). Are parties and governments at the local level doing better than parties and governments at the national level? Again, a brief look demonstrates a wide variety of opinions across and within countries. A Cedatos poll of citizens in provincial capitals in Ecuador in 2000 one year into mayoral termsnoted wide variation in mayoral approval ratings, with a high of 78 percent in Guayaquil and a low of 31 percent in Machala (Quitos mayor, Paco Moncayo, won only a 37 percent approval rating). Overall, however, the poll concluded that the level of approval of the mayors exceeds the Presidents, and that the population looks at these authorities as actors close to their daily problems, their neighborhoods, and their communities. A 1998 poll of Bolivians residing in the capital city of La Paz found only 9.3 percent who agreed with the statement that they had much condence in the municipal government; 46.9 percent claimed to have little condence; while

Decentralized Politics and Political Outcomes

185

nearly 42 percent responded that they had no condence in their municipal government (IINCIP 1998, 17). In the same poll, 13.3 percent claimed to have much condence in the [national] government; only 7 percent had much condence in political parties (IINCIP 1998, 16). Lack of condence in particular municipal administrations contrasts sharply with support for the idea of decentralization. A 1994 Bolivian poll asked respondents if they were in agreement or in disagreement with the Law of Popular Participation (the decentralizing law); 44.8 percent said they were in agreement with it, while 26.4 percent disagreed with it and 28.8 percent either did not respond or said they did not know. At the same time, however, only 9.6 percent responded that they knew much about the reform, while 62.4 percent said they knew little about it and 15 percent said they knew nothing about it. Asked why they were in favor or against the reform, 39.9 percent said they did not know; this was the most frequent response, with other positive reasons garnering 20.5 percent, and the next highest response (with 11.7 percent) community participation (Instituto de Encuestas 1994, 3133). A similar pattern emerges in the Peruvian case. When asked whether they approved or disapproved of their district mayors management, respondents in ve major cities varied in their responses: approval was voiced by 55.9 percent in Lima, 51 percent in Cusco, 40.1 percent in Junin, 37.9 percent in Loreto, and 50.5 percent in Piura (Grupo Propuesta Ciudadana 1996, 135). In contrast, over 80 percent of respondents in all ve cities believed decentralization would improve agricultural development, employment, and education (Grupo Propuesta Ciudadana 1996, 121). In a poll undertaken during Colombias 1991 constitutional convention, 82.2 percent of respondents said they believed that the revised constitution should allow for the popular election of governors (Semana 1991). Citizens in decentralizing countries appear enthusiastic about decentralization but only sometimes nd their particular local governments to their liking. How have particular parties done in decentralized contests? In particular, how have traditional parties fared in subnational competition? It is conceivable that decentralization might have the effect of strengthening political parties by bringing them closer to their constituents and rebuilding the connections between them through the election process, which generates information at the local level about citizen preferences. Another scenario suggests that decentralization might undermine traditional parties by creating the conditions under which minor parties or popular independents could win power on a small scale. Launching an independent campaign or trying to create a nationally viable party can be prohibitively expensive. Local and regional elections create non-trivial seats of power that might be attainable for independents and incipient parties. Where traditional parties do not adequately represent their constituents and where barriers to entry are low, decentralization may lead to the growth of new parties or encourage a barrage of independent candidacies. Some scholars view this as a major benet for

186

ONeill

democracy, while others caution that these ssures at the local and/or regional level may lead to a drop in party coherence and in policy coherence across levels of government over time. If decentralization is creating arenas of power where new parties are forming and building strength for a run at national ofce, we should expect to see a decline in traditional party vote share at the regional/local level rst, followed by a fall in the traditional party vote share in national elections. The data in Figure 6.2, which graphs the support for traditional parties at both the national and subnational level in each country, do not show a clear pattern of this type in any of the cases. The country that comes closest may be Bolivia, where a decline in support for traditional parties 20 at both national and local levels between 1991 and 1995 is followed by a decline in their support at the national level between 1997 and 2002; however, there is an increase in the support for traditional parties in the 1999 local contests. In Colombia, prior to the 2002 presidential contest, both national and subnational support for Liberals and Conservatives stays fairly stable. In Ecuador, traditional parties 21 tend to do better in subnational than in national elections; this is also true in Venezuela.22 In Peru prior to 1985, traditional parties 23 did equally well in national and subnational contests; the 1989 municipal results foreshadowed the 1990 national results, with the traditional parties losing support at all levels. Since 1990, the traditional parties remained equally weak at all levels, until Alan Garcas respectable showing in the 2001 presidential contest. The experiences in Ecuador and Venezuela suggest that, to the extent that traditional parties are losing vote shares, they are doing worse at the national level than they are at the subnational level exactly the opposite of what a theory of party building from local or regional roots would suggest. Instead of breeding grounds for new voices, subnational contests may be the last outpost for traditional party politicians. Changes in the electoral fortunes of traditional parties documented in Figure 6.2 cannot be wholly linked either to the onset of decentralization in these countries or to the particular contents of decentralizing reforms. In Peru and Venezuela, in particular, the precipitous fall in support for the traditional parties had much more to do with failed economic policies and perceived corruption at the national level. In addition, Fujimori changed the rules governing local election registration and candidacy to encourage the proliferation of independent candidates in local contests, perhaps in a bid to undermine the creation of rival parties at this level of government. Given the collapse of traditional parties at the national level in Peru and Venezuela, what is most surprising is the ability of these parties to win a fair number of subnational victories. As previously noted, when Perus APRA experienced a stunning defeat in the 1990 national elections, it still managed to elect its partisans to all but one regional executive position. Likewise, as the AD (Accin Democrtica, or Democratic Action) and COPEI (Comit

Figure 6.2

Percent of the vote won by traditional parties, by level


Colombia 10 8 6 4 2 0 2000 1980 Peru 10 8 6 4 2 0 1985 1990 Year 1995 National 2000 Local 1980 1985 Regional 1990 Year 1995 2000 1985 1990 Year 1995 2000 0 1980 1985 1990 Year 1995 2000 2 4 6 8 10 Ecuador

Bolivia

10

1980

1985

1990 Year

1995

Venezuela

10

1980

188

ONeill

de Organizacin Poltica Electoral Independiente, or the Social Christian Partyliterally, Committee of the Independent Political Electoral Organization) have fallen on hard times in national contests, they have been able to win a fair number of gubernatorial and mayoral contests in Venezuela. While it is tempting to focus on the cases where traditional party support has declined precipitously, comparing the cases where support for traditional parties has remained high at the subnational level reveals some interesting insights as well. It is particularly interesting to compare Colombia and Bolivia. In the Bolivian case, institutions played a crucial role in allowing Bolivias major parties to retain their hegemony in local elections: electoral laws forced candidates to run on established party platforms, disqualifying independent candidates. This law was changed early in Snchez de Lozadas second term. In the rst municipal elections allowing non-party candidates to run (December 2004), political parties still won approximately 77 percent of the vote; however, traditional parties did very poorly.24 In Colombia (and Peru, also), in contrast, independent candidates have multiplied at a quick pace. The ability of Colombias parties to continue to dominate local elections may owe to their exibility; it is not unusual to see several different candidates from the same party competing against each other. Whereas Perus system encourages candidates to run on independent platforms and Bolivias forbids this, Colombias system encourages independent candidates to run within the major political parties, perhaps contributing to their vitality. It must be noted that this electoral vitality may come at the cost of jeopardizing the internal coherence of parties and diluting the meaningfulness of party labels (see Pachano and Pizarro Leongmez in this volume). At the same time that national political trends have affected the ability of traditional parties to dominate subnational politics, decentralization also creates the conditions under which subnational politics can affect politics at the national level. The next section explores this direction of causality; in particular it examines the extent to which decentralization has affected the career paths of politicians.

Career Paths
The decentralization of real scal resources and the opening up of subnational leadership positions to electoral contestation has a third plausible impact on political practice in Latin America that is worth further exploration: it introduces new career paths to ambitious politicians. Looking through history at the typical political trajectory toward the presidency in Latin American countries, one would nd very few executive biographies listing gubernatorial or mayoral positions. Instead, the political rsums of Latin American presidents typically included either prominent positions within the legislature or inuential ministry

Decentralized Politics and Political Outcomes

189

positions or both. Today, the number of presidents and presidential hopefuls with a history of executive service at either the regional or the local level is multiplying at a rapid rate. The turn-of-the-millennium round of presidential elections in the Andes illustrates that candidates with experience in elected executive positions at the regional or local level have won a signicant percentage of the vote in many countriesand have won the presidency in some. Colombias president, lvaro Uribe, elected in 2002, served as the governor of Antioquia. Bolivias 2002 presidential elections saw the multi-term mayor of Cochabamba come in a close third in the voting. The winner of Ecuadors 1998 presidential election had been the mayor of the capital, Quito. In Ecuador also, in 2002, one ex-president actually sought a subnational executive position after serving as head of state: Leon Febres Cordero became mayor of Guayaquil after serving as president. Though Hugo Chvez in Venezuela had no previous governing experience, his main competitor in the 2000 elections, Arias Crdenas, was the two-term governor of Zulia. Of the Andean countries, only Peru did not have a major candidate in its 2001 presidential election with experience as a mayor or governor. Both of these trendsgovernors and mayors running for higher ofce, as well as national ofceholders turning to regional and local executive positions for career advancementhave potentially profound consequences for the evolution of political parties in decentralized countries. Where independents can achieve power in a subnational contest and use their performance in those positions to launch a bid for presidential power,25 traditional parties lose their ability to control access to the national electoral arena. More importantly, where popular party members can use subnational elective positions to leverage their possibility of winning nominations within parties, traditional party elite lose their inuence over political succession and, perhaps, over party coherence. On the one hand, this trend has the positive effect of creating new avenues to power and reining in the party elite. It may also have the effect of allowing more peripheral parts of the country to play a greater role in national politics if they can elect one of their own to the presidency on a strong record of subnational governance. On the other hand, this diminishes a partys ability to discipline its members within legislatures once new executives have been selected. Furthermore, this trend may shorten the time horizons of key policymakers as executives become less accountable to party organizations in general. To evaluate these claims, this section presents a great deal of empirical evidence, charting whether and how the road to the presidency has changed for major presidential candidates between approximately 1980 and 2002. As in the other sections of this chapter, there is a wide range of experiences across these ve countries. These differences tend to be strongly related to the ability of parties to control access to the executive and to subnational ofces, and to the power wielded by subnational ofcials. As one might expect, where mayors and

190

ONeill

governors have become powerful, there has been an increase in the number of former mayors and governors running for president and gathering a signicant percentage of the vote (I limit my analysis to presidential candidates who have won at least 5 percent of the national vote). The ability of parties to play a strong role in choosing candidates also plays a critical role, but it is a more uneven one: it is not always true that weak party control over nomination leads to an increase in the numbers of mayors and governors running for national ofce over time. In the ensuing paragraphs the record will be carefully examined. Prior to decentralization, the political biographies of major presidential candidates in Latin America tended to look much like that of Carlos Andrs Prez, who was president of Venezuela between 1974 and 1979 and was voted in for a second term in 1988. Carlos Andrs Prez was rst elected to a state legislature (1946); one year later, he was elected to serve in the national legislature. After a military coup interrupted his career, he returned to the national legislature (1959) and was soon (1960) named the rst director general of the Ministry of Interior Relations and later its head. A short while later, he became the partys congressional leader (1964), then its national secretary (1968) and a member of its powerful National Executive Committee. Finally, he was chosen as the partys candidate (1973) and won the presidency after years of service to the party in national politics. Contrast this with the political rsum of Colombias president, lvaro Uribe Vlez. In 1976 he began his political career as the Benets Chief of Medellns public enterprises. Just one year later, he became the secretary-general of the Ministry of Labor and, from 1980 to 1982, he served as the director of Civil Aeronautics. In 1982, he was appointed mayor of Medelln and was a councilman in that city between 1984 and 1986. From 1986 to 1994, he served as a senator in the national legislature and was then elected governor of Antioquia in 1994, a post he kept until 1997. While these vignettes arise from different countries, they represent two very different trajectories to the national executive. One begins with a short period of political service at the subnational level and moves quickly into the national legislature and party service; the other begins with more extensive local service, moves on to national legislative service, and then returns to a subnational position before launching his candidacy for president. This comparison alone cannot support the contention that decentralization has fundamentally altered career paths. What is interesting is the extent to which career trajectories of the rst type used to be much more common in the region, and the extent to which, more recently, career trajectories of the second type have become the norm throughout the region. Venezuela provides an excellent example of this trend. In 1983 and 1988, before decentralization took hold in the country (the laws were passed in 1989), none of the major candidates had served in a subnational executive position as

Decentralized Politics and Political Outcomes

191

either a mayor or governor. In the 1993 presidential elections, while the winner (with 30.45 percent of the vote) had not served as a mayor or governor, the other three candidates (accounting for 68.28 percent of the vote) had all served as subnational executives. This trend continued into the 1998 and 2000 elections, where the winner had not served in this role, but his major challengers (winning 39.97 percent of the vote in 1998 and 37.5 percent of the vote in 2000) both had. In addition, it is interesting to note that the subnational positions held by major candidates in the 1993, 1998, and 2000 elections varied quite a bit. Two of the major candidates in 1993 were former governors (Andrs Velsquez was governor of Bolvar and polled 21.95 percent of the national vote; Oswaldo Alvarez Paz was governor of Zulia and polled 22.73 percent). A third major candidate in that race, Claudio Fermn, was a former mayor of the capital, Caracas. In 1998, the major contender to Hugo Chvez was Henrique Salas Rmer, who won 39.97 percent of the vote and had been the governor of Carabobo from 1995 to 1998. Finally, in 2000, Chvez beat Francisco Arias Crdenas (37.5 percent), the former governor of Zulia. Over time, it appears that most of the countrys candidates emerge not from national political service in the legislature, ministries, or the party, but from subnational executive positions. In addition, these positions are spread throughout the country to some extent; there is not one key ofce that must be obtained to launch a plausible presidential campaign. Although Colombia has one of the strongest records of decentralization in the Andean region, the rising importance of subnational executive positions as a presidential credential has not been as stark as it has been in Venezuela. In 1986, prior to the direct election of mayors and governors, the victorious presidential candidate, Virgilio Barco, had served in state as well as local government. He began his political career in 1937 as Secretary for Housing and Public Works in the state of Norte de Santander. He moved on to become a departmental assembly member (1945 47) in Norte de Santander, and then moved on to the Ccuta city council (1947 49). He then moved into the national legislature in 1949. Later, he would work on several presidential campaigns; he served several times as a national senator, was appointed mayor of the capital, Bogot, served in several ministerial positions, and as ambassador to Great Britain and the United States. The result in 1990 was similar, with the victor, Csar Gaviria, having served as mayor of Pereira (a post he was appointed to in 1970), in the course of a political career that also involved service in the national legislature, various ministries, earlier presidential campaigns, and service to the Liberal Party itself. Two of his opponents (polling a combined 37.7 percent) had not served in such subnational positions; however, a third opponent (Rodrigo Lloreda Caicedo, who won 12.4 percent of the vote) had also served as an appointed governor (Valle del Cauca) early in his career (1968 79). In 1994, the competition centered around two politicians, Ernesto Samper, who eventually won 44.98 percent in the rst round of voting, and Andrs

192

ONeill

Pastrana, who won 45.3 percent in the rst round, but lost the second round. Sampers rsum looks much like that of Barco, Gaviria, and other successful Liberal presidential candidates: he served as the elected deputy to Cundinamarcas departmental assembly and on the city of Bogots council as well. He served in the national legislature, as Minister of Economic Development, and as an ambassador to Spain and to the UN. Pastrana, on the other hand, had begun his political career in the city council of Bogot and had become the citys rst elected mayor, in 1988. At the end of his term, he won a Senate seat, but turned it down to launch his rst, ultimately unsuccessful, presidential campaign. This pattern would become more common in the elections to come. In 1998, Pastrana again ran for ofce and, this time, won. His main competitor, Horacio Serpa, has a more traditional rsum: he was mayor of Barrancabermeja early in his career, but had gone on to such national posts as Minister of Government (the highest cabinet position) and Minister of Interior. He had been a senator and a member of the Constitutional Convention that rewrote Colombias Constitution in 1991, and the National Director of the Liberal Party. While this is a very traditional background, two of Pastranas other competitors came from less traditional backgrounds, including Noem Sann, who has largely a business background and no executive ofce holding at subnational levels, and Antanas Mockus, whose highest elected position was mayor of Bogot. The 2002 campaign brought back Serpa and Sann, but was won in the rst round by lvaro Uribe Vlez, a politician whose major experience was obtained in key subnational executive positions. Summing up the Colombian experience, it seems that subnational ofceholding has long been a part of most presidential candidates credentials. What is new in recent years is that these positions are now seen as capping ones career instead of as minor ofces held to vault one into the national legislature, a variety of ministerial positions, and direct positions within the party leadership. Todays most successful presidential candidates gain a great deal of their experience in high-prole mayoral and gubernatorial positions and lack the extensive experience in legislative, ministerial, or party service that was once common. In contrast to the Venezuelan and Colombian cases, Perus recent presidential contests do not show a gradual increase in the number of major candidates with experience as mayors and governors. This is not to say that political career trajectories have remained stuck in a traditional mold; rather, the prole of Perus major presidential candidates has changed drastically over time, but subnational ofceholding has not become an important milestone on the pathway to the presidential campaign. The importance of mayoral experience in seeking the presidency peaked in the 1985 presidential campaign, and then declined. Also, Perus experience differs notably from Colombias and Venezuelas in that all of its major candidates with subnational experience have come from the same subnational position: mayor of Lima.

Decentralized Politics and Political Outcomes

193

In 1980, the two major presidential candidates (Fernando Belande and Armando Villanueva del Campo, who together polled 72.6 percent of the vote) had no experience as elected mayors or governors. Belande had founded his own political party and had been elected to the presidency in 1963; Villanueva had held several ministry positions, and served in and been the president of both houses of the legislature. The third-place nisher, with 9.6 percent of the vote, Luis Bedoya Reyes, had been the mayor of Lima. In 1985, Alan Garca won the presidency with 45.74 percent of the vote. He brought a very traditional record to the campaign: he had served in the constitutional assembly in 1978 and then as a national legislator from 1980 onward. His two major competitors, Luis Bedoya (21.3 percent) and Alfonso Barrantes (10.23 percent), had both been Lima mayors. Javier Orlandini (6.25 percent), the fourth-place candidate, had not. This was to be the height of subnational ofceholder success in national contests. The 1990 contest between Alberto Fujimori and Mario Vargas Llosa pitted two relative newcomers to politics against one another. Neither had subnational government experiences, nor did either have experience in national politics. Polling just 7 percent of the vote was Henry Pease Garca, another former Lima mayor. In 1995, the contest between Fujimori and Javier Prez de Cuellar can also be described as a contest of nontraditional politicians; neither major candidate in this contest had served in subnational government. Politicians with neither traditional political career trajectories nor subnational government experience also dominated contests in both 2000 and 2001. In 2001, the fourth-place candidate (9.85 percent), Fernando Olivera, chose as his running mate Ricardo Belmont, who had been mayor of Lima from 1990 to 1993 and again from 1993 to 1995. In Peru, the political experience of presidential candidates changed drastically from 1980 to 2001; however, the decline of candidates with experience in the legislature, government ministries, and high party positions was not combined with an increase in the number of candidates with subnational political experience, as it was in several other cases. While several former mayors of Lima contested the presidency in 1980 and 1985, their presence in presidential contests declined in the 1990s and beyond. What is also interesting in the Peruvian case is the lack of variety in the kind of subnational political experience that various presidential candidates brought to their candidacies: all of those who had served in subnational government served in the capital, Lima, as its mayor. Given the weakness of subnational governmentsparticularly after Fujimori came to powerthe inability of former subnational ofcials to run on an impressive record does not seem too surprising. Bolivias experience from 1980 to 2002 differs from all three of the cases just outlined. In presidential contests from 1980 to 1997, no former mayors or governors launched presidential campaigns netting them more than 5 percent of

194

ONeill

the national vote; however, in the 2002 elections, the third-place candidate had been the mayor of CochabambaBolivias third-largest cityfor several terms. None of the four top candidates in the 1980 contest (polling a combined 84.4 percent of the vote) had served in subnational government; none of the four top candidates in 1985s contest (polling a combined 78.9 percent of the vote) had served in subnational government. Similarly, 1989s top ve candidates (polling 93.6 percent) did not include any subnational ofceholders; neither did 1993s top ve candidates (90.2 percent of the vote). Finally, none of 1997s six major candidates (96.8 percent) were former mayors, although the vice-presidential running mate of fourth-place candidate Ivo Kuljis had been mayor of Cochabamba. The 2002 contest suggests that career paths in Bolivia may also be changing, despite the relative newness of subnational elections (mayors were rst elected in 1995) and the tight control Bolivian parties retain over choosing candidates for both presidential and mayoral positions. It is interesting to note that 2002 candidate Manfred Reyes Villa did not run on a traditional party labelhe headed the NFR (Nueva Fuerza Republicana, or New Republican Force) listand that polls consistently showed him in the lead as the election neared. Comparing across the four countries discussed here, a few trends become evident. First, there appears to be a dramatic change in the career trajectory of major presidential candidates in most countries in the region. Major candidates in the early 1980s tended to have a great deal of experience in the national legislature; they often held several ministerial positions, ambassadorships, and positions of leadership within their national political parties. To the extent that they served in subnational levels of government, they did so early in their careers and often held elected positions in subnational governments with few resources, or held appointed positions at the subnational level. In the 1980 elections in Peru and Bolivia, and in the 1983 election in Venezuela and the 1986 election in Colombia, nearly all of the major candidates biographies roughly t this prole. Shortly after decentralization took hold in Peru, Colombia, and Venezuela, several of their most successful presidential candidates began to list election to subnational ofce high on their rsums. In Venezuela and Colombia, major candidates came from a variety of governorshipsnot just from the states in which the capitals resided. In very few cases did mayors come from cities other than the capitals; in Peru, the only candidates with subnational government experience had been former mayors of the capital city. Peru is distinct in two additional ways. First, it demonstrates a decline in the number of candidates with subnational governing experience as power is recentralized in the 1990s. Second, it demonstrates the importance of real power in vaulting subnational ofceholders to national prominence: in Peru, where decentralization extends to local governments and not regional governments, only mayors (not governors) run for national ofce. In Venezuela and Colombia, where decentralization empowered both levels of subnational government, former mayors and former governors have both run

Decentralized Politics and Political Outcomes

195

credible campaigns for president. Still, there is variation related to power and resources even here: in Venezuela, where regional governments are given many more resources than local government, the number of governors running credible national campaigns (two in 1993 and one each in 1998 and 2000) dwarfs the number of former mayors running equally credible campaigns (one in 1993). A third consideration in trying to understand the rise in the numbers of mayors and governors running for president is the role played by political parties. In both Peru and Venezuela, the ability of a strong, insulated party leadership to choose national candidates has been eroding over time for reasons unrelated to decentralization. The extent to which this party decline has led to an increase in mayoral and gubernatorial candidacies for president differs starkly across the two cases: a weakening party system is not enough to vault subnational ofceholders into national prominence. In addition, it appears that the strength of those ofceholderstheir access to sources of nance and their ability to build a strong record of performance at the subnational levelalso plays a major role. Where subnational governments have been strengthened, candidacies of this type are more likely than in cases where they have been weakened. The Colombian case also sheds light on this issue. While Colombias major parties continue to play a key role in most major elections, these parties have never had strict control over choosing presidential candidates. Even without a faltering party system, it is possible for politicians with a strong record of subnational government management to rise to prominence within the ranks of presidential candidates. Finally, in the Bolivian case, where political parties remained both strong and in control of the presidential (and mayoral) nomination process through 2002, there was little change from the traditional career trajectory to presidential candidacy. Even with signicantly strengthened local governments, popular mayors have not been able to make the jump to successful presidential candidacies. This seems to suggest that strong party control over nomination plays a key role. It is too early to draw anything concrete from this experience, however. At the same time that experience as a mayor or governor has become a coveted credential in a presidential run, the ofces of mayor and/or governor have become more attractive in themselves. The increase in scal resources available to ofcials serving in state and local government has made positions at these subnational levels more attractive to ambitious politicians. According to a recent study (Campbell 2003, 12), a new generation of leaders sought and won ofce as a result of decentralization. Looking at a set of Latin American countries that is broader than the Andean region, he notes that mayors are four times more likely to have professional backgrounds than a decade ago (Campbell 2003, 3). In addition, while better educated and more accomplished individuals were being drawn to local and regional government, they were also professionalizing the bureaucracies at this level. A World Bank study (1995) of Colombia documents a steep decline in the ratio of employees to professionals in fteen randomly

196

ONeill

selected municipalities between 1988 (the rst direct election of mayors) and 1994. In almost all of the cases, this ratio dropped by more than half during this six-year period; in dramatic cases, it fell from 62.3 to 7.3 and from 52 to 5.2 (World Bank 1995, 19). While a well-educated and professional workforce should improve local government, the attraction of a more professional workforce to more powerful local and regional ofces may also have made it more difcult for typically excluded groups to win subnational ofces. In Bolivia, one of the striking effects of increasing the power of local governments has been the decrease in the number of women elected to serve in local governments. In the 1993 municipal elections, which occurred before the Law of Popular Participation and which elected a total of 858 city councilors, 231 women were elected, representing roughly 27 percent of the total. In 1995, after Popular Participation, only 135 women were elected out of a total of 1,625 councilorsrepresenting only 8.3 percent of all councilors elected (SNPP 1996, 10). This picture becomes even more complex if one examines the electoral success of indigenous and peasant candidates in Bolivias local elections. Based on data from a series of questionnaires undertaken in 1996 and 1999, Xavier Alb (1999) notes that 464 indigenous and peasant candidates won election in 170 municipalities in the 1995 elections, as either primary candidates or as alternates (suplentes). Indeed, 55 percent of municipalities elected at least one indigenous or peasant candidate as a member of the council or as an alternate (Alb 1999, 16, 22). Taken together, these signs point to the complex reality of representation at the local level in decentralizing countries: access to powerful local ofces leads to an inux of talent and an immediate increase in the quality of the candidates from the perspective of education and professional development. It may or may not lead to an opening of public positions to a broader range of society. In Bolivia, women have not fared well in the new environment, while indigenous and peasant candidates appear to be doing quite well. To some extent, local governments are becoming more professional and also more broadly representative of the wider public. These ndings tend to support the idea that decentralization improves representation and also the ability of local government to provide better services than they had previously.

By Way of Conclusion
There is no simple answer to the question, Has decentralization improved democratic representation in the Andes? Looking across the three categories of democratic representation explored herevoter turnout, the ability of traditional parties to win subnational elections over time, and the effect of decentralization on independent presidential candidacies decentralizations effect on the crisis

Decentralized Politics and Political Outcomes

197

of democratic representation appears mixed. However, mostly the crisis of democratic representation seems to have occurred despite decentralization, rather than because of it. While a crisis of democratic representation is characterized by low voter turnout, the turnout in local and regional elections has been relatively high where subnational ofcials enjoy real access to power. Moreover, public opinion polls show most Andeans optimistic in their outlook toward subnational governments. A crisis of democratic representation is also characterized by high party volatility across elections; while this is true at the national and subnational levels, traditional parties seem to do slightly better in subnational than in national elections, suggesting that decentralization is not the cause of eroding support for traditional parties. Finally, independent presidential candidacies are hypothesized to be a hallmark of a crisis of democratic representation. Since decentralization provides a base of experience and public exposure from which independents might launch a presidential bid, decentralization would seem to play a supporting role in the crisis of democratic representation. The results in this area are mixed, however. While more candidates are running for president from subnational executive positions in the Andean countries, many independent campaigns have not been borne from subnational experience and thus one might expect independent candidacies from nontraditional sources even in the absence of decentralization. For example, it is notable that the nontraditional rsums of several presidents and presidential candidates in Venezuela (Chvez), Peru (Fujimori, Toledo), and Ecuador (Gutirrez) have not included subnational positions. This suggests that the trend toward nontraditional presidential candidacies, while inuenced by the availability of subnational positions within the public eye, has occurred despite decentralization in many countries. Exculpating decentralization as the primary cause of a crisis of democratic representation does not mean its effects on democratic representation have been uniformly positive. On the contrary, its effects are mixed. In some places, where subnational governments are freely elected and have access to signicant scal resources; where elections create a level of competition that leads to accountable public servants; or where well-educated professionals are drawn into the race and win, democracy has been improved. Elsewhere, where subnational governments control few resources; where they must depend on the central governments favor for the disbursement of those resources; where elections remain controlled by strong, undemocratic forces (be they large landowners or armed guerrilla movements); and where capable candidates are kept out of the race through institutional rules, the machinations of hegemonic parties, or by other means, democracy has not been improved by decentralization. In general, the idea of decentralization remains popular in the region, regardless of what the outcomes have been across countries. Polling evidence suggests that respondents make a distinction between their favor or disfavor of particular

198

ONeill

subnational administrations (which varies widely over place and time) and their support for decentralizing reforms (which is strong almost everywhere). Not only do Andean citizens pay lip service to the concept of decentralization, they turn out to vote in large numbers, in many countries matching the participation in national contests. While subnational elections that are concurrent with national elections elicit a greater turnout, the number of people that turn out even in staggered subnational elections is often a high percentage of the number who turn out in national contests proximate in time. While this is generally true, different countries experience different patterns in their subnational turnout over time. Where subnational governments are weak, as in Ecuador, participation at this level grows at a declining rate. In Bolivia, where local governments are perceived to be unstable and only indirectly accountable, turnout has also declined. Increasing subnational power seems to correlate with sharp increases in subnational turnout, as in Venezuela and Colombia. At the same time that citizens are turning out to vote, they are often voting for nontraditional parties in subnational contests, bringing new voices into the political arena. The only exception to this trend is Bolivia, where independent candidacies were not allowed in elections before 2002. In all other Andean countries, traditional parties have seen their ability to win subnational contests decline since decentralization was instituted. Where traditional parties have declined precipitously in national contests, the decline at subnational levels has often been less severe than at the national level (Venezuela and Peru). This suggests a slow opening to new voices at the subnational level. Colombia stands out as a case where the traditional parties have maintained electoral strength in both national and subnational contests, perhaps suggesting that decentralization has had the desired effect of generating information from base constituencies that has allowed political parties to change with the times and revitalize themselves. Finally, decentralization has created a layer (sometimes two) of powerful elected ofces that has attracted a growing number of well-trained and ambitious politicians. Not only has this increased the capacity of these governments and arguably improved public service delivery, it has also changed the career paths of politicians who seek the presidency. In countries where governors and mayors wield signicant resources, mayoral and gubernatorial positions allow politicians the opportunity to build a strong record of achievement that can be turned into a powerful credential in seeking the presidency. As a result, the number of presidential hopefuls that have high-prole gubernatorial or mayoral experience on their rsums has climbed dramatically in recent years. This is particularly notable in Venezuela and Colombia, but extends throughout the region. In Peru, where subnational governments have been somewhat weak and became weaker in the 1990s, the number of former mayors running for president and polling more than 5 percent of the vote peaked in 1985 and then declined. Furthermore, all of the candidates with subnational experience had been mayors of Lima; in

Decentralized Politics and Political Outcomes

199

Colombia and Venezuela, presidential contenders came from a variety of gubernatorial and mayoral positions. The relationship between decentralization and political representation is complex. Where decentralization has created strong subnational governments, it has largely improved democratic participation and representation; where it has been weak, it has not done as much good. Decentralization has attracted more qualied managers to subnational government positions and given them incentives to serve their constituents. At the same time, these reforms have increased the ability of independents and dissident party members with high-prole subnational government experience to launch presidential campaigns that weaken the ability of national parties to create and promote policy coherence over time. This may be the largest unforeseen danger of decentralization to the operation of democratic politics.

Notes
1. There is a large literature in economics. For a good recent review article, see Ter-Minassian (1999). 2. Not all economists expected decentralization to yield uniformly good results on the scal side. Notable criticisms include Prudhomme (1995) and Tanzi (1994). 3. Similarly, not all scholars of politics expected decentralization to be uniformly positive. Dahl (1971), for instance, worried about the effects on democracy if subnational identities became an important cleavage in national politics. 4. Perhaps this is because political outcomes are harder to measure or because it is expected that decentralizations effects must be felt in the longer term. 5. This chapter will not explore the scal and economic consequences of decentralizing reforms in this region. A great deal of scholarship has begun to focus on this area; many scholars nd that scal decentralization as practiced in the region has adversely affected scal balance at the national level. 6. For an excellent discussion of success cases in Latin America, see Campbell (2003). 7. Attempts to typologize types of decentralization through assigning particular meanings to words such as devolution, deconcentration (Rondinelli 1981), and so forth have not been widely embraced. 8. Allowing for the popular election of mayors and/or governors and municipal, provincial, or regional councils occurs as a discrete event and is therefore easy to locate temporally. Increases in the scal resources available to elected subnational governments are a bit trickier to differentiate: How much money needs to be available to these governments to make them meaningful? How would we dene autonomous scal power? I include a scal component in this denition because electing ofcials who are completely dependent upon scal resources controlled by the center effectively severs the accountability to local constituents created by the local election of these ofcials. Instead of trying to dene how many scal resources is enough to consider a reform as decentralizing, I will simply refer to reforms that allow for greater resources to be

200

ONeill

transferred to subnational governments as more scally decentralized, and those that allow for fewer resources to be transferred as less scally decentralized. See IDB (1997) for a more careful treatment of these distinctions. 9. In practice, this has led to a high rate of turnover at the local level. 10. This gure was obtained from a conversation with Jonas Frank, Ph.D. candidate and worker at a Quito-based NGO, October 2000. 11. In 1999 2000, debate over decentralization shifted from increasing transfers to provincial and municipal governments to a debate over provincial autonomy, particularly focused on the potential autonomy of Guayas, the coastal province that includes the city of Guayaquil. Led by former president and PSC member Febres-Cordero, the movement succeeded in getting the issue on a national referendum held during the summer of 2000. Sufcient support was not gained for this move, which would have turned tax bases in Guayas that had formerly contributed to the national tax base into the sole ownership of the state, disadvantaging other states in the country that are net recipients of national tax redistribution. 12. Bernalillo County, which includes the city of Albuquerque, New Mexico. 13. Beginning in 1994, presidential elections went to a two-round runoff system, but up until then, they had occurred in a single round of balloting. My choice of looking at rst- versus second-round turnout in national elections ends up making very little difference. 14. The indirect nature of elections and the lack of independent candidates have been cited as reasons for low turnout by Molina, Garca, and Landvar (2001, 14 15). This law, barring independent candidacies, was changed early in Snchez de Lozadas second term. 15. Comparing 1998 national turnout and 1997 municipal turnout (instead of 2000 municipal turnout, due to the unusual nature of those elections). 16. Comparing national and municipal results from 2000. 17. Comparing 1998 municipal elections with 2000 national elections. 18. Comparing 1993 municipal elections with 1995 national elections. 19. Comparing 1997 national elections with 1999 municipal elections. 20. MNR (Movimiento Nacional Revolucionario), MIR (Movimiento de Izquierda Revolucionario), and the ADN (Accin Democrtica Nacionalista). 21. CFP (Concentracin de Fuerzas Populares), PSC (Partido Social Cristiano), ID (Izquierda Democrtica), and the PRE (Partido Roldosista Ecuatoriano). 22. AD (Accin Democrtica) and COPEI (Comit de Organizacin Poltica Electoral Independiente). 23. APRA (Alianza Popular Revolucionaria Americana) and AP (Accin Popular). 24. See Partidos cuadruplicaron los votos de las agrupaciones, December 23, 2004, http://bolivia.com/noticias/AutoNoticias/DetalleNoticia24248.asp. 25. My focus here is on the presidency, but this analysis could be extended to cover career paths to key legislative or cabinet positions.

References
Alb, Xavier. 1999. Ojotas en el poder local: Cuatro aos despus. La Paz, Bolivia: CIPCA (Centro de Investigacin y Promocin del Campesinado) and PADER (Proyecto de Apoyo al Desarrollo Econmico Rural).

Decentralized Politics and Political Outcomes

201

Alesina, Alberto, Alberto Carrasquilla, and Juan Jos Echavarra. 2002. Descentralizacin en Colombia. In Reformas institucionales en Colombia: Una agenda reformista para los desafos del nuevo siglo, ed. Alberto Alesina, 95 134. Bogot: Fedesarrollo, Alfaomega. Alesina, Alberto, Ricardo Hausmann, Rudolf Hommes, and Ernesto Stein. 1999. Budget Institutions and Fiscal Performance in Latin America. Journal of Development Economics 59, no. 2: 253 73. Barr, Robert. 2001. Parties, Legitimacy, and the Motivations for Reform: Devolution and Concentration in Latin America. Paper delivered at the 2001 Annual Meeting of the American Political Science Association, San Francisco, August 30 September 2. Bird, Richard M. 1990. Fiscal Decentralization in Colombia. In Decentralization, Local Governments, and Markets, ed. Robert J. Bennett, 394 410. Oxford: Clarendon Press. Buller, Eduardo. 1993. Regionalizacin y municipalizacin en el proceso de descentralizacin administrativa: El caso de Peru. In Descentralizacin y gobiernos municipales, ed. CORDES, 135 77. Quito: Corporacin de Estudios para el Desarrollo. Cameron, John. 2000. Municipal Decentralization and Peasant Organization in Ecuador: A Political Opportunity for Democracy and Development. Paper presented at the Latin American Studies Association Conference, Miami, March. Campbell, Timothy. 2003. The Quiet Revolution: Decentralization and the Rise of Political Participation in Latin American Cities. Pittsburgh: University of Pittsburgh Press. Dahl, Robert. 1971. Polyarchy: Participation and Opposition. New Haven, CT: Yale University Press. Dammert Ego Aguire, Manuel. 1999. Desborde territorial descentralista: Replanteando la reforma descentralista peruana: Territorios sociales, estado con regiones y municipios. Selfpublished. de la Cruz, Rafael. 1992. La estrategia de la descentralizacin en Venezuela. In Descentralizacin, gobernabilidad, democracia, ed. Rafael de la Cruz, 1773. Caracas: Nueva Sociedad. Delgado Silva, Angel. 1995. Municipios, descentralizacin y democracia: Una propuesta democrtica. Lima: Servicios Grcos. Diamond, Larry. 1999. Developing Democracy: Toward Consolidation. Baltimore: Johns Hopkins University Press. Diamond, Larry, with Svetlana Tsalik. 1999. Size and Democracy: The Case for Decentralization. In Developing Democracy: Toward Consolidation, by Larry Diamond, 117 60. Baltimore: Johns Hopkins University Press. Eisenstadt, Todd. 1999. Electoral Federalism or Abdication of Presidential Authority? Gubernatorial Elections in Tabasco. In Subnational Politics and Democratization in Mexico, ed. Wayne Cornelius, Todd Eisenstadt, and Jane Hindley, 269 93. La Jolla, CA: Center for U.S.-Mexican Studies, University of California at San Diego. Fisman, Raymond, and Roberta Gatti. 2000. Decentralization and Corruption: Evidence across Countries. World Bank Policy Research Working Paper, no. 2290. World Bank, Washington, DC. Fox, Jonathan. 1994. Latin Americas Emerging Local Politics. Journal of Democracy 5, no. 2: 105 16. Garman, Christopher, Stephan Haggard, and Eliza Willis. 2001. Fiscal Decentralization: A Political Theory with Latin American Cases. World Politics 53, no. 2: 205 36. Grindle, Merilee. 2000. Audacious Reforms: Institutional Invention and Democracy in Latin America. Baltimore: Johns Hopkins University Press. Grupo Propuesta Ciudadana. 1996. Descentralizacin, participacin ciudadana y reforma del estado: Encuesta en cinco departamentos del Per. Lima: Grupo Propuesta Ciudadana.

202

ONeill

Huther, John, and Anwar Shah. 1998. Applying a Simple Measure of Good Governance to the Debate on Fiscal Decentralization. World Bank Policy Research Working Paper, no. 1894. World Bank, Washington, DC. IDB. 1997. Economic and Social Progress of Latin America. Washington, DC: Inter-American Development Bank. IINCIP. 1998. Encuesta de percepcin poltica interna y externa en la ciudad de La Paz. La Paz: Instituto de Investigaciones en Ciencia Poltica, Universidad Mayor de San Andrs. Instituto de Encuestas, Universidad Catlica Boliviana. 1994. Encuesta de percepcin poltica V: Ciudad de La Paz. La Paz: Fundacin Hanns-Seidel. Kay, Bruce. 1995. Fujipopulism and the Liberal State in Peru, 1990 1995. Chapel Hill: Duke University/ University of North Carolina Program on Latin American Studies. Kraemer, Moritz. 1999. One Decade of DecentralizationAn Assessment of the Venezuelan Experiment. Unpublished paper, Inter-American Development Bank, Washington, DC. Molina, Carlos Hugo, Martha Garca Ferruno, and Mara Elisa Landvar. 2001. Evaluacin de los consejos departamentales y provinciales: Estudios de caso. La Paz: Friedrich Ebert StiftungInstituto Latinoamericano de Investigaciones Sociales. Musgrave, Richard A. 1959. The Theory of Public Finance: A Study of Public Economy. New York: McGraw-Hill. Oates, Wallace E. 1998. The Economics of Fiscal Federalism and Local Finance. Northampton, MA: Edward Elgar Publishing. ONeill, Kathleen. 2003. Decentralization as an Electoral Strategy. Comparative Political Studies 36, no. 9: 1068 91. . 2005. Decentralizing the State: Elections, Parties, and Local Power in the Andes. New York: Cambridge University Press. Prudhomme, Rmy. 1995. The Dangers of Decentralization. World Bank Research Observer 10: 20120. Rodden, Jonathan. 2002. The Dilemma of Fiscal Federalism: Grants and Fiscal Performance around the World. American Journal of Political Science 46, no. 3 ( July): 670 87. Rondinelli, Dennis. 1981. Government Decentralization in Comparative Perspective: Theory and Practice in Developing Countries. International Review of Administrative Science 47: 133 45. Rubinfeld, Daniel L. 1987. The Economics of the Local Public Sector. In Handbook of Public Economics, vol. 2, ed. Alan A. Auerbach and Martin Feldstein, 571 645. New York: North-Holland. Semana. 1990. La gran encuesta del 89. Semana, no. 399 400, December 26, 1989 January 8, 1990, 83 97. .1991. La gran encuesta. Semana, no. 453, January 8 15, 1991, 20 31. . 1996. La gran encuesta del 96. Semana, no. 714, January 9 16, 34 39. SNPP. 1996. Las primeras elecciones: Directorio de alcaldes y concejales de la participacin popular. La Paz: Unidad de Investigacin y Anlisis, Secretara Nacional de Participacin Popular. Stein, Ernesto. 1998. Fiscal Decentralization and Government Size in Latin America. Working Paper, no. 368. Inter-American Development Bank, Washington, DC. Tanaka, Martn. 2002. La dinmica de los actores regionales: El despertar del Letargo? Unpublished paper, Instituto de Estudios Peruanos, Lima. Tanzi, Vito. 1994. Corruption, Governmental Activities, and Markets. IMF Working Paper, no. 94/99, International Monetary Fund, Washington, DC.

Decentralized Politics and Political Outcomes

203

Tendler, Judith. 1997. Good Government in the Tropics. Baltimore: Johns Hopkins University Press. Ter-Minassian, Teresa. 1999. Fiscal Federalism. Washington, DC: International Monetary Fund. Thedieck, Franz, and Eduardo Buller. 1995. Descentralizacin de la administracin en el Peru. In Descentralizar en Amrica Latina? ed. Sociedad Alemana de Cooperacin Tcnica / Programa de Gestin Urbana, 195 236. Quito: Sociedad Alemana de Cooperacin Tcnica / Programa de Gestin Urbana. Tiebout, Charles M. 1956. A Pure Theory of Local Expenditures. Journal of Political Economy 5: 416 24. Treisman, Daniel. 1999. Decentralization and Corruption: Why are Federal States Perceived to be More Corrupt? Paper presented at the Annual Meeting of the American Political Science Association, Atlanta, GA, August. Willis, Eliza, Stephan Haggard, and Christopher Garman. 1999. The Politics of Decentralization in Latin America. Latin American Research Review 34: 756. World Bank. 1995. Local Government Capacity in Colombia: A Country Study. Washington, DC: World Bank.

7 The Nature of Representation in Andean Legislatures and Attempts at Institutional Reengineering

Brian F. Crisp
he Andean democracies share recent efforts to redesign their political institutions. Disappointment and frustration with existing institutional arrangements led Bolivia, Colombia, Ecuador, Peru, and Venezuela to recast the structures governing their policymaking processes. Citizens perceived politicians to be self-serving and the policymaking process to be opaque and corrupt. What is more, major issues confronting the countries, such as economic stagnation and even open armed conict, seemed to drag on without an effective governmental response. Reformers sought to reshape the terms of the relationship between governments and citizens by adopting new constitutions. Legislatures have gured prominently in these efforts to rene how representation works. Despite fears of excessive power concentrated in the hands of presidents, legislatures have proven to be important in determining policy outcomes across Latin America (Mainwaring and Shugart 1997; Morgenstern and Nacif 2002; Johnson and Crisp 2003). However, legislators receive little trust from the public, and their behavior is often singled out as part of the justication for institutional reform. What role did legislatures play in provoking institutional overhauls? Is there even preliminary evidence that the overhauls have changed the way legislators carry out representation? If the way legislators behave in ofce is in part a function of the institutions through which they are nominated and elected, what does the experience of the Andes in the 1990s tell us about engineering representation? The relationships between parties and among members of the same partisan delegation tend to take extreme forms in Andean legislatures. The way in which legislators went about representation helped motivate reformers. Rather than balancing national and parochial concerns, Andean legislators either blindly followed national party elites but failed to establish any personal connection with the voters who elected them, or they acted as purveyors of particularistic

Nature of Representation and Institutional Reengineering

205

rewards within their districts but failed to take positions on issues of national importance. Unfortunately, episodes of institutional reform often target the symptoms of extreme behavior rather than its root causethe electoral incentives of legislators. As a result, new constitutions and electoral systems have done relatively little to change legislator behavior. Because of their failure to yield tangible results, episodes of constitutional reform have generated disillusionment about the efcacy of institutional engineering and the ability to modify the nature of representation.

Legislatures and Representation


Before we can adequately trace this chain of events, we need to consider more fully the concept of representation itself. When conceptualizing the link between voters and their elected representatives, scholars have emphasized the idea of responsiveness (Pitkin 1967). Eulau and Karps (1977) elaborated that responsiveness to the substantive interests of constituents could take four forms: policy responsiveness, allocations responsiveness, service responsiveness, and symbolic responsiveness. What constitutes the substantive interests of constituents has been captured on many different dimensions, the left/right ideological continuum being the most common. An important dimension of substantive interestsand representation of themthat is often overlooked is what we might call the national/parochial continuum. One of the most perplexing issues surrounding representation is that legislators are almost always elected by a geographically dened sector of the population to represent its interests and yet their job as representatives is to govern the nation as a whole. This, according to Pitkin, is the classical dilemma of representation (1967, 215). Pitkin argues that the alternatives are not mutually exclusive: a representative should be responsive to both parochial and national concerns (218). However, institutional designs are not neutral on the question of which strategyfocusing on parochial concerns or national concernsis the most efcient or rational manner for legislators to get reelected. In the abstract, the median voter may have some mix of preferences for policies targeted at the nation as a whole, at a particular geographic region or socioeconomic sector, or at a particular locality or individual entity (Taylor-Robinson and Diaz 1999; Crisp et al. 2004). Yet, the representation options offered by competing politicians may gravitate toward a single point on the national/parochial continuum or dimension. Where the electoral system is party-centered, partisan legislative delegations will behave in a disciplined fashion in the pursuit of national, programmatic goals that efciently enhance the reputation of the party as a whole. On the other hand, where the system is more candidate-centered, individualistic members of Congress will focus on targeted or parochial issues that can enhance their individual reputations.

206

Crisp

The institutional incentives that determine whether a system is likely to be relatively more party- or candidate-centered or the intraparty dimension of legislative representation (Shugart 2001)include the process by which one achieves access to the ballot, the nature of the ballot itself, the degree of vote pooling, and/or the level at which votes are cast (Carey and Shugart 1995). The potential for variation even in the Andean cases is quite extensive when we consider that across cases might mean not only across countries but also across chambers, across members of the same chamber elected under mixed-member rules or from districts of different magnitude, and across parties with varying candidate selection procedures competing under any of these rules. The intraparty dimension of legislative politics is central to the concept of representation because it helps determine the policy goalsnational to parochial representatives will pursue. For example, if a voter wants a publicly funded neighborhood clinic from her national government, the likelihood of obtaining a clinic in her particular neighborhood, and thus feeling well represented, is in large part a function of whether legislators are motivated to provide particularistic goods to a well-dened constituency that can determine their prospects for renomination and reelection. Where the electoral system is party-centered, the legislature might take up serious questions related to healthcare policy, but it is relatively less likely that individual legislators will feel beholden to delivering the benets of that policy to a particular constituency. Even where elected representatives are disciplined members of a partisan legislative delegation, their ability to deliver on the programmatic, national campaign promises that enhanced the reputation of their party will obviously be a function of that partys size in Congress. The effective number of parties or the interparty dimension of legislative representationand relatedly the size of any given party are conditioned by institutional characteristics including the seat allocation formula and district magnitudes. The timing of presidential elections can also inuence whether members of the successful candidates party can ride his coattails into ofce. Thus, a party may have a clearly articulated ideological program but remain insufcient in size to implement it. In other words, it may faithfully represent the substantive interests of its constituents but not be able to deliver policy or allocation responsiveness, given its less than majority status (though legislators may still be able to deliver effectively on service and symbolic responsiveness while in the minority). The size of partisan delegations seems less important where intraparty institutional incentives encourage legislators to focus on their individual reputations. Where individual legislators behave in an entrepreneurial fashion, the number of party labels is unlikely to effect how they carry out the task of representation because members of the same delegation would be less likely to behave in a homogenous fashion diminishing the usefulness of considering party as a unit of analysis.

Nature of Representation and Institutional Reengineering

207

In order to evaluate the importance of legislative politics and the national/ parochial continuum of representation on the demand for and effectiveness of institutional overhaul, I will proceed as follows. I will operationalize the intraparty and interparty dimensions of legislative politics and describe the prereform Andean legislatures in these terms. In a third section, I will characterize the constitutional reforms undertaken in the ve countries, focusing particularly on the institutional characteristics that promote party-centered or candidate-centered behavior among legislators. Next, I will evaluate whether the reforms undertaken led legislators to carry out representation differently in the earliest case of constitutional reform Colombia. After showing that legislator behavior varied only modestly as a result of political reform, I look at the scant evidence available on the impact of institutional reform on public opinion regarding support for democracy. I will conclude by drawing some lessons regarding the efcacy of institutional engineering in the Andes and the potential for disillusionment that failed reform attempts may generate.

Pre-Reform Andean Legislatures


In carrying out representation, legislators can tend exclusively to the particular needs of their individual districts; they can obediently support the party line regarding major policy initiatives; or they can combine these two sets of priorities to offer a more balanced form of representation. Assuming that most politicians want to perpetuate their careers, the way in which they engage in representation or balance national and parochial concernswill in large part be a function of the institutional rules that structure their prospects for reelection, election to another ofce, or even appointment to a political post. These rules govern the relations within partiesthe intraparty dimension of representationand the relations among partiesthe interparty dimension of representation.

The Intraparty Dimension of Representation


Intraparty politics is central to the way representation is conducted. When partisan delegations are relatively unied, they can promise broader programmatic policies and expect to be able to deliver on those promises if voters give them a legislative majority. However, where partisan delegations are only loose federations of reelection-seeking individuals, they cannot consistently deliver on promises that require unity on a single or limited number of issues.1 Instead, legislators from all parties are more likely to logroll and deliver particularistic rewards to their individual constituencies. The cooperation that does occur in such legislatures will assure that individual constituencies get served (and that legislators deliver legislative pork for which they get credit), but this form of representation does not revolve around intraparty cooperation and interparty conict.

208

Crisp
Table 7.1 Pre-Reform Intraparty Characteristics
Centralized nomination? Party-line voting? Party-level vote pooling? Intraparty incentives

Bolivia Colombia Ecuador Peru Venezuela


SOURCE :

yes no no yes yes

yes no yes no yes

yes no yes yes yes

hyper-centralized hyper-personalistic not extreme not extreme hyper-centralized

Shugart, Crisp, and Moreno (2002).

Candidate selection procedures and general election rules serve as incentive structures for legislators seeking reelection (see Table 7.1). Where party leaders control access to the ballot under their partys label, legislators are more likely to behave as members of a cohesive unit. They must be responsive to the leaders who exercise control over their future opportunities to run. On the other hand, where decentralized procedures, such as primaries or collecting a minimal number of signatures, are involved, the incentive to be a disciplined member of a party faction is diminished. While parties vary within a nation, major parties often share candidate selection procedures. Most parties in Bolivia, Peru, and Venezuela used elite-dominated nomination procedures, while in Colombia and Ecuador party leaders did not restrict the use of their labels. In terms of general election rules, we must examine at what level votes are cast and to what level they pool. In Bolivia, Ecuador, and Venezuela voters could cast only one ballot for a closed party list, and the votes were pooled at the level of the party. In Colombia and Peru, on the other hand, it was possible to distinguish among candidates from the same party. In Peru this was accomplished with open lists and in Colombia with sub-party lists. In Peru, votes were pooled to the level of the party, while in Colombia they were only pooled to the level of the sub-party list. Looking at Table 7.1, we see that there were three systems that did virtually nothing to encourage legislators to balance national and parochial concerns when conceiving of how best to represent. Bolivia and Venezuela were characterized by intraparty centralizationparty leaders dominated the candidate selection process in most parties and voters could not disturb those party-prepared slates. Venezuelan parties in particular were known for their high levels of party discipline (Rey 1972; Coppedge 1994; Crisp 2000). Party leaders in Congress could negotiate among themselves and with the executive with every condence they could deliver the votes of their copartisans. Personalism and pork barrel politics did not plague budgetary decision making. On the other hand, citizens felt virtually no connection to individual representatives. The move to a mixedmember system in 1993 was heralded as personalization because the singlemember districts (there were a few multi-member districts in the nominal tier)

Nature of Representation and Institutional Reengineering

209

were meant to more directly connect voters to legislators. However, the move was too little and/or too late to stave off a major revamping of the governmental institutions and party system. Colombia anchored the other end of the spectrum, giving legislators little incentive to think about the reputation of their parties when engaging in representation. Use of party labels went virtually unrestricted, and voters could choose from among multiple sub-party lists in every district. Partisan delegations in Congress were notably undisciplined, and patron-client links with voters were cultivated through the distribution of targeted pork-barrel programs. Both representatives and senators spent much more of their time traveling to maintain relations with the constituents (Ingall and Crisp 2001; Crisp and Desposato 2004). In Eulau and Karpss terms, where parties were weak, legislators emphasized allocations, service, and symbolic responsiveness relative to policy responsiveness. Presidents had to cobble together coalitions by negotiating with sub-party factions and individual legislators. This coalition building came at the expense of diluting programmatic goals.

The Interparty Dimension of Representation


When we look at relations among partisan delegations in Andean legislatures, Ecuador stands out for its relatively high effective number of partiesthe number of parties weighted by their size. Nonconcurrent presidential and legislative races and the Hare allocation formula contributed to a single chamber with more partisan actors. Colombia and Peru, on the other hand, had an average effective number of parties less than 2.5 (see Table 7.2). All other things being equal, it should be easier to construct legislative majorities where there are fewer parties, though this might come at the expense of representativenessthe proportionality between votes and seats. Polarization, the dispersion of the vote away from the relative center of the party system on a leftright continuum, was greatest in Ecuador, with Bolivia and Venezuela also showing a good deal of ideological diversity. The index ranges from 1 to 1 and can reach its maximum only when half of the vote goes to the right and half to the left; if all of the vote went to just one extreme, polarization would be zero because the relative center would be the extreme as well and there would be no dispersion (Coppedge 1998, 557). Not surprisingly, there was clearly a strong relationship between the effective number of parties, size of the presidents party, and ideological polarization. As the effective number of parties increases, the size of the presidents party (or any single party) generally decreases (R2 .93). Where the effective number of parties is high, so is the degree of polarization (R2 .78). Thus, legislatures like the ones in Colombia and Peru were characterized by a relatively low effective number of parties, little ideological diversity, and relatively great support for the executive. Ecuadors legislature

210

Crisp
Table 7.2 Partisan Composition of Andean Legislatures Prior to Constitutional Reform
Effective number of parties Mean ideology Ideological polarization Size of the presidents party

Bolivia 1985 1989 1989 1993 1993 1997 Colombia 1982 1986 1986 1990 1990 1992 Ecuador 1979 1984 1984 1986 1986 1988 1988 1990 1990 1992 1992 1996 Peru 1980 1985 1985 1990 Venezuela 1979 1984 1984 1989 1989 1994 1994 1999 Rest of Latin America

3.98 4.32 3.92 3.71 2.21 1.99 2.45 2.18 5.62 3.63 5.77 7.39 3.79 6.55 6.61 2.39 2.46 2.31 3.16 2.65 2.42 2.83 4.74 3.25

.30 .36 .24 .31 .18 .20 .17 .16 .02 .13 .05 .07 .23 .00 .35 .22 .07 .50 .18 .11 .23 .17 .22 .02

.44 .54 .58 .19 .14 .13 .17 .12 .55 .51 .56 .62 .43 .61 .54 .39 .51 .26 .45 .51 .44 .50 .36 .35

32.83 33.10 25.40 40.00 50.07 41.20 49.20 59.80 25.60 43.30 12.90 20.40 43.10 18.30 15.60 56.90 54.40 59.40 40.08 42.20 56.50 48.30 13.30 42.20

SOURCES : Johnson and Crisp 2003; my own calculations based on data available at http://www .electionworld.org/.

held down the other end of the interparty spectrum. Venezuela and Bolivia fell in between, with Venezuela more closely resembling the concentrated cases and Bolivia more closely approximating the dispersion in Ecuador. Bolivia and Colombia, on average, had the most conservative legislatures, while Peru and Venezuela had relatively leftward-leaning mean ideology scores (Coppedge 1997, 1998). In sum, in strictly interparty terms the Andean legislatures were quite diverse. Institutional rules interacted with underlying preference structures to generate variations in the partisan composition of legislative chambers across the region.

Extreme Electoral Systems and Representation


Shugart has juxtaposed the interparty and intraparty dimensions of electoral systems to capture their efciencythe extent to which they permit the articulation of policy-based electoral majorities (2001, 28). In order to be efcient,

Nature of Representation and Institutional Reengineering


Figure 7.1

211

Interparty and intraparty incentives of legislators in the pre-reform Andean countries


Hyper-centralized

Venezuela Bolivia

Hyper-representative

Ecuador

Peru Colombia

Hyper-personalistic

electoral systems must institutionalize incentives that avoid the extremes of being hyper-representative and pluralitarian on the interparty dimension and hyper-centralized and hyper-personalistic on the intraparty dimension. In intraparty terms, hyper-centralized systems give individual legislators virtually no incentive to cultivate a personal connection to constituents and instead leave them excessively obedient to party leaders. Party labels rather than individual deeds and reputation constitute the core of the representative connection. A hyperpersonalistic system does just the opposite. A hyper-representative system is one that encourages a high effective number of parties and therefore relies on postelection bargaining to form governing coalitions. A pluralitarian set of institutions, on the other hand, frequently translates a plurality of popular support into a clear governing majority. Using a very simplied version of Shugarts scheme, I depict the pre-reform incentives of Andean legislators on both the interparty and intraparty dimensions. These are very rough placements, but they capture some of the diversity just discussed (see Figure 7.1). While Peru and Colombia on average had the lowest effective number of parties in the Andes, they did not turn a plurality of support into a majority of seats avoiding an extreme classication on the interparty dimension. In interparty

Pluralitarian

212

Crisp

terms, Ecuador was exceptional for never generating a majority party in the legislature, and therefore requiring post-electoral bargaining to generate policy. The presidents party averaged less than 26 percent of the seats and never had more than 43 percent. In intraparty terms, Colombia and Venezuela stand out for opposite reasons. Colombia was exceptionally personalistic, while Venezuela was extremely party-centered. In sum, no legislature approached efcient representation at the center of Figure 7.1. In intraparty terms, where party leaders exercised nearly ironclad discipline over their legislative delegations, we would expect legislators to focus almost exclusively on national concerns that would enhance the reputation of the party as a whole. Conversely, where a lack of discipline results from legislators need to think about their personal reputations, representation or responsiveness will gravitate toward the parochial end of the spectrum. Given the rather extreme nature of representation as it was carried out in the Andes, we might expect constitutional reforms to focus on the electoral incentives of legislators. Where partisan delegations ignored parochial concernsVenezuela and Boliviainstitutional changes that loosened the grip of party leaders would bring the systems into greater balance. On the other hand, where legislators behaved as individualistic entrepreneurs focusing on pork barrel and patronage Colombia and to a lesser extent Peruwe might expect reforms that would encourage legislators to mix in concern for national, programmatic issues. In the next section, I will show that institutional reforms rarely, if ever, revamped the incentives at the heart of the intraparty dimension of representation.

Constitutional Reforms
Reformers often focused primarily on the constitutional allocation of powers across branches.2 Rather than adjusting the incentives of legislators themselves, reformers chose to simply strengthen the presidents hand in dealing with Congress. For example, the presidents powers of veto and agenda control were signicantly enhanced in Ecuador. If a president vetoes legislation on constitutional grounds, only the judicial branch can override his veto. In addition, legislators lost their ability to amend the amount of expenditure proposed by the president in his budget. In Peru, agenda powers were likewise enhanced by limiting the legislatures ability to introduce new taxes or add to the proposed budget. The new Venezuelan constitution allows the legislature to delegate decree authority on any matternot just economic and nancial matters as was previously the caseand it expands the presidents authority over states of exception. There were very limited changes to the electoral incentives that would inuence the intraparty dimension of representation. In Venezuela, the upper house was eliminated, but the lower house is elected under mixed-member rules, as has been the case since 1993. The new constitution appears to mandate the use

Nature of Representation and Institutional Reengineering

213

of primaries or caucuses for nominating legislative candidates. Such a change would pit copartisans against one another, heighten the importance of ones personal reputation, and perhaps diminish the party-centered behavior of legislators in ofce. However, thus far, political parties have chosen to ignore the provision without repercussions. Bolivia adopted a mixed-member electoral system similar to the one used in Venezuela. In an effort to get the best of both worlds, the goal was to make Congress more accountable to local constituencies through geographic representation while maintaining the overall proportionality of the system (Shugart and Wattenberg 2001). This change could diminish the hypercentralized nature of intraparty relations if legislators elected in the nominal tier have the liberty to respond differently to constituents than to their copartisans on closed proportional representation lists at the department level. Two factors do not bode well for dramatic changes in behavior. First, candidate selection procedures were left untouched, meaning that candidates who stray too far from the wishes of the party leaders may be denied access to the ballot in the future. What is more, the mixed-member proportional system was not adopted by a new regime devoid of common practices. It was grafted on to a highly centralized system with disciplined practices. Given the existing norms, anything short of a clear signal to dramatically change ones behavior is likely to have muted effects. In Peru, the new unicameral Congress was elected in a single nationwide district, abolishing the regional districts of the former lower chamber. A similar reform was adopted for the Colombian Senate. Reformers appear to have reasoned that a nationwide district would encourage legislators to focus on programmatic issues that would generate votes across the entire country. Peru maintained its open list system and Colombia its sub-party list system. Given the preference voting, candidates are still encouraged to gather the bulk of their votes regionally, based on their personal reputations (on Colombia, in particular, see Crisp and Ingall 2002; Crisp and Desposato 2004). In Ecuador, they moved from closed to open list rules. Open list rules could enhance the personal vote-seeking behavior of legislators. The pre-reform system limited the inuence of party leaders by not centralizing control over ballot access, and the adoption of an open list allowed voters to disturb any ranking of candidates established by party leaders. Thus, the one case that was relatively balanced between hyper-centralized and hyper-personalistic incentives on the intraparty dimension adopted rules that could skew behavior in the personalistic direction. Thus, if we were to revise Table 7.1 to reect post-reform intraparty characteristics, only one cell would change. The entry for Ecuador on party-line voting would not be no. So, rather than balancing the need to represent national and local concerns in the four of the ve cases that were extreme, ve instances of constitutional reform managed to take the one efcient case and make it extreme.3 The effective number of parties represented in the legislature is inuenced by seat allocation formulas, district magnitude, and the pull effect of presidential

214

Crisp

races. Increasingly proportional seat allocation formulas and relatively larger district magnitudes are associated with a higher effective number of parties. Concurrent presidential races can exert a downward pull on the number of parties, especially where the presidential race is decided by plurality rather than majority runoff. When legislative delegations are likely to behave in a relatively unitary and rational manner, the effective number of parties is critical for understanding the number of actors who must nd a policy proposal acceptable (where parties are very undisciplined, we must think about individual legislators as the most relevant unit of analysis within a legislature). Electoral systems that make it very difcult for any party to achieve a majority of legislative seats are hyperrepresentative, while those that consistently translate a plurality of votes into a majority of seats are hyper-majoritarian. Recall from Figure 7.1 that of the Andean cases, only Ecuador with its large effective number of parties tended toward an extreme on the interparty dimension of representation. Very few changes were made to Andean constitutions and electoral laws that we should expect to have a systematic impact on the partisan fragmentation of legislatures (see Table 7.3).

Table 7.3 Partisan Composition of Andean Legislatures after Constitutional Reform


Effective number of parties Mean ideology Ideological polarization Size of the presidents party

Bolivia 19972001 Pre-reform average Colombia 1992 1994 1994 1998 1998 2002 Pre-reform average Ecuador 1996 1998 1998 2000 Pre-reform average Peru 1995 2000 2000 2005 Pre-reform average Venezuela 1999 2005 Pre-reform average

5.62 5.62 3.98 2.57 3.00 2.61 2.09 2.21 5.23 5.03 5.43 5.62 4.78 2.91 6.64 2.39 3.77 3.77 3.16

.20 .20 .30 .12 .04 .16 .16 .18 .16 .28 .04 .02 .03 .03 NAa .22 NAb NAb .18

.47 .47 .44 .14 .25 .13 .05 .14 .45 .42 .47 .55 .14 .14 NAa .39 NAb NAb .45

25.40 25.40 32.83 56.07 54.00 53.30 60.90 50.07 25.60 23.20 28.00 25.60 41.05 55.80 26.30 56.90 46.10 46.10 40.08

SOURCES : Johnson and Crisp 2003; my own calculations based on data available at http://www.electionworld.org/ and in Coppedge 1997. a Six of the thirteen parties that won seats are new parties, so Coppedges database includes no ideology scores for them. Four others were not new but were not scored for the 1995 elections. b The majority of the parties are new and therefore not classied by Coppedge (including the MVR).

Nature of Representation and Institutional Reengineering

215

The move to nonconcurrent elections in Venezuela should tend to increase the effective number of partiesthe number of parties weighted by their size by shortening the presidents electoral coattails, but the only post-reform Congress elected thus far was elected concurrently with the president (compare gures in Table 7.2 and Table 7.3 for pre- and post-reform measures, respectively). The abolition of the Senate, with its district magnitude of two, eliminates another source of pull that might have kept the effective number of parties low prior to reform. The Bolivian Congress must now select the president from only the top two contenders, rather than the top three, when no candidate receives a majority of the vote. There is some chance that this will encourage parties to coalesce for presidential campaigns and thus exert a downward pressure on legislative races, but any effect is likely to be very minimal, and there is no evidence of such an effect thus far. The adoption of a mixed-member electoral system for the legislature does reduce the average magnitude, but the statewide races still exist in one tier of the voting. Finally, Colombias adoption of a single, nationwide district for the Senate with a magnitude of 100 is good for low vote getters, seemingly enhancing the prospects of small parties. However, given that votes are pooled at the level of the sub-party list, early evidence indicates that low vote-getting lists from the traditional Liberal and Conservative Parties have been the beneciaries (Botero 1998). Given the limited amount of time that most reforms have been in place and their half-hearted nature, any differences we see in Table 7.3 between the pre-reform and post-reform effective number of parties should not be attributed to institutional incentives. Looking at the indicators of interparty relations summarized in Table 7.2 for the pre-reform era and in Table 7.3 for the post-reform era, there is a decidedly noticeable lack of changeindicating both the moderate nature of institutional reforms and the importance of other explanatory factors, especially in periods of crisis. The institutionally determined patterns that characterize interparty and intraparty dimensions of representation identied in the extant literature are probabilistic and based on repeated observations across time and space. Both time and space are limited when the analysis is restricted to post-reform countries in the Andean region, and periods of reform and crisis are hardly when we should expect to observe normal causal connections. Unfortunately, despite the level of frustration that provoked reform and the difculty of actually carrying it out in many cases, reformers failed to address in a thoroughgoing manner the institutional characteristics that would have most directly affected the incentives of legislators to offer more balanced forms of representation. On the other hand, popular expectations that the effects of the new constitutions would be dramatic and far-ranging made it almost inevitable that disappointment or disillusionment would be the likely outcome. Given their limited and haphazard nature, should we or the citizens of Bolivia, Colombia, Ecuador, Peru, and Venezuela have expected reforms to change the way legislators engaged in representation?

216

Crisp

The Colombian Experience with Political Reform as a Cautionary Tale


The experience of Colombia seems particularly troubling for the possibility that institutional reform is likely to lead to legislator behavior that garners condence and trust. Support for democracy averaged the lowest score in the Andes (see the next section). Political overhaul did not translate into lasting public support because the reforms implemented were watered down and off the mark, given the source of the woes. Colombia was the rst Andean case to seek to reengineer representation through constitutional reform, and it is the one country on which I have sufcient data to look for pre- and post-reform changes in legislator behavior. As the Andean case staking out the extreme of personalistic representation, it seems like a prime candidate for being able to see some moderation in the way elected representatives carry out their duties. Virtually nothing was done to encourage different forms of behavior by members of the Chamber of Deputies who were and continue to be widely perceived as corrupt purveyors of particularistic rewards. Because deputies personal vote-seeking incentives were not reduced, the nature of representation in Colombia has changed less than citizens probably expected it to. Presidents continue to bring forward national programs, only to spend most of their time trying to preserve even the remnants of a coherent set of policies. Building support for their policies, including a legislative majority, requires bargaining, compromises, and payoffsincluding illegal payoffs in the form of corruption. The incentive structure of members of the Senate was potentially changed by the adoption of a single, nationwide district. Colombian reformers sought to create an upper house that was less clientelistic in its interaction with constituents. By adopting a nationwide district for the Senate, the political reform was intended to encourage senators to focus on large, programmatic concerns that would win them votes across the entire country. Returning to Eulau and Karpss (1977) theoretical work on representation, the goal was to reward senators for policy responsiveness as opposed to allocation (including pork barrel rewards), service, and symbolic responsiveness. As Carey and Shugart (1995) reason, increasing magnitude in a system with intraparty competition (sub-party lists) should enhance personal vote-seeking incentives, and personal vote seeking is frequently associated with pork barrel, particularist rewards. However, a nationwide district should make enhancing ones personal reputation by claiming credit for large, programmatic proposals a viable means of winning reelection. For example, offering to defend the environment may not win you sufcient votes to get elected in any normal-sized district, but in a nationwide district such a candidate could earn the votes of all the citizens throughout the country for whom the environment was the single most important issue. The incentives were only potentially changed because a nationwide district allows voters from all over

Nature of Representation and Institutional Reengineering


Table 7.4 Bill Targets in the Colombian Senate before and after Electoral Reform
Pre-reform Congress (1986 1990) Post-reform Congress (1994 1998)

217

Nationally targeted bills Sectorally targeted bills Regionally targeted bills Locally targeted bills Individually targeted bills Total number of bills
SOURCE :

219 137 58 75 18 507

417 187 32 75 18 729

Crisp and Ingall 2002.

the country to support candidates whose platforms are of broad, national appeal. However, it also allows candidates to seek all of their votes in a single department, just as they did prior to the reform. As a result, the Colombian Senate has a very diverse set of memberstraditional politicians dependent upon geographically concentrated patron-client networks and a new breed of senator who receives support from geographically dispersed voters motivated by a policy issue of primary concern to them. Though most Colombians would probably conclude that the adoption of a new constitution generally, and a nationwide district for the Senate more specically, was unsuccessful, it has changed the nature of representation in the Senate. Much more of the chambers time is now spent considering bills targeted at the nation as a whole (see Table 7.4). What is more, it can be shown that a senators probability of targeting a bill at the nation as a whole increases as his or her electoral support base becomes more dispersed (Crisp and Ingall 2002). Is greater attention to issues of national concern associated with increased party discipline that would facilitate adopting coherent programs of government? In Colombia, roll call procedures are used, though not as frequently as one might expect in a system with such purportedly low party discipline. The vast majority of votes are cast by a show of hands, but only the number of yeas and nays are recordednot who cast them. In the nearly twenty-ve-year period between the end of the power-sharing National Front and the constitutional replacement of 1991, only sixteen roll call votes were taken in the Senate (twenty were taken in the Chamber of Representatives). During the rst full legislative term after reform, lasting from July 1994 to June 1998, seventeen roll call votes were cast in the Senate (only four in the Chamber of Representatives). This is relatively scant information on which to base any judgments, and the mere fact that the very uncustomary roll call procedures were invoked leads to some doubt about their representativeness (see Ames 2001 on the potential pitfalls when using roll call votes to measure discipline). However, they are the only data available on which to estimate party cohesiveness.

218

Crisp

Excluding highly consensual votes (where 90 percent of each major party voted the same way), Liberals and Conservatives showed virtually the same propensity to break discipline as they did in the pre-reform era. When roll call procedures were used when senators were elected in department-wide districts, 20 percent of Liberals and 19 percent of Conservatives were likely to dissent from the rest of their partisan delegations. In the post-reform era, with senators having to make a name for themselves in a single, nationwide district where 100 seats are up for grabs and sub-party slates are still competing against one another, 31 percent of the Conservative delegation and 26 percent of the Liberal delegation were likely to defect on any given vote (during the 1994 98 Congress). Thus, institutional incentives can change behavior, but Colombian reformers failed or were unable to carry out more far-reaching changesincluding similar changes for the Chamber of Deputies and the elimination of sub-party lists for the election of either chamber (and thus intraparty competition). The adoption of a nationwide district for the Senate made it possible for candidates to focus on programmatic concerns and win votes across the country from voters with whom these positions resonated. However, increasing the magnitude of the district without eliminating intraparty competition made it even more difcult for partisan delegations to act as coherent units. Combined with little effort to change the way deputies carry out representation in the lower chamber, the modest change in behavior by senators has not been enough to satisfy Colombians. In late 2002, political reform was once again at the top of the agenda.4

Public Perception of Legislatures (and Democracy More Generally)


Pre- and post-reform indicators of how legislators carry out representation are scant. As I will detail below, using indicators of citizens perceptions of legislatures to help us evaluate the effectiveness of institutional reforms in addressing the crisis in the Andes also poses several challenges. Thus far we have learned that institutional incentives regarding how Andean legislators should carry out representation were diverse in terms of intraparty politics but tended to push legislators toward one extreme or the otherthinking only in terms of their parties reputation or only in terms of their personal reputations. Hyper-personalistic systems should be associated with bills and laws focused on local and individual targets. Party vote-seeking incentives on the other hand should be associated with bills and laws focused at broader targets including the nation as whole (Crisp et al. 2004). Partisan delegations in Venezuela occupied the party-centered extreme, with Bolivian partisan delegations nishing a close second. Colombia held down the opposite end of the spectrum, with very undisciplined partisan delegations, and Peruvian legislators were the second most personalistic. Despite new constitutions in every Andean case, very little was done to change these incentives. In interparty terms, Ecuador stood out for its representativeness, with

Nature of Representation and Institutional Reengineering


Table 7.5 Public Condence in Congress, 1996
1996 2000

219

Bolivia Colombia Ecuador Peru Venezuela Andean average Rest of region


SOURCE :

22 15 27 33 19 23 29

16 14 9 23 37 20 ??

Latinobarmetro 1996, 2000.

Bolivia being the only other system to have an average effective number of parties greater than 3.6. Tables 1.3 and 1.4 report public condence and trust in Andean legislatures. Unfortunately, more complete time-serial data on these questions is not widely available (or at least not at a price most individual academics can afford), and the surveys were not begun until 1995 offering only pre- and post-reform observations in one or two Andean cases. Before the surveys were administered, new constitutions were adopted in Colombia in 1991, in Peru in 1993, and in Bolivia in 1994. Ecuador was in the midst of reform and adopted a new constitution in 1996, the rst year for which I have data. Venezuela promulgated its new constitution in 1999, midway between the two observations reported here. It does not appear that incentives to pursue exclusively one extreme of representation or the other were a recipe for public support. Neither Colombian nor Venezuelan legislators generated much condence from the general public. They were the only countries where fewer than 20 percent of respondents expressed much or some condence in Congress in 1996 (see Table 7.5). The Venezuelan Congress generated signicantly greater condence in the postreform 2002 surveys. It would be tempting to argue that institutional reforms were responsible for the boost, but this is only tenable in the most general terms. In Venezuela, the 1999 Constitution created a unicameral legislature and mandated the use of participatory candidate selection procedures. The latter change was blatantly ignored by all parties. Rather than condence due to specic alterations in legislators incentives, it seems more likely that the increased condence was due to a more general sense that political, economic, and social changes were under way. Thus, the increased condence would be more accurately attributed to President Chavezs challengeincluding, but certainly not limited to, legislative reformto the traditional political and economic elites. Ecuador, the one case where intraparty incentives were relatively balanced between hypercentralized and hyper-personalistic extremes, was the only case to generate much condence among more than 6.75 percent of the population. The leg-

220

Crisp
Figure 7.2

Public satisfaction with democracy


60

50

Percentage satised

40

30

20

10

0 1995 1996 1997 Peru Venezuela 1998 Years Ecuador Bolivia 2000 2001 Colombia 2002

islature received relatively high marks in 1996 when reform was underway. However, by 2002 the legislature had the lowest level of condence of any country in the Andes. In my conclusion, I will return to the idea of institutional reforms creating false expectations, generating only limited changes in behavior, and ultimately leading to disillusionment with the idea that institutions matter at all. While the Andean countries average of much condence was equal to or slightly greater than the rest of the region in 1995, the Andean average was substantially lower than the rest of the region on some condence in Congress. Any conclusions drawn from these gures must be considered very tentative, but one possible explanation for the differences in condence is the extreme nature of the electoral system. The systems most extreme on the intraparty dimension generally fare worse. Peoples frustrations with the system are reected in the lower than average levels of condence and trust in the legislature. Figure 7.2 traces the percentage of respondents who were very satised or somewhat satised with democracy in their country between 1995 and 2002. Clearly, a longer time series that included data predating all our cases of reform would be preferable, but comparable annual data does not exist. In addition, these results report responses to a question about democracy as a form of

Nature of Representation and Institutional Reengineering

221

government, not the legislature in particular. While the behavior of parties in the legislature is undoubtedly key to most conceptualizations of representative democracy, we are making an additional assumption when we use feelings about democracy to discuss satisfaction with legislators behavior. With those cautions in mind, it is the case that in three countriesPeru, Ecuador, and Venezuelathe rst observation taken after political reform showed the highest levels of satisfaction in the countrys time series. For example, satisfaction with Venezuelan democracy peaked with the 1999/2000 observation5the period when the new constitution was drafted and adopted. Satisfaction reached 55 percent in that year, while it averaged 19 percentage points less in the other six observations. Two years after the adoption of the new constitution, the percentage of satised respondents had dropped by 15 percent. Similar to the trend for condence in the Ecuadorian legislature cited above, if the bump in popularity was generated by the promise of change, it was eeting. Colombia, the Andean country furthest from its experience of reform when these data were collected, had the lowest average level of satisfaction, dipping to just 11 percent by 2002.6 In sum, Andean legislatures do not earn much condence, and Andean democracies, more generally, have trouble sustaining satisfaction. Figures for public condence in the Congress suggest that legislators who focus either predominantly on parochial concerns or predominantly on national concerns fail to earn the respect of their citizens. Promising to use institutional reform to elicit better behavior can lead to a temporary spike in popular support, but when the expected new behavior fails to materialize, the public quickly becomes disillusioned.

Future Prospects of Andean Legislatures


Thus, in the end, the institutional reforms undertaken thus far have been rather ineffective. Striking a balance between programmatic and parochial forms of representation is not an easy task. Legislators in the Andes were prone to offer rather extreme forms of representation prior to reform, and the tentative evidence available indicates that their citizens have not been impressed with the degree to which the legislators behavior changed as a result of adopting new constitutional structures. The cynicism generated by the failure of one round of reforms makes additional changes even more difcult to achieve. Because institutions by denition involve the formalization of a set of practices or norms, they are not merely the reection of underlying social or economic forces. A disjuncture may emerge between voters preferences and the behavior encouraged by a set of institutional structures. Unfortunately, just because the disjuncture is sufciently severe to motivate reform, there is no guarantee that the reforms undertaken will be optimal.

222

Crisp

This is not to say that institutional reform generally is not a viable route for changing the nature of representation.7 The scholarly literature has identied cross-national differences in legislator behavior that follow the regular patterns we would expect given the nature of the electoral system. The evidence presented here on the Colombian Senate indicates that legislators in a given system will change their behavior over time if the rules under which they are elected are changed. The lesson to be drawn from the Andes is that institutional changes have not systematically revamped electoral systems to encourage legislators to offer a mix of national and parochial policy promises. I suspect that reforms have failed to dramatically change the nature of representation in Andean legislatures for a combination of two reasons. First, constitutional reformers may nd it easier to identify the symptoms of extreme representation than its root causes (Shugart, Crisp, and Moreno 2002). For example, where presidents have a hard time getting their proposals through Congress, it may appear more obvious to strengthen the presidents constitutionally allocated powers than to change the incentives of legislators. Second, the legislators and parties that are expected to change their behavior are typically deeply involved in the reform process itself. They may be hesitant to implement sweeping changes to the rules of the political game because they are unsure whether they will prosper electorally under a new set of institutions. Continued comparative research on the motivations of legislators as they provide alternative forms of representation can only further inform practitioners as they attempt to improve the relationship between elected ofcials and those they govern.

Notes
1. Unfortunately, we do not have systematic information on the cohesiveness of partisan delegations in Andean legislatures. Data collection of this sort is underway for many countries, but comparable data is still not available. Some legislatures simply do not use the roll call procedures necessary to measure party discipline. Instead, we must, at least for the time being, rely on the incentives that encourage partisan or individualistic behavior to evaluate intraparty politics in the Andes. 2. For a complete review of political reforms in Latin America during the 1990s, see Shugart, Crisp, and Moreno (2002). 3. In the Conclusion, Mainwaring argues that Andean cases underwent important electoral system reforms in recent years in order to enhance the direct accountability of representatives to voters and to enhance mechanisms of representation. He then reasons that if electoral reforms were extensive and yet the perception of crisis persists, then institutions offer us little leverage on the causes of the crisis or its likely solutions. However, as I noted in the text, changes to the incentives legislators face on the intraparty dimension of representation were reformed very little, and/or in the wrong direction.

Nature of Representation and Institutional Reengineering

223

Thus, we have no evidence to evaluate whether reforms that would generate electoral efciency (Shugart 2001) on the intraparty dimension would help ameliorate the crisis. No one should expect the creation of a few set-aside seats for indigenous candidates to solve hyper-personalistic or hyper-centralized extremes. Certainly, they are not the only factors to be considered, but without relevant reforms it seems hasty to conclude that institutions hold no explanatory power. 4. Interestingly, in an effort to pre-empt President Uribes referendum on a wide array of political reformsthat would eventually go down to defeat due to a lack of turnoutthe Colombian Congress passed a bill mandating a single list per party in each district. Assuming no further reforms, the rst congressional elections held under these rules are scheduled for early 2006. 5. The Latinobarmetro was not done in each country in both 1999 and 2000. Instead, some were surveyed in 1999 and others in 2000, and the results for the two years over which the region was completely surveyed are reported together. 6. By the end of the year, newly elected president lvaro Uribe had placed political reform rmly back on the agenda, hoping to make the regime more efcient and less corrupt. His administration was received with a great deal of enthusiasm, and it seemed likely that promises of change would generate optimism to be recorded in the 2003 survey. If the boost in condence was generated, would it be as eeting as it had been in Venezuela? 7. See my earlier note taking exception to Mainwarings conclusion that electoral reforms were extensive.

References
Ames, Barry. 2001. The Deadlock of Democracy in Brazil. Ann Arbor: University of Michigan Press. Botero, Felipe. 1998. El Senado que nunca fue: La circunscripcin nacional despus de tres elecciones. In Elecciones y democracia en Colombia 19971998, ed. Ana Mara Bejarano and Andrs Dvila, 285 335. Bogot: Universidad de los Andes. Carey, John M., and Matthew S. Shugart. 1995. Incentives to Cultivate a Personal Vote: A Rank Ordering of Electoral Formulas. Electoral Studies 14: 41739. Coppedge, Michael. 1994. Strong Parties and Lame Ducks: Presidential Partyarchy and Factionalism in Venezuela. Stanford, CA: Stanford University Press. . 1997. A Classication of Latin American Political Parties. Kellogg Institute Working Paper, no. 244. Helen Kellogg Institute for International Studies, University of Notre Dame, Notre Dame, IN. . 1998. The Dynamic Diversity of Latin American Party Systems. Party Politics 4, no. 4: 547 68. Crisp, Brian F. 2000. Democratic Institutional Design: The Powers and Incentives of Venezuelan Politicians and Interest Groups. Stanford, CA: Stanford University Press. Crisp, Brian F., and Scott W. Desposato. 2004. Constituency Building in Multimember Districts: Collusion or Conict? Journal of Politics 66, no. 1: 136 56. Crisp, Brian F., Maria C. Escobar-Lemmon, Bradford S. Jones, Mark P. Jones, and Michelle M. Taylor-Robinson. 2004. Electoral Incentives and Legislative Representation in Six Presidential Democracies. Journal of Politics 66, no. 3: 823 46.

224

Crisp

Crisp, Brian F., and Rachael E. Ingall. 2002. Institutional Engineering and the Nature of Representation: Mapping the Effects of Electoral Reform in Colombia. American Journal of Political Science 46, no. 4 (October): 733 48. Elections around the World. http://www.electionworld.org/. Eulau, Heinz, and Paul D. Karps. 1977. The Puzzle of Representation: Specifying the Components of Responsiveness. Legislative Studies Quarterly 2, no. 3: 233 54. Ingall, Rachael E., and Brian F. Crisp. 2001. Determinants of Home Style: The Many Incentives for Going Home in Colombia. Legislative Studies Quarterly 26, no. 3: 487511. Johnson, Gregg B., and Brian F. Crisp. 2003. Mandates, Powers, and Policies. American Journal of Political Science 47, no. 1 ( January): 128 42. Latinobarmetro. Various years. Press Releases (Prensa). http://www.latinobarometro .org. Mainwaring, Scott, and Matthew Soberg Shugart, eds. 1997. Presidentialism and Democracy in Latin America. New York: Cambridge University Press. Morgenstern, Scott, and Benito Nacif, eds. 2002. Legislative Politics in Latin America. New York: Cambridge University Press. Pitkin, Hanna Fenichel. 1967. The Concept of Representation. Berkeley: University of California Press. Rey, Juan Carlos. 1972. El sistema de partidos venezolanos. Politeia 1: 175 230. Shugart, Matthew Soberg. 2001. Extreme Electoral Systems and the Appeal of the Mixed-Member Alternative. In Mixed-Member Electoral Systems: The Best of Both Worlds? ed. Matthew Soberg Shugart and Martin P. Wattenberg, 25 51. New York: Oxford University Press. Shugart, Matthew Soberg, Brian F. Crisp, and Erika Moreno. 2002. Re-Constituting Democracy: Institutional Patterns of Political Overhaul in Latin America. University of California San Diego, University of Arizona, University of Iowa. Typescript. Shugart, Matthew Soberg, and Martin P. Wattenberg, eds. 2001. Mixed-Member Electoral Systems: The Best of Both Worlds? New York: Oxford University Press. Taylor-Robinson, Michelle M., and Christopher Diaz. 1999. Who Gets Legislation Passed in a Marginal Legislature and Is the Label Marginal Legislature Still Appropriate? A Study of the Honduran Congress. Comparative Political Studies 32: 590 626.

Part III

POPULAR POLITICS AND THE CRISIS OF DEMOCRATIC REPRESENTATION

8 Urban Citizen Movements and Disempowerment in Peru and Venezuela

Daniel H. Levine and Catalina Romero

his chapter addresses a core puzzle: Why is continued citizen mobilization accompanied by growing disempowerment of those same citizens? Why do movements fail, leaders burn out, and members disperse, and what are the implications of this organizational failure for democratic representation? Citizen involvement in such movements arises precisely because of the failure of conventional vehicles of representation to provide trustworthy and effective means of connecting new citizen groups and their needs with public institutions. That these new movements should also regularly fail raises important questions about the quality and durability of democracy. Our consideration of the issues is rooted in a close examination of urban movements, mobilization, and empowerment and disempowerment in the recent experience of Venezuela and Peru. The puzzle that concerns us is of course not limited to these two countries: it is common to all the Andean republics, and in different ways, to much recent experience of urban mobilization in Latin America and beyond. After a brief account of urban citizen movements and politics in our two cases, we outline general reections on the nature of empowerment and disempowerment, on the peculiar combination of strengths and weaknesses that mark many contemporary movements. A close examination of types of movements and their links with political parties and protest follows. We close with analysis of two waves of mobilization: in Peru (which sparked the ouster of President Alberto Fujimori) and in Venezuela (both for and against President Hugo Chvez Fras), and with reections on the likely future of empowerment and disempowerment for urban citizens and the implications of this perspective for democratic representation.

228

Levine and Romero

The Puzzle
The puzzle is set up by three key facts that situate Venezuela and Peru in a meaningful comparative perspective while providing grounds for comparison between the social and political processes each country has experienced over the last twenty years. The rst points to the decay, decline, and eventual disappearance of once powerful political parties, and of the system of organizations and political norms built around them. The second addresses the creation, expansion, rise to prominence, and decay (often after specic goals were met) of networks comprised of civic organizations, sometimes referred to as civil society or popular movements, depending on the country and circumstances. The third is a trajectory of mobilization, activism, and sustained protest (associated with the trajectory of new movements and networks) rising to peaks at moments of crisis and dissipating thereafter. In both countries, and through extended periods of time, huge numbers of people were mobilized for sustained, repeated, and often risky collective actions, including rallies, campaigns to collect and deliver signatures, marches, demonstrations, sit-ins, and the like. A profusion of new and often short-lived groups combined with established organizations such as trade unions or business federations, political parties, and professional groups to manage and sustain the effort. Relevant moments of crisis, examined in more detail below, include the movement to reject Fujimoris 2000 reelection, or the waves of mobilizations and counter-mobilizations (centered on the Chvez government and its survival) that began in spring 2002 and culminated in the remarkable civic strike touched off in December 2002 and stretching into February of the next year. The recent experience of Peru and Venezuela has sparked an extensive literature.1 The key point to underscore, and the real value of comparing these otherwise very different societies, is how much the comparison sheds light on a common effort to grapple with similar problems in the construction, defense, and deepening of democracy. Central to this effort has been a continuing, and not always successful, struggle to enhance participation, broaden access to politics through linked institutional reforms (including reforms in systems of representation), and to strengthen the accountability of politicians and public institutions. In each case, the effort was spurred by the emergence of new capabilities and groups outside the net of state and established political parties who have sought to open and energize politics. Facing institutions and leaders they rejected as corrupt and unresponsive, citizens in both countries turned to civil society as an arena for participation and a platform for demands about representation. In this light, the mobilization by citizen movements that we examine here involves more than simple demand making that then nds representation through established conventional channels: it is a claim to representation that politicizes new spaces and groups in national life. The failure or short-circuiting of the movements raises

Urban Citizen Movements, Disempowerment in Peru, Venezuela

229

questions about the possibilities of constructing enduring representation that starts and ends in civil society. Peru and Venezuela confront their common dilemma from very different starting points. Peru is a poor, ethnically divided country whose modern history displays a series of short-lived bouts with electoral democracy. The military government (1968 80) and the protests accompanying the reinstitution of civilian democratic politics marked an important new beginning on several critical dimensions. Enormous numbers of people were brought to political action by the popular movements of the late 1970s and enfranchised thereafter. This means that starting with the elections of 1980 there was perhaps for the rst time in the history of Peru a genuine mass public for electoral politics. The restoration of party and electoral politics coincided with the beginnings of massive violence by Shining Path. The popular movement would soon be caught between the violence of two armies: Shining Path and the repressive forces of the Peruvian state, bent on eliminating them. Throughout the same period, Peru experienced a rapid urbanization that profoundly altered the character and dynamics of urban life. A host of new groups and spaces came into being, and movements and protests became a part of the daily social and political scene (Dietz 1998; Stokes 1995). This was also a time of profound economic decay, which hit the popular movement hard, making collective action of any kind difcult. Democratic institutions were undermined by the governments of Alberto Fujimori (1990 2001), whose growing authoritarianism, isolation, and corruption ultimately triggered his downfall. The combination of violence, institutional and economic decay, and leadership betrayal was deadly for urban movements. Although the capacity and will for mobilization remains, as visible in the massive protests against Fujimoris fraudulent reelection in 2000, movements lack sustained organization and means to ensure continuity and accountability. Venezuelas experience of civic movements starts later in time (mid-1980s) and arises within a well-established democratic system: the goal was to democratize democracy by broadening citizen access and loosening the grip of the countrys powerful political parties and dominant state apparatus. The existing democratic system was much stronger, richer, and more deeply established than was the case in Peru. From this position of strength, the fall is all the more notable. Beginning in the late 1980s, economic and institutional decay began to bite, popular dissatisfaction with established institutions (especially the dominant political parties) grew sharply, and the political system entered an extended crisis that continues to this day. As was the case in Peru, these developments were accompanied by the emergence of a wide range of citizen movements, centered in the cities, that demanded more authentic and accountable representation. Venezuelan movements differ from their Peruvian counterparts in many ways, not least the fact that from the beginning their membership base and agenda have been predominantly middle class. The initial demands of movements found expression in

230

Levine and Romero

decentralizing reforms that devolved power to states and cities, expanded the number of ofces open to election, and reduced barriers to participation. But these reforms were swept away by the continuing crisis of the country, and marginalized by the victories of Hugo Chvez, who came to ofce with a wholly different agenda of total change (Kornblith 1999; Salamanca 2004; Levine 2002). The initial power of Chvezs movement was enhanced by the collapse of older political structures; as these recovered ground and citizen movements began to emerge again, opposition mounted, mostly in the form of civil society mobilizations, once again seeking political redress and accountability outside the formal structures of the political system. To summarize, in both countries the discredit and decay of established leaders and parties combined with institutional failure and sustained economic crisis opened the way, at different points and with country-specic nuances, for a wide range of movements to emerge and claim a voice as civil society. Participants in these movements sought, by their activism, not only to satisfy immediate demands (say for housing or services), but also to express, by their action, a claim to citizenship and equal status apart from established, conventional structures of representation. They project not only their demands, but also their image of themselves as citizens, forcefully onto the public stage. Their activism politicized urban spaces in the two countries in new ways: creating new forms of action and building (often literally) new spaces for such activism. In both cases, the longterm results of such activism, in terms of sustained benets, new policies, or accountable leaders, have been problematic. The weakness, reversibility, and often open failure of the effort requires us to reconsider the possibilities and limits of democratic representation, and to search for possible solutions in ways that go beyond tinkering with electoral machinery. The relation of empowerment and, by extension, disempowerment with democratic representation is central to our inquiry. Most discussions of empowerment have a people-friendly character. They underscore the need to provide people with the skills and capabilities that make access to power possibleto empower them and to enhance the quality and authenticity of representation. As typically used in these discussions, quality and authenticity of representation involve more than simply assuring that electoral results reect votes more or less accurately and fairly (according to whatever electoral rules are in use). Assuming universal suffrage and relatively free and open elections, representation that is authentic and of high quality entails lowering barriers to organization, multiplying instances and arenas of political action and representation, making voting easier, and ensuring that representatives are more accountable and more accessible to ordinary citizens. The goal of such reforms is to link new urban spaces, groups, and networks to the institutional structures of the political system in ways that allow social energies to bubble up and nd representation. Our concern is that the link has been problematic and the record at best mixed.

Urban Citizen Movements, Disempowerment in Peru, Venezuela

231

The line of analysis we follow here requires that the concept of representation (and democratic representation) be situated in a broad analytical and social context that extends beyond the details of electoral rules and institutional structuring to address the potential links of these public spaces with the new social spaces, movements, and claims to representation being advanced from society. In both our cases, massive numbers of citizens have repeatedly joined together and sought representation of their interests through public, often risky, mobilizations of all kinds. Politics and systems of representation should be capable of linking together these new networks and spaces, but with rare exceptions this has not happened. Political leaders remain wedded to a top-down vision in which it is they who know what to do and how to do it. They either do not make the links or they use them for a time and move on. Citizen movements are too often left stranded and divided, lacking enduring channels of contact or control into the political sphere. The theoretical and practical challenge is to rethink the relation between social movements and political representation in ways that preserve the energy and openness of both. That is our agenda here.

Movements and Politics in Peru and Venezuela


The decay of political parties and the rise of an explicitly anti-party politics is common to the recent experience of both Venezuela and Peru. In Venezuela, an entire political system built around powerful, permanent political parties weakened under long-term economic pressure, exacerbated by massive corruption and ineffective leadership, and further undermined by reforms set in motion in the mid-1980s. Although it is not easy to date the start of the decline with precision, most observers agree that by the early 1990s, the two dominant parties AD (Accin Democrtica, or Democratic Action) and COPEI (Comit de Organizacin Poltica Electoral Independiente, or Independent Political Organizing Committeea term hardly ever used in subsequent years; the party has long been known only by the acronym)were shadows of their former selves (Molina and Prez 2000; Crisp 2000; Crisp, Levine, and Molina 2003). Their weakened condition undermined the ability of leaders to respond effectively to the crisis created by the two attempted coups of 1992, and by the continued economic crisis. Once-legendary party discipline weakened, making secure interparty deals in the legislature much more difcult to manage. COPEI divided, and its founder, Rafael Caldera, waged a brilliant anti-party campaign to win the presidency in a four-way race in 1993. This was the rst presidential election since the restoration of democracy in 1958 that was not won by either AD or COPEI. Although the two parties continued to do well in regional and local voting, the whole period was marked by continued intraparty divisions, by an explosion of citizen organization (including insurgent unionism) escaping from party control, by rising levels of abstention, and by growing anti-party sentiment.

232

Levine and Romero

The election of 1998 and subsequent national, regional, and local votes have conrmed the deathbed status of the established political parties and the entire political system constructed around them. President Chvez has moved strongly against the parties, and subsequent national voting has been dominated by personalist coalitions, both pro- and anti-Chvez. The voting system invented for elections to the Constituent Assembly, which wrote the countrys new Bolivarian Constitution, gave supporters and allies of President Chvez a disproportionate share of seats (95 percent of seats, with 66 percent of the vote). Subsequent legislative elections returned to the old system, with results (in terms of seats) that were more proportionate to votes received. In Peru, political parties (with the sole exception of APRAthe Alianza Popular Revolucionaria Americana, or American Revolutionary Popular Alliance) were not as powerfully structured or deeply organized as in Venezuela. A political system hinged on electoral competition between well-established parties, each with its afliated movements and organizations, made a tentative appearance in 1955, and again, with the restoration of democracy and civilian politics, after 1980 (Tanaka, this volume). In this system, APRA was joined by AP (Accin Popular, or Popular Action, founded several decades earlier by Fernando Belande Terry), the PPC (a Christian Democratic Party) and by Izquierda Unida (United Left), a loose coalition of leftist parties. The fortunes of these parties rose and fell through the 1980s as the economic situation deteriorated and the insurgency (led by Shining Path) grew and extended its reach across the country. AP won the presidency with Fernando Belande Terry in 1980, and then plummeted in support; APRA won with Alan Garca in 1985, and then lost support; and the Izquierda Unida gained steadily in municipal elections through to the mid- to late 1980s, only to collapse in division. The 1990 election completed the decline of the parties as central political organizations. This election ended up as a contest between two coalitions led by independentsthe writer Mario Vargas Llosa and the unknown Alberto Fujimori, the eventual winner. Only two years into his term, President Fujimori dissolved Congress and began rewriting the rules of the political game. He was elected for a second term in 1995, running against Javier Prez de Cuellar, former Secretary General of the United Nations. After Fujimoris ousterfollowing his reelection in 2000, which was contested as fraudulent and boycotted by possible opponentsthe presidency was won by another independent, Alejandro Toledo, a leader of the anti-Fujimori movement. His opponent in the second round was former president Alan Garca, running again for APRA. In both countries, the decay of parties and of a party system (strong or weak) was accompanied and pushed or pulled along by an explosion of citizen organization and new movements of all kinds. We understand the emergence of new groups and their sustained presence on the public scene as an effort to create spaces and connections where democracy can be practiced and interests aggregated

Urban Citizen Movements, Disempowerment in Peru, Venezuela

233

and pressed in political encounters of all kinds, from constitutional debates and legislative discussions to petitions, referenda, marches, rallies, and demonstrations. Representation is clearly at issue, but the phenomena we address here are not well captured by conventional principal-agent discussions of the matter. This is representation both as claim to voice and a legitimate place at the political table and as a challenge to the terms of representation enshrined in existing public institutions. The process of social participation was visible earlier in Peru, where movements emerged in the 1970s in opposition to military rule and as an expression, above all in the cities, of grassroots organizing to meet social and environmental needs. The Peruvian Catholic Church played a key role in promoting and protecting many such movements, training activists and providing invaluable connections among them. By the end of the 1980s, and into the Fujimori period, the combination of economic decline (which made collective action of any kind more difcult) with increasing violence, both from Shining Path and the government, undermined the ability of many groups to survive and renew themselves.2 Mobilizations continued, of course (Dietz 1998; Stokes 1995; Tovar 1991; Levine and Stoll 1997), but became more short-lived and more limited and specic in focus. The transition to a democratic regime combined with generational changes in the leadership of the Catholic Church also removed key allies from the scene. Parties further weakened because urban organizations developed a sense of autonomy, looking in a democratic way to their own collective interests and goals, which seemed to be different from those of the political parties. The return to democracy in the 1980s reestablished elections at the municipal level, providing public spaces for participation and the expression of demands. In Venezuela, the power of party organizations and their ability to colonize civil society and monopolize access to resources long inhibited the growth of independent civic associations. As we have seen, these began to appear in the mid1980s, with roots in movements in the business sector and, above all, groups of urban property owners opposed to unrestricted development. Motives of neighborhood defense soon expanded into a broad agenda aimed at the creation of more autonomous urban governments with independent elected, not appointed, mayors and governor. This reform was put into effect in 1989, and combined with other decentralization measures, began to reshape the dynamics of party leadership and campaigning in the 1990s. At the same time, the countrys longterm economic decline, which continued throughout the 1990s, undermined the ability of party leaders to distribute patronage and thus hold loyalties. Autonomous professional groups appeared, private foundations and new business groups consolidated their position, and independent union movements began to gain ground. The latter, most successful in the steel mills of Guyana, spawned a successful political movement, La Causa R (the Radical Cause). The term civil society appeared as a regular feature of Venezuelan political discourse, and

234

Levine and Romero

efforts to forge some kind of unied position were made (Gmez Calcao 1998; Salamanca 2004; Levine 1998). The political trajectory of Hugo Chvez Fras, his election to the presidency in 1998 (afrmed in subsequent votes under a new constitution and with new electoral rules), and his overall political project challenged the legitimacy of the core political arrangements of the past four decades, and looked to build a new and supposedly more democratic society and political system. Fiery populist and class-based rhetoric has been the daily bread of the Bolivarian revolution from the beginning, and mobilization of masses has been its core claim to legitimacy. Like Fujimori earlier in Peru, Chvez looked to destroy existing political parties (and associated groups, notably the trade unions), with the difference that Chvez wanted to rebuild politics in a revolutionary and participatory style, with a broad range of arenas and groups in direct contact with the leader and the state. In practice, this has meant attacking and dismantling old structures, restlessly inventing and reinventing new ones, including, notably, the regimes own political party, and diverting state resources into vaguely dened Bolivarian circles. The decay and rout of the old system was so complete that it took several years for opposition to begin to regroup. Early steps (2001) came with the defeat of government-sponsored efforts to take the Central University in Caracas for the people, and with the defeat of a government-sponsored referendum to renew the leadership of the trade-union federation. These were followed by a massive series of work stoppages, strikes, and marches that became a regular feature of the calendar in Caracas and, to a lesser extent, in other cities. Protest techniques common in other countries, such as cacerolazos (or banging of pots and pans, creating a truly deafening noise) and caravans of cars honking horns, were put to use, and massive marches (long since abandoned in favor of televisioncentered campaigning) returned to center stage. A regime claiming legitimacy on the basis of its ability to mobilize was now running into massive countermobilizations. Fearing the appearance of weakness and the prospect of losing control of the street, the regime began to put on its own massive marches. There was continuous escalation in this process from early December 2001 through to the tragic events of April 1113, 2002, when a huge march, heading through downtown Caracas to the presidential palace, was attacked by snipers. Many were killed, and in the ensuing crisis the government was replaced and then retook power as the military divided and different coalitions of citizens took and retook the streets. All sides then pulled back from the brink for a while, but after about six weeks, the rhythm of marches and countermarches began again, accelerating through the fall of 2002 and culminating in the remarkable civic strike of late 2002 and early 2003. It is instructive to compare the mobilizations that forced out Fujimori with those competing to oust, support, or restore Chvez. The former were managed

Urban Citizen Movements, Disempowerment in Peru, Venezuela

235

by a loose coalition of groups from across the country, knit together by local- and national-level activists with prior experience in mobilizations, energized by the OAS ndings of fraud and irregularities in the 2000 reelection of Fujimori, and by growing revelations of corruption linked to Vladimiro Montesinos, Fujimoris director of the National Intelligence Service. Mobilizations were sparked rst by students and womens groups, who began with symbolic acts such as sweeping the plaza of the Congress and regular washings of the national ag (to cleanse them of corruption). As protests expanded, they were joined by NGOs and then by political parties, which added nancing and organizational reach.3 The campaign itself combined enormous marches (such as the Marcha de los Cuatro Suyos4 in Lima, on July 28, 2000) along with a series of sustained regional mobilizations, and innovations such as the previously mentioned weekly public washings of the national ag. The organizations and political parties so prominent in the 1980s had disappeared from the political scene following the coup of 1992, losing their legal status after failing to win seats in elections for the Democratic Constituent Congress in 1993 and later in the presidential elections of 1995. They resumed a role only when protests were well under way. In Venezuela, by contrast, as the opposition to President Chvez recovered and began to gather force, the organizational backbone for sustained action did not rest on groups formed over the past ten or fteen years. An unexpected but highly effective anti-government alliance was formed between the trade-union federation, the business federation, the Catholic Church, and the mass media. The rst two provided organizational resources, the latter two, legitimacy and an amplied public voice. That this coalition was able to put so many people into the street on such a regular basis depended less on the groups own members than on the motivation of a loosely linked net of neighborhood and human rights groups.5 Despite continuous reference to the role of civil society, in neither country did the specic membership organizations of the previous decade, once seen themselves as the potential foundation for a new kind of politics, play a central role. Different kinds of organizations emerged to take the lead. Apart from human rights groups, which have grown throughout the region in the last fteen years in response to dictatorship (Sikkink 1993), the key organizational players were either occasional coalitions gathered for a particular purpose around a specic leaderfor example, Alejandro Toledo and Per Posible (Possible Peru) or old-line organizations such as trade unions, business federations, or the Church. Mobilization and commitment were sustained not so much by group structures themselves as by the presence of numbers of loose or weak ties among groups and individuals that facilitated connections and the exchange of information, support, and resources across groups, social sectors, and physical spaces (Granovetter 1973; Smith 1996). If this is correct, mobilization even massive and sustained mobilizationis compatible with the absence of an organizational underpinning like that commonly provided by political parties. But at the same

236

Levine and Romero

time, the absence of a continuing organizational structure can undermine the potential consolidation of gains and make it all but impossible for citizens to demand and achieve accountability from leaders without a new round of massive, institution-challenging mobilizations. This deserves closer examination.

Considerations on Empowerment, Disempowerment, and Representation


Empowerment is a notoriously plastic concept, often used in conjunction with equally protean terms such as civil society or social capital. Like accountability, empowerment has no easy equivalent in Spanish, and neologisms such as empoderamiento ll the linguistic gap. The elasticity of these concepts reects their multidimensional character: they point to processes that involve organizational growth, personal and collective identity, specic leadership skills, trust, the ability to secure goods and services, and the like, and operate simultaneously on a range of social levels. Of these concepts, empowerment is perhaps the most people-friendly. Empowerment denotes a kind of social and political process and a pattern of structure and organization that provides citizens with a growing range of arenas for access to the public sphere, reduces barriers to action, and creates conditions that enhance a sense of self worth and recognized personal as well as collective identity.6 In this light, the relation between empowerment and a sense of citizenship seems clear enough. Those women and men who come to see themselves as citizens with rights equal to others are in that measure set on the road to individual and collective action as normal and possible. The emphasis on identity, however, masks considerable ambiguity around the relation of empowerment to organization. Organization can further empowerment by linking individual and group capacities together and moving action to larger arenas. But at the same time, by subordinating group efforts to leadership concerns and stiing independent decision, overarching organization can also disempower. In his work on religion in the United States, Warner (1993, 1070) states, It is to be expected that the empowerment functions of religion are latent. At an individual level, those who seek well-being in religion tend not to nd it; those who gain well-being from religion are not those who seek it. The logic of Warners apparent paradox rests on an argument that locates empowerment (like social capital) in the long-term construction of community, trust, and the skills and disposition required for working togethernot just in creating or joining organizations, and much less in simply getting the goods. This is a lot for any social process to deliver, and many movements have not been able to ll the bill. A review of recent theoretical and empirical work on urban social movements, empowerment, and representation in Latin America reveals a slow recovery from a hangover brought on by exaggerated expectations,

Urban Citizen Movements, Disempowerment in Peru, Venezuela

237

laced with a heavy dose of idealization of the new movement. The autonomy of movements (vis--vis institutions such as political parties, state institutions, or the Church) was overdone, and a romantic image of the small is beautiful kind made many observers anticipate that a totally new kind of politics would arise from the seeds provided by these movements. This in turn would provide the basis for a different pattern of representation with new kinds of political parties, and altered institutions that would hopefully be more democratic and more fully empowering of citizens than what had hitherto existed (Hellman 1992; Lander 1995; Levine and Stoll 1997; Lora 2002; Ortner 1995; Oxhorn 2001; Tovar 1991). This did not happen: in case after case, the new politics was easily absorbed into the old, and movements split or simply fell apart. That movements fail and empowerment does not endure should come as no surprise. Movements often fail or run out of steam: activism is costly and antinomian and the day-to-day pressures of economic and family survival make organization difcult to sustain (Piven and Cloward 1977, 1998). Anyway, as Stokes and others have shown for Peru, the development of supposedly more participatory (and therefore empowering) styles of organization among the urban poor does not necessarily replace older self-concepts and forms of action. People are practical, and new styles of action take their place as an alternative to be weighed and perhaps used, as circumstances seem to indicate. What does disempowerment mean, and what is the path from empowerment to disempowerment? There is withdrawal from activism, often prompted by burnout, sometimes by family pressures (commonly gender specic and affecting women). There is also a failure of leadership replacement. Groups that campaign for democracy may of course remain authoritarian within, and leaders may nd it difcult to let new generations come to the fore. The problem is notorious in groups linked to the Catholic Church (as many have been), where dependence on clergy makes for enormous vulnerability if and when more conservative clergy arrive on the scene. Finally, of course, with the opening of new political spaces (through transitions to democracy or reforms within democratic systems) younger activists easily nd other, perhaps more rewarding and less costly outlets for their energies. We do not suggest that empowerment is necessarily illusory. Many men and women have indeed acquired new skills and self-images and imparted these to others in their communities. The central point here is that the concept is incomplete, and the reality fragile. The difculty lies more with links to organization and the reliable construction of representation, which may undermine the consolidation of gains. The linkage between the civic spaces of empowerment and the public spaces of political representation and state power remains problematic. The absence of stable links to larger structures also undercuts the visibility of groups in the public sphere, which is essential to their gaining recognition as legitimate actors and claimants of rights and goods.

238

Levine and Romero

The theoretical problem is to discern what there is about the way in which empowerment was sought, representation constructed, or connections built by urban movements that has self-limiting or perhaps self-destructive qualities. Our working concept of representation must be broad enough to encompass both groups and formal political structures. We also need to understand how the fate of groups and protest is related to the issue of formal, electoral mechanisms of representationin other words, how elections, electoral mechanisms, and preelection politics (candidate selection, district boundaries, voting systems) are related to, and perhaps reinforce, patterns within groups. Other institutional matters, most notably the impact of judicial and penal systems, are also vital, especially for considerations of security of property and persons. Our earlier review of movements and politics in our two cases showed that although the party-focused model of organization was clearly stronger in Venezuela than Peru, in both countries, the decay (or, in Peru, the failed consolidation) of that model (and of its controlling norms) had contradictory effects. The long process of organizational deterioration in Venezuela set many potential clients free from party controls while opening the eld for new kinds of groups operating in newly created political spaces. Cases in point include the expansion of urban neighborhood movements, the impact of new electoral rules on the development of different styles of representation, and the emergence of a range of groups and federations self-consciously identied as civil society. In Peru, where parties were never that strong to begin with, the surge of urban growth (fueled by internal war) overwhelmed older structures and spawned a proliferation of urban groups of all kindsunied by their common need to solve urgent and immediate problems of housing, food, transport, education, and violence. Lacking reliable interlocutors and regular access to channels of inuence and state resources, the connection between particular causes and concerns and more general political afliations is hard for most people to identify, much less sustain. In both countries, new urban citizen movements arose to address very specic needs created by the urban context and the deteriorating economic situation. Satisfying needs required some rearrangement of the relevant institutions and political spaces, and led to campaigns for political and electoral reform. Building these connections and sustaining these campaigns requires allies and patrons: leaders and groups who can provide and manage access. There is a ne line here between sustaining empowerment and falling into time-honored clientelist patterns, and the line is easily blurred. One need not have the complex pattern of dependence of PRI-controlled Mexico at its height (Eckstein 1977) to recognize that groups and communities need allies in the state and the larger political arena and that these allies may and likely do have other agendas. State or party control of resources is critical here, hence the critical role often played by NGOs with autonomous resources in freeing groups from dependence on parties. The middleclass character of many of the neighborhood movements in Venezuela provides a roughly equivalent independence.

Urban Citizen Movements, Disempowerment in Peru, Venezuela

239

The institutional reforms put in place in Venezuela (in the 1990s) and Peru (in the 1980s) opened new possibilities for organization, representation, and action. These possibilities were taken up with great vigor in both countries: local governments were energized, and a profound process of political de- and realignment got under way. But gains proved short-lived, and both Fujimori and Chvez worked to recentralize politics, curtailing and limiting the reforms that had gathered force in earlier periods. The post-1992 Fujimori governments moved more and more into a populist mode, making citizen groups dependent on the state, and restricting independent access to resources. The Chvez regime made an effort (successful for a while) to bypass formal processes of interest mediation or representation in favor of a direct relation between the leader and the people (Kornblith 1993; Levine 2003; Salamanca 2004). This was a setback for the autonomy of social movements, and for decentralization, which had provided them with viable arenas for mobilization and action. In both countries the presence of NGOs weakened, as many transnational groups turned their attention and resources to newly opened elds of action in central and eastern Europe. Much of the reevaluation of work on urban movements and empowerment has been linked to the literature on transitions to democracy and democratization. But more is at issue than regime change. We believe that the issues can be more effectively situated in a broad context of thinking about activism and social movements, and institutions. Notable cases of transitions to democracy present the following anomaly: citizen mobilization and new citizen groups that were prominent in campaigning for democratization declined, split, and often simply disappeared with the restoration of democracy. The anomaly lies not only in decline, which makes sense, given the availability of channels of action and of competition for resources and for supporters. Although decline was in all likelihood inevitable, the process was accelerated in key cases by naive and unworkable understandings of politics, and by untrustworthy and unreliable political allies. With the possible exception of Brazil, where the PT (Partido dos Trabalhadores, or Workers Party) has clear roots in the popular movements and has grown steadily at all levels, the common experience has been one of division and betrayal (cf. Blondet 1991; Lander 1995; Levine and Crisp 1998; Levine and Stoll 1997). If we reframe the problem in terms of activism and social movements, the anomaly presented by activism with disempowerment is easier to understand. Two points are critical. First, movements commonly emerge, grow, succeed or fail, and decline, moving through what Tarrow (1994, 156) terms a cycle of protest. What is distinctive about such periods, he writes, is not that entire societies rise in the same direction at the same time [they seldom do]; or that particular population groups act in the same way over and over, but that the demonstration effects of collective action on the part of a small group of early risers triggers a variety of processes of diffusion, extension, imitation, and reaction among groups that are normally quiescent. In this light, the proper question is not so much why groups do not survive, but what, if any, legacy they leave in new rules,

240

Levine and Romero

expectations, or capabilities. The second point is connected and has to do with the opportunity structure that urban citizens faceresources and institutional channels available, accountability, and access. Writers like Castaeda have argued that a focus on local-level organization and the delivery of good government offers the most promising path to a rebirth of the left and sustained empowerment of popular sectors in Latin America. The record is mixed on both counts. We nd a clear legacy of norms about rights and activism, but weakness at making enduring and representative connections. More often than not, surges of activism leave activists, at the end, at the mercy of a different charismatic leadernew face, same dependence. The record of institutional reform is promising but incomplete, with a reversal of many reforms. Despite widespread attention to institutional design and institutional engineering, failures of accountability are more the norm than the exception. The institution of provisions for referendum and recall of ofcials holds possibilities, but does little to address candidate selection, electoral rules, or the all too common impunity of police and lack of access to courts.

Urban Spaces and Urban Citizen Movements


The preceding considerations bring us to a closer look at urban spaces and urban citizen movements: to the city as a stage or arena for action, and to its citizens as actors. Both dimensions are important. As in much of Latin America (and the Third World as a whole), the urban context in Venezuela and Peru is marked by dominant capital cities and explosive growth, with bigger cities growing faster than smaller ones and all cities faster than rural sectors. In recent years, regional cities have experienced substantial expansion. Internal migration is the predominant motor of urban growth in both countries. In Venezuela, rural poverty, road construction, and urban investment paid for by petroleum sparked a process of migration, beginning in the 1930s, that has substantially emptied the countryside. Explosive urban growth came later in Peru, but when it came, it was magnied by extreme rural poverty and internal war that drove refugees to seek safety in the cities. Rural-to-urban migration in Peru produced a mixing of ethnic groups on a scale unknown in the past: people of highland Indian culture came to Lima, bringing cultural expressions (such as Andean music or the Quechua language) with them. In both countries, the new presence of migrants overwhelmed urban infrastructures (particularly in the capital cities of Lima and Caracas), creating urgent needs for water, transport, education, and other services, and, of course, for representation.7 The spatial conguration of urban expansion, and the availability of transport within the city, has had notable impact on organization, citizen movements, and empowerment. Our analysis of the emergence and problems of urban movements is a structural one: following Ecksteins pathbreaking work on Mexico City (1977), we situate movements in a context created by the political

Urban Citizen Movements, Disempowerment in Peru, Venezuela

241

opportunity structure of nation and city, and given specic form and content by the availability of resources and by those (NGOs, unions, political parties) present and competing to provide services, orientation, and leadership. This strategy makes theoretical sense. It also makes practical sense, given the difculty of arriving at reliable estimates of the numbers of movements and associations existing and active at any given time.8 We provide estimates where possible, but urge caution in relying solely on the numbers. In Peru, the career of urban movements did not follow the track of other social movements in the sense of a steady accumulation of forces. Rather, after each successful mobilization, the movements seemed to fade away. As one local leader said in despair, once electricity was obtained and public lighting was in place, They buy a TV set and stay at home. The same thing happened after struggling to get water and sewage for the neighborhood and getting their houses connected to the main service. Urban movements gained signicance and presence in Peru during the 1970s, when the public space was reduced by the presence of a military regime, with elections possible only within the private sphere with voluntary organizations free to assemble and elect their leaders. The state regulated these elections and acknowledged the right of elected leaders to negotiate for public services. These electoral practices and the experience of representation were important for the creation of an autonomous public space within the authoritarian regime. Toward the end of the decade, attempts were made to centralize neighborhood organizations in Lima. How can we best understand the empowerment of urban actors in the 1970s in the context of a changing political system? What was the meaning of the power that was being generated in these neighborhoods? From the perspective of class theories of accumulation of power, this was clearly a process of gradual upward social mobility, not a major transformation of power relations. But from the perspective of building citizenship, there was indeed a signicant change in terms of power: once subjects or clients, members of the movements became citizens with rights. The pursuit and exercise of political rights in the cities is conditioned on refashioning the cities as political arenas not only for protest (claiming spaces) but also as venues for classic kinds of representation, including the creation of relatively autonomous units of government. With the end of military rule and the advent of democratic politics in 1980, municipal elections opened a public electoral space for movements. Many former movement leaders became mayors or city council members.9 During the 1980s, the core agenda of major urban movements underwent a notable change in Peru. Housing and public services eroded, and former migrants had new issues of concern, new demands to press. Tanaka (1998, 117) notes that struggle was centered increasingly within the private sector: Achieving basic services and the consolidation of the urban scene has changed in a radical way the priorities of the urban settlers (pobladores), giving rise to a new pattern of

242

Levine and Romero

meaning regarding participation, collective action, and membership in organizations. Attaining public goods lost its centrality and needs related to private goods became more important. Currently, most new land invasions lead to confrontations with groups of working-class landholders associated in cooperatives, instead of the earlier struggles against big urban landholders or the state leading to faceto-face struggles more than initiating social movements sustained in collective beliefs. In some sense, urban demobilization responds to urban development and to municipal administration of once self-managed neighborhoods. Cities grow slowly into the margins of the old invasions or climb higher into the hills, in the process transforming single-family houses into multi-family dwellings. In Lima, mobilizations have typically had very concrete goals (garbage collection, security, housing and land titles, water, electricity, parks and green spaces) that were easily assumed by municipal governments. Candidates for municipal elections are now commonly seen as potential experts on city management rather than as mobilizers or politicians, and provision of public services has become a core issue of campaigns. This has contributed to the multiplication of candidates and the short lives of many local movements. Once the neighborhood is converted into a municipality, the local voluntary organizations resemble those of any other part of the city: sport clubs, cultural associations, school parents associations, Christian communities, market vendors, teachers unions, and so on. As political and economic crises became more acute throughout the 1980s, the more dynamic movements became those around survival: those dedicated to providing food, resisting unemployment, and literally defending life from both terrorism and an arbitrary, repressive state. Social actors that were organized and could participate in public demands were mostly middle class: teachers, nurses, medical doctors, public employees, and public transportation workers. The sustained economic crisis, ties to foreign debt, and repeated structural adjustment packages weakened both businesses and unions. The result, throughout the 1980s, was a growing demobilization of the masses, aided by a deadly mix of terror and repression that began to rise sharply after 1980 with the sudden appearance of Shining Path on the national scene. The particular character of this crisis helps explain the prominence of survival organizations in city life. The Glass of Milk municipal program, formed under Alfonso Barrantes, the Lima mayor elected as a member of United Left in 1983, distributed a million glasses of milk every day nationwide to preschool children, mothers with newborns, and later to tuberculosis patients. There were also soup kitchens, known as popular dining rooms (comedores populares), some self-managed, and others sponsored by Catholic parishes and party-inuenced organizations such as the APRA-linked Mothers Clubs and the popular kitchens sponsored by AP. The most common pattern was that a group of women got together, cooking in one of the members homes and selling meals

Urban Citizen Movements, Disempowerment in Peru, Venezuela

243

for a nominal price to their members, who would then pick up the meals and take them home to eat with their families. Complementary aid came from NGOs or the state, sometimes through the donation of cooking equipment such as stoves or pots and pans, and also through the regular provision of food, including oil, rice, or wheat. Voluntary work by members, and their own contributions in nancing the food, is central to the operation of this kind of organization, and may be a reason why they often do not reach the poorest families in neighborhoods. These and similar organizations changed how politics was conducted and representation conceived. In September 1988, after the rst wave of structural adjustment policies, they organized a huge mobilization under the slogan Protesta con propuesta (Protest with Proposals). They demanded support from the government to buy food from local producers instead of importing it from abroad. This linked aid for the poor to rural development, joining the agenda of urban movements to peasant demands. Soon after, the Church started the Emergency Social Program (PSE), offering resources to channel international support for the poor. The program opened a public space where the leadership of the comedores, the entrepreneurial association CONFIEP, NGOs, and the Catholic Church worked together to elaborate an emergency program. This program continued under the Fujimori government as the Social Emergency Program (PES). In 1990, the leadership of the comedores decided to institutionalize their right to receive public funding to feed the poor. The various organizations (independent and related to political parties) joined together toward this aim, lobbied legislators from different parties, and achieved their goal at the end of the year with a law, promulgated by Fujimori in 1991, that recognized the responsibility of the state to feed the poor. A successful organization of comedores led the women into the streets to oppose the terrorist movement Shining Path. In a mobilization at El Agustino, they marched against the general strike called by Shining Path, chanting ni con hambre ni con balas (neither hunger nor bullets) to proclaim their autonomy and courage. Events like this cost many people their lives, either during the protests or in their aftermath. Once the distribution of food became centralized by the government in the mid-1990s, the leadership, which was usually elected or rotated among the members of the associations, was replaced by personnel from the same organizations but loyal to the Fujimori regime. There were over two thousand self-managed comedores, three thousand Mothers Clubs, and more than seven thousand Glass of Milk committees in Lima alone. Many political cadres from these organizations joined the political movements Fujimori created for each new election as independents. Women candidates, leaders of the soup kitchens, were elected as council members in the municipalities, aiming to represent their organizations own interests, but ending up as part of the political establishment, dragging their former constituency with them.

244

Levine and Romero

Throughout the 1990s, continued violence and economic crisis undercut the vitality of urban organizations, making it difcult to hold open assemblies and discussions in the neighborhoods, and hard to elect new leaders. Urban citizen movements were caught between terrorists, on the one hand, and a repressive and controlling government, on the other. The main goal of terrorist organizations, most notably Shining Path, was to control territory within the city, and neutralize, co-opt, or eliminate competing groups and leaders. For the same reasons, in the name of national security, a central goal of the regime was to control the neighborhoods and establish secure ties with the population. Authorities were suspicious of autonomous organizations; disempowerment and control were the dominant state strategies. Major mobilizations during the early 1990s were linked to the killing of grassroots leaders and to massive demonstrations of solidarity, often around funerals, as was the case following the public assassination of Mara Elena Moyano by Shining Path. Later, with the coming of peace, there was a notable political vacuum since no political parties were working among grassroots groups and only the Catholic Church, the evangelicals, and NGOs remained to organize what was left of civil society. It was only toward the end of the second Fujimori government (1995 2000) that the people recovered the streets and public squares as arenas for assembly and protest. In Venezuela, as in Peru, urban space (above all, the streets, plazas, and neighborhoods of the capital city) is a prime arena for political action of all kinds: from rallies, demonstrations, and marches, to street ghting. Urban mobilizations played a central role in the overthrow of the countrys last dictator, Marcos Prez Jimnez, in January 1958. Urban land invasions and the formation of vast new shantytowns remained a prominent feature of city life through the early 1960s, but have since faded. As noted earlier, a different kind of urban movement came onto the national scene decades later with the emergence of civil society as an actor in national politics and the concerted drive to create spaces and vehicles for that action. Neighborhood associations (vecinos) were formed, with the initial goal of urban development and defending property rights. Their agenda soon expanded to include pressure for greater municipal autonomy, and the scal and electoral reforms this entailed. Early neighborhood associations began in the 1970s in a series of middle-class areas of Caracas. FACUR (the Federacin de Associaciones de Comunidades Urbanas, or the Federation of Urban Community Associations) was established in 1971 as a coordinating body for these associations. FACUR provided a model for associations and similar regional federations that soon began to spring up all across the country. By the early 1990s, there were federations in every state, which together grouped an estimated total of about fteen thousand associations. In 1987 the neighborhood movement succeeded in gathering 140,000 signatures on petitions asking for a reform of the basic law governing municipalities (Ley Orgnica del Regimen Municipal, or LORM). This was one of the most important nonviolent mobilizations to that

Urban Citizen Movements, Disempowerment in Peru, Venezuela

245

date in Venezuela. Changes to the law included the election of governors, the election of mayors, the creation of parish councils, and the possibility of recalling ofcials. The impact of the movement was magnied by the school for neighborhood groups, the Escuela de Vecinos de Venezuela, or EVV. The EVV arose out of classes within FACUR, and consolidated on a national level in the mid-1980s with important support from business and from national and international NGOs. Since that time, the EVV has established regional ofces, mounted regular programs of courses for associations and local public ofcials, and maintained a range of correspondence courses, periodic meetings, and media presentations. EVV leaders have generally resisted pressure to form a political party, preferring instead to spin off a series of pressure groups, each devoted to a specic issue. Examples include Queremos Elegir (We Want to Elect), a group devoted to electoral reform; Fiscales Electorales de Venezuela (Electoral Ofcials of Venezuela), dedicated to promoting citizen involvement in supervising voting sites; and Venezuela 2020, an organization that promotes workshops and roundtables concerned with the shape of the countrys future. In other words, not a party but something more like civil society (Garca Guadilla and Silva Querales 1999; Gmez Calcao 1998; Lander 1995; Levine 1994; Levine, Crisp, and Rey 1996; Salamanca 2004). The term civil society came into wide use in Venezuela only in the 1990s. Until then, the political parties founded in the 1940s and the political system consolidated around them after 1958 encapsulated the expression of organized social life through party-controlled networks. Much contemporary theorizing (Escobar and Alvarez 1992) depicts the emergence of civil society as, above all, defensive. The neighborhood movement, which began as uncoordinated efforts by urban middle-class citizens to resist unplanned city growth and to defend their neighborhoods, is a case in point. The emergence of the human rights movement is another. Human rights organizations began to appear in the 1980s in response to specic abuses and to challenge long-standing practices of ofcial (especially police) impunity.10 They gained national stature and impact in the wake of the Caracazo (massive riots in Caracas) of February 27, 1989. Mounting violence throughout the 1990s kept them in the public spotlight. What these groups had in common was an effort to mobilize opinion (and people) outside the existing network of organizations controlled by the countrys political parties.11 Through the 1990s, as the political crisis grew and political parties were blamed by many Venezuelans for all the countrys problems, civil society became a catchall banner for reform and right-thinking activism. Once in power, the Chvez government made an effort to put its rhetoric of participatory (as opposed to representative) democracy into practice through a series of provisions in the constitution that make a place, at least in theory, for the active participation of civil society in politics. The 1999 Bolivarian

246

Levine and Romero

Constitution, for example, provides that the legislative councils of the different states consult with civil society on matters of interest to the states (Art. 206) and that civil society nominate three members of the national Electoral Council, charged with managing elections (Art. 296). Similar provisions are scattered elsewhere in the constitutional text. As a practical matter, disputes about how to dene civil society made it almost impossible to gure out who could and should be recognized as speaking in its name. The steadily increasing polarization of the country has also made civil society a highly contested term: both pro- and anti-Chvez groups claim to speak in its name, denying legitimacy and authenticity to the other. The results are occasionally anomalous: for example, although in theory civil society was to participate in evaluating candidates for the citizen power, in practice the president took this task upon himself, on the grounds that because a majority of the population had voted for him, he was the proper representative of civil society (Salamanca 2004). More often, lately, the results are confrontational and too often deadly, with groups clashing in the streets. It is not easy to come up with reliable estimates of the scale of the phenomenon. One review estimated that the total number of civic associations in Venezuela ranges from about twenty-ve thousand to about fty-four thousand (Salamanca 2004). Of these, the largest proportion are neighborhood associations, with a substantial number of groups that specialize in promotion and development, working with government and international resources. There is also a strong but regionally concentrated cooperative movement, and a signicant although numerically small network of human rights organizations. Not all civic associations are mobilizational in character or intent. There are music groups, civic theaters, cooperatives, and sports clubs, and a host of related groups whose logic and daily life need have little to do with mobilization and political confrontation. But these and other groups are linked to politics (and thus to the state) in two important ways that draw them into the partisan arena. First, many if not most seek and receive resources from the state. Even in times of economic decline, the Venezuelan state remains a powerful source of nancing and material resources for groups of all kinds. Second, the steadily increasing rhythm of mobilization and polarization since the late 1990s has made it difcult for groups to keep apart from political division and on the margins of confrontation. Indeed, the years from 1989 to the present are arguably the most protest-lled period in the last one hundred years of Venezuelan history: one massive urban uprising, two attempted coups, the impeachment and removal of one president, and a rising tide of violent actions in the universities and on the streets. Protest surged following the Caracazo in February 1989, and after a short respite under the second government of Rafael Caldera (1993 98), the rhythm of demonstrations, marches, and street protests picked up again as the country entered a new electoral cycle. To be sure, urban protests, often violent, had never

Urban Citizen Movements, Disempowerment in Peru, Venezuela

247

completely gone away. Student activism, sparked by regular violent actions under the leadership of encapuchados (literally, hooded ones, students with hoods to shield their identities), regularly spilled over from campuses into the streets. Protest and the scope of confrontation broadened with the election of Chvez as president. Lpez Maya (2002) shows that, among the kinds of protest, confrontational actions showed the strongest increase in 1999. Her gures of course do not include events beginning in late 2001, when protest grew and mobilizations, marches, and clashes became the daily bread of urban life, not only in the capital city of Caracas, but throughout the country: mobilization and countermobilization, rally and counter-rally, with massive marches following one another at ever shorter intervals. Protests, occupations of buildings, and coordinated actions involving banging of pots and pans (cacerolazos) or blowing of whistles or car horns (bocinazos) became everyday occurrences. Events reached their rst crisis point in the bloody confrontations of April 1114, 2001, when snipers red on a huge march making its way through Caracas to the presidential palace. The president was ousted and returned to ofce a few days later. After a brief respite while all sides stepped back from the brink, protests, marches, and countermarchesthis time all over the countrybegan again, coming to a second crisis point in the civic strike that began at the end of 2002. The leadership and organizational backbone for the opposition evolved quickly, starting with a pact between the trade-union movement, the business federation, and the Catholic Church. Union leaders, fresh from defeating the government in a referendum, played a critical role in the day-to-day organizing of protest activities. They were soon joined by political party activists, human rights groups, and others, as a range of new coordinating groups were put together (e.g., the opposition Coordinadora Democrtica, or Democratic Coordinator, in the summer of 2002). For the present purposes, the most striking features of this whole process are the central roles played by old-line organizations such as business and union federations, how efforts to resolve protest are undermined by the weakness of leaders on all sides, and the predominance of extremists, free to act given the utter demise of a professional political class used to negotiation and compromise.

Religion, Mobilization, and Public Space in Peru


The Catholic Church in a number of South American countries opened a public space to meet, associate, and participate under dictatorial regimes. When other public spaces were closed and forbidden, assemblies or Christian prayer groups were open for the faithful, and were used by citizens to exchange information, listen to others, form opinions, and circulate rumors, ironic stories, and hopes for the future. In Peru, this role has been secondary. Even in the worst times of the dictatorship, between 1968 and 1980, freedom of association and of public

248

Levine and Romero

assembly were accepted, although state-sponsored organizations regularly competed with autonomous ones for members and public voice. In this context, Christian communities offered a complementary space for association and critical reection, contributing to the quality of leadership and stimulating involvement in other organizations and in the political realm. There was a clear distinction between the public space of religion and the public space of politics. Committed Christians were, for the most part, careful to act on their own and not in the name of their particular church, Catholic or evangelical. This experience underscores the role religion can play in empowerment and disempowerment, but not necessarily in building representation. Until 1980, Catholicism was the established church in Peru and the state was confessional. The Catholic Church expected to play a prominent public role. These expectations were not abandoned with the ofcial separation of church and state in the 1980 Constitution. The Churchs prominent role in the struggle against poverty, and later in the promotion and defense of human rights, opened new areas for common action with other organizations in civil society, as well as with international agencies sharing the same goals. In 1988 the Church created a space for bringing together different actors in the Emergency Social Program, including international cooperation agencies, business entrepreneurs, and grassroots leadership. Later, when repression and terrorism continued throughout the nation, the Church supported human rights organizations, putting its newly gained religious legitimacy behind its pastoral agents clergy, nuns, and laityto care for the relatives of people missing and tortured and the innocent in prison. Drzewieniecki (2001, 4) writes:
In many parts of the country, clergy and Catholic lay workers developed new, more egalitarian ways to work with the poor through parishes and the expanding network of Christian base communities. These Catholic activists as well as CEAS (Comisin Episcopal de Accin Social or Episcopal Commission for Social Action), whose human rights department was founded in 1976, played a very important role in the development of human rights work in many different areas of the country. CEAS became one of the most important human rights organizations in the country and played an important role in the creation and institutionalization of the Coordinadora Nacional de Derechos Humanos [National Human Rights Coordinator].

During the Jubileum campaign, called by Pope John Paul II and joined by an ecumenical movement to lobby for forgiveness of the foreign debt to the poorest countries, the network of base Christian communities in Peru under the leadership of CEAS (Comisin Episcopal de Accin Social) collected the largest number of signatures among the participating countries around the world. In Peru, the same network provided volunteers for Transparency, an NGO formed to promote fair elections through election observation. In the 2000 and 2001 elections, the volunteers played a key role.

Urban Citizen Movements, Disempowerment in Peru, Venezuela

249

Commitment and involvement come together in compromiso (commitment), a key word for Christians throughout the Catholic Church in Peru. The kind of compromiso at issue here works more for participation in general than for involvement in politics specically. A clear distinction is drawn between the political public sphere in which people can participate as citizens, and the social public sphere created through the creation and practice of organizational life. This distinction is reinforced by the current experience of participation in church communities and parishes, and in social movements that remain alienated from institutional channels for representation. The Catholic Churchs option for the poor has contributed to a growing awareness of identity embedded in common interests and culture that cuts across different classes and ethnic divides, and even across different parties. But this religious awareness has not had a similar intellectual elaboration in other elds such as literature or politics. Institutional politics and modern culture remain distant from the recently included citizens who often feel themselves marginalized or alienated from public agendas. Demands for cultural representation at the institutional levelin Congress, government, the arts, and mass media have been added to those of economic interests. And newly appointed bishops in such important cities as Lima, Arequipa, and Trujillo are not helping to bridge the gap or to link the elite with the citizenry.

Disempowerment as the Future of Empowerment


The combination of activism and massive citizen mobilization with disempowerment joins the social and political trajectories of Venezuela and Peru in an unexpected convergence. From different starting points and the most varied social, organizational, and political traditions, these two nations have arrived at a shared space that does not augur well for citizen empowerment or representation. The decay of political institutions, including but not limited to political parties, has left Venezuelans and Peruvians with space for the creation of civil society a space they have lled, as we have seen, with great energy and creativity. But in the absence of reliable and trusted political intermediaries either formal institutions or political partiesthese energies are rarely converted into sustained and authentic representation. Is disempowerment the future of empowerment? Reading back from recent waves of massive, mostly urban mobilizations in each country (anti-Fujimori in Peru, and pro- and anti-Chvez in Venezuela) may provide some clues. The decay of institutions is arguably greater in Peru, where the allencompassing corruption of the Fujimori-Montesinos system of rule spread discredit very widely. Taking a longer view, the very idea of citizenship remains uncertain in Peru: although voting is obligatory, the regular rhythm of alternation between authoritarian and democratic regimes (with change coming about

250

Levine and Romero

every ten years or so) has meant that as Peruvian citizens approach legal voting age they cannot be sure if elections will be held at all, or what the electoral rules will be. Individual or civic forms of citizenship began to take signicant form in Peru as a result of the struggle against Shining Path and the MRTA (Movimiento Revolucionario Tupac Amaru, or Tupac Amaru Revolutionary Movement). Earlier movements have conquered citizen rights in the areas of economic and social rights, and later in politics, in the struggle to guarantee rights of association and public demonstrations. Individual rights have been third-generation, reversing the order outlined by Marshall (1983). With strong links to transnational human rights networks, civic activists developed networks and actions within Peru that soon gained international impact (see Burgerman 2001 on El Salvador and Guatemala). Although twenty years of activism have earned these groups considerable social recognition, such that they are now considered an important part of civil society, such organizations remain geared to working with victims, and not to building a membership base. The end of authoritarian rule and the return of democratic institutions and the rule of law has not cleared the slate of human rights issues. Much remains on the agenda, as the creation in Peru of a Truth Commission to review the past twenty years demonstrates. Human rights groups have begun to broaden their agendas to include social and economic rights as an integral component of human rights. Activism and pressure has also continued, with specic concern for the countrys political transitionthe dismantling of the Fujimori regime and the reconstruction of democracy and political rights. Explicitly democratic groups have been formed, rst among university students in opposition to Fujimoris reelection and in defense of the Constitutional Tribunal that rejected his (and the Congresss) efforts to provide a basis for his reelection (Tanaka, this volume). These struggles were reinforced by the efforts of womens collectives (Mujeres por la Democracia), as well as by artists and people from the media, in such groups as Resistencia, with a creative adaptation of forms of protest from other countries, such as the weekly washing of the national ag, sweeping the area in front of the Congress, or the mounting of Walls of Shame in various places, where passersby could post their ideas, photos, drawings, or commentary. There was also a series of street actions, rallies, and demonstrations in the plazas of Lima. Many of these elements came together under the direction of opposition political parties in the massive Marcha de los 4 Suyos, held on the very day on which Fujimori was sworn in for his third, and short-lived, presidential term. To this point, the effort to construct democratic institutions in Peru has placed emphasis more on institutions than on resources or actors. The new cutting edge of urban struggles and mobilizations is taking place outside Lima. In Iquitos, the Regional Front, strengthened during the border negotiations with Ecuador carried out under Fujimori, is advancing claims for resources for development and

Urban Citizen Movements, Disempowerment in Peru, Venezuela

251

demanding that benets go to Peruvians before they go to Ecuadorans. There are also active regional fronts in Tacna (on the frontier with Chile), in Puno, and in Madre de Dios (on the frontier with Bolivia). These are not urban movements because they include peasants, but they are organized primarily in provincial cities by businesspeople, academics (from the local universities), and local ofcials. Recently initiated processes of decentralization have added dynamism and resources to these organizations, and in this way have opened spaces for the emergence of new groups and leaders. In Venezuela, new citizen movements and forms of protest, indeed the very idea of civil society as an autonomous space for organization and action, appeared within an already established democracy. Their goal was not to challenge or overturn authoritarian rule, but rather to broaden or deepen that democracy by loosening the constraints imposed by a moribund but still all-controlling set of institutions and tacit rules centered on the political parties. The historical track thus differs from Peru, but the resulting situation is surprisingly similar. The creation of a movement around Hugo Chvez Fras, his rise to power, and the implementation of a Bolivarian revolution drew strength from the discredit of the old system and the implicit association of the movement with civil society, at least in rhetorical terms. The whole process makes sense as part of a general onslaught on the old system, its institutions, and its operative rules. The very word representative barely appears in the Bolivarian Constitution of 1999. Instead, Venezuelan democracy is and always will be democratic, participative, elective, decentralized, alternative, responsible, pluralist, and with revocable mandates (Art. 6). The results have been meager. New institutions have either foundered or never made it off the drawing board. In both countries steep economic decline and political deadlock following close on the heels of apparent euphoria (the resignation of Fujimori, the election of Toledo, the victory of Chvez, his removal and restoration in 2002, surging opposition, and growing violence) have combined to make sustained activism harder for many ordinary people. As a result of the extraordinary Venezuelan civic strike, individuals, businesses, movements, and national nances suffered major costs. Although the decay of parties sets groups in both countries free, in the same measure it sets them adrift and leaves them at the mercy of a supposedly direct relation to the leader, whoever that may be. Civil society constructed in this way is unlikely to yield enduring organization, and all too likely to be dependent on, and ultimately betrayed by, personalist elites as unaccountable as their populist predecessors. Without strong and durable organization, civil society is unlikely to provide coherence and direction for a complex and conictridden society. The preceding observations underscore two points. First, it is clear that a conventional focus on representation is not adequate to capture the experience of social movements, empowerment, and disempowerment that we have recounted

252

Levine and Romero

here. These movements operate on a different terrain, a terrain they help to create precisely because conventional vehicles of democratic representation are difcult to access and fail to satisfy their expressed needs. At the same time, it is important to acknowledge that the failure of groups self-dened as civil society to negotiate the transition from mobilization to formal political representation weakens the groups, weakens the representation (including the formal system), and weakens democracy by disillusioning many who put their hopes in organization as a way of creating something authentically democratic. It will not be easy to solve the puzzle of mobilization with disempowerment. Part of the difculty is practical: obstacles of all kinds litter the path of those who try. There are also theoretical problems to address. Much thinking about empowerment, citizenship, and representation is caught somewhere in between reection on social movements and analysis of institutional design, on the one hand, and examination of efforts to expand citizen access and participation in already existing arenas, on the other. We believe that the problem needs to be refocused on specic ways to provide enduring and legitimate form to this new participationin other words, to institutionalize it. This is beginning to happen in Peru through the Round Tables for the Fight against Poverty, established at district, provincial, and departmental levels. There has also been the National Accord, with participation of political parties, regional fronts, and the Mesas de Concertacin (Negotiation Roundtables); and the Truth Commission. All these can be understood as kind of an end run around existing political spaces, institutionalizing empowerment in new places. In Venezuela, despite provisions in the 1999 Bolivarian Constitution for referenda and citizen-managed forums of all kinds, none of this has been institutionalized. The mounting polarization and the acute political crisis of the country has, if anything, exacerbated the fragmentation of political forces to such a point that agreement on such spaces (let alone on who should participate in them) is very unlikely. As we noted earlier, the decentralization initiatives of the 1990s have been cut short, and some institutional reforms (e.g., of the electoral system) have aggravated, rather than ameliorated, problems of representation. The future of urban citizen movements does not look as good as it used to, and it is not easy to be optimistic, at least not in the short or medium term. It is important to be clear about the core problem. That organizations fail and leadership is unreliable or manipulative is itself nothing new. The difculty for movements, and for the potential for democratic representation in and through them, is not so much in the survival of any given organization as such, but rather in the creation of a kind of institutional safety net, something for groups to fall back on when times are hard. This is the role played, and played well, throughout Latin America by foreign-nanced NGOs, whose monetary and organizational autonomy provides an invaluable cushion. Such a role is long-standing in Peru and only now getting under way in Venezuela. Scholars also need to learn

Urban Citizen Movements, Disempowerment in Peru, Venezuela

253

what the members of movements know very well: that the agenda of urban citizen movements has changed. At issue is less the traditional range of urban demands: land, water, housing, transport, education, and security. Such concerns of course remain, but now and in the foreseeable future the pursuit and exercise of political rights in the cities are conditioned on refashioning the cities not only as political arenas for protest (claiming spaces), but also as venues for classic kinds of representation. Only with these in place, and viable connections to other levels of government, can the puzzle begin to be solved.

Notes
1. On Peru, see, among others, Tanaka (1998, this volume); on Venezuela, among others, see Coppedge (2002), Molina (2002), Levine (2002, 2003), Lpez Maya (2002), and the sources these authors cite. 2. The problem of violence and its aftermath is a central issue for Peruvian politics over the last quarter century. On the scale of the violence and its aftermath, see the nal report of the Commission of Truth and Reconciliation (Comisin de la Verdad y Reconciliacin 2004). On the scale of recent patterns of violence and protest in Venezuela, see Hernndez (2002) and Lpez Maya (2002). 3. The process is reminiscent of the rebirth of mobilization and protest in Chile, which Garretn calls the invisible transition that led up to the referendum that ended the Pinochet regime (Garretn 1989). 4. The name comes from the four regions, or Suyos, of the Inca empire, or Tahuantinsuyo. 5. Unionized workers are only a small proportion of the total workforce, and the business federation is of course not a mass organization. The two made common cause, drawing on the union federations successful defeat of Chvez forces in the union referendum of fall 2001 and the business federations strong opposition to a package of decree laws announced around the same time. Relations between the church and the government had been tense for some time, inamed by the presidents own erratic rhetoric (calling priests devils in cassocks, e.g.) and by his program for control and inspection of private education. The mass media have been a favorite target of the government since the beginning, and with rare exceptions, have responded in kind. 6. See Oxhorn (2001). Efforts must be systematically undertaken at the grassroots level to begin to empower people by helping them to be proud of who they areregardless of their social class, gender, ethnicity, religion, and so on. Studies have already demonstrated the success of such efforts to overcome peoples symbolic exclusion (14 15). 7. The pressure was such in Peru that a well-known book by Jos Matos Mar (1984) is entitled Popular Overow and the Crisis of the State in Peru (Desborde popular y crisis del estado en el Peru). 8. Contested denitions of what counts as civil society mean that such numbers remain in dispute (Oxhorn 2001; Salamanca 2004). 9. Representation was more effective when there were multiple electoral districts, as was the case in 1985 and 1990, since a single nationwide electoral district (1992, 1995,

254

Levine and Romero

and 2000) restricts electoral possibilities to the elites, leaving aside the new leadership emerging from recent movements. 10. Groups include PROVEA, COFAVIC, and Red de Apoyo Para la Justicia y la Paz; see Levine (1998) for details. 11. See the 1991 Annual Report of PROVEA, a major human rights group, which states: In contrast to earlier years, and basically during and after the National Protests of February 1989, it was possible to conrm that the social spectrum participating in protests is widening. Now participation in organized protests has opened elds of action for new groups: along with students and workers one nds a range of professional associations and social groups: doctors, nurses, peasants, Indians, remen, police, cultural workers, housewives, and neighborhood groups actively joining in movements in defense of basic rights (PROVEA 1992, 114 15).

References
Blondet, Cecilia. 1991. Las mujeres y el poder: Una historia de Villa el Salvador. Lima: Instituto de Estudios Peruanos. Burgerman, Susan. 2001. Moral Victories: How Activists Provoke Multilateral Action. Ithaca, NY: Cornell University Press. Comisin de la Verdad y Reconciliacin. 2004. Hatun Willakuy: Versin abreviada del informe nal de la Comisin de la Verdad y Reconciliacin del Per. Lima: Comisin de la Verdad y Reconciliacin. Coppedge, Michael. 2002. Soberana popular versus democracia liberal en Venezuela. In Venezuela: Rupturas y continuidades del sistema poltico (1999 2001), ed. Marisa Ramos Rolln, 69 96. Salamanca, Spain: Ediciones Universidad. Crisp, Brian F. 2000. Democratic Institutional Design: The Powers and Incentives of Venezuelan Politicians and Interest Groups. Stanford, CA: Stanford University Press. Crisp, Brian F., Daniel H. Levine, and Jos E. Molina. 2003. The Rise and Decline of COPEI in Venezuela. In Christian Democracy in Latin America: Electoral Competition and Regime Conicts, ed. Scott Mainwaring and Timothy R. Scully, 275 300. Stanford, CA: Stanford University Press. Dietz, Henry. 1998. Urban Poverty, Political Participation, and the State: Lima, 1970 1990. Pittsburg: University of Pittsburg Press. Drzewieniecki, Joanna. 2001. Coordinadora nacional de derechos humanos: Un estudio de caso. In Cuadernos de investigacin social. Cuaderno no. 17. Departamento de Ciencias Sociales, Ponticia Universidad Catlica del Per, Lima. Eckstein, Susan. 1977. The Poverty of Revolution: The State and the Urban Poor in Mexico. Princeton, NJ: Princeton University Press. Escobar, Arturo, and Sonia Alvarez, eds. 1992. The Making of Social Movements in Latin America: Identity, Strategy, and Democracy. Boulder, CO: Westview Press. Garca Guadilla, Maria P., and Nadeska Silva Querales. 1999. De los movimientos sociales a las redes organizacionales en Venezuela: Estrategias, valores e identidades. Politeia 23: 727. Garretn, Manuel Antonio. 1989. Popular Mobilization and the Military Regime in Chile: The Complexities of the Invisible Transition. In Power and Popular Protest: Latin American Social Movements, ed. Susan Eckstein, 259 77. Berkeley: University of California Press.

Urban Citizen Movements, Disempowerment in Peru, Venezuela

255

Gmez Calcao, Luis. 1998. Civic Organization and Reconstruction of Democratic Legitimacy in Venezuela. In Reinventing Legitimacy: Democracy and Political Change in Venezuela, ed. Damarys Canache and Michael Kulisheck, 169 86. Westport, CT: Greenwood Press. Granovetter, Mark S. 1973. The Strength of Weak Ties. American Journal of Sociology 78, no. 6: 1360 80. Hellman, Judith Adler. 1992. The Study of New Social Movements in Latin America and the Question of Autonomy. In The Making of Social Movements in Latin America: Identity, Strategy, and Democracy, ed. Arturo Escobar and Sonia Alvarez, 52 61. Boulder, CO: Westview Press. Hernndez, Tosca. 2002. El desafo de la violencia en el actual sistema poltico venezolano. In Venezuela: Rupturas y continuidades del sistema poltico (1999 2001), ed. Marisa Ramos Rolln, 289 311. Salamanca, Spain: Ediciones Universidad. Kornblith, Miriam. 1999. Agenda de reformas y crisis sociopoltica en Venezuela: Una difcil combinacin. Politeia 22: 83 120. Lander, Edgardo. 1995. Neoliberalismo, sociedad civil, y democracia: Ensayos sobre Amrica Latina y Venezuela. Caracas: Universidad Central de Venezuela, Consejo de Desarrollo Cientco y Humanstico. Levine, Daniel H. 1994. Goodbye to Venezuelan Exceptionalism. Journal of Inter American Studies and World Affairs 36, no. 4: 145 82. . 1998. Beyond the Exhaustion of the Model: Survival and Transformation of Democracy in Venezuela. In Reinventing Legitimacy: Democracy and Political Change in Venezuela, ed. Damarys Canache and Michael Kulisheck, 187214. Westport, CT: Greenwood Press. . 2002. The Decline and Fall of Democracy in Venezuela: Ten Theses. Bulletin of Latin American Research 21, no. 2 (April): 248 69. . 2003. El consenso democrtico venezolano en dos tiempos, 1972 2002. Politeia 30: 21 40. Levine, Daniel H., and Brian Crisp. 1998. Democratizing the Democracy? Crisis and Reform in Venezuela. Journal of Interamerican Studies and World Affairs 30, no. 2: 27 62. Levine, Daniel, Brian Crisp, and Juan Carlos Rey. 1996. El problema de la legitimidad en Venezuela. Cuestiones Polticas (Venezuela), no. 16: 5 44. Levine, Daniel, and David Stoll. 1997. Bridging the Gap between Empowerment and Power in Latin America. In Transnational Religion and Fading States, ed. Susanne Hoeber Rudolph and James Piscatori, 63 103. Boulder, CO: Westview Press. Lpez Maya, Margarita. 2002. Venezuela after the Caracazo: Forms of Protest in a Deinstitutionalized Context. Bulletin of Latin American Research 21, no. 2 (April): 199 218. Lora, Carmen. 2002. Sobre lo siniestro en el movimiento de mujeres. Pginas (Lima), no. 173 (February): 55 63. Marshall, T. H. 1983. Citizenship and Social Class. In States and Societies, ed. David Held et al., 248 60. New York: New York University Press. Matos Mar, Jos. 1984. Desborde popular y crisis del estado: El nuevo rostro del Per en la dcada de 1980. Lima: IEP. Molina, Jos Enrique. 2002. The Presidential and Parliamentary Elections of the Bolivarian Revolution in Venezuela: Change and Continuity (1998 2000). Bulletin of Latin American Research 21, no. 2 (April): 219 47. Molina, Jos Enrique, and Carmen Prez. 2000. Venezuela ratica el cambio: Elecciones de 2000. In Venezuela: Rupturas y continuidades del sistema politico (1999 2001), ed. Marisa Ramos Rolln, 143 76. Salamanca, Spain: Ediciones Universidad.

256

Levine and Romero

Ortner, Sherry. 1995. Resistance and the Problem of Ethnographic Refusal. Comparative Studies in Society and History 37, no. 1: 173 93. Oxhorn, Phillip. 2001. When Democracy Isnt All That Democratic: Social Exclusion and the Limits of the Public Sphere in Latin America. Miami: North South Center. Piven, Frances Fox, and Richard Cloward. 1977. Poor Peoples Movements: Why They Succeed, How They Fail. New York: Vintage. . 1998. The Breaking of the American Social Compact. New York: New Press. PROVEA. 1992. Situacin de los derechos humanos en Venezuela: Boletn de derechos humanos y coyuntura. Caracas: Programa Venezolana de Educacin y Accin en Derechos Humanos. Salamanca, Luis. 2004. Civil Society: Late Bloomers. In The Unraveling of Representative Democracy in Venezuela: Toward a New Model of Participation, ed. Jennifer McCoy and David J. Myers, 93 115. Baltimore: Johns Hopkins University Press. Sikkink, Kathryn. 1993. Human Rights, Principled Issue Networks, and Sovereignty in Latin America. International Organization 47 (Summer): 411 41. Smith, Christian. 1996. Resisting Reagan: The U.S. Central America Peace Movement. Chicago: University of Chicago Press. Stokes, Susan C. 1995. Cultures in Conict: Social Movements and the State in Peru. Berkeley: University of California Press. Tanaka, Martn. 1998. Los espejismos de la democracia: El colapso del sistema de partidos en el Per, 1980 1995, en perspectiva comparada. Lima: Instituto de Estudios Peruanos. Tarrow, Sidney. 1994. Power in Movement: Social Movements, Collective Action, and Politics. Cambridge: Cambridge University Press. Tovar, Teresa. 1991. El discreto desencanto frente a los Actores. Paginas (Lima) 111 (October): 25 39. Warner, R. Stephen. 1993. Work in Progress towards a New Paradigm in the Sociological Study of Religion in the United States. American Journal of Sociology 98, no. 5 (March): 1044 93.

9 Indigenous Politics in the Andes: Changing Patterns of Recognition, Reform, and Representation

Deborah J. Yashar
Latin Americas ruling classes, unable to wish Indians away, were quite happy to build nations without Indians, and this they have been trying to do for almost two centuries. To their chagrin, as the new millennium dawns, not only are indigenous peoples still presentand their numbers are rising, but they are actually challenging the very model of the nation-state that ruling groups have tried so conscientiously to build up. Rodolfo Stavenhagen, Indigenous Peoples and the State in Latin America

he third wave of democratization profoundly raised hopes and shaped opportunities for political representation. Yet in the wake of authoritarian regimes, the creation of new electoral institutions, the revival (and formation) of political parties, and renewed respect for human rights, much of Latin America appears to be suffering from a crisis of representation. This is evident not only in a diverse set of new democracies (e.g., Ecuador, Bolivia, Guatemala, Peru, and Argentina), but also in an older and smaller group of once stable, if limited democracies (Colombia and Venezuela) (Hagopian and Mainwaring 2005). As this volume has highlighted, this crisis of representation is particularly striking in the Andean region.1 More established political party systems have collapsed (with dominant political parties suffering a decline in support, credibility, and legitimacy) in Colombia, Venezuela, and Peru. Weak party systems have remained weak in Ecuador and Bolivia. Coups have occurred in Peru (Fujimoris 1992 autogolpe) and Ecuador (2000), and attempted coups have taken place in Venezuela (2002)although in each case civilians took ofce shortly thereafter. A former military dictator was elected president in Bolivia (1997), and those afliated with

258

Yashar

past military regimes have successfully won electoral ofce at subnational levels in Bolivia, Venezuela, and Argentina. Popular mobilizations, moreover, have toppled presidents in Ecuador (1997) and Bolivia (2003). State institutions are weakly and unevenly institutionalized, often subverting the norms, rules, and practices that allow for participation and representation in formal political circles. And the fate of more traditional social movements in the workplace, shantytowns, and countryside remains uncertain. Surveys of the region have highlighted, unsurprisingly, the thin legitimacy of Latin Americas contemporary democratic institutions.2 In short, the institutional foundations, sociological organization, and political imagination required for democratic representation appear weak indeed. This bleak political picture is tempered, however, by unprecedented organizing, claim making, and even representation by and for indigenous people. While the rest of the region appears stymied by weak political parties, emasculated social movements, and corrupt political institutions, indigenous people have forged national and international movements. Over the past three decades, these movements have proposed a set of reforms that include legal recognition, representation, autonomy, and bicultural education, among other things. Increasingly, these movements have given life to indigenous politicians, political parties, and consultants that have shaped political debates and sought to push through political reforms. In this regard, indigenous people have become a politically organized force and emerged as new claim makers in the political arena. Identity politics and democratic participation have thus intersected recently in unexpected ways in Latin Americagiving indigenous activists a powerful voice in civil society and indigenous politicians a new (although not always powerful) voice in political circles. This chapter discusses the changing terms and scope of ethnic representation in the Andes. While making reference to the region as a whole, it focuses particularly on the three countries with the largest indigenous populations in the Andean region: Ecuador, Bolivia, and Peru. The chapter is divided into four main sections. The rst section provides a brief comparative historical overview highlighting how Latin American states wrote indigenous people out of formal politics only to be confronted at the end of the twentieth century with the rise of signicant indigenous movements demanding recognition, representation, and reform. The second section steps back to problematize the concept of indigenous representation. Next, contemporary institutional reforms that have formally opened up new channels for indigenous people to seek representation are evaluated. The nal major section discusses the challenges faced by indigenous movements as they turn to partisan politics as a means to promote and represent an indigenous agenda. These four sections collectively highlight the phenomenal advances in indigenous claim making and the signicant obstacles to indigenous representation in the Andes.

Indigenous Politics in the Andes

259

Writing Indians Out of Politics, Bringing Indigenous People Right Back In


It is no surprise to argue that Latin American states vastly underrepresented indigenous peoples just as politicians consciously misrepresented them.3 What is perhaps less well understood are the ways in which twentieth-century modernizing projects did the same. Indeed, twentieth-century politicians attempted to write indigenous people out of politics through nation-state projects designed to create (or at least project) homogeneous nation-states (Stavenhagen 1992). Latin American politicians presumed (as did most scholars) that ethnic identities and cleavages were secondary and ephemeral. As such, politicians enacted reforms that not only sought to disguise the presence of indigenous people but actively sought to assimilate them. Scholars in turn presumed that ethnic politics were ultimately inconsequential for the study of Latin American politics in general, and political representation in particular. As such, Latin America was portrayed as a region wracked more by class than ethnic cleavages. In this section, I look below the proverbial political surface and briey highlight how corporatist projects of the mid-twentieth century essentially wrote indigenous people out of politics, all the while providing indigenous communities with the foundations to subsequently become a political force. At midtwentieth century, Latin American governments promoted corporatist projects of interest intermediation. As an entire literature has highlighted, these projects fundamentally set out to restructure society into functional groups that would access the state along class-based lines. While in some cases these class-based federations had ties to political parties that ostensibly represented them during democratic periodsMovimiento Nacionalista Revolucionario (MNR, or National Revolutionary Movement) in Bolivia, Alianza Popular Revolucionaria Americana (APRA, or American Revolutionary Popular Alliance) in Peru, Partido Institucional Revolucionario (PRI, or Institutional Revolutionary Party) in Mexico, and the Peronists in Argentinain others (Ecuador, Chile, Brazil, etc.) these institutions were created without any clear links to existing political parties.4 Studies of these corporatist projects provided lasting insight into the ways in which corporatism provided both inducements and constraints for Latin Americas labor and peasant movements. However, these same studies largely neglected the impact of these reforms on indigenous communities and ethnic cleavages. This oversight is surprising. Indeed, as part of these corporatist national projects, corporatist citizenship regimes attempted to turn Indians into national peasants in several of the Andean casesin the 1950s and 1960s in Bolivia, in the late 1960s and 1970s in Peru, and in the 1970s in Ecuador. In Bolivia, indigenous people in the Andes were mobilized into peasant federations in the 1950s, with the presumption that they

260

Yashar

would vote for (and be represented by) the MNR, the party that oversaw the corporatist project. During the military period of the 1960s, the Bolivian military subsequently sought to displace the MNR and to tie the peasantry to them, in what became known as the Pacto Militar-Campesino (Military-Peasant Pact). In Peru, the populist military government of Juan Velasco Alvarado (1968 75) also set out to mobilize indigenous people into peasant federations and other corporatist organizations, such as the Sistema Nacional de Apoyo a la Movilizacin (SINAMOS, or National Support System for Mobilization). Accordingly, Velasco declared the social death of Indiansreferring to indigenous peoples from that point on as Peruvian peasants, renaming the Day of the Indian as the Day of the Peasant, and encouraging indigenous people to refer to themselves as peasants. In Ecuador, the populist military government of General Guillermo Rodrguez Lara (1972 76) in particular also set out to create these corporatist ties, although his efforts to forge a powerful peasant federation were less successful than in Bolivia and Peru. Indeed, throughout Latin America, indigenous people gained access to land reform, social services, and other kinds of state-organized reforms only insofar as they joined peasant organizations and channeled their demands through peasant federations. Hence, indigenous people had strong incentives to publicly forsake their ethnic identities and to assume a class-based identity in union organizations and exchanges with political ofcials. In exchange for mobilizing into these peasant federations, indigenous communities in fact did gain access to land (some of which was communal), subsidies, services, and other benets. These were important resources that provided a modicum of political and material autonomy for communities that often maintained indigenous networks and forms of governance at the local levelbeyond the gaze of state ofcials and peasant federations. Corporatism, therefore, did advance material demands for many (although hardly all) indigenous people. However, it would be hard to declare that these corporatist projects represented indigenous people in any meaningful sense of the word. Indeed, given the rare and short-lived efforts to organize indigenous people into political parties, as in Bolivia in the late 1960s, it became commonplace to assume that indigenous identity was not an important political identity. The assumption among politicians and scholars alike was that politically mobilized people did so along non-ethnic linesin populist parties, class-based social movements, and the like. Latin American politicians complemented these corporatist measures with educational programs designed to promote assimilation alongside Indian institutes, which had been designed to study indigenous cultures and promote assimilation (although in many cases they simply languished).5 Such assimilationist programs were put into place in Peru and Bolivia, as well as Mexico and Guatemala, to incorporate people perceived as backwards into the ranks of a new, and presumably more civilized, nation.6 States encouraged indigenous men and women to discard

Indigenous Politics in the Andes

261

any public display of indigenous identity, encouraged the adoption of mestizo identities, and, consequently, publicly encouraged miscegenation to whiten the population. According to state ofcials and intellectuals, mestizaje would allow for social mobility as ones ethnic status changed from indigenous (other) to mestizo (one of us); this process would presumably depoliticize ethnic cleavages.7 What politicians and political scientists alike did not explicitly acknowledge was that these corporatist and assimilationist projects did not do away with indigenous identities. Rather they simply provided dual incentivesby encouraging indigenous people to assume a class-based identity in national public forums while at the same time providing the communal land base (via land reforms) to sustain ethnic ties and governance at the local level. Hence, class-based forms of organizing and representation masked, and in some cases nurtured, the survival of ethnic enclaves. Indigenous communities subsequently started to mobilize along ethnic cleavages beginning in the 1970s and continuing through the 1990s in Ecuador, Bolivia, Colombia, and elsewhere. Indigenous mobilization in the Andes was complemented by indigenous mobilization in the Amazon.8 Elsewhere I have analyzed why, where, and how indigenous movements developed in the late twentieth centurywith quite powerful organizations emerging in Ecuador and Bolivia, and quite weak ones emerging in Peru (Yashar 1998, 2005). In general, indigenous movements throughout the region initially developed to defend local autonomy. Local autonomy was increasingly placed in question as a result of state reforms that included the slowing down of land reforms, colonization, the cut in social services, and the opening up of land markets that had previously recognized the inalienability and indivisibility of communal landholdings. With community spaces in question, indigenous movements began mobilizing to defend their autonomy and to gain a stronger political voice. However, indigenous communities were not equally capable of mobilizing and were successful in doing so only where two additional factors were present: transcommunity networks and political associational space.9 Transcommunity networks (those left in place by unions, churches, and non-governmental organizations) enabled indigenous leaders to mobilize across disparate communitiesthose separated by geographic distance and language; the rst generation of indigenous movements emerged only where they could capitalize on these existing networks. In turn, political associational allowed indigenous movements to organize relatively free from repression. Both transcommunity networks and political associational space were present in various forms in Ecuador and Bolivia by the 1970s and 1980s; they were nearly absent in Peru by the 1980s, as a result of the civil war that destroyed transcommunity networks and foreclosed spaces for political organizing.10 As a result, strong regional and national movements emerged in Ecuador and Bolivia, and comparatively weak (some would argue largely nonexistent) movements emerged in Peru. In this regard, Perus weak and localized indigenous movements stand out in the 1980s

262

Yashar

and 1990s as a continental anomaly, given the widespread pattern of indigenous organizing in most other countries in the Americas. Indeed, in the last third of the twentieth century, Latin American indigenous movements emerged to defend local autonomy and to challenge the idea that indigenous identity and culture are anachronisms. For while many indigenous men and women outwardly assimilated into mestizo culture (leading to an ofcial decline in the absolute numbers of self-identied indigenous peoples), selfidentied indigenous communities have survivedalbeit, as all communities do, they have changed over time. Although the data on indigenous populations are problematic, the following data (gathered from 1979 to 1991)11 on the Andean countries are widely used to approximate the size of this population: Bolivia, 60 70 percent; Peru, 38 40 percent; Ecuador, 30 38 percent; Colombia, less than 2 percent; and Venezuela, less than 2 percent.12 The political challenge has become more than a question of demography and numerical survival. It has become a question of setting an agenda for indigenous people. By the mid-1990s, signicant indigenous movements had formed throughout the Americas, with the strongest organizations in Ecuador and Bolivia. In Ecuador, the most important group is the Confederacin de Nacionalidades Indgenas del Ecuador (CONAIE, or Confederation of Indigenous Nationalities of Ecuador). In Bolivia, there are several important groups: the Confederacin Sindical Unica de Trabajadores Campesinos de Bolivia (CSUTCB, or Unied Peasant Workers Trade Union Confederation of Bolivia); the Confederacin Indgena del Oriente, Chaco, y Amazona de Bolivia (CIDOB, or Indigenous Confederation of the East, Chaco, and the Amazon); and more recently, the cocalero, or coca growers movement. Moreover, there are prominent organizations in Colombia as well, and (as noted) the weakest organizational forms exist in Peru. Indigenous movements throughout the region have demanded formal recognition, local autonomy, legal pluralism, additional land reforms, and bicultural education, among other reforms. These movements have largely voiced their demands through classic social movement politics. They have organized unprecedented marches from the Amazon to the Andes (particularly in Bolivia and Ecuador), staged highway disruptions, occupied government buildings, and organized street protests against various political and economic reforms. From a comparative perspective, indigenous mobilizational capacity stands in marked contrast to the weakness of other social movements in Latin America, which have declined in strength since the heady days of anti-authoritarian protests. The groups ability to mobilize explicitly on behalf of indigenous people and to take the lead in more general societal protests against various neoliberal reforms has undoubtedly led them to assume the mantle as the central and most powerful social movement actors in Ecuador and Bolivia. Like social movement actors before them, they have largely mobilized outside of the halls of the state seeking to represent those that have been shunned by formal politicsand have used the power of their numbers to voice their claims. As such, they have made

Indigenous Politics in the Andes

263

the issue of indigenous people, organizations, and claims a part of political debate. And they have highlighted the inadequacy and injustice of efforts to erase indigenous identity as a basis for political mobilization and representation. While the movements and their demands as a whole are quite varied, the point to emphasize here is that indigenous people are organizing and articulating ethnic-based agendas in unprecedented ways. As such, indigenous people have increasingly emerged as a political force to be reckoned with. It is this contemporary mobilization of indigenous people that is so striking. Not only have indigenous people demanded recognition as such, but they have also demanded reform (including autonomy demands) and representation. This is happening precisely at the moment that the rest of the region appears to be undergoing a crisis in representation. So, the question becomes, to what extent has indigenous representation advanced in recent years? The following two sections take up that question.

Problematizing Indigenous Representation


This chapter has thus far highlighted the historic obstacles to indigenous representation in the Andes. Before turning to the question of if/how indigenous representation has advanced in the contemporary period, we must rst address the conceptual and analytical ambiguities that surround the question of indigenous representation in the Andes. The editors of this volume have discussed democratic representation as the process by which an agent expresses the stated interests of a principal within and before democratic institutions. With this as a conceptual base, one can conclude that indigenous representation refers to a political context in which indigenous people can elect ofcials and/or appoint delegates to act on their behalf. While on the face of it, such a denition appears to be self-evident, it is in fact inherently open to interpretation. In practice, this denition is better at indicating negative cases. That is to say, it is easier to delineate when indigenous representation is not taking place than when it is occurring or, rather, when it is occurring in a meaningful sense. This conceptual ambiguity results because the denitions of the noun (representation) and the adjective (indigenous) themselves beg important and interrelated questions about the principal (who is indigenous), interests (who denes them), agents (are they in fact representing those in whose name they speak), and locales (where does representation take place). Let me elaborate on the conceptual problematic.13

Principals and Interests


When we speak about indigenous people, who in fact is the principal? This seemingly simple question elicits no simple answer. First, there are no easily observable and agreed-upon measures for indicating who is in fact indigenous.

264

Yashar

Ethnicity is a conceptually slippery categoryall the more so now that scholars have come to recognize its constructed and changing boundaries. But of equal importance, indigenous people in Latin America have embraced their indigenous identities in some places while denying them in others. As such, we must keep in mind that identity is an elective and uid concept that is sometimes, but not always, primary. While one cannot credibly claim to be indigenous if one does not have indigenous ancestry, one can claim not to be indigenous; as noted earlier, this latter position was, in fact, advocated by state policies throughout the region. Given the uid nature of ethnic identity in the region, indigenous people should be dened as those people who self-identify as such; such selfidentication asserts shared ascriptive characteristics and a common history as the original inhabitants of what we now call the Americas. But even this denition confronts problems of measurement; indeed, basic demographic data on indigenous populations remain highly disputed. Even if we could denitively agree on who is indigenous and how many indigenous people exist, the question of what is meant by the term indigenous interests remains unanswered. And how one identies and measures indigenous interests presents yet another methodological (and politically charged) challenge.14 One might do so by aggregating individual indigenous preferencesthe stated interests of indigenous elders, the ndings of indigenous intellectuals, the agendas of indigenous movements, and/or the platforms of indigenous politicians. But denitions of indigenous interests are bound to vary, depending on the methodology used. For, as scholars of black politics and representation in the United States have debated for many years, there is no simple answer to this question (see, e.g., Swain 1995 and Tate 2003). Similarly, as Warren (1998) has so eloquently shown for Guatemala, debates among indigenous spokespeople about how to dene and pursue indigenous interests can be quite heated. I do not intend to settle this debate here. Rather, I highlight these numerous questions with two goals in mind. First, on a cautionary note, I do not want to suggest that there is some mechanical and universal way of identifying indigenous people, dening indigenous interests, and evaluating indigenous representation. To the contrary, the topic is complex and would require many booklength treatments. Second, and following from this rst point, it is incumbent on the author, therefore, to indicate how she is using the term indigenous representation. Here, I use two (admittedly imprecise, narrow, conict-laden, and non-exhaustive) measures to discuss indigenous principals and interests. On the one hand, I have used a micro-conception of indigenous principals and interests: to analyze when individual indigenous people choose to identify as indigenous and to select and elect representatives accordingly (without speaking to whether these representatives can and do defend indigenous interests). I implicitly use this micro-measure in the previous section to show how indigenous

Indigenous Politics in the Andes

265

representation was historically denied. In the next two sections, I use it to highlight how contemporary institutional reforms increase opportunities for indigenous individuals to self-consciously take part as Indians in their countries democratic process, and to highlight the rise in the number of indigenous people who have been elected to political ofce in recent years. On the other hand, I have also used a macro-conception of indigenous representation to analyze indigenous organizations and social movements as they voice collective interests within the formal political process. I have largely used this macro-conception when analyzing the challenges for indigenous representation, but it also has some relevance for the subsequent discussion of institutional reforms and how they advance indigenous representation. This macro-conception of indigenous principals and interests is in some senses thicker than the micro-concept, insofar as the relevant indigenous actors publicly claim their indigenous identity and self-consciously seek to promote some version of indigenous interests. While it is possible in this second scenario to identify a set of indigenous interests sanctioned by a particular movement, it is important to also bear in mind that movements themselves are not universally sanctioned, internally consistent, or made up of unied groups.

Agents and Interests


In discussing indigenous representation, one must also problematize the agent all the more so since it is often argued that elected and appointed ofcials are not the real or true representatives of indigenous people. There is no agreement over whether the agents/representatives of indigenous people must belong to the same social group (however dened) as the principals; whether he or she will act in the best interests of the principals, even when the principals disagree; and/or whether they are even accountable to the principals.15 Hanna Pitkin (1967), in particular, has discussed these issues in her classic work on representation, as has Jos Antonio Lucero (2002), who analyzes the complexity of indigenous movement representation in the Andes and puts forth a pragmatic constructivist understanding of representation.16 Drawing on (and greatly simplifying) the work of these scholars, I highlight two aspects of representation here: descriptive, or mirror, representation (which asks: Are the representatives from the same group as I am?); and functional, or guardian, representation (which asks: Are the representatives pursuing my interests?). While we commonly discuss identity-based representation in terms of the former, we tend to equate ideological and partisan-based representation with the latter. Both measures, of course, provide different sorts of insight. In my discussion, I make use of both aspects of representation: I discuss the fact that more indigenous people are gaining ofce (although this measure does not necessarily indicate whether these new political representatives are consciously

266

Yashar

or actively pursuing the interests of indigenous people); and I discuss the fact that more elected ofcials and authorities (regardless of whether they are indigenous themselves) are self-consciously pursuing agendas that have been articulated by indigenous social movements.17

Locale
Finally, in talking about indigenous representation, the question of the site of representation is particularly relevant. As discussed previously, indigenous people have historically been denied formal access to the seats of state power. However, this does not mean that representation has not occurred in other ways. Indeed, as highlighted below, indigenous people have often operated simultaneously in two spheresin political spheres legally sanctioned by the state and in political spheres recognized by customary law. This chapter focuses on the advances and setbacks of the former (in line with the focus of this volume), while recognizing the centrality of the latter to indigenous claim making. In sum, I cannot hope to adequately answer all of the complex questions associated with indigenous representation in this chapter. Given the complexity of discussing indigenous representation, I therefore delimit the scope of the questions discussed in this chapter. The next section highlights the contemporary political reforms that have created new opportunities for political participation by and for indigenous people. I largely discuss indigenous representation in its micro-analytic guise by looking at institutional changes that have promoted the opportunity for indigenous individuals to speak out and to be elected. The nal section largely discusses the dilemma of indigenous representation in its macroanalytic guise, as indigenous movements seek to take part in formal electoral politics.

New Institutional Opportunities for Advancing Indigenous Representation


The third wave of democracy has forged new institutional opportunities for indigenous people to participate in political debates, demand political inclusion and autonomy, and seek a greater voice and representation.18 This section outlines some key institutional and constitutional reforms that have manifestly opened up opportunities for indigenous people to seek political representation.19 Indigenous interests have found greater voicewhether these interests are dened in their micro-analytic or macro-analytic version. So, too, indigenous representation has found more institutionalized spaces for securing political ofcewhether these representatives are evaluated as mirrors or guardians of indigenous interests.

Indigenous Politics in the Andes

267

An underlying theme for this section, however, is that such institutional opportunities are fundamentally constrained by weak states and by political institutions that work against sustained representation of indigenous individuals, communities, and movements. Indeed, the very conditions that work against representation for all people in the Andes (the subject of this volume) also work against institutionalizing meaningful and sustained representation for those indigenous people and movements that seek to take part in the formal political arena.

Suffrage
The third wave of democracy obviously re-extended the rights of citizens to vote for their elected ofcials in Ecuador, Bolivia, and Peru. On the face of it, therefore, the current period of democratization has opened up channels for indigenous people to take part in electionsa micro-foundation for delegating authority to representatives in national politics. But the current period is more than a simple period of redemocratization. With the third wave, universal suffrage was extended in Peru (1979) and Ecuador (1980), a policy that overturned earlier literacy restrictions. Given the high rates of illiteracy among indigenous populationsparticularly those living in the countrysidethese measures had essentially prevented indigenous people from voting in the past. In this regard, the most recent round of democratization literally enfranchised indigenous people, thereby advancing opportunities for them to voice their individual preferences for political representation. As indigenous people started to organize as indigenous movements, political parties increasingly turned to indigenous communities to secure their votes. Indigenous people have found, therefore, that they have a greater political voice in elections than ever before. Obstacles, however, remained. In Peru, the escalation of civil war in the 1980s, with Sendero Luminoso (Shining Path) and the Movimiento Revolucionario Tupak Amaru (Tupak Amaru Revolutionary Movement), placed serious limits on the ability of people to freely take part in electoral and partisan politics. The widespread violence in provinces with large populations of indigenous people (particularly Ayacucho) essentially foreclosed the ability of people to come together independently, to publicly articulate political platforms, and to vote freely in elections. Registration, moreover, was less than universal. Registration (and access to voting booths) has posed problems, particularly in the Amazonian regions of the Andean countries discussed in this volume. In point of fact, registration was (and is) spotty and unreliable, most notably in the Amazonian basinwhere the state is weak and some communities remain itinerant. Finally, voting for established political parties does not mean that once in ofce these elected ofcials

268

Yashar

will represent those who voted for them. Indeed, in many cases there is considerable desencanto (disillusionment), as elected ofcials ignore communities between elections or try to placate them with minor clientelistic rewards. This observation is, of course, not peculiar to indigenous peoples.20 Rather, it is part of a larger dynamic whereby Latin Americans who cast their votes have disproportionately little faith that those elected (at the very least, those elected to national ofce) will pursue their interests (Lagos 2003a, 2003b). How else to explain efforts to force presidents out of ofce before they have completed their terms in Bolivia (2003 and 2005) and Ecuador (1997, 2000, and 2005); these anti-executive mobilizations are a testament to the weak faith in the electoral system as a vehicle for advancing interests and securing worthy representation.

Recognizing Multiethnic Citizenries


A second signicant reform included the explicit recognition of multiethnic citizenries. Constitutions now recognize the multiethnic and multicultural makeup of Colombia (1991), Mexico (1992), Peru (1993), Bolivia (1994), Ecuador (1998), and Venezuela (1999), among others in the region.21 Indigenous people participated in the rewriting of these constitutions, albeit in single-digit numbers, including seven indigenous representatives in Ecuador, three indigenous delegates in Colombia, and three appointed delegates in Venezuela (Van Cott, forthcoming). In several countries, constitutions and/or corresponding legislation recognize indigenous legal systems and authorities, although the actual terms of recognition remain ill-dened and largely unimplemented (Stavenhagen 2002, 32 33). These reforms are an important symbolic victory for indigenous peoples, who have worked to undermine myths of national unity.22 As Yrigoyen Fajardo shows, constitutional recognition is more than a symbolic victory since it represents a fundamental shift away from the monocultural premise of Latin American judicial institutionalism: The norms of such a system (indigenous customary law) were considered admissible only in the absence of law, but never as an alternative to it, in which case it would be dened as crime (Yrigoyen Fajardo 2000, 197). Hence, the constitutional recognition of ethnic heterogeneity in some Latin American states has broadened the public imagination and established a legal precedent for discussing ways to accommodate and represent a diverse citizenry, as discussed next.

Indigenous Seats
In a few cases, Andean countries have apportioned seats for indigenous people. The most recent Colombian and Venezuelan constitutions have guaranteed two and three indigenous seats in the national legislatures, respectively. Other reforms have given the nod toward securing some indigenous representation in municipal and regional assemblies and councils. Venezuela now guarantees

Indigenous Politics in the Andes

269

a seat on state-level assemblies and municipal councils in districts where they [indigenous] have a signicant presence; and Perus 2002 Ley de Elecciones Regionales (Regional Elections Law) included a 15 percent indigenous quota on regional party lists for elections in the overwhelmingly indigenous Amazon, although the law did not indicate where those names would be placed on the lists (Van Cott, forthcoming). These are historically unprecedented changes in these countries, for they recognize indigenous people as political actors with guaranteed rights to compete in elections and occupy political ofce. These reforms advance the mirror image of indigenous representation. These reforms, while revolutionary in concept, will not revolutionize the policy output or the scope of indigenous representation. The set-aside seats are few in number, making it hard to imagine that indigenous representatives can do more than join existing coalitions and voice their opposition to those policies deemed disadvantageous to indigenous peoples. Moreover, Van Cott (forthcoming) convincingly cautions against presuming that these seats will necessarily translate into the ability of indigenous people to forge their own organizations and choose their own representatives. Indeed, these mechanisms could provide yet another venue for more established political parties to make their mark at the expense of more local efforts to cultivate, vet, and propose indigenous candidates. If signicant and enduring indigenous representation in the state is to occur, indigenous people will have to mobilize for more than a handful of seatswhich, on their own, can be seen as only concessionary tokens.23

Decentralization
As Kathleen ONeill argues in her chapter in this volume, decentralization has also recongured the spaces and terms of participation and representation in the region. Among the Andean countries, decentralization in Colombia, Bolivia, Ecuador, and most recently Peru has increased opportunities for electing ofcials that are responsible to local constituencies, thereby changing the location or spaces in which representation can take place. I will not repeat ONeills argument here, other than to underscore the following key points. Clearly, decentralization has increased the opportunities for indigenous adults to elect representatives that will be responsible to local constituencies. In many places, the new reforms have increased the numbers of indigenous politicians elected to ofce. The rst elections in Bolivia, in 1997, led to a marked rise in indigenous councilors and subsequently to the election of a number of indigenous mayors (see Alb 2002, 82ff.). The same can be said for Colombia and Ecuador (see Van Cott 2002, 51, 65). With the 2004 reform in Bolivia that now allows candidates from social movements and indigenous movements to run for elected ofce (even when they do not have ties to a national political party), we will perhaps see that number rise even more.

270

Yashar

Also, it is worth noting that in all cases except Bolivia there is the legal possibility of creating some form of indigenous municipality (see Van Cott, forthcoming). However, in practice, these indigenous municipalities have largely remained legal prospects rather than realized forms. In Bolivia, the Law of Popular Participation does not allow for indigenous municipalities but does allow for indigenous communities to negotiate with existing municipalities to forge indigenous districts; these districts, however, have no prior claim to resources or authorityall of which must be negotiated with the corresponding municipalities. As such, there is a recognition of the right to choose indigenous authorities, but their purview is open to debate (Yashar 1998, 2005). Indeed, as ONeill, Gray Molina (2003), and others have shown, it is hard to generalize about the impact of decentralization on democratic representation, given widespread variation across cases, within countries, and over time. Indeed, several cautionary ags have been raised. While decentralization has increased the opportunity to elect local ofcials, this has not necessarily translated into an increase in indigenous voice, participation, and representation. Indeed, localizing politics has meant, in some places, giving greater opportunities for local, non-indigenous elites to assert control over local regions. Moreover, even in cases like Bolivia, where decentralization measures included oversight committees (comits de vigilancia) that institutionalized a role for territorially based communities, in practice, the committees have limited training and resources to effectively perform that job. As ONeill and Gray Molina show, more research is needed to chart out and explain this variation in outcomes. Finally, not all local governments have in fact received the kinds of nancial resources that would enable indigenous communities to elect ofcials. ONeill shows that there is a signicant variation in resource bases for different subnational governments. Citing 1995 IDB data, she notes that subnational spending as a percentage of total spending in 1995 ranged from a comparative high in Colombia (39 percent) and Bolivia (26.7 percent) to a middle range for Venezuela (19.6 percent), to a low in Peru (10.5 percent) and Ecuador (7.5 percent). In short, decentralization has not everywhere meant greater spaces for indigenous people to assume ofce and/or elect those who will have the resources to effectively act on their behalf.

Autonomy
Finally, several states have recognized some form of indigenous autonomy thereby creating new spaces for indigenous people to seek ethnic representation and jurisdiction. In these autonomous spheres, indigenous people can choose authorities and governing systems outside of the jurisdictional purview of central states, local municipalities, and national law.

Indigenous Politics in the Andes

271

New constitutions in Colombia (1991), Peru (1993), Bolivia (1994), and Ecuador (1998) have gone a long way toward recognizing (although not necessarily implementing) indigenous laws and norms, indigenous authorities and authority systems, and jurisdictional functions (Yrigoyen Fajardo 2000; Van Cott 2000, 2002). Moreover, national legislation has further institutionalized certain kinds of autonomy in several of these countries, particularly in the Amazon. In Ecuador, for example, following the thirteen-day, two-thousand-person march from Puyo to Quito organized by the Organizacin de Pueblos Indgenas del Pastaza (OPIP, or Organization of Indigenous Peoples of Pastaza) in 1992, the government eventually conceded nineteen different territorial blocs, encompassing 138 legally recognized communities and 1,115,475 hectares.24 In Bolivia, the state ultimately conceded four autonomous indigenous territories following the 1990 March for Territory and Dignity, organized by CPIB (Central de Pueblos Indgenas del Beni, or Central for Indigenous Peoples of the Beni).25 In 1996 the Bolivian government passed a new agrarian reform that provided indigenous communities with the legal basis for appealing for territorial recognitionincluding the right to vast expanses of land and the political autonomy of indigenous authorities. By August 1997 the state had recognized seven distinct territories totaling 2.6 million hectares, and it was processing thirty-four more demands totaling about 20 million hectares.26 And in 1998, the Bolivian government created a program to title 10 million hectares of indigenous lands (Plant 2002, 217). However, the Banzer administration (19972002) delayed the process of titling these lands (Van Cott 2002, 56), and third-party colonist and forestry concessions have also slowed down the process (Plant 2002, 209). Demands for territorial autonomy in Bolivia have been complemented by efforts to establish indigenous districts (not municipalities); with the 1994 Bolivian Law of Popular Participation (largely a municipalization and decentralization law), indigenous communities gained the right to request indigenous districts albeit with mixed results. The terms of these indigenous districts in Bolivia are underspecied; the law creates the possibility (but neither the obligation, terms, nor mechanisms) for establishing the districts. Consequently, indigenous districts are hard to negotiate and are not autonomous administrative units; they might, but do not necessarily, institutionalize the right to customary law, bilingual education, communal property, state resources, etc. The future and fate of these municipal districts depends on the mayor, who has the power to recognize them and to determine resource allocation, but has no legal obligation to do one or the other.27 According to Luz Mara Calvo, the former head of the Subsecretara de Asuntos Etnicos (SAE, or Subsecretariat of Ethnic Affairs), these districts have not emerged unblemished. To the contrary, most have confronted signicant problems due to poorly delineated legal rights and limited skills within the community.28 Moreover, it should be noted that the law does not delineate

272

Yashar

the precise terms of the relationship between the municipality and a hypothetical indigenous district. Indeed, the nancial and political relationship is largely up for negotiation between those proposing the indigenous district and the individual municipality. For example, it is up to the goodwill of the municipal government to decide if it will transfer any nancial and social resources (and if so, how much) to the indigenous district; indeed, the municipal government has no legal responsibility to distribute part of the co-participation funds to pay local ofcials salaries, or to cover administrative costs (see Balslev 1997, 35 41, 53 58, 86). According to documents from the Secretara de Participacin Popular (Direccin Nacional de Organizacin Territorial Administrativa), there were 127 such districts by April 1997; 29 by 2000, 138 had been formed (Velasco 2001, 42). In the 1991 Colombian Constituent Assembly, moreover, indigenous peoples also negotiated reforms that granted territorial autonomy (Clavero 1999, 187 89; Van Cott 2000). The 1991 Colombian Constitution referred to indigenous lands as territorial entities in Article 286; according to this article, existing political authority structures assume governing capacity, including criminal and civil jurisdiction, in these territories; moreover, the territories are responsible for determining their own development strategy and for administering public resources as if they were municipalities. We also nd states recognizing some version of autonomous regimes (reserves) in the Amazonian Basin in Brazil (Brysk 2000, 201). While concessions of territorial autonomy in each country have confronted serious obstacles in implementation (Plant 2002, 209), they constitute a signicant symbolic and legal precedent for indigenous movements as a whole. Beyond the Amazon, as well, the state has recognized indigenous communities as politically autonomous units, which can choose their own representatives according to customary law. Bolivias 1997 agrarian reform law, for example, opened up the possibility to recognize the ayllus (communal kinship organizations) that dot the Andean countryside. Ecuadors 1998 Constitution also created Circunscripciones Territoriales Indgenas y Negras (Indigenous and Black Territorial Circumscriptions), although the legislature had not acted on this possibility at the time of this writing. We have seen similar efforts beyond the Andean region, including in Mexico, Guatemala, and Nicaragua. Latin American states, therefore, have initiated reforms that recognize some degree of political autonomy for indigenous people. Indigenous movements have demanded this autonomy as a means of securing indigenous political jurisdiction over that land, including the right of indigenous legal systems and authorities to process and adjudicate claims. In this regard, autonomy reforms have provided a legally recognized space in which indigenous people can choose alternative locales, forms, and modes of representation. State autonomy reforms, however, must be assessed cautiously. Stavenhagen (2002, 34 35) concludes

Indigenous Politics in the Andes

273

that many of the autonomous agreements remain ill dened and weakly implemented in Colombia and Nicaragua, among others.

Challenges of Implementation, Questions for Representation


So far in this section, I have highlighted signicant and unprecedented constitutional and legislative reforms. Without a doubt, these reforms have created greater opportunities for indigenous individuals to vote for representatives, and have designated new formal political entities (e.g., the municipality and the autonomous region) where they can do so. Moreover, these reforms have understood the need to ofcially recognize indigenous people as citizens and to secure some spaces for participationboth in the state and within indigenous communities. These changes are without parallel, historically speaking. However, as I have argued throughout this section, one must temper these observations with a healthy dose of skepticism. There are serious obstacles to representation overall obstacles that indigenous people also confront. First, party systems remain weak at best, and absent at worst. In this context, the vote for indigenous representatives (however dened) is a tall order; for even where indigenous people overwhelmingly choose who they want to represent them, these new representatives face political battles, in which partisan lines, coalitions, and policy making remain nebulous, ever-changing, and often corrupt. While all is not bleak, the hurdles are high. Second, and related, indigenous representatives have often inserted new items onto the policy agenda. However, their ability to push these initiatives through is far from certain. As such, the national distrust of politicians is likely to reverberate back to them, a point I will come back to in the nal section. Finally, where states are weak (as they are to varying degrees throughout most of the Andes and Amazon), policy remains poorly implemented, political and civil rights are poorly upheld, and even the institutional reforms noted here are placed in jeopardy. Indeed, progress on this institutional front has been slow and uneven, as noted by Stavenhagen.
While these legal advances are surely important in themselves, the open question is how the new legislation will be implemented and how Indian communities will benet. The answer is not at all clear. Complaints are increasingly heard that the new laws are not being implemented as they should be, or that secondary legislation has not been adopted after general principles were laid down in the new constitutions. (2002, 33 34)

In short, while there are de jure opportunities, de facto obstacles remain, not least of all because political parties remain weak in all of these countries and states remain incapable of implementing many of the reforms upon which politicians and citizens can agree. In this context, it is difcult to imagine how indigenous people can get their interests meaningfully and consistently represented in the

274

Yashar

state and in government. For these reasons, indigenous people have largely mobilized outside of the state through social movementsalthough, as the next section explains, they have increasingly calculated that they must also enter into the partisan political arena. This nal section takes up the question of the challenge of indigenous social movements as they seek to articulate indigenous agendas and to push them, not only in the streets but also in the halls of the state.

The Democratic/Electoral Challenge: The Sirens Call? From Street Politics to State Politics
To date, indigenous movements have primarily demanded political representation and political change through social-movement politicsin the streets, in protests, in documents, and in international fora.30 They have achieved some notable political successes, including the negotiation of territorial autonomy, bicultural education, a chance to help run state ofces, and a voice in public debates. In other words, key demands have found their way into policy (although rarely solely because of indigenous mobilization), even as most indigenous movements initially eschewed party politics, electoral campaigns, and the like.31 With the advent of increasing political clout, increasing overtures from the existing political parties (concerned to tap into a mobilized indigenous constituency), and an increasing desencanto with existing political parties (which fail to promote the agendas that indigenous communities and movements demand), many indigenous movements began to question their principled opposition to participating directly in electoral politics. By the end of the 1990s, many indigenous leaders and movements began to run for ofce, form political parties, and engage in partisan alliances. Indigenous movements in Bolivia and Ecuador increasingly turned to political parties. As Mainwaring and Scully (1995) have observed, these two countries (along with Peru and Brazil) have developed comparatively inchoate party systems.32 In the 1990s, indigenous movements in Ecuador and Bolivia still confronted heavily patronage-driven political systems that appeared to offer little political space for representation (however dened). Weak political institutions continued to foreclose the ability of new and less powerful groups to gain a political toehold in political debates; existing political parties often appeared to disregard their constituencies and to engage in personalistic and clientelistic behavior. Despite the odds, indigenous activists and movements in Bolivia and Ecuador have successfully placed indigenous leaders into ofce. Indeed, by the end of the 1990s, we nd more indigenous people in elected and appointed ofce than ever beforealthough, as we shall see, the degree to which this notable change has allowed them to represent indigenous people is an open question. Bolivias Andean indigenous movements became a recognized electoral force in the 1990s although they started irting with party politics in the 1970s. In this

Indigenous Politics in the Andes

275

earlier period, the Katarista and Indianista movements were largely unsuccessful at elding candidates against a backdrop of internal and cross-movement division. While a few indigenous leaders were elected, they remained isolated both from the broader indigenous movement and from formal party politics in general. Collectively, these parties rarely achieved even 3 percent of the total national vote (Van Cott 2005, chap. 3). Following Katarista Vctor Hugo Crdenass 1993 election to the vice presidency (an election that further divided the existing Katarista parties), other indigenous activists started to organize more forcefullyparticularly in the aftermath of the reforms that President Gonzalo Snchez de Lozada and Vice President Crdenas oversaw in the areas of decentralization, municipalization, and agrarian reform. Crdenas sought to promote greater recognition, pluralism, dialogue, local autonomy, and bicultural education (so that class- and ethnic-based concerns would be addressed). And with the decentralization and popular participation laws that municipalized the country, more indigenous people won electoral seats at the local level. In this regard, the 1990s witnessed a radical increase in the mirror-version of indigenous representationwith more indigenous people gaining elected ofce. Ironically, the Katarista movement that had spearheaded the indigenous movement and rst elded indigenous candidates, including Crdenas, did not remain a competitive political party. The Amazonian movement, CIDOB, also fared miserably (Yashar 2005, chap. 5). A new generation of Andean indigenous leaders, led by Evo Morales and Felipe Quispe, later took the lead organizing new political movements behind them. Both Evo Morales and Felipe Quispe have become prominent political actors within Bolivia, and both have mobilized within the CSUTCB and on behalf of indigenous people. However, they mobilized different constituencies and articulated different visions. Evo Morales came to represent the increasingly powerful cocalero movement, largely composed of Quechua-speaking indigenous people. As a CSUTCB leader, he appropriated the nearly defunct Movimiento al Socialismo (MAS, or Movement toward Socialism) and refashioned it as a pro-cocalero, anti-neoliberal, and anti-imperialist party one that has challenged neoliberal policies and U.S.-afliated eradication programs that limit the rights of indigenous people to grow coca (portrayed as central to an indigenous cosmovision). Morales has had a meteoric political career. Not only did he and several of his colleagues win national seats in the legislature in 1997, but he came in second in the 2002 presidential election and his party became a dominant actor in the current legislature claiming to speak both on behalf of indigenous people and the anti-neoliberal cause. MAS won 8 of 27 senate seats and 27 of 130 seats in the lower chamber, where it assumed the second vice-presidency and leadership of key committees (Van Cott 2005, chap. 3). Morales won the 2005 presidential election in an unprecedented landslide; twelve MAS senators and seventy-two MAS deputies were also elected.

276

Yashar

Moraless vision of indigenous representation does not stand alone. Several indigenous people have tried to run for ofce, although here I focus on those who come out of the indigenous movement itself.33 Notably, Felipe Quispe emerged as Moraless rival, both within the CSUTCB and in electoral politics.34 Quispe, a former politician, guerrilla leader, and secretary general of the CSUTCB, has articulated what appears to be a more millennial form of Aymaran nationalism, at times using what is perceived to be a militant discourse. He also ran for president in 2002 on the ticket of the Movimiento Indgena Pachakuti (MIP, or Pachakuti Indigenous Movement). In that election, he received 6 percent of the vote.35 He was seen as a formidable force and Moraless most important rivalalthough Morales has thus far bested him in movement and electoral politics. Seen as a whole, the Bolivian case is remarkable, both because numerous indigenous men and women have gained signicant access to local ofce, and because prominent indigenous leaders have had an important impact on national-level electoral races. Indigenous leaders have come to occupy the presidency and vice presidency, and they have guratively and metaphorically given the other candidates a run for their money. As such, one nds even traditional parties seeking to include indigenous running mates on their tickets. In this context, the face of Bolivian politics has become more diverse. However, while there has been a rise in ethnic-oriented politicians in the state (both at the local and national level), it remains less certain whether ethnic political parties as a whole are developing new ways of doing politics or whether they are largely replicating the clientelist politics of the broader party system. Indeed, the rise of prominent indigenous leaders has coincided with ongoing political fragmentation and internal divisions among indigenous politicians, which makes it difcult to unambiguously evaluate indigenous representation in Bolivia. It is true that the number of indigenous representatives in the national legislature has increased signicantly (although not all of them necessarily see their primary role as representing indigenous people). Yet, given their competing visions and their partys organizational structure, their capacity to represent and promote a common or broader indigenous agenda is less clear. Accordingly, the strength of political parties for the collective pursuit of indigenous interests (however dened) remains uncertain. Ecuadors indigenous movements did not have the same historic pattern of engaging in party politics as Bolivias. While individual leaders had some personal experience, CONAIE, Ecuadors largest indigenous organization, originally rejected the electoral process, protesting during the 1990 election and encouraging followers to cast a null vote in 1992 (Andolina 1999, 210). In the 1996 elections, following a complicated set of internal debates, CONAIE changed course. It decided to become part of a national coalition, the Movimiento de Unidad Plurinacional PachakutikNuevo Pas (MUPPNP, or Pachakutik Plurinational Unity MovementNew Country). The coalition was comprised of

Indigenous Politics in the Andes

277

Pachakutik (the recently formed indigenous party associated with CONAIE), the Coordinadora de Movimientos Sociales (Coordinator of Social Movements), and the Movimiento de Ciudadana por un Nuevo Pas (Citizenship Movement for a New Country).36 The coalition won a total of eight seatsincluding those for provincial and national deputies (Mijeski and Beck 1998, 4). Among those elected were Luis Macas, former president of CONAIE, and Miguel Lluco, former president of ECUARUNARI (Ecuador Runacunapac Riccharimui, or Awakening of the Ecuadorean Indian). The electoral success was striking, as MUPPNP elected over seventy candidates at the local and national levels and won seven of every ten races it entered.37 Former CONAIE leaders were also considered and chosen for national-level political appointments, during both the short-lived Bucarm administration and the Alarcn administration. While this entry into formal politics was not free of conicts, it represented a noteworthy achievement.38 Moreover, as an elected member of the legislature, Miguel Lluco oversaw the signing of International Labor Organization Convention 169, which advocates many of the indigenous collective demands proposed but not passed in the Constituent Assembly. And former CONAIE leader Nina Pacari was a second vice president in the national legislature. CONAIE also participated actively in the Constituent Assembly that was held from December 1997 to May 1998. The party won 10 percent of the assembly seats and was the third largest political force in the assembly (Andolina 1999, 231, 313). While assembly members did not achieve all their goals, their partial successes were striking. The resulting constitution included an article on the collective rights for indigenous people, although the document did not outline guarantees for these rights, nor did it recognize the country as a plurinational state (Andolina 1998, 1999; Mijeski and Beck 1998). Despite these initial inroads, Pachakutiks electoral performance has been uneven since 1998. In 1998 it lost standing in the legislature, going from 10 to 5 percent of the congressional seats (Mijeski and Beck 1998, 12). Later, when CONAIE played a major role in the mobilization against then-president Mahuad, CONAIE and Pachakutik had to address internal divisions and to contend with charges that CONAIE was not committed to representative democracy. The 2002 presidential elections did not resolve these internal divisions. Pachakutik eventually allied with Colonel Lucio Gutirrez, who then successfully ran for the presidency (under the banner of the Sociedad Patritica, or Patriotic Society, a newly formed political party). Gutirrez had collaborated with CONAIE in the 2000 coup against Mahuad. He assumed ofce in 2002 with support from Ecuadors largest indigenous organization, as well as the presumed allegiance of the 14 indigenous candidates who had been elected to the Congress (100 seats in total) and the various indigenous leaders elected to subnational ofce. Indeed, Gutirrez appointed Luis Macas and Nina Pacari (two of CONAIEs most visible indigenous leaders) to two important ministerial positions. However, CONAIEs

278

Yashar

support soon turned to opposition after Gutirrez implemented austerity measures. In the ensuing months, divisions within CONAIE and with Pachakutik became even more pronounced. In April 2005, the Congress ousted Gutirrez from ofce. In other words, as indigenous people have assumed ofce in Bolivia and Ecuador, less organizational and programmatic unity (rather than more) has been the norm. For while all indigenous leaders would probably agree on a core set of ideasmore political equality, inclusion, and respect; more social services; and more territorial autonomythey do not necessarily share a vision of how to prioritize and achieve these goals. Insofar as there is now a recognition of the heterogeneity of indigenous interests, this recognition is a political advance; however, insofar as divisions make it difcult to push for political changes, these divisions are politically problematic. In this context, the type of indigenous representation that is being advanced remains somewhat ambiguous. In each of the cases discussed, high-prole indigenous politicians presented a mirror image of representationliterally changing the face of national politics. So, too, these same leaders advanced a guardian image of representationadvancing many claims voiced by indigenous movements, including increased access to state resources, the creation of new state ofces, the monitoring of land and autonomy reforms, and increased public accountability for legislation (not) passed. In many ways, the ability of these politicians to push these reforms rested on the recognition by larger national publics and politicians that there was a large, mobilized indigenous force. In some cases, mainstream politicians saw value in indigenous demands for recognition, representation, and autonomy; in others, they simply saw the need to capture votes and defuse protest. In either case, elected indigenous ofcials gained the ability to inform agendas (although much less frequently to pass legislation) because of the existence of a politically organized indigenous movement and indigenous politicians able to mobilize on their behalf. Ironically, perhaps, these successful electoral campaigns have not translated into the growth of indigenous movements and an ability to inuence more fundamental and sustained policymaking on behalf of a broader, more comprehensive indigenous agenda. For, despite early optimism and fanfare, electoral participation has posed some (perhaps short-term) challenges to the existing indigenous movementsa dynamic that is also identied in the broader literature on social movements and democracy.39 Piven and Cloward (1979) and Tarrow (1998) have noted that the advent of political parties, more interaction with state ofcials, and reform policies can undermine once-vibrant movement organizations that were founded to protest and articulate new agendasas their struggles are subsumed or displaced by these formal institutions and sites of political negotiation. Accordingly, Latin American social movements have historically voiced concern about the destructive impact that political parties and

Indigenous Politics in the Andes

279

alliances with state ofcials can have on movement autonomy and integrity (see Hellman 1992; Foweraker 1995, chap. 4).40 At this early stage in the game, it would be foolhardy to draw conclusions about the fate of Latin Americas indigenous movements, their decision to take part in elections, and the consequences for all forms of indigenous representation since the movements are just beginning to move into electoral and party politics. Moreover, given the varied national contexts different histories with democracy, clientelism, party systems, electoral rules, and the likewe should be wary of simple generalizations for the Latin American indigenous movements as a whole. As Eckstein has stated,
The relationship between democratization and social movements is, in essence, historically contingent. If and when political parties get the upper hand, social movements tend to lose their vitality; however, if they do not or before they do, political parties and social movements may nurture each other. (Eckstein 2001, 398)

This contingency requires a greater span of time in which to observe these relationships. For democracy, in general, and electoral participation, in particular, can pose both opportunities and constraints for social movements, in ways aptly delineated by Eckstein (2001, 398 400). That said, a few cautionary observations are in order with respect to the challenges that electoral participation (taken to mean the decision to run in elections) can pose for the unity and integrity of the movements themselves and their subsequent prospects for advancing some form of indigenous representation. Four dynamics are highlighted here (drawn from Yashar 2005), the last of which is likely to pose the most difcult challenge to Latin Americas indigenous movements. First, as indigenous leaders are elected and appointed to political ofce (bringing along with them an advisory staff ), they often leave indigenous movements with less experienced leadership to take their place. For relatively young movements, this can be a particular problem. While this would not necessarily be problematic if movements were better institutionalized, in the short run it has challenged movements to identify new actors who can assume leadership positions and to institutionalize the mechanisms for doing so. This challenge was particularly noteworthy in Ecuador, as the key executive leadership (which had visibly dominated the movement since the 1980s) chose to run for seats in the legislature beginning in 1996. With the ofcial departure from CONAIE of Luis Macas, Nina Pacari, Jose Mara Cabascango, Miguel Lluco, and others, Ecuadors national indigenous confederation had to quickly identify new leaders who could manage the confederation, command the loyalty of its diverse constituency, and formulate agendas that would speak to the demands and concerns of their mass base. Antonio Vargas was elected in an extremely contentious process. While Vargas subsequently made a name for himself as the new president of the movement, several

280

Yashar

of his actions were severely questioned, and have, at least in the short term, weakened the power and inuence of the movement. Vargas led CONAIE when it decided to form an alliance with the military to overthrow the constitutionally elected government of Jamil Mahuad in January 2000. Not only did participation in the coup raise questions about the democratic credentials and credibility of Vargass leadership, but it also led to cuts in external aid to the movement as a whole (Lucero 2001). Moreover, the fact that CONAIE was quickly sidelined in the aftermath of the coup raised questions about Vargass political skill. It is debatable if a more experienced leader would have made similar choices, but it is commonly suggested otherwise. In other words, movements that have been identied with the same leadership confront (short-term) obstacles when that leadership chooses to move into party politics. These obstacles pose challenges for indigenous movements to credibly and forcefully articulate indigenous agendas (i.e., for them to serve as principals in the representative process) and to serve as desirable coalition partners for those newly elected indigenous leaders who are now trying their hand in the game of electoral politics. Strikingly, Luis Macas returned to movement politics and was sworn in again as CONAIEs president in 2005, highlighting not only his immense popular support but also the general commitment to bring back more experienced leadership. Second, those indigenous movement leaders who are elected to political ofce confront a Herculean task. Given their small numbers, it is nearly impossible for them to deliver on the major demands they once made as movement leaders. Legislative action requires numbers of votesit cannot just be mandated once in ofce. In this context, elected indigenous leaders are confronted with what appears to be a choice between (a) maintaining their ideological purity and hence appearing ineffective (because they cannot achieve concrete goals), or (b) working to deliver on some issues via legislative compromise, logrolling, and coalition buildingpotentially appearing to betray the ideals of the movement. The ability to navigate these two extremes is no easy taskparticularly in a context of prevailing economic crisis, weak party systems, and patronage politics. This in turn can have negative consequences for the movement from which these leaders emerged (Wade 1997, 17). The Bolivian case is telling in this regard. When Vctor Hugo Crdenas was elected as Bolivias rst indigenous vice president, he achieved national and international kudos for his role in creating greater spaces for indigenous voices and advances in important legislation. Nonetheless, he was widely critiqued by the Kataristas in the Andes for betraying the ideals of the movement. While this criticism stemmed in part from older debates and divisions within Katarismo, it was not limited to this; indeed, Crdenas ended up working most closely with CIDOB, from the Amazon (with whom he developed cordial and productive relations), rather than with CSUTCB, from the Andes (with whom relations were conictual). The Ecuadorean case is also suggestive on this score.

Indigenous Politics in the Andes

281

As noted earlier, CONAIE had an impressive rst showing in the 1996 elections. Just two years later, however, CONAIE witnessed a decline in electoral support, from 10 to 5 percent of congressional seats (Mijeski and Beck 1998, 12). The reasons for this decline in support are up for debate. But the data unequivocally indicate that one cannot assume that indigenous electoral participation translates into constant and ongoing support. To date, we have no indication that this is the case. To the contrary, indigenous ofcials confront even more difcult tasks: not only must they dene their electoral constituency in a national context of weak parties, apathetic electorates, and economic downturn, but they must maintain their image as political warriors for the indigenous movement from which they emerged. As a case in point, in 2003, Ecuadorean president Lucio Gutirrez forged a governing coalition in the legislature with Pachakutik and the Movimiento Popular Democrtica (Popular Democratic Movement). After a little more than two hundred days, Pachakutik pulled out (following CONAIEs condemnation of the government), and tensions between Pachakutik and CONAIE ensuedwith leaders from both groups denying that the rupture was serious or noteworthy. Third, highly respected leaders of indigenous movements are not necessarily embraced as ideal elected ofcials, particularly in better-established political party systems. We cannot assume that ethnic identication translates into votes for those who share that ethnic background; mirror representation is not always what people seek. In countries with weak party systems, we do nd some notable examples where indigenous leaders have won elections. As in Ecuador, nationally recognized indigenous leaders in Guatemala were elected to the national legislaturefor example, Rosalina Tuyuc, the indigenous spokeswomen of the Coordinadora Nacional de Viudas de Guatemala (National Coordinator of Guatemalan Widows, or CONAVIGUA). However, in those cases where there existed a comparatively stronger and older party system, as in Bolivia, we nd a more checkered history.41 Evo Morales emerged as an important presidential candidate in 2002 and won the 2005 presidential election. Moreover, CSUTCB indigenous activists from Bolivias coca-growing region won legislative races in the 1990s. However, these electoral examples are the exceptions rather than the rule in Bolivia. Indeed, Bolivias movements in the Andes (CSUTCB) and Amazon (CIDOB) suffered miserable failures when they initially entered the electoral arena, in races at both the executive and legislative levels. These electoral failures followed on the heels of remarkable movement successes in mobilizing indigenous people and negotiating favorable policy outcomes with the government in place. In this context, CSUTCB leaders in the mid-1980s and CIDOB leaders in the mid-1990s had expected to perform admirably in the electoral arena.42 Not only did the leadership perform poorly in national elections in both cases (as was largely expected), but perhaps even more surprising, they performed miserably at

282

Yashar

the legislative level as well (which was unexpected). At the very least, indigenous candidates thought that they would do well in their home districts or states. Yet in the 1997 elections, CIDOB (which allied with the Movimiento Bolivia LibreMBL, or Free Bolivia Movement) did not manage to elect a single legislative candidate. Hypotheses about these failed showings aboundincluding the failure of the leadership to consult with their base and the internalized racism that leads many indigenous people to question if their own leaders will and can perform adequately in the formal, white world of electoral politics. Either way, the simple fact is that one cannot predict indigenous electoral success from indigenous movement success. The case of Bolivia is an important reminder of this.43 In the short-term, these disastrous electoral showings weakened indigenous movements and their indigenous movement leaders cum electoral candidates, as broader constituencies came to question their political choice to enter elections and the failed outcomes, having done so. Fourth, as indigenous leaders engage in partisan politics, indigenous movements are more likely to fall prey to partisan competition, thereby exposing themselves to the kinds of political cleavages that can divide movements and weaken demands for recognition, representation, and reform. As with the third observation, this is particularly problematic in countries with relatively stronger political party systems. This kind of partisan competition is likely to happen anyway in a competitive electoral system. But as indigenous leaders search for partisan afliations or coalitions, they accelerate latent partisan divisions within a given movement. As long as an indigenous movement does not take a formal political stance, it is possible for the same movement to house multiple partisan afliations. Where and when movements formally decide to forge and/or ally with a given party, however, these political divides are made manifest. This dynamic was illustrated by indigenous movements that consciously chose not to take a partisan stand to avoid divisions within the movement (as with CONAIE in Ecuador, prior to 1996), and by movements that engaged directly in elections and confronted internal inghting (as with CSUTCB and CIDOB in Bolivia, and CONAIE following its decision to take part in elections in 1996). Indeed, the CSUTCB has been wracked by partisan divides since the 1980s. And CONAIE has confronted increasing internal conict in the rst decade of the twenty-rst century over who to eld for president and how to keep the different regional organizations together under one confederational umbrella. Indeed, in the 2002 presidential election, CONAIE members divided over whom to supportwith some supporting former CONAIE president Antonio Vargas and others supporting Auki Tituana. Moreover, the relationship between CONAIE and Pachakutik has become strained as the former has sought to prevent the latter from changing from a movement to a party. The ways in which partisan conicts have played out within social movements have led scholars such as Sieder (2002), Alb (2002), and Calla (2000) to observe that political parties and partisan competition can further divide rural indigenous communities.

Indigenous Politics in the Andes

283

The dangers of the co-optation of leaders and the fragmentation of indigenous movements in the post-constitutional phase of reform are high. In part this explains why many indigenous activists have rejected political parties all together. (Sieder 2002, 9)

Conclusion
In sum, politics has been highly contested in recent years in the Andes. In a context of weak states, parties, and party systems, much of democratic politics in the Andes has taken a downward spiralthe focus of this volume. One of the few areas where new voices are being heard and novel legislation is being passed is in the realm of ethnic politics. Rising and powerful indigenous movements have been effective claim makers and have compelled politicians to debate an emerging indigenous agendaalthough not always at a pace and with a content deemed acceptable to indigenous people. In this regard, movements have forced public debate, although they have not always been able to select their representatives in that process. The strength of these new indigenous movements in tandem with new institutional reforms (extending suffrage, decentralization, etc.) poses new and important opportunities for indigenous social movements to try to translate their street power into state power. As the previous section has shown, this is a tall order, posing challenges both for the integrity of the movements as well as for the newly elected indigenous leaders now in ofce. The last section highlighted that in the short run this dual strategy has tested the unity and integrity of indigenous movements, just as it has stymied most new indigenous politicians confronted by weak states, parties, budgets, and economies. The weakness of Andean states is likely to be the greatest obstacle of all to political representationas voting is unevenly institutionalized, policies are poorly implemented, corruption is rampant, and the rule of law is spotty. Hence, while the crisis of democratic representation is certainly a function of weak competitive party systems in the countries discussed at greatest length in this book,44 it is more profoundly a function of the weak reach of the state (ODonnell 1993; Yashar 2005; Mainwaring, this volume). Indeed, even laudable political reforms have been compromised by the inability of the state to implement them, the resistance of authoritarian social forces, and the weak ability and presence of the state (in particular, in the countryside). Thus, indigenous movements and leaders confront particularly high hurdles to representing those in whose name they have been elected. For while the current period has advanced micro-analytic and mirror conceptions of indigenous representation (getting more indigenous people to vote and be elected), it has demonstrated the ongoing weakness of macro-analytic conceptions of indigenous representation (being able to sustain indigenous movements and parties, and pass and implement legislation advancing collectively dened indigenous interests).

284

Yashar

It is in this context, then, that efforts to organize within civil and political society are so striking and so necessary for advancing democratic representation even if at times organizing parties and entering elections can marginalize social movements in the process. For while particular indigenous movements and indigenous political ofcials might not survive in the short-term, it is clear that interaction between movements and parties has fundamentally created a new political imperative. Indigenous mobilization in both realms has solidied indigenous peoples as political actors whose interests are, at least now, part of the national dialogue; other political parties must at least take a stand on some of the issues associated with agendas articulated by indigenous social movements. While we cannot be so sanguine about the degree to which indigenous interests will or will not be institutionalized, we can be certain that indigenous peoples are part of the citizenry and electorate and that their political issues (diverse as they may be) are part of the national dialogue. Latin Americas indigenous movements thus have forged a fundamental but unresolved political debate about how best to design political representation in multiethnic polities.

Notes
I thank the editors of this volume and Jos Antonio Lucero for their valuable suggestions and insights. Of course, all the normal caveats apply. 1. This chapter uses the terms Andean region in a very loose sociological, and only partially geographic, sense. I use the term to refer colloquially to those Latin American countries with portions of their national territories in the Andes, excluding Chile. Many of these same countries also straddle the Amazon, including Ecuador, Bolivia, Peru, Colombia, and Venezuela. 2. Lagos (2003a, 2003b). She reports that while support for democracy remains relatively high, support for existing democratic institutions is weak. 3. This section draws on arguments in Yashar (1996, 1998, 1999, 2005). 4. See Collier and Collier (1991) for a conceptual, theoretical, and empirical treatment of the ties between corporatist projects, labor, political parties, and the state. 5. See Stavenhagen (1988, 105; 2002, 27) and Maybury-Lewis (1991) for a discussion of these institutes. While Brazil formed an Indian ofce in 1910, other Latin American countries largely founded these ofces in the 1930s and 1940s. 6. See Stavenhagen (1992) and Wade (1997) for an overview of Latin America. See Mallon (1992) for a discussion of the varied contexts and forms that this policy took in Mexico, Peru, and Brazil. This attempt to create a more homogeneous population contrasted with U.S. history, where more rigid social lines were drawn between Indians, blacks, and the colonial population. 7. Even if ethnic identity was understood as uid, states and landlords often continued to repress these same communities (when rebellious in the face of state colonization, development plans, and repressive rural labor relations) according to a rigid understanding of the appropriate class status of the heretofore indigenous population.

Indigenous Politics in the Andes

285

8. Indigenous movements mobilized in defense of autonomy in the Andes and the Amazon. However, the process by which Andean and Amazonian indigenous movements formed variedwith Andean communities seeking to defend autonomy in the context of eroding corporatist citizenship regimes and Amazonian indigenous communities seeking to defend autonomy in the context of development reforms (some tied to corporatist land reforms) that promoted colonization of their lands. See Yashar (2005) for an explanation of why, where, and how indigenous movements have emerged in Latin America. 9. My theoretical thinking on social movements and the issue of networks and opportunities draws from the social movement literature, in particular Tarrow (1998), McAdam, McCarthy, and Zald (1996), and McAdam, Tarrow, and Tilly (2001). 10. Indigenous organizing in Peru has been relatively weak in the contemporary period. However, subnational organizing did occur in Puno and in parts of the Amazon, where political associational space and transcommunity networks remained stronger than in the rest of the country. In other words, the same three factors (state reforms that challenge local autonomy, transcommunity networks, and political associational space) used to explain the emergence of strong movements in Bolivia and Ecuador can be used to explain subnational variation in Perus overall weak history of indigenous organizing. 11. Sources: 2001 Statistical Abstract of Latin America 37, Table 532: 104; Statistical Abstract of Latin America 30, Part 1: 150; Mayer and Masferreer (1979, 220 21); and Varese (1991). 12. Demographic data about the numbers of indigenous people residing in each country are based on (not terribly reliable) estimates rather than precise calculations. The numbers do not reveal the ways in which indigenous communities have changed with respect to the meaning, content, scope, and form of identities, practices, or goals of indigenous peoples. Nor do these gures intend to stipulate a shared identity among indigenous peoples. That said, there is little dispute that large indigenous populations reside in Bolivia, Peru, and Ecuador, with signicantly fewer numbers in Colombia and Venezuela. Sources for data: 2001 Statistical Abstract of Latin America 37, Table 532: 104; Statistical Abstract of Latin America 30, Part I: 150; Mayer and Masferreer (1979, 220 21); Varese (1991). 13. The following does not presume to be an exhaustive discussion of debates about representation. Its goal is much less ambitious. Rather, it hopes to delineate some of the conceptual questions that arise as ones tackles questions about indigenous representation. 14. This observation recalls old Marxist debates about identifying a class in and for itself. How does the analyst determine the interests of those who themselves do not agree on their primary identication, goals, and priorities? 15. My understanding of representation has been most clearly shaped by Pitkins classic book, The Concept of Representation (1967). I will not engage explicitly with her book here, although my debt to her is great. 16. See Lucero (2002, chap. 2) for a provocative analysis of representationhighlighting representation as mirror, lter, and producer of identities and interests. 17. This last indicator does not address if these representatives are accountable and, if so, to whom. 18. This section on institutions draws heavily on the primary work of Rodolfo Stavenhagen, Xavier Alb, Ricardo Calla, Donna Lee Van Cott, Jos Antonio Lucero, Rachel Sieder, and Raquel Yrigoyen Fajardo, all listed in the References; and Kathleen ONeill, who contributed Chapter 6 to this volume. 19. This volume has primarily focused on representation via elected ofcials. In line with the themes of the volume, this chapter focuses on the changing institutions and norms that have facilitated the election of indigenous leaders and/or people. However,

286

Yashar

it is worth noting that several states have attempted to respond to the question of representation and inclusion by forming indigenous ofces within the executive branch. On the whole, these ofces have been largely dependent on executive whim. Hence, in Bolivia, former president Gonzalo Snchez de Lozada (1993 97) decided in his rst term to form super-ministries, one of which included an Ofce of Indigenous Affairs. When Hugo Banzer became president of Bolivia in 1997, he created a new Ministry for Peasants, Native Peoples, and Indigenous Affairs. 20. See Stokes (2001), in particular, for her discussion of bait-and-switch politics. The observations made here also apply to societal opposition to Chvez in Venezuela although indigenous people are not at the forefront of these protests and, therefore, this development is not listed in the text. 21. Dandler (1996), Stavenhagen (1992, 2002), Van Cott (2000, 265 68; 2002), and Yrigoyen (2000). In a striking Guatemalan referendum in May 1999, the voting population (18 percent of the eligible electorate) rejected proposed reforms to amend the constitution and acknowledge the multiethnic composition of the country (as outlined in the peace accords). For one of the few academic discussions of this surprising episode, see Warren (2002). 22. Indigenous movements have appealed to norms, laws, and organizations operating in the international arena. As discussed by Brysk (1994, 1996, 2000) and Wilmer (1993), the international arena has provided a new discourse, funds, and forums that have often shaped debates about indigenous rights. In particular, indigenous movements have lobbied Latin American states to ratify the International Labour Organization (ILO) Convention 169 on Indigenous and Tribal Peoples in Independent Countries. Convention 169 outlines the rights of indigenous peoples and the responsibilities of multiethnic states toward them. At a minimum, it calls on states to recognize ethnic heterogeneity where states had advanced nationalist aspirations of mestizo homogeneity. The following Latin American states have ratied ILO Convention 169: Mexico (1990), Bolivia (1991), Colombia (1991), Costa Rica (1993), Paraguay (1993); Peru (1994), Honduras (1995), Guatemala (1996), Ecuador (1998), and Argentina (2000). Ratication provides a mechanism for advocating reforms to accommodate ethnically diverse populations. While these Latin American states have yet to live up to the terms of the convention, it has provided a language, legitimacy, and set of transnational advocacy networks to continue work on these issues at home. 23. Van Cott (2005) argues convincingly that where countries have relaxed the requirements for candidate registration (such that one does not need to be a member of a formal political party to be on the ballot), indigenous people and movements have greater prospects of running in and winning elections. Of course, the same requirements that can increase the chance of indigenous representation can simultaneously open the doors to more easily elect catchall politicians with fewer institutional mechanisms for holding them accountable to their constituencies. 24. Interviews in 1997 by the author with Ecuadorean indigenous leaders Leonardo Viteri (March 6, 1997, Quito) and Csar Cerda (May 6, 1997, Quito); and Ecuadorean politician and consultant Gonzalo Ortiz Crespo (February 27 and March 11, 1997, Quito). Also see Selverston (1994, 146) and Selverston-Scher (2001, 45). 25. Interviews in Bolivia by the author with indigenous leaders Marcial Fabricano of CPIB and CIDOB ( June 13 and 20, La Paz) and Ernesto Noe of CPIB ( July 25, 1997, Trinidad); with researchers Zulehma Lehm (August 1, 1997, Trinidad) and Wilder Molina ( July 29, 1997, Trinidad) at CIDDEBENI; and with lawyer Carlos Romero Bonifaz of CERES ( July 1, 1997, Santa Cruz, and July 29, 1997, Trinidad). See Libermann and Godnez (1992); Navia Ribera (1996); Molina (1997); Van Cott (2000).

Indigenous Politics in the Andes

287

26. Interviews by the author in Bolivia with Isabel Lavadenz, former national director of the Bolivian National Institute of Agrarian Reform (August 4 and 5, 1997, La Paz), and Jorge Moz, researcher at UDAPSO (May 31, 1997, La Paz). See also Moz and Lavadenz (1997). 27. Interview by the author with Bolivian lawyer Alcides Vadillo ( June 11, 1997, La Paz). 28. Interviews by the author in Bolivia with Luz Mara Calvo ( July 9, 1997, La Paz) and George Gray Molina (May 23, 1997, La Paz). 29. Also see Balslev (1997, particularly Annex 2, 11721). 30. This section draws on Yashar (2005, chap. 7). 31. Of course some indigenous movements did forge political parties in an earlier period. The Colombian indigenous movement has been engaging in party politics since the 1990s, when it mobilized for the 1990 Constituent Assembly. In Bolivia, one nds even earlier efforts to forge indigenous political parties in the 1970s, although these were overwhelmingly unsuccessful until the late 1990s. 32. While both Bolivia and Ecuador have comparatively weak, patronage-driven political systems, Bolivia maintained a considerably more stable political party system than Ecuador. In Ecuador, political parties have largely unintelligible programmatic differences, weak roots in society, little party discipline, and scant institutional endurance. Bolivia, by contrast, has had a history of signicant and enduring political parties. Most notable among them is the MNR, dating back to the 1940s, and the once-socialist MAS, among othersalthough even in Bolivia most of these long-standing political parties now have little ideological coherence and unpredictable electoral support. 33. CONDEPA, a populist party that has reached out to Bolivias cholo community, is one such example. However, I will not comment on this party insofar as it is not an example of an indigenous movement that then turned to party politics. Indeed, while including indigenous discourse and prominent indigenous gures, and while gaining a signicant electoral toehold among Bolivias cholo community, it does not parallel the examples discussed in the text of indigenous movements turning to party politics. 34. In the 1997 Bolivian elections, the rivalry between Evo Morales and Alejo Velizboth of whom wanted to assume the mantle of the cocalero movementwas also palpable. 35. Electoral data comes from Bolivia: Elecciones Presidenciales de 2002, Political Database of the Americas (Georgetown University and the Organization of American States, 1999), http://www.georgetown.edu/pdba/Elecdata/Bolivia/pres02.htm. 36. Diario Hoy, August 24, 1996. 37. Mijeski and Beck (1998, 4), and Pallares (1997, 544)both cite the Washington Post, July 23, 1996. 38. With the entry into formal politics, several conicts emerged over political alliances and appointments to ministries; charges of corruption and opportunism, particularly leveled against Amazonian leaders; tensions between Pachakutik, CONAIE, and Nuevo Pas; etc. For a discussion of some of these conicts, see Interpress Service, November 1996; Diario Hoy, June 13, 1996, 3A; Mijeski and Beck (1998, 5); and Andolina (1999, 225 32). 39. For a discussion of social movements and democracy in Latin America, see, in particular, Alvarez and Escobar (1992, particularly chapters by Hellman and Canel), and Eckstein (2001). For a more general discussion of movement cycles (including their demise), see Piven and Cloward (1979) and Tarrow (1998). 40. There is also a signicant literature that has analyzed what has become of antiauthoritarian movements in post-transition settings. Alvarez and Escobar (1992), Canel

288

Yashar

(1992), and Schneider (1992) have noted that democratization (and the decline in human rights abuses) can take away the raison dtre of movements that once dened their mission as anti-authoritarian. Schneider also notes that different kinds of political parties can displace these movements once they regain the ability to negotiate in the political sphere. 41. Bolivia has a weak party system, as noted by Mainwaring and Scully (1995). However, compared to Ecuador, Guatemala, and Peru, its system has been relatively more institutionalized. 42. Several 1997 interviews by the author in Bolivia with indigenous leaders Marcial Fabricano ( June 13 and 20, 1997, La Paz) and Jos Uraabi ( July 2, 1997), as well as collective interviews led by CIDOB ( June 2730, 1997) in Camiri, Villamonte, and Monteagudo. 43. Importantly, indigenous representation at the municipal level did increase with decentralization, as noted earlier. 44. Mainwaring and Scully (1995) observed that Ecuador, Bolivia, Peru (and I would add Guatemala) have inchoate party systems. They are neither institutionalized nor stable. Mexico, by contrast had (until recently) a hegemonic party systempreventing effective and meaningful forms of competitive democratic participation. Scholars of these cases commonly refer to a crisis of representationparticularly in the Andean cases.

References
Alb, Xavier. 2002. Bolivia: From Indian and Campesino Leaders to Councillors and Parliamentary Deputies. In Multiculturalism in Latin America: Indigenous Rights, Diversity and Democracy, ed. Rachel Sieder, 74 102. London: Institute of Latin American Studies, Palgrave Macmillan. Alvarez, Sonia, and Arturo Escobar. 1992. Conclusion: Theoretical and Political Horizons of Change in Contemporary Latin American Social Movements. In The Making of Social Movements in Latin America, ed. Arturo Escobar and Sonia Alvarez, 31729. Boulder, CO: Westview Press. Andolina, Robert James. 1998. CONAIE (and Others) on the Ambiguous Spaces of Democracy: Positioning for the 199798 Asamblea Nacional Constituyente in Ecuador. Paper presented at the 21st International Congress of the Latin American Studies Association, Chicago, September 24 26. . 1999. Colonial Legacies and Plurinational Imaginaries: Indigenous Movement Politics in Ecuador and Bolivia. Ph.D. diss., University of Minnesota. Balslev, Anne. 1997. Distritos municipales indgenas en Bolivia: Las primeras experiencias en el Chaco. La Paz: Subsecretara de Asuntos Etnicos. Brysk, Alison. 1994. Acting Globally: Indian Rights and International Politics in Latin America. In Indigenous Peoples and Democracy in Latin America, ed. Donna Lee Van Cott, 28 51. New York: St. Martins Press. . 1996. Turning Weakness into Strength: The Internationalization of Indian Rights. Latin American Perspectives 23, no. 2: 38 57. . 2000. From Tribal Village to Global Village: Indian Rights and International Relations in Latin America. Stanford, CA: Stanford University Press.

Indigenous Politics in the Andes

289

Calla, Ricardo. 2000. Indigenous Peoples, the Law of Popular Participation, and Changes in Government: Bolivia, 1994 1998. In The Challenge of Diversity: Indigenous Peoples and Reform of the State in Latin America, ed. Willem Assies, Gemma Van der Haar, and Andr Hoekema, 7794. Amsterdam: Thela Thesis. Canel, Eduardo. 1992. Democratization and the Decline of Urban Social Movements in Uruguay: A Political-Institutional Account. In The Making of Social Movements in Latin America, ed. Arturo Escobar and Sonia Alvarez, 276 90. Boulder, CO: Westview Press. Clavero, Bartolom. 1999. Derecho indgena y cultura constitucional en Amrica. Mexico City: Siglo Veintiuno. Collier, David, and Ruth Berins Collier. 1991. Shaping the Political Arena: Critical Junctures, the Labor Movement, and Regime Dynamics. Princeton, NJ: Princeton University Press. Dandler, Jorge E. 1996. Indigenous Peoples and the Rule of Law in Latin America: Do They Have a Chance? Paper presented at a workshop on the Rule of Law and the Underprivileged in Latin America, Helen Kellogg Institute for International Studies, University of Notre Dame, November 9 11. Eckstein, Susan. 2001. Where Have All the Movements Gone: Latin American Social Movements at the Millennium. In Power and Popular Protest: Latin American Social Movements, ed. Susan Eckstein, 351 406. Updated and expanded version. Berkeley: University of California Press. Foweraker, Joe. 1995. Theorizing Social Movements. London: Pluto Press. Gray Molina, George. 2003. The Offspring of 1952: Poverty, Exclusion, and the Promise of Popular Participation. In Proclaiming Revolution: Bolivia in Comparative Perspective, ed. Merlee S. Grindle and Pilar Domingo, 345 63. London: Institute of Latin American Studies, University of London; Cambridge, MA: David Rockefeller Center for Latin American Studies, Harvard University. Hagopian, Frances, and Scott Mainwaring, eds. 2005. The Third Wave of Democratization in Latin America: Advances and Setbacks. Cambridge: Cambridge University Press. Hellman, Judith. 1992. The Study of New Social Movements in Latin America and the Question of Autonomy. In The Making of Social Movements in Latin America, ed. Arturo Escobar and Sonia Alvarez, 51 61. Boulder, CO: Westview Press. Lagos, Marta. 2003a. Public Opinion. In Constructing Democratic Governance in Latin America, ed. Jorge I. Domnguez and Michael Shifter, 137 61. 2nd ed. Baltimore: Johns Hopkins University Press. . 2003b. A Road with No Return? Journal of Democracy 14, no.2 (April): 163 73. Libermann, Kitula, and Armando Godnez, eds. 1992. Territorio y dignidad: Pueblos indgenas y medio ambiente en Bolivia. La Paz: ILDIS; Caracas: Editorial Nueva Sociedad. Lucero, Jos Antonio. 2001. High Anxiety in the Andes: Crisis and Contention in Ecuador. Journal of Democracy 12, no. 2 (April): 59 73. . 2002. Arts of Unication: Political Representation and Indigenous Movements in Bolivia and Ecuador. Ph.D. diss., Princeton University. Mainwaring, Scott, and Timothy R. Scully. 1995. Introduction: Party Systems in Latin America. In Building Democratic Institutions: Party Systems in Latin America, ed. Scott Mainwaring and Timothy R. Scully, 134. Stanford, CA: Stanford University Press. Mallon, Florencia. 1992. Indian Communities, Political Cultures, and the State in Latin America, 1780 1990. Quincentenary Supplement, Journal of Latin American Studies 24: 35 53. Maybury-Lewis, David. 1991. Becoming Indian in Lowland South America. In NationStates and Indians in Latin America, ed. Greg Urban and Joel Sherzer, 20735. Austin: University of Texas Press.

290

Yashar

Mayer, Enrique, and Elio Masferreer. 1979. La poblacin indgena de Amrica en 1978. Amrica Indgena 32, no. 1 (AprilJune): 220 21. McAdam, Doug, John D. McCarthy, and Mayer N. Zald, eds. 1996. Comparative Perspectives on Social Movements: Political Opportunities, Mobilizing Structures, and Cultural Framings. Cambridge: Cambridge University Press. McAdam, Doug, Sidney Tarrow, and Charles Tilly. 2001. Dynamics of Contention. New York: Cambridge University Press. Mijeski, Kenneth J., and Scott H. Beck. 1998. Mainstreaming the Indigenous Movement in Ecuador: The Electoral Strategy. Paper presented at 21st International Congress of the Latin American Studies Association, Chicago, September 24 26. Molina, Wilder M. 1997. El movimiento social indgena del Beni en el contexto del proceso de consolidacin de la movilizacin intercomunal hasta la Marcha por el Territorio y la Dignidad (19871990). Manuscript, Centro de Investigacin y Documentacin para el Desarrollo del Beni (CIDDEBENI), Trinidad, El Beni, Bolivia. Moz, Jorge A., and Isabel Lavadenz. 1997. Reforming the Agrarian Reform in Bolivia. Development Discussion Paper for the Harvard Institute for International Development, Harvard University, and UDAPSO, Bolivia. Presented in Cambridge, MA, and La Paz, Bolivia. Navia Ribera, Carlos. 1996. Reconocimiento, demarcacin y control de territorios indgenas: Situacin y experiencias en Bolivia. Working Paper, no. 34. Centro de Investigacin y Documentacin para el Desarrollo del Beni (CIDDEBENI), Trinidad, El Beni, Bolivia. ODonnell, Guillermo. 1993. On the State, Democratization, and Some Conceptual Problems: A Latin American View with Glances at Some Postcommunist Countries. World Development 21, no. 8 (August): 1355 69. Pallares, Amalia Veronika. 1997. From Peasant Struggles to Indian Resistance: Political Identity in Highland Ecuador, 1964 1992. Ph.D. diss., University of Texas at Austin. Pitkin, Hanna Fenichel. 1967. The Concept of Representation. Berkeley: University of California Press. Piven, Frances Fox, and Richard A. Cloward. 1979. Poor Peoples Movements: Why They Succeed, How They Fail. New York: Vintage Books. Plant, Roger. 2002. Latin Americas Multiculturalism: Economic and Agrarian Dimensions. In Multiculturalism in Latin America: Indigenous Rights, Diversity, and Democracy, ed. Rachel Sieder, 208 26. London: Institute of Latin American Studies, Palgrave Macmillan. Schneider, Cathy. 1992. Radical Opposition Parties and Squatter Movements in Pinochets Chile. In The Making of Social Movements in Latin America, ed. Arturo Escobar and Sonia Alvarez, 260 75. Boulder, CO: Westview Press. Selverston, Melina H. 1994. The Politics of Culture: Indigenous Peoples and the State in Ecuador. In Indigenous Peoples and Democracy in Latin America, ed. Donna Lee Van Cott, 13152. New York: St. Martins Press, in association with the Inter-American Dialogue. Selverston-Scher, Melina. 2001. Ethnopolitics in Ecuador: Indigenous Rights and the Strengthening of Democracy. Coral Gables, FL: North-South Center Press. Sieder, Rachel. 2002. Introduction to Multiculturalism in Latin America: Indigenous Rights, Diversity and Democracy, ed. Rachel Sieder, 123. London: Institute of Latin American Studies, Palgrave Macmillan. Stavenhagen, Rodolfo. 1988. Derecho indgena y derechos humanos en Amrica Latina. Mexico City: El Colegio de Mexico, Instituto Interamericano de Derechos Humanos. . 1992. Challenging the Nation-State in Latin America. Journal of International Affairs 45, no. 2 (Winter): 421 40.

Indigenous Politics in the Andes

291

. 2002. Indigenous Peoples and the State in Latin America: An Ongoing Debate. In Multiculturalism in Latin America: Indigenous Rights, Diversity, and Democracy, ed. Rachel Sieder, 24 44. London: Institute of Latin American Studies, Palgrave Macmillan. Stokes, Susan. 2001. Mandates and Democracy: Neoliberalism by Surprise in Latin America. New York: Cambridge University Press. Swain, Carol M. 1995. Black Faces, Black Interests: The Representation of African Americans in Congress. Cambridge, MA: Harvard University Pres