Reorganizing Popular Politics: Participation and the New Interest Regime in Latin America 9780271058924
A historic shift has occurred in the organizational structures through which the lower classes in Latin America express
196 110 6MB
English Pages 408 [398] Year 2015
Polecaj historie
![The Distributive Politics of Environmental Protection in Latin America and the Caribbean (Elements in Politics and Society in Latin America) [New ed.]
9781009263429, 9781009263436, 1009263439](https://dokumen.pub/img/200x200/the-distributive-politics-of-environmental-protection-in-latin-america-and-the-caribbean-elements-in-politics-and-society-in-latin-america-newnbsped-9781009263429-9781009263436-1009263439.jpg)
Citation preview
00 Front.qxd
11/10/2009
9:25 AM
Page i
REORGANIZING POPULAR POLITICS
.
00 Front.qxd
11/10/2009
9:25 AM
Page ii
00 Front.qxd
11/10/2009
9:25 AM
Page iii
REORGANIZING POPULAR POLITICS .
and the
in
, edited by
Ruth Berins Collier and Samuel Handlin
,
00 Front.qxd
11/10/2009
9:25 AM
Page iv
-- Reorganizing popular politics : participation and the new interest regime in Latin America / edited by Ruth Berins Collier and Samuel Handlin. p. cm. Includes bibliographical references and index. Summary: “A comparative analysis of lower-class interest politics in Argentina, Chile, Peru, and Venezuela. Examines the proliferation of associations in Latin America’s popular-sector neighborhoods, in the context of the historic problem of popular-sector voice and political representation in the region”—Provided by publisher. –-–– (cloth : alk. paper) –-–– (pbk. : alk. paper) . Working class—Latin America—Political activity. . Social movements—Latin America. . Pressure groups—Latin America. . Political participation—Latin America. I. Collier, Ruth Berins. II. Handlin, Samuel. JL.R .—dc Copyright © The Pennsylvania State University All rights reserved Printed in the United States of America Published by The Pennsylvania State University Press, University Park, PA - The Pennsylvania State University Press is a member of the Association of American University Presses. It is the policy of The Pennsylvania State University Press to use acid-free paper. Publications on uncoated stock satisfy the minimum requirements of American National Standard for Information Sciences—Permanence of Paper for Printed Library Material, .–.
00 Front.qxd
11/10/2009
9:25 AM
Page v
CONTENTS
. Acknowledgments vii : Introduction: Popular Representation in the Interest Arena Ruth Berins Collier and Samuel Handlin Situating the Analysis: Analytic Approach, Cases, and Historical Context Ruth Berins Collier and Samuel Handlin Logics of Collective Action, State Linkages, and Aggregate Traits: The UP-Hub Versus the A-Net Ruth Berins Collier and Samuel Handlin : Direct Action and Associational Participation: Problem-Solving Repertoires of Individuals Thad Dunning Political Participation and Representational Distortion: The Nexus Between Associationalism and Partisan Politics Jason Seawright : - Targeting State and Society: Strategic Repertoires of Associations Diana Kapiszewski Three Forms of Scaling: Embeddedness, Nodal NGOs, and Flexible Fronts Samuel Handlin and Diana Kapiszewski Associational Linkages to Labor Unions and Political Parties Candelaria Garay
00 Front.qxd
11/10/2009
9:25 AM
Page vi
vi
.
: General Patterns and Emergent Differences Samuel Handlin and Ruth Berins Collier Appendix A: Selection of Focus Districts Appendix B: Survey of Associations Appendix C: Survey of Individuals List of Contributors References Index
00 Front.qxd
11/10/2009
9:25 AM
Page vii
ACKNOWLEDGMENTS
. This project is collaborative in the deepest sense. Unlike most edited volumes, which are bound more loosely by a common topic and perhaps some common guidelines, this volume, which had its genesis in a Berkeley graduate seminar, was conceived from the beginning as a unified whole. In the argot that developed among the collaborators, we came to call the project , an acronym for Comparative Infrastructure of Representation in Latin America, and the participants cirelistas. Although the chapters are associated with specific authors, they are based empirically on data from the same questionnaires, written collaboratively by the original cirelista team, which consisted of Ruth Berins Collier, Diana Kapiszewski, and Jason Seawright, in addition to Christopher Cardona, Sebastián Mazzuca, and Sally Roever. All the authors thank the last three for the way they gave tirelessly to this endeavor, not only joining the others in endless hours into the night over untold numbers of pizzas (even in the Berkeley gourmet ghetto, excellent pizza can get tiresome), but also participating in the survey administration in the field. The cirelistas are grateful to Laura Stoker, who in her inimitable way discussed issues of sampling and survey design with unparalleled clarity and enthusiasm. We also acknowledge Peter Houtzager, John Harriss, and Adrián Gurza Lavalle, who contributed to the survey construction. The empirical analysis is based on data from two surveys, a survey of individuals, dubbed the Qaire, and a survey of associations, or the Agram. Each was administered in four capital cities in Latin America: Buenos Aires, Argentina; Santiago, Chile; Lima, Peru; and Caracas, Venezuela. We have relied on the help of many people in each country, without whom the data collection would not have been possible. The Qaire was administered by a professional survey firm in each country: Gallup of Argentina, Statcom in Chile, in Peru, and in Venezuela. The Agram was administered by a cirelista country coordinator (see appendix B) along with local researchers, who were indispensable in sharing their expertise in general and in helping to select and oversee a local team of interviewers. We want to thank the association leaders who devoted time to answering the survey questions and were usually enthusiastic about sharing their activities, organizational goals, and plights with our research team. In Argentina, we benefited greatly from the generosity of the Centro de
00 Front.qxd
11/10/2009
9:25 AM
viii
Page viii
.
Estudios de Estado y Sociedad, , in Buenos Aires, which provided space, access to their information, and continual support for the implementation of the project. We especially thank Inés González Bombal and Mario Roitter for their generous support and advice as well as ’s director, Silvina Ramos. Several sociologists at the Universidad Nacional de General Sarmiento also provided advice and support, especially Maristella Svampa and Pablo Bonaldi. Several people participated in the implementation of the survey of association leaders and in the collection of secondary data. Among the interviewers, we would like to acknowledge the outstanding work provided by Gabriela Alach, Liliana Alegre, Arturo A. Fernández, María Gioaninni, Julieta Grimberg, Carolina Kostenbaum, Mónica A. Marra, Ulises Pallares, and Juan Ignacio Vallejos. A team of interviewers trained at the Universidad de General Sarmiento, composed of Mariana Barattini, Lorena Bottaro, Lucía Canel, Marina García, and Fabiana Leoni, provided outstanding work and was key in implementing the survey of association leaders in half of the focus districts. Gabriela Alach worked as assistant field coordinator during the first part of the implementation of the survey of association leaders. Alejandro Bonvecchi provided help coordinating the data entry. Juan Ignacio Vallejos and Julieta Grimberg provided research assistance and collected all the secondary data gathered for the project. In Chile, the field coordinator for the Agram was Professor Marco A. Fernández (Instituto de Ciencia Política de la Universidad Católica de Chile), and the field assistant was Francisco Javier Carreras Vicuña (who was, at the time, pursuing an MA at the Universidad Alberto Hurtado). The members of the team of student interviewers who administered the Agram were Alejandro Maureira Matta, Carlos Vásquez Bustos, Carolina Bascuñán Escobar, Carolina Muena Moyano, Carolina Olsen Hernández, Claudia Casal Cordero, Esteban Ferreira Urrea, Felipe Andrés Padilla Poblete, Francisca Laguardia Sotomayor, Gloria Alejandra de la Fuente González, Javier Durbalin Estrada, Jazmín Voigth Vallejos, Karen Contreras Orellana, Luz Cecilia Soto Mendoza, Mauricio Leone Araya, and Patricio Orlando Ibarra Cifuentes. We gratefully acknowledge Paulina Valenzuela and the entire staff of Statcom Estadísticas Consultores Ltda. who did an excellent job administering the Qaire, and doing data input and database construction for both the Qaire and the Agram. Katty Kauffman provided generous hospitality, encouragement, and expertise on Chile. Professor Gabriel Davidovics (Centro de Estudios de Opinión Pública, Universidad de Chile) kindly gave patient assistance. Finally, we would like to thank the staff of various municipalities, including Cerro Navia, Estación Central, Huechuraba, La Pintana, Peñalolén, and San Ramón, who were helpful in various ways as we conducted our research in their areas.
00 Front.qxd
11/10/2009
9:25 AM
Page ix
,
ix
The Peru research team received guidance from Julio Calderón, as local adviser, and Luis Vicuña Muñoz, who served as field coordinator. We thank them both for their help across many aspects of the fieldwork. Two Berkeley undergraduates were very helpful in the field, Azrael Guevara and Camilo Romero. We also acknowledge the contributions made by the Agram interviewers: Masha Vásquez Luque, Yerson M. Guarníz Aranda, Dania Franco Vaisman, María del Rosario Castro Bernardini, Félix Lossio, Juan Alberto Rocha, Saúl Vicuña Montañez, Verónica Boggio. Finally, a number of individuals from the survey firm Apoyo S.A. deserve special thanks for their valuable role in helping to pretest the Qaire in Lima: Sandra Manrique Becerra, Jorge Morales, Luis Caceda Zavala, Silvia Raquel Durand, Silvia Mas Lozada, Agustín Asenjo Valdivia, and Patricia Esther Quiróz Ruíz. In Venezuela, the field coordinator for the Agram was Mailet Carolina Alen Fajardo, of the Departamento de Ciencia Política de la Universidad Simón Bolívar. We thank her and also extend our appreciation to Olga María Avilés Arjona, Viviana González, Anais Colmenares, David Gil, and Wilmer Jordán Flores Arévalo for their work on the Agram. Alfredo Torres Uribe and the entire staff of Como S.A. carried out the enormous task of administering, doing data input, and completing database construction for the Qaire. Professor Adolfo Vargas of the Universidad Simón Bolívar offered invaluable assistance, including access to collateral data sets. Gustavo Mata, then an undergraduate student of political science at Berkeley, provided both hospitality and extensive research assistance. Finally, we would like to thank the staff of the Instituto Nacional de Estadística in Venezuela, who provided several maps and demographic data sets that made the completion of this project possible, as well as the local governments of various municipalities and parishes within the Caracas area, including Petare, San Agustín, La Vega, La Candelaria, de Enero, and Sucre, which were helpful in various ways as we conducted our research in their areas. This project would not have been possible without funding from a number of sources. We greatly appreciate the generous support of Mick Moore and the Center for the Future State of the Institute of Development Studies, University of Sussex. The Berkeley Institute for Research on Labor and Employment provided a welcome home during much of this project in addition to financial support. We also benefited from funding from the Berkeley Center for Latin American Studies and a grant from the Committee on Research as well as the University of California Institute for Labor and Employment. In the period after the data collection, a number of Berkeley’s talented graduate students helped in many ways, from data management and research
00 Front.qxd
11/10/2009
9:25 AM
x
Page x
.
assistance to insightful comments on the analysis. Taylor Boas, Maiah Jaskoski, and Lindsay Mayka gave invaluable assistance at various stages. We also would like to acknowledge Mauricio Benítez, Tasha Fairfield, Bea Gurwitz, and Wendy Sinek. We also express our thanks to two undergraduates, Maritza Trejo and Daniel Katz, who tirelessly provided computer and technical support. In addition, we thank Margaret Keck, who commented on earlier drafts of several chapters. An earlier version of parts of chapters and were published as part of “Democratization and the Popular Interest Regime in Latin America,” in The Diversity of Democracy: Corporatism, Social Order, and Political Conflict, ed. Colin Crouch and Wolfgang Streeck (Cheltenham: Edward Elgar, ).
01 Chapter 1.qxd
11/9/2009
11:48 AM
Page 1
PART I
01 Chapter 1.qxd
11/9/2009
11:48 AM
Page 2
01 Chapter 1.qxd
11/9/2009
11:48 AM
Page 3
1 : Ruth Berins Collier and Samuel Handlin
Latin America has experienced a historical discontinuity in the last quarter century in the context of two global macrosocial processes—international economic restructuring, often referred to under the rubric of globalization, and the third wave of democratization. Democracy became pervasive in the region, with virtually all countries having competitive electoral regimes after a period of widespread military rule. The regime transitions opened new possibilities for societal demand making and new expectations for the accountability of state officials, holding out the promise of more inclusive and integrative polities in countries that throughout the twentieth century had been marked by exclusionary political dynamics, difficulties incorporating the working classes into mass politics, and trouble sustaining democratic regimes in the face of popular demands. In this context, new associations were founded, creating a more vibrant civil society than had previously existed in the region, including a proliferation of organizations around lower-class interests. Yet the adoption of market-oriented economic models entailed profound socioeconomic change that in many ways made this democratic victory a problematic one for the working or lower classes. These groups constitute the majority of the population, who were presumably empowered by the new democra-cies, yet remained losers under the economic reforms, as inequality and often even poverty increased. By the turn of the twenty-first century, a reaction began that in some countries saw the election of “leftist” presidents of various stripes, who in office began to follow through on their campaign promises with expanded social
01 Chapter 1.qxd
11/9/2009
.
11:48 AM
Page 4
programs to reach the lower classes, who had been largely excluded from the benefits of economic growth. This initial unresponsiveness and the new attention to the plight of the lower classes underline the question of popular political representation as a pivotal issue in Latin America’s new democracies. Central to exploring this question is the fact that the dramatic changes in economic models and political regimes have been accompanied by a major shift in the urban popular interest regime in the region, the organizations through which the urban popular sectors, or the lower and lower middle classes, have sought to pursue their interests.1 Through the twentieth century, the most important organizations through which the urban lower classes framed and attempted to promote their interests were labor unions. They were not the only popular-sector organizations, but they were politically privileged both by their own resources and capacity to undertake collective action and typically by their affiliation to political parties (notwithstanding the often double-edged implications of that affiliation). The new economic models have challenged this privileged position of unions. The shift from state-led to market-oriented economic models has produced significant changes in both the state and the world of work. Changes in the role of the state have redrawn the public-private boundary, shifting the arena in which people seek solutions to social problems. Changes in labor markets and labor processes have made work-related solidarities and collective action more difficult to construct and maintain. The move to labor market flexibility and the relative rise of informal workers have challenged the position of unions and put them on the defensive. With the new economic model, unions have become a problematic support base for governing parties that oversee policies of economic reform. As the position of unions has been challenged, a new interest regime has emerged with the proliferation and activation of a broad array of urban popular associations, including community-based associations and nongovernmental organizations (). We refer to this shift in interest regime as one from the -Hub (unionparty hub) to the A-Net (associational networks). The labels emphasize three main points of contrast. First, the two interest regimes are constituted by the predominant role of different primary organizations as “base units”—unions 1. The term “popular sectors” is widely understood in Latin America to refer to groups within the lower strata of the income hierarchy, which constitute the majority in most Latin American countries. We use the term “popular sectors” interchangeably with the terms “working classes” and “lower classes.” Correspondingly, the adjective “popular” refers to these sociologically defined groups, so that its use in terms such as “popular participation,” “popular associations,” “popular neighborhoods,” or “popular interests” refers specifically to the lower classes. In this study we use these terms to refer specifically to the urban popular sectors, which are the focus of analysis.
01 Chapter 1.qxd
11/9/2009
11:48 AM
Page 5
,
in the -Hub, as opposed to a diverse array of urban popular associations in the A-Net. Second, the role of parties differs. Parties were central in the Hub, as unions were typically affiliated to and constituted the core support base of different forms of labor-based parties (s), either populist or Marxist inspired. Parties play a much less central role in the A-Net, as associations typically have more distant, intermittent, instrumental relations to parties, if they have any at all. Third is a contrast between the structure or internal order of the two interest regimes. The -Hub was constituted by the central, privileged, and dominant role of unions as organizations of interest intermediation, although nonunion organizations of course also existed. Furthermore, the union hub was structured hierarchically into federations and national peak confederations. In contrast, the A-Net does not have a privileged, clearly defined organizational hub, nor is it hierarchically structured in the same way, although it includes many organizations that are oriented to coordinating others. Instead, its structural form is the network, an ordering that is more horizontal and fluid, and in which no particular type of organization is privileged. While the present analysis focuses explicitly on contemporary patterns, a cross-temporal comparison, based on the idea of the shift from the -Hub to the A-Net, underlies the book analytically and is explicitly addressed in chapters and . The emergence of a -Hub in the first half of the twentieth century and its subsequent decline and replacement by the A-Net toward the end of the century reflects a more general historical transition that is not limited to Latin America. The -Hub was constituted by a set of institutions that arose, albeit in quite different forms, in many regions of the world among countries undergoing early industrialization. It emerged from political contestation over the incorporation of the urban working class into mass politics, and was sustained by a world economy whose dynamism was based on leading economies oriented primarily toward industrial production for the domestic market. The transition to more market-oriented economies, which has corresponded to closer integration of the international economy, has generally challenged the -Hub, though the extent of its decline has varied across world regions and countries. The rise of urban associationalism is likewise a more general pattern, widely noted in many other developing countries, and the phenomenon has some similarities with the rise of the new social movements in the advanced industrial countries (Touraine ; Melucci ; Laclau and Mouffe ; Kriesi et al. ; Habermas ).2 2. The present analysis of Latin America, of course, refers to a universe of cases that is bounded in a particular way; that is, they share economic models, a similar timing and pattern of development,
01 Chapter 1.qxd
11/9/2009
.
11:48 AM
Page 6
Within Latin America, where the -Hub has been severely challenged, the fundamental changes we are witnessing may constitute a “new critical juncture.” In an earlier critical juncture of “labor incorporation” in the first half of the twentieth century, the challenge of popular participation and the transition to mass politics led to the legalization of the labor movement as a legitimate political actor and to its partisan affiliation (Collier and Collier ). These party-affiliated unions, even in those countries where they never came to encompass a large percent of the working classes, thereby became the major structures of interest articulation and interest intermediation of the popular sectors, constituting the -Hub as the first popular interest regime. The present analysis looks at a potential new critical juncture in the contemporary period in which these structures have been challenged, and new patterns of participation and new structures of representation are emerging. These structures potentially encompass a greater segment of the popular sectors: if the Hub privileged the newly created formal working class, the A-Net is more inclusive of the informal sectors, groups likely to be especially relevant in contemporary Latin American politics given the dramatic expansion of the informal sector in the past twenty-five years. However, the extent to which these structures of representation attract the participation of the popular sectors is a crucial empirical question for examination, as are how they represent the popular sectors and how their activities intersect with the electoral/partisan arena. Analysts disagree about the implications of the new pattern for popular representation. Some have seen in the new urban associationalism new forms of citizenship, the emergence of new actors, and new sources of citizen activation and participatory processes that, compared to hierarchical, bureaucratized party-affiliated unions or clientelistic patterns, have more potential for authentic representation and accountability. Others have seen instead a crisis of popular representation: resource-poor, fragmented associations that have limited reach or political influence and that may be limited and ineffective in their capacity to represent popular interests. Of course, still others have sought to find a middle ground between these contrasting images.3 Our goal is to explore this middle ground further. Beyond noting that reality falls between broadly distinctive social structures, and other regional traits. All these factors have shaped interest regimes and given them particular characteristics. In other world regions with different historical experiences, such as the post-Communist or Asian cases, these economic and political factors would be quite different, with distinct consequences for the resulting interest regimes. 3. See, for instance, Jelin (), Alvarez, Dagnino, and Escobar (), Oxhorn (a), Panfichi (), Foweraker, Landman, and Harvey (), and Levine and Romero ().
01 Chapter 1.qxd
11/9/2009
11:48 AM
Page 7
,
extremes, we seek to examine traits and patterns characterizing that middle ground along a series of dimensions and to undertake a comparative analysis that indicates where countries fall within that multidimensional space. What do these patterns suggest about channels for the expression of popular voice? To what extent does the new interest regime provide a potentially effective organizational infrastructure for expressing popular interests? To what extent do associations provide an effective channel either in advancing popular demands in the interest arena or in connecting with political parties? Existing literature provides insufficient empirical evidence for addressing these questions. Several studies explore individual participation in the interest arena, looking at associational participation or other kinds of problem-solving strategies among the poor.4 Others have examined the activities of associations, looking at specific neighborhood associations, s, or social movement organizations.5 Still others have examined novel forms of policy-making institutions in which citizens and associations have an influential role, most notably “participatory budgeting” institutions.6 However, these studies are generally limited to a single country—indeed sometimes to one or two neighborhoods— making broader generalization difficult and providing a limited basis for explaining variation. Despite the new attention that civil society, social and human capital, social movements, and participatory governance have received in the comparative politics literature and in policy debates in multilateral organizations, no cross-national studies bring systematic evidence to bear on the set of questions on which this project focuses. This book attempts to map and explore these issues by undertaking a systematic analysis of the metropolitan areas of the capital cities of four South American countries—Argentina, Chile, Peru, and Venezuela. These four countries display substantial variation in terms of socioeconomic level, timing and conditions of economic reform, historic strength and political mobilization of the urban working classes, traditional union-party relations, and party system stability. The analysis is based on data from the project (Comparative Infrastructure of Representation in Latin America) generated from two original surveys conducted in the capital cities of each country, primate cities containing
4. Among others, see Stokes (), Dietz (), Anheier and Salamon (), Salamon et al. (), Thornton (), Auyero (), and Powers (). 5. Examples include Foweraker and Craig (), Gay (), Reilly (a), Jelin (), Alvarez, Dagnino, and Escobar (), Ibarra Guëll (), and Blondet (a). 6. For examples of the literature on participatory budgeting, see Abers (), Baiocchi (), and Goldfrank (); for scholarship on other forms of participatory governance, see Coelho (), Keck and Sikkink (), and McNulty ().
01 Chapter 1.qxd
11/9/2009
.
11:48 AM
Page 8
to percent of the national population. A survey of individuals (n = ,) employed a random sample of the entire population in each capital city, as well as an oversample within eight popular-sector focus districts selected to produce variation in terms of income level and history of or leftist voting among the focus districts. A survey of associations (n = ) was also conducted within the same focus districts of each capital city, employing a chainreferral technique and using associational leaders as respondents.7 The rest of this chapter accomplishes four tasks. First, it introduces the concepts of the interest arena and of “the political,” particularly as it pertains to activity within the interest arena. Second, it raises the question of how to approach the concept of representation in the interest arena, an issue that has not received much attention, since most discussions of representation concern the electoral arena. Third, given the focus on representation of and participation by the lower classes, the chapter discusses the concept of the popular sectors as a socioeconomic category or “class” group defined with reference to a materialist dimension. This issue arises in the context of recent strands in the literature that problematize class and emphasize the rise of postmaterialist issues. Finally, the analysis is situated in the political science literature on political participation and interest groups.
Political Activity in the Interest Arena Approaches to exploring representation and participation may focus on two different sites in the democratic polity. The first is the electoral arena, in which recruitment to state office is contested. In this arena, participation through the ballot box and representative relationships between elected officials and constituencies are structured by a set of constitutionally specified institutions, formal electoral laws, and rules regulating political parties, the main organizational actors in this arena. Though participation in the electoral arena may be related to the goals of affecting policy and pursuing interests, it has been recognized as a blunt mechanism for signaling policy preferences or pursuing specific interests (Lowi ; Verba, Nie, and Kim, ). A second site of participation and representation is the interest arena, the considerably more informal locus of specific interest articulation and problem solving, in which both individuals and organizations are important actors. The two arenas do not operate in isolation from each other—parties and politicians with an electoral 7. For more information on the surveys and sampling procedures, see appendixes.
01 Chapter 1.qxd
11/9/2009
11:48 AM
Page 9
,
calculus in mind operate in the interest arena, and interest groups may attempt to shape electoral competition—but they are constituted by different activities and sets of institutions, and the distinction between the two arenas has traditionally proven to be useful for analyzing different processes of representation or attempts of citizens to solve problems politically. The primary focus of this study is the interest arena. The interest arena is the site in which specific interests are articulated and pursued by individuals and organizations—activities that are conceptualized in this study as forms of political problem solving. In exploring patterns of problem-solving activity, we take two complementary approaches that focus respectively on the actions of individuals and on the operation of interest organizations. First, then, we focus on the pattern of individual participation: who participates, how do they do so, and toward what ends? A second focus is the popular interest regime, or the organizations that represent the popular sectors in the interest arena and their pattern of relations, both among themselves and with states and parties. What kinds of organizations exist for advancing popular-sector interests and what strategies do they pursue? Further, what are the salient aggregate traits of the interest regime? Of particular concern are its scope, level of scaling, access—or the degree to which associations seem able to engage in activities in which they seek to access the state in order to present claims—and autonomy. This final trait, which has become central in many discussions of structures of state-society intermediation, is fraught with difficulties in terms of both conceptual clarity and observation or measurement. In the present analysis, therefore, we approach the issue of autonomy primarily through the lens of associational dependence on the state, which may challenge autonomy (see chapter ). Whether undertaken by individuals or organizations, we define political action in the interest arena in terms of a wide, but delimited, range of activities. We extend the rubric to cover state-targeted problem solving as well as some but not all forms of society-targeted activity. Excluded are problemsolving activities pursued through the market, through family connections, or through “private” patron-client relationships—that is, through influential people outside the state who directly solve problems rather than serving as intermediaries between individuals and the state. As such, we do not exclude a prevalent and politically important form: clientelistic or patronage-based relationships, which involve government actors or patrons whose power derives from their access to state resources and largesse. What, then, do we include as political activity in the interest arena? Many studies of individual participation (e.g., Verba, Nie, and Kim ; Rosenstone
01 Chapter 1.qxd
11/9/2009
.
11:48 AM
Page 10
and Hansen ; Verba, Schlozman, and Brady ) have tended to limit the purview of political activities to those that involve interacting with, petitioning, and otherwise making claims on the state to influence policy. However, participation studies have often resorted to an ad hoc list when operationalizing political action in the interest arena, and one that is not consistent with the restrictive definition of the political as state-targeted acts. For instance, the influential studies by Verba and his colleagues (:; :) adopt a statetargeted definition of the political, but then include activities through which participants act cooperatively to “deal with some social issue” (Verba, Nie, and Kim :) or “work with others on [a] local problem” or participate in a “community problem-solving organization” (Verba, Schlozman, and Brady :). Dietz (:) “solves” the inconsistency by adopting a broader definition of the political in which he includes both what he labels “formal” and “informal” political participation. The former consists of attempts to influence the selection and/or policy decisions of state personnel, while the latter is defined as “individual or collective efforts to supply oneself and neighbors with public goods, sometimes but not always through attempts to obtain a favorable distribution of government resources.” Thus, informal participation may consist of communal problem solving. Although Rosenstone and Hansen use a more restricted operationalization in their study, they adopt a similar, broad definition of political participation: “political participation is action directed explicitly toward influencing the distribution of social goods and social values” (:). The present study adopts a similarly expansive approach to political participation in the interest arena. In addition to directly engaging in activities that target the state, an individual may participate in associations, which in turn may be oriented to either state-targeted claim making or the society-targeted provision of benefits. At the level of interest organizations, conceptions of the political have varied considerably. The mainstream literature on interest groups, parallel to the literature on political participation, also tends to adopt a conception of the political that is restricted to state-targeted activities. A related literature on civil society, on the other hand, sometimes takes a very broad, inclusive approach, adopting a conception of voluntary associations as inherently political. Brysk (:), for instance, defines civil society as “public and political association outside the state . . . [whose] political role is not just to aggregate, represent, and articulate interests, but also to create citizens, to shape consciousness, and to help define what is public and political.”
01 Chapter 1.qxd
11/9/2009
11:48 AM
Page 11
,
The present approach to delimiting political organizations lies between these extremes. Organizations considered political are not so broadly defined as to include those that empower citizens or promote skill acquisition virtually by their very existence; nor, parallel to the above discussion of individual political participation, do we limit the definition of political organizations to those engaged in state-targeted action. Rather, the present analysis includes organizations through which popular-sector groups seek to solve collective problems through society-targeted strategies of provisioning (distributing goods, services, and information to the community and organizing community events) as well as capacity building (financing and creating other organizations or training leaders). These strategies of solving collective problems without necessarily going to the state are parallel, for example, to the collective bargaining role of unions vis-à-vis employers, which has traditionally referred to activities within the interest regime, quite apart from demands unions target at the state. Thus, at the level of both individual participation and organizational activity, the present analysis adopts an approach that does not restrict the political to state-targeted activity but also includes certain forms of society-targeted activity that attend to areas of social need that the state might address but, for a variety of reasons, does not (or not sufficiently). The present conceptualization of the political is thus not held hostage to the set of policies that at any particular moment or place define the limits of state action. This conception includes group or organizational activity with the goal of collective solutions to public or shared problems that, as Verba, Schlozman, and Brady (:) put it, “are also undertaken by governments [in Latin America] and abroad”; it is a conception that is drawn with reference to “functional overlap with public institutions.” It is particularly relevant to Latin America, where, with the recent retreat of the state, important areas of economic and social policy are no longer under state coordination, leaving people to seek solutions collectively through cooperation and the pooling of effort and resources. Examples might include a neighborhood watch organization formed to increase public security because this service is poorly provided by the state, or an formed to provide information on public health or awareness to low-income communities. In sum, various types of activities in the interest arena will be considered as political in the present study. Individuals can try to solve particularistic problems through state institutions (courts, for example, presenting claims before a state agency or office) or appealing to a political actor—that is, a party or elected official. Individuals can also get involved in presenting claims for
01 Chapter 1.qxd
11/9/2009
.
11:48 AM
Page 12
collective or social issues, through petitions, protest, or through participation in associations that engage in claim-making strategies. Associations for their part may engage not only in these state-targeted activities of claim making, but also in society-targeted activities of distributing social goods or services.
Political Representation in the Interest Arena Underlying the analysis in this volume is a concern with political representation in the interest arena, a subject not often tackled by political theorists pursuing the issue of representation, who have tended to focus instead on the electoral arena. Electoral politics is the site in which citizens participate in the recruitment and selection of authoritative policy makers, yet it is not an arena where precise, unambiguous, or even identifiable interests or preferences are communicated. As Riker (:xviii) stated, “outcomes of voting are, or may be, inaccurate or meaningless amalgamations [so that] what the people want cannot be known.” The interest arena, by contrast, is the site in which specific interests are articulated and pursued by individuals and organizations—activities that are conceptualized in this study as forms of problem solving. How, then, might one think about representation in relation to the interest arena? This question has often been a central analytic and normative issue in empirical studies of the interest arena. The literature reveals a persistent tension. Although the interest arena allows for more refined and specific “inputs” into the decision-making organs of the state, it is an arena of representation that is often understood as endemically biased, at the level of both individual participation and the infrastructure of interest organizations in the aggregate. At the individual level, studies have demonstrated a positive relationship between socioeconomic status and many forms of political participation (Verba, Nie, and Kim ; Rosenstone and Hansen ; Verba, Schlozman, and Brady ). Of particular relevance to the interest regime, Olson () has argued that individuals whose interests are more general and shared with more people will likely be free riders rather than participants in collective action. The implication is that special interests will organize more readily than those with more general interests, a result that has implications for bias, at the systemic level of the interest regime, in favor of smaller, concentrated special interests. Empirically, many studies of American politics have demonstrated a persistent bias toward the interests of big business or capital in general (Lindblom ; Cigler and Loomis ; Baumgartner and Leech, ). As Schattschneider famously said, “the flaw in the pluralist heaven is that the heavenly chorus
01 Chapter 1.qxd
11/9/2009
11:48 AM
Page 13
,
sings with a strong upper-class accent” (:). In the Latin American context, this kind of bias has been much explored in the context of both post–World War II (O’Donnell ; P. Evans ; B. Schneider ) and contemporary politics (Shadlen ; Kurtz b). As mentioned above, however, another strain of literature has suggested that the growth of associationalism in Latin America provides the potential that systemic bias, while not disappearing, may be ameliorated.8 The question of bias or “equality” in the interest arena has thus been central, but it has been a difficult one to analyze and conceptualize. Despite its importance, many political theorists of representation pay scant attention to the interest arena and instead focus on the electoral arena and specifically on the act of representing, particularly on the behavior of an elected agent in relation to his or her constituency. The axis of debate within this approach revolves around what Pitkin (:–) identifies as the classic “MandateIndependence Controversy” and Mansbridge () as the Mandate-Trustee distinction. Does representation primarily entail fulfilling the mandate of one’s constituents as expressed by their demonstrated preferences, or does it require an agent to make decisions independently in order to further the constituency’s underlying interests? The divergent interpretations of O’Donnell () and Stokes () concerning whether Latin American presidents enacting “neoliberalism by surprise” were representing their constituencies is an example of this debate. These different conceptions are not often presented as such stark alternatives in practice and rarely are portrayed as such by political theorists. In that sense, Pitkin (:) suggests that agents must have the latitude to act independently, but that divergence between constituent preferences and the agent’s actions must be few, far between, and justifiable on the part of the agent. Often unanalyzed in these theoretical discussions, but more routinely recognized in empirical studies, is the central idea that representatives are influenced by the expression of preferences through the interest system. Mansbridge introduces the idea that the interest system is important for political representation. In delineating types of representation she explicitly draws on the empirical literature, and the type she conceptualizes as “anticipatory representation” makes room for the expression of preferences and interests by organized groups (:–). Representation of this type is based on two-way communication and “mutual education between legislator and 8. Whether or not this bias has been ameliorated by new citizens organizations, especially by non-self-interested advocacy groups that counterbalance self-interested groups, is contested in the literature on the United States; see Baumgartner and Leech ().
01 Chapter 1.qxd
11/9/2009
.
11:48 AM
Page 14
constituents.” In this communicatory approach the interest arena becomes central, along with other channels and structures of communication, as it is a component of the representative-constituency relationship. In this respect, Mansbridge points to the importance of “the entire representative process— including political parties, political challengers, the media, interest groups, hearings, opinion surveys, and all other processes of communication.” This conception of representation is broader than the behavior of the agent, so that the key question is “how well the entire representative system contributes to ongoing factually accurate and mutually educative communication” (:–). For Mansbridge, the notion of equality is not an important component of representation. The problem for—or threat to—this type of representation is not unequal voice among societal interests, but manipulation of information by the representative. The idea of equal access or voice is discussed, but more important in this conception is the quality of deliberation, which does not require equal access and is instead more dependent on communication by the representative in a way that educates the represented rather than obscures the real issues. The notion of equality, however, underlies another approach to representation: descriptive representation. Descriptive representation and its emphasis on egalitarianism and proportionality is most useful for our present consideration of the interest arena. In everyday language, people often employ a notion of representation that centers on the congruence of characteristics of the representative and the represented. At the center of this conception is the question of whether the composition of the population is mirrored by the composition of government. As Pitkin (:) notes, this approach was deeply rooted in the ideas of many of the founders of the American republic, such as John Adams, who wrote that a representative legislature “should be an exact portrait, in miniature, of the people at large, as it should think, feel, reason, and act like them.” This notion is the one that underlies the idea of a “representative” sample in methodology. Descriptive representation is centrally concerned with the distribution of interests or traits. It is not just that X represents Hispanics or women in the Congress, but that the proportion of those interests or traits in Congress should be roughly that in the population at large, just as PR systems are designed to give roughly proportional legislative representation to interests organized as political parties. This conception is based on an implied egalitarianism; hence the consequences for proportional distribution in either the electoral or interest arena. In the electoral arena, this underlying notion of equality is institutionalized in
01 Chapter 1.qxd
11/9/2009
11:48 AM
Page 15
,
modern representative democracies in the provision for universal suffrage, which establishes the legal equality of all citizens in the electoral arena. Distortions in the way this fundamental equality may be translated into unequal political influence are generally considered politically and normatively problematic. Bias is most commonly examined in terms of the translation of votes into seats, and institutions such as the electoral law and federal arrangements have received much attention in this respect. Distortions in the relative power of constituencies produce, for example, the “overrepresentation” of rural voters in many Latin American legislatures or the “underrepresentation” of more progressive parties.9 Some scholars also see distortions in patterns of registration and turnout (Verba, Schlozman, and Brady ).10 Parallel considerations arise in the interest arena, where an underlying concern of bias runs throughout much literature. This egalitarian conceptualization of representation is implicit in the early pluralist tradition, which largely dismissed concerns of bias, assuming an equilibrium notion in which interest groups would form as needed to present demands and counterdemands. It is also implicit in subsequent research that found and problematized bias in the interest arena, whether focusing on mechanisms at the individual level, such as differential “group” collective action problems or resources (Olson ; Verba, Nie, and Kim ; Verba, Schlozman, and Brady ), or elaborating on a more systemic bias (Schattschneider ; McConnell ; Lindblom ; Walker ). This kind of bias in descriptive representation was the focus of analysis of Verba, Schlozman, and Brady in their arguments about representational distortion—the differential rates of political participation by richer and poorer groups in associational life as well as in voting. But an analysis of representational distortion in the interest arena must examine more than the rate of participation by individuals belonging to different social groups: it must of course also take into account the role of organizations in the interest arena. At this level, assessing representational distortion in the interest regime is particularly difficult precisely because of the nature of the “units” represented in the interest regime. The group unit presents two particular difficulties. First, as Baumgartner and Leech (:) have pointed
9. On the overrepresentation of rural voters, see Snyder and Samuels (). The vote/seat distortion of the Chilean electoral system has been the subject of debate; see, for instance, Garretón (), Guzmán (), Rabkin (), Navia (), and Zucco (). On Brazil, see Mainwaring (). 10. For these authors, the issue is not problematic if the behavior is voluntary; however, one should note that distortions may also arise from the “supply side” of the electoral system, in that constituent groups may stay away if they think none of the choices reflects their interests.
01 Chapter 1.qxd
11/9/2009
.
11:48 AM
Page 16
out, no clear baseline or reference point—no uncontested set of “societal groups”—exists from which to observe the existence or magnitude of representational distortion of actual organizations. Second, unlike the formal equality of the vote in the electoral arena, no institutional mechanisms provide a basis for or even the pretence of equality of organizations in the interest arena (Baumgartner and Leech :–). In the electoral arena, the weighting of the vote is clear and transparent, and distortions due to malapportionment, a constitutional “preference” for representing subnational units, or other factors such as the electoral law are easy to discern and even measure. In contrast, groups in the interest arena are assumed to wield greatly divergent influence over state officials and the policies they make. As many scholars have noted, describing relative power may be possible with regard to specific policies, but observing and assessing relative levels of influence by social groups at a systemic level is exceptionally difficult (McAdam, McCarthy, and Zald :). The present study deals with this latter problem by putting aside the question of influence (and notions of representation based on responsiveness) and by focusing on descriptive representation, following a more pragmatic approach in line with the advice of Baumgartner and Leech and similar to the approach of Verba, Schlozman, and Brady. Specifically, this project contributes to an analysis of representation in the interest arena through a more limited and tractable focus on () patterns of popular participation, comparing that of the popular sectors with that of the middle classes, and on () traits of the organizational infrastructure through which the popular sectors attempt to solve collective problems. Participation in the interest arena constitutes a step prior to influence and effectiveness; it is the first stage in the process of representation. The question of whether the popular sectors are capable of affecting meaningful policy decisions is one of the dominant themes of the debate on popular representation in contemporary Latin America. But to begin to explore this question, we must first have a broad empirical picture of the pattern of participation of the popular sectors in politics and the nature and activities of popular organizations that further their interests and intermediate between them and the state.
The Popular Sectors and the Question of Class Politics This volume is motivated by the historically problematic nature of popularsector inclusion in politics in Latin America. Mass political inclusion has been a source of conflict in the region since the decline of oligarchic dominance.
01 Chapter 1.qxd
11/9/2009
11:48 AM
Page 17
,
Throughout the twentieth century most countries were unable to make the transition to mass politics successfully. Rather, two dominant patterns emerged: either institutions were founded that effectively controlled and co-opted the political voice of the working classes, or, in the face of what the elite viewed as the political threat of excessive lower-class influence, democracies were overturned in favor of coercive authoritarian regimes that repressed representative structures and eliminated mass participation in politics (Collier and Collier ). Given this historical tension between democratic stability and mass representation in the region, the study focuses on the working classes, or the urban popular sectors, exploring and mapping the relatively new patterns and structures through which they participate in politics and pursue their interests. To inquire about the inclusion of the popular sectors in mass politics is to ask about a socioeconomic category or “class” group defined along a materialist dimension. This section discusses the conceptualization and operationalization of the popular sectors in light of recent orientations of scholarship away from class as a dominant category of analysis and from materialist concerns as the predominant cleavage. The shift in the predominant organizations of popular interest representation from labor unions, which are prototypically class organizations, to urban associations, with heterogeneous and primarily territorially based constituencies, resonates with the suggestion that the twentieth-century pattern of class politics has been superseded, and post- or nonmaterialist issues and alignments have become salient. While this assertion has been made principally for the advanced countries,11 some Latin American analyses have suggested a similar pattern (Torcal and Mainwaring ; Roberts ). To what extent does the new configuration of interest organizations reflect—and advance—this trend? It is the present perspective that while “postmaterialist” issues have become important in Latin America, the current period during which this shift in interest regime is occurring is nevertheless a profoundly materialist one. And the salience of materialist issues justifies a focus on the popular sectors as a group defined in materialist terms, as a heterogeneous fuzzy set located at the bottom of the socioeconomic hierarchy of inequality. The literature on the rise of postmaterialist issues has its origin primarily in analyses of advanced industrial countries, initially reflecting the emergence of the civil rights, feminist, gay, environmental, and antinuclear movements. These movements have also become a feature of Latin American politics, along 11. See Inglehart and Rabier (), Clark, Lipset, and Rempel (), Eder (), Manza, Hout, and Brooks (), and G. Evans ().
01 Chapter 1.qxd
11/9/2009
.
11:48 AM
Page 18
with the human rights and pro-democracy movements that became salient toward the end of the authoritarian period in the region. As may be expected from the European analysis, many of these postmaterialist issues have their core base of activists among the Latin American middle and upper class. However, rights and identity issues have also been important among the popular sectors, which have participated particularly in the human rights, pro-democracy, women’s, and indigenous rights movements.12 Particularly relevant for the present project, with its urban focus, are women’s activation and participation. Women have assumed new political roles and prominence in the interest arena, mobilizing, along with indigenous movements, previously marginalized groups at least partly on the basis of new identities. Yet, to a substantial extent, newly activated popular-sector women have mobilized around the distribution of goods and services, such as food, child care, and family provisioning, which are fundamentally materialist concerns. Thus, by giving primacy to a group defined along material lines, we do not dismiss the importance of other issues or other sources of group identity.13 The point is rather that, overall, materialist issues continue to dominate the political agenda, even when they have been differently framed in terms of gender or indigenous claims. The prominence of material issues can be seen in the way they are privileged in electoral campaigns and the consistently high salience respondents ascribe to material concerns in survey data. In direct relevance to the concerns of this volume, it has also been widely recognized that, despite the importance of postmaterial concerns for some, associationalism among the popular sectors overall has tended to revolve around material demands “as they must deal with the daily hardships caused by poverty and material need” (Foweraker, Landman, and Harvey :). The current conjuncture in Latin America thus can hardly be conceptualized as postmaterialist. 12. From the Latin American perspective, at least to some extent, the postmaterialist nature of these issues might be challenged. First, the timing of materialist and postmaterialist issues is not sequential but contemporaneous; that is, the latter are not postmaterialist, not a “luxury” to which attention is paid by those who have first achieved a certain level of economic security. Second, they are not orthogonal to materialist issues, but instead some of these issues or values may be tightly intertwined with materialist goals. For example, the working class as well as the business class championed democracy as a means for material improvement in a situation in which the authoritarian regime was not—or no longer—providing it (Conaghan and Malloy ; R. Collier ). Similarly, human and indigenous rights cannot be divorced from class or materialist issues in a context in which the lower classes were particular targets of the abuse and infringement of those rights, and some of the demands are fundamentally economic, such as indigenous claims to land rights (Yashar ). 13. While not a central component of our analysis, the analysis in chapter documents the participation of women in the interest arena and the substantial feminization of associationalism.
01 Chapter 1.qxd
11/9/2009
11:48 AM
Page 19
,
On the contrary, the current period is widely considered a major turning point precisely in material terms, with a fundamental transformation in the model of economic accumulation and in basic patterns of wealth generation and distribution. Materialist issues have been of central concern for many social groups. Technocrats, politicians, and business actors have been deeply engaged in the policy-making process of economic reform and restructuring. Of course, material issues are not only salient for elites. Analyses of political economy and market reforms, even those that express medium- to long-term optimism about the effects of these reforms, have widely recognized that the lower classes have not captured much of the benefit of what has generally been a spotty record of economic growth. After a postwar period of steady growth and some progress toward reducing the world’s worst income distribution, inequality in Latin America increased in the post-reform period, and the earlier pattern of poverty reduction has become interrupted and erratic. We are thus in a period when materialist concerns remain at the forefront both for those who are reforming or refashioning economic relations and for the majorities that have been economically squeezed. Indeed, the ongoing salience of materialist issues in Latin America has been accompanied—and supported—by a shift in social structure that is quite opposite the social-structural shift that is often invoked to explain the rise of postmaterialist issues in the advanced countries. In the latter, a long-term decline in the blue-collar workforce and the growth of white-collar occupational categories has blurred the traditional divisions that formed the basis of class politics (Dalton b). In addition, blue-collar workers generally have come to enjoy substantial prosperity, so that, as Clark and Lipset (:) put it “class is ceasing to be a major determinant of life-chances.” If Western Europe presents a picture in which class polarization has declined, relative income and equality have risen, and material issues in that setting have become less salient, being reduced to just one set among many, Latin America in the last two decades presents the opposite pattern: a similar relative decline in the bluecollar working class has been attended by a burgeoning of the informal sector, increasing polarization of income, and downward mobility of the working classes. If these changes in economic policy and social structure motivate and justify a focus on the popular sectors, the question remains how one demarcates both popular-sector individuals and popular-sector associations. The popular sectors, as a heterogeneous category in the lower part of the socioeconomic hierarchy, constitute a large percent of the population in highly
01 Chapter 1.qxd
11/9/2009
.
11:48 AM
Page 20
unequal countries.14 Those it embraces are heterogeneous in many ways, as the plural form of the label indicates. The concept does not define a class in Marxian terms, as it is not delimited in terms of a particular position in a mode of production. Indeed, the concept of the popular sectors embraces substantial diversity in this respect. It includes both proletarians who sell labor and the self-employed in the lower strata of the income hierarchy. It also covers both workers in the formal economy and a burgeoning group of earners in the informal economy, now up to about half of the workforce. To the extent the popular sectors can be considered a class category, it would be in terms of a Weberian approach to class, referring to the material position of the popular sectors resulting from their position in multiple markets, not just the labor market. Pinpointing a line of demarcation that bounds this category is difficult. Pakulski () has suggested the criteria of “distance” and “clustering” as a means for demarcating classes along the socioeconomic hierarchy. In this historically most unequal region of the world, income structures are characterized by what Birdsall, Graham, and Pettinato (:) have referred to as “top-driven inequality.” The data reveal a very sharp distinction between the small upper and upper-middle class, indeed between the top percent, and the rest of the population.15 However, it is difficult to identify specific cutpoints, or specific clusters of individuals, within the remaining percent, which constitutes a very heterogeneous category but is nevertheless characterized by a more equal distribution of income than in the generally more equal, developed countries (Székely and Hilgert :). Other aspects of socioeconomic inequality, such as educational attainment, may thus offer an appropriate approach to demarcation within the non-elite percent. This choice has strong theoretical justification. In the United States, for example, Verba, Schlozman, and Brady (:) find that education is a basic component of social status, in that it affects income and occupation (as well as other individual determinants of political participation). Furthermore, education affords substantively defensible lines of demarcation. The high school diploma is an important credentialing mechanism for skill level, job, and social status. Its importance in this regard is perhaps suggested by its subjective 14. The inclusion of the lower-middle classes among the popular sectors is standard usage. As Birdsall, Graham, and Pettinato () point out, the “middle” classes defined in terms of a location near the median of the income distribution does not enjoy a level of economic security or “middleclass” standards of living. 15. Data on income distribution in Latin America is fraught with problems of reliability. In their discussion of these problems, Székely and Hilgert (:) show that this conclusion probably understates the gap between the rich and the rest, as it is based on data that “grossly” underestimates the income of the top decile.
01 Chapter 1.qxd
11/9/2009
11:48 AM
Page 21
,
evaluation, as indicated by the fact that the modal value of educational attainment in the data corresponds to attaining a high school degree. This study thus primarily employs a dichotomous measure, in which the popular sectors are operationalized as those who have not completed high school. However, to reflect the uncertain boundaries of the popular-sector category, we also employ other operationalizations—at times analyzing two cut-points, at incomplete and complete secondary education, and at times using a continuous measure, allowing a more nuanced assessment of social status, within the popular sectors as well as between that group and those of higher status. A second question pertains to defining and delimiting “popular” associations from the many other associations in civil society. Popular associations are defined as those that attract primarily popular-sector participants, serve popular-sector constituencies, or have the primary goal of advancing popularsector interests. Although popular associations are often “by” and “of ” the popular sectors, they may also be simply “for” them, as are many s that work on behalf of popular-sector interests.16 Operationally, we drew a chain-referral sample of associations in eight “focus districts” of each city, chosen to exclude the wealthier districts. In addition, the starting points of each chain were purposively selected for types of associations of particular importance to the popular sectors in each city, such as those distributing food, those addressing problems of unemployment, or neighborhood associations (see appendix B).
Political Participation and Interest Organizations As is clear from the foregoing, the present focus on political activity in the interest arena and its consequences for popular representation reflects a set of concerns common to many currents in political analysis, particularly the two strands of scholarship, which correspond to the activities of individuals and of organizations in the interest arena. The first, the literature on political participation, has been most extensively developed in the context of advanced democracies, especially the United States. The second, a diverse literature on interest organizations, has been explored more extensively across a broader set of empirical contexts and has a particularly robust tradition in Latin American 16. A substantial debate in the literature has addressed the question of whether are indeed popular organizations if they are not “by” or “of ” the societal constituency. We choose to consider as popular associations those that are dedicated to addressing popular-sector problems without making this distinction. See, for example, Escobar (), Reilly (a), Fisher (), Brysk (), Friedman and Hochstetler (), and Peruzzotti (b).
01 Chapter 1.qxd
11/9/2009
.
11:48 AM
Page 22
analysis. Discussing these bodies of scholarship allows us to contextualize the concerns of the current volume. Political Participation At the individual level, our focus is on patterns of political participation in the interest arena and the political structures through which individuals pursue their interests and solve problems. Assessing various problem-solving strategies gives us analytic purchase on the microlevel foundations of the popular interest regime. The present approach to delineating these problem-solving strategies has much in common with the analytical tradition that has explored political participation, with some important departures. Because it is so central to democratic regimes, voting and vote choice has been the subject of an entire subfield of studies on political participation (e.g., Berelson, Lazarsfeld, and McPhee ; Campbell et al. ; Key ; Miller and Shanks ). Clearly related to the actual decision at the ballot box is a wider set of activities related to the process of recruitment to elected office, and a broader literature on political participation has correspondingly examined a variety of campaign activities (e.g., Huckfeldt and Sprague ; Wielhouwer and Lockerbie ). The focus of all these studies is on the constitutionally specified electoral arena of politics as a site of participation. The present study, by contrast, is primarily concerned with participation in the interest arena, although chapter also examines participation in party/electoral activities and its intersection with participation in the interest arena. As indicated above, some studies of political participation have cast a wider net and included various activities within the interest arena as well. Studies of participation have tended to focus on the social-psychological and resource-based description of what might be called the “participant personality” and seek to explain levels or degrees of participation. Consequently, this approach often relies on a somewhat ad hoc list of participant activities, extracting dimensions of such activities through factor analysis (Verba, Nie, and Kim ; Shi ) and/or employing an additive index of overall participation (Converse ; Verba and Nie ; Hansen ; Verba, Schlozman, and Brady ). Explanations are often rooted in motivations (especially different types of benefits), levels of engagement or interest in politics, and resources or capacity, especially, education, psychological resources (efficacy), age or experience, financial resources, social position and networks, and civic skills. Because of the focus on the participant personality, these studies tend to emphasize individual-level explanatory variables rather than sociotropic ones.
01 Chapter 1.qxd
11/9/2009
11:48 AM
Page 23
,
Although some scholars have tried to integrate sociotropic explanations, those that emphasize individual traits are dominant. This focus is common to both the literature on vote choice and that on participation more generally. Regarding the former, this approach is evident in the funnel of causality of the socialpsychological approach of the Michigan school (Campbell et al. ; Klingemann and Wessels ), which has sociological (or “structural”) traits at the distal, broad end of the funnel, but tends to emphasize political attitudes at the proximate, narrow end. In the broader political participation literature, the emphasis has also been on personal factors. Verba, Schlozman, and Brady () examine three factors, which they label resources, recruitment, and issue engagement. However, they “convert” these last, more macro factors into individual traits, by looking at whether or not individuals have been contacted or have been exposed to attempts at mobilization, rather than traits of, for instance, the political campaigns, parties, or party systems. Rosenstone and Hansen (), however, pay more attention to the “other half ” of the explanation in their concern to make intertemporal comparisons. Like most studies, they examine what they refer to as personal resources (including social, economic, psychological, and experiential resources) and political engagement or involvement. However, they argue that these factors may help to distinguish among individuals, but are less useful in accounting for change over time, and therefore they add “political” factors, looking at strategic mobilization by politicians, parties, and interests, and the macro logic or incentives in the political system that lead actors to mobilize participation. This orientation points to the importance for comparative analysis of explanations centered in local or national factors that relate to issues such as the party system, the organization of interests, decentralization, and regime differences. The present study builds on the extant literature on participation at the same time that it develops certain other themes. We are concerned with looking at political participation not only to understand who participates, but also to build an aggregate picture of patterns and structures of political problem solving. This orientation requires three subtle but important departures from the more traditional participation literature just described. First, rather than focusing overwhelmingly on the question of why people participate, we pay equal attention to the issue of how they participate. Second, instead of focusing on an aggregated “score” of individual participation that combines several different types, we develop a framework for distinguishing and analyzing diverse forms of participatory activity in the interest arena. Finally, we aggregate individual participation into country-level scores as a basis for macro-comparisons. A further extension would be to employ a comparative analysis not only to
01 Chapter 1.qxd
11/9/2009
.
11:48 AM
Page 24
uncover similarities and differences across national contexts, but to push further an analysis of macrosocial factors that may illuminate these cross-national patterns. Interest Organizations Another set of questions concerns societal organizations in the interest arena. Relevant are both the properties of individual organizations and more systemic traits of the interest regime, or the aggregate nature of these organizations and their interrelations. A brief look at three strands within this literature can help situate the current analysis. Interest organizations and the interest regime were the empirical referents for both postwar pluralistic group theory and analyses of corporatism that proliferated beginning in the s. In addition, under different guises interest organizations have also been the subject of analysis for more recent studies of both social movements and civil society, a diffuse literature grouped here under the rubric of “postcorporatism.” Each of these perspectives examines the features of interest organizations, analyzes their representational role, and discusses other political traits of the aggregated interest regime. The present approach to examining popular associations shares some similarities with all three of these approaches, while also departing in significant respects. Pluralism has often provided a foil for subsequent theorizing but has been revived in the context of new realities. Corporatism is particularly relevant for the analysis of the -Hub, as postcorporatism is for the A-Net. Together these literatures put a set of perspectives on the table for looking at issues such as organizational formation, representational distortion, and relations to the state that are among the central concerns of this volume. Pluralist theories of interest representation emerged out of post–World War II analyses of interest group behavior in American politics. A central tenet within the literature was a nearly reflexive relationship between societal interests and the formation of claim-making organizations—as Berger () described the pluralist perspective, organizations emerged as “spontaneous emanations of society.” As a result, the formation of and participation within interest organizations was seldom problematized, with the consequence that issues of representational distortion—whether some societal groups organized effectively and others did not—were rarely explored. Instead, the literature tended to assume that any interest demands would produce a supply of effective claim-making organizations. The relationship between interest groups and the state was left relatively undertheorized, as interest groups were typically defined as organizations
01 Chapter 1.qxd
11/9/2009
11:48 AM
Page 25
,
integrated into politics as claim makers on the input side, pressing state officials for policy outputs. Indeed this one-way relationship was reflected in the labels typically used: “pressure politics” and “pressure groups.” Bentley (, ) and Truman (), the founding analysts in this tradition, conceived of policy making in terms of a “vector-sum” model in which groups pressed competing demands on the state, and the result was metaphorically seen as the outcome of a parallelogram of forces, a resultant of the direction and intensity of the array of interests pressed upon the state. In this conception, the state was a black box that simply “resolved” the force vectors with a policy output corresponding to a kind of geometric logic. Pluralism, then, did not problematize the formation of interest organizations, their relationship to the state, or representational distortion or bias in the aggregated interest regime. Subsequent theorizing would raise each of these issues as a response to both empirical and theoretical developments. Empirically, the civil rights and antipoverty movements in the United States highlighted years of weak organization among groups with few resources. Olson () was particularly important in problematizing collective action and theorizing the differential capacities of diverse groups to form organizations. As attention turned to the ways in which groups were differentially endowed with resources and to how some groups faced greater difficulties overcoming obstacles to collective action, such as the free rider problem, the implications for bias within or representativeness of the interest system were highlighted. In addition to calling into question the representativeness of the interest system at an aggregate level, scholars attentive to Michels’s iron law of oligarchy began to highlight the issue of representation within organizations, showing that endemic dynamics between members and leaders tended to have negative consequences for the ability of organizations to represent the interests of constituents (Michels ). Scholars working in other empirical contexts, witnessing very different patterns of state-society relations, began to question the pluralist model even more fundamentally, developing a second literature on interest organizations. Brazil provided a case that led analysts (Schmitter ; Erickson ; Mericle ) to focus on the primacy of class organizations, particularly unions, and to theorize a model of interest politics that, in its state-centric nature, resembled the corporatist arrangements of fascist models. The corporatist model was soon elaborated and broadened to include Western Europe, and two types were distinguished: “state” or “authoritarian” corporatism in Latin America and “societal” or “liberal” corporatism in Western Europe (Schmitter ; Lehmbruch ).
01 Chapter 1.qxd
11/9/2009
.
11:48 AM
Page 26
The corporatist model drew attention to certain features of the interest regime. The predominant groups were class or productionist organizations of labor and capital, which scaled to the national level, forming peak associations. This interest regime was not based on voluntary, fragmented, unstable societal emanations of the pluralist conception, but stable, often bureaucratized or professionalized, hierarchically organized groups in close interaction with the government. Unlike pluralist theories, scholarship on corporatism located the phenomenon historically in a particular political economy that lent itself to the primacy of certain types of organizations, such as labor unions. While scholars of Latin America analyzed the subsequent, post–World War II emergence of other types of popular associations, these were largely seen as playing a peripheral role in the interest system as a whole. Relations with the state became central in corporatist analysis. Corporatist organizations of labor and capital, while acting in the “private” sphere of industrial relations, were also understood as claim-making organizations interacting with state officials. In a departure from the pluralist model, scholars saw these relationships between organizations and the state as considerably more complex than the simple exertion of pressure on government for a particular policy response. Particularly in the Latin American context, analyses of “state corporatism” stressed the ways in which interest organizations served to intermediate state-society relations, not only representing their constituencies as “input” mechanisms to affect state policy, but also serving as “output” mechanisms through which the state “structured” and controlled interest organizations through laws, regulations, discretionary subsidies, and co-optation (Schmitter , ; Erickson ; Stepan ). The resulting trade-offs were especially salient regarding unions. The corporatist system offered some degree of institutionalized access to politics and legitimated unions as actors within the political arena, but they also served as controls, both through outright constraints and double-edged inducements (Collier and Collier ). Autonomy and the representational effectiveness of the union-based interest regime in Latin America thus became important analytical issues. In sum, corporatist analyses complicated and enriched pluralist analysis by conceptualizing the interest system as an interaction of bottom-up and topdown, or society-centric and state-centric, dynamics, with goals of not only pressure politics but also of exerting state control over societal interests and channeling social conflict. This dual perspective also allowed analysts to theorize the different types of collective action problems confronted by labor compared to capital, and the role of the state in partially solving them. It was further recognized that the hierarchical and bureaucratic nature of unions posed its
01 Chapter 1.qxd
11/9/2009
11:48 AM
Page 27
,
own difficulties for the representation of constituent interests, as did the privileging of labor unions within the interest system, which led to the relative exclusion of other groups such as the informal sector. A third approach to the interest system might be called postcorporatist, a rubric covering two strands in the literature that have examined nonunion forms of organization, often lauding them as a welcome corrective to the representational problems of corporatism. The first is the new social movements () literature that initially responded to the postmaterialist issues that arose in the advanced countries, especially the peace, feminist, civil rights, and gay rights movements, and has since been the basis for theorizing contention and “new” forms of organizing in Latin America during and after third wave transitions. This analytic focus on new social movements has been joined by a related body of literature reinvigorating the concept of civil society, a sphere defined not in terms of an oppositional relation to the state as in the literature, but as specifically separate from both the state and the market, implying an autonomy from the state very different from corporatism, particularly in its Latin American form. Although the two perspectives differ in important respects, they share certain conceptions of the interest system. The literature in both traditions has broadened the focus beyond organizations formed around the class cleavage and, when examining Latin America, has returned to more society-centric perspectives. In their approach to organizational formation and participation, each tradition has moved beyond both the “spontaneous emanation” approach of pluralist visions and the highly constrained approach to collective action problems emphasized in the Olsonian tradition. The new social movements framework provided an analytic perspective that emphasized social networks, framing, mobilizing structures, and social benefits in facilitating collective action (McCarthy and Zald ; McAdam, McCarthy, and Zald a; Tarrow ; McAdam ). Both literatures have viewed base organizations (social movement organizations or civil society organizations) as participatory, often uninstitutionalized, nonbureaucratic, and hence unstable, and as interrelated in nonhierarchical, shifting, multinodal networks. In the Latin American context, both perspectives have been invoked to interpret the “resurrection” of organizing and the great proliferation of associations targeted toward both the postmaterialist themes of the s of the advanced countries as well as materialist concerns, particularly those brought about by economic crisis and adjustment. Initial analyses of Latin America in the postcorporatist perspective were optimistic that new forms of popular organization would enhance representation
01 Chapter 1.qxd
11/9/2009
.
11:48 AM
Page 28
in the region, because they did much to include popular groups beyond formal workers and were considered internally participatory and democratic, and were able to escape from bureaucracy, Michelsian oligarchy, hierarchy, and cooptation. Such optimism was also a product of the times: the initial empirical referent for these literatures was contestation over regime change in Latin America or cooperative strategies to confront economic crisis, rather than problems regarding coordination among associations and everyday relations with state actors in “ordinary times” (Jelin ; Hellman ; Eckstein b). In later literature within both traditions, scholars have taken a closer look at the representational capacity of social movements and civil society and at the obstacles interest organizations face in influencing state actors and in creating a critical mass of coordinated organizations. In this context, a more pessimistic approach has become common, with some scholars seeing greater autonomy from corporatist constraints in terms of a trade off in the new interest regime between autonomous, nonbureaucratic, nonhierarchical organizations and lack of influence on or access to the state. The result has been seen as a “neopluralist” pattern (Oxhorn a), a “dilemma of democracy” (Kurtz a), or a situation of “democratization without representation” (Shadlen ). Others have remained more optimistic about the potential for popular organizing and the linking of popular organizations through network structures, but have cautioned that the effectiveness of this new model of popular representation was difficult to discern and would depend upon institutional characteristics of the state and the stability of democratic regimes (Chalmers, Martin, and Piester ; Korzeniewicz and Smith ). These literatures highlight a number of issues and questions that will be taken up in the present analysis. Many of these have implications for the effectiveness of interest organizations individually and collectively as an organizational infrastructure. Like the corporatist and postcorporatist perspectives, this study problematizes the formation of and participation in associations, examines the organizational and financial resources of associations, and relates these topics to the theme of bias or representational distortion in the interest system. Rather than treating the associational world as largely autonomous from the state, as in the pluralist and to some extent the postcorporatist literature, the analysis instead follows the corporatist tradition of paying attention to ways that states and associations are linked and the potentially doubleedged nature of these relationships for popular representation. To what extent are they on the “input” side of politics (pluralism and social movements), are they independent of the state (civil society), or do they serve an intermediating role (corporatism) which, at least to some extent, relates to the “output”
01 Chapter 1.qxd
11/9/2009
11:48 AM
Page 29
,
side of participating in policy implementation? Finally, the analysis shares with all three traditions a focus not just on examining the characteristics of associations, but on examining the nature and aggregate traits of the interest regime more generally.
Goals of Analysis and the Chapters to Follow This book pursues a number of goals. First, it conceptualizes political activity in the interest arena and the notion of a popular interest regime. Second, it describes a historic shift in the urban popular interest regime and, distinguishing between individual and organizational collective action, it lays out the differing logics of collective action of the new A-Net compared to those of the -Hub which characterized most of the twentieth century. Third, it analyzes the ways in which citizens participate in the interest arena both to address particularistic grievances and to advance collective interests. Fourth, it examines the nature of the emergent interest regime by describing characteristics of popular associations, the problem-solving strategies they adopt, and the ways they coordinate. Finally, it lays out four analytic dimensions for comparing interest regimes—scope, scaling, access, and autonomy—and compares the four national cases along these dimensions. As a brief guide, we summarize the ground covered in each of the chapters. Two more introductory chapters immediately follow, which further contextualize the subsequent empirical analysis. Chapter , by Ruth Berins Collier and Samuel Handlin, introduces the approach of this volume in more concrete terms; it lays out some key distinctions, concepts, and overarching themes, discusses the four countries in which the surveys were fielded, and provides historical background on the shift in interest regime from the -Hub to the A-Net. Chapter , also by Collier and Handlin, then undertakes a comparison between the two types of interest regimes, exploring differences in the logics of two types of collective action based on two factors: the nature of base organizations (unions in the -Hub and popular associations in the A-Net) and the orientation of the state (the state that prevailed at the time of the -Hub and the neoliberal state corresponding to the A-Net). It also compares the Hub and the A-Net along four aggregate traits of the popular interest regime: scope, scaling, access, and autonomy. Against this historical contrast of the ANet and the -Hub, the rest of the book presents a more empirically grounded analysis of the A-Net. Part explores patterns of individual participation in the interest arena.
01 Chapter 1.qxd
11/9/2009
.
11:48 AM
Page 30
Chapter , by Thad Dunning, analyzes two types of individual problem solving: direct action, in which citizens themselves make claims to the state, and associational participation, exploring the incidence and correlates of different problem-solving strategies, as well as cross-national differences. It suggests that associational participation is a critical component of the popular problemsolving repertoire, outpacing direct action in three of four countries and especially outpacing the incidence of collective forms of direct action: associationalism appears to be the most important channel for solving problems collectively and pursuing collective or group interests. In the -Hub, unions played an important role in mobilizing members into partisan politics and affiliating them to labor-based parties. Chapter , by Jason Seawright, asks if associations play a similar role in mobilizing participants into electoral politics and if they may thereby counterbalance class bias in the electoral/partisan arena. Relying on a notion of “shared-member linkages,” the chapter finds evidence that the new associationalism may instead provide a mechanism that reinforces and sustains patterns of representational distortion. Part turns from individuals to popular associations. Chapter , by Diana Kapiszewski, examines the repertoire of strategies employed by popular associations. She demonstrates that they tend to engage in both “society-targeted” and “state-targeted” activities and finds an affinity between the use of contentious and institutionalized claim-making activities. The chapter further explores the relationship among the utilization of all these strategies and characteristics such as organizational and financial resources scaling, and state dependence. Chapter , by Handlin and Kapiszewski, pursues the issue of associational scaling in the A-Net. The analysis distinguishes three forms of scaling: interassociational ties, and the presence of two kinds of coordinating associations, labeled Nodal s and Flexible Fronts. Both kinds of coordinating associations appear to play special roles in directing claims to national-level authorities while inter-associational ties may be especially conducive to local-level claim making and protest activity. Chapter , by Candelaria Garay, explores the relations popular associations have with the main actors of the -Hub, unions and parties. The analysis suggests that, in Argentina and Chile, associations tend to develop links to social movement and public-sector unions, and those with such ties have somewhat distinctive strategic profiles. It also distinguishes types of party linkages and suggests that popular associations generally tend to have more contingent and instrumental relations with political parties than existed under the -Hub.
01 Chapter 1.qxd
11/9/2009
11:48 AM
Page 31
,
In concluding, chapter , by Handlin and Collier, emphasizes cross-national variation. It argues that, when interest regimes are analyzed in terms of the four aggregate traits distinguished in chapter , distinct national configurations have emerged. The chapter offers a preliminary explanation of these differences, paying particular attention to the timing of the shift in interest regime, the effects of economic crisis, the continuity of party systems, and the projects of elite actors to mobilize associational support. The chapter also emphasizes the fluidity of the A-Net relative to the -Hub and provides some “updating” about changes in each case, particularly in Venezuela, where the A-Net has been dramatically reconfigured by the chavista project.
02 Chapter 2.qxd
11/9/2009
11:48 AM
Page 32
2 : , , Ruth Berins Collier and Samuel Handlin
This chapter provides a more concrete guide to the orientation and context of the analysis in the empirical chapters. Those chapters address the question of interest politics from the dual perspective of individual participation and interest organizations, corresponding to parts and , respectively. The initial section of this chapter presents the main dimensions and distinctions the study will draw upon and the main themes central to the analysis. As the data are drawn from four specific Latin American countries, a second section of this chapter offers some more detailed information about these country cases, discussing the historical background of each country and the range of variation they represent in terms of economic development and political trajectories. As surveys necessarily reflect a particular moment in time, it also summarizes the specific social, political, and economic milieu in each country when the surveys were fielded. This last point is particularly important given that our research was conducted at times of significant political and/or economic crisis in two of the cases, Venezuela and Argentina. A final section of the chapter offers a more general summary of one of the fundamental ideas underlying the volume, that over the last decades a shift has occurred in the popular interest regime, or the set of organizations that represent the popular sectors in the interest arena, from one centered on a union-party hub (-Hub) to one based in more diffuse associational networks (A-Net). While the substantive chapters of the book examine the A-Net, it is important to explore briefly the decline of the -Hub and the flourishing
02 Chapter 2.qxd
11/9/2009
11:48 AM
Page 33
,
of associations that has led to the new interest regime. The following chapter then builds on this historical narrative by contrasting the two interest regimes more explicitly in terms of characteristics fundamental to their functioning as representative structures for the popular sectors.
Analytic Approach of the Book This book examines political activity in the interest arena, in which individuals and interest organizations are the two primary types of actors. Below, we lay out the basic distinctions that will be employed in the analyses of individual problem solving and of popular associations. We then introduce a set of five major themes for analyzing the interest arena that cut across the chapters of the book, four of which directly bear upon aggregate traits of the interest regime. Individual Problem Solving One of the most fundamental questions regarding the interest arena concerns the degree to which individuals use it as a site for pursuing their interests. To what extent do they participate in what kinds of activities, and how do the popular sectors differ from the middle class in this regard? The survey of individuals examines a wide range of activities. We conceptualize these as problem-solving strategies—defined as types of political action in the interest arena—and develop a framework for their assessment that is based around four analytic distinctions. The first is between direct action and associational participation. We use the term “direct action” to refer to a discrete activity in which the participant himself or herself seeks to address a problem or issue by making some type of state-targeted claim or demand. As is common in studies of political participation, this category includes a set of in-person contacts in which individuals present claims to elected officials or state agencies, as well as other activities in which voice is expressed more distantly, such as engaging in protests and signing petitions. Also included are participation in judicial procedures, an activity that is unexpectedly prevalent, and contacting political parties or influential individuals specifically for the purpose of enlisting them as intermediaries for state-targeted claim making. As suggested by Verba, Schlozman, and Brady (), direct-action activities often require many types of resources—material (time and money), human (knowledge, experience, and “civic skills”), social (contacts), and psychological (a sense of efficacy)—and as such they may, in
02 Chapter 2.qxd
11/9/2009
.
11:48 AM
Page 34
the aggregate, be marked by a pattern of representational distortion to the disadvantage of the popular sectors. Associational participation is by its nature a more sustained and intermediated activity than direct action. Associations typically work in relation to a set of general concerns or issue areas, and their activities are more varied compared to the more one-off types of direct action. Nevertheless, of course, individual participation in associational activities may be more temporally limited and ebb and flow according to the level of interest. One of the most fundamental questions regarding the emergence of the A-Net concerns the nature and extent of popular-sector participation in the new structures, as well as how the popular sectors compare to the middle classes in this respect. The second distinction is between state-targeted and society-targeted strategies. The former consists of claim-making activities directed toward state authorities. The latter consists of provisioning strategies, in which resources are pooled and goods and services distributed to participants or constituencies. While direct action by definition consists of only state-targeted claimmaking activity, associational participation may involve either state-targeted or society-targeted activities. When state-targeted activities are being undertaken, we invoke a third distinction between unmediated and intermediated forms of problem solving. Associational participation is by definition an intermediated activity—individuals participate, but associational leaders are generally responsible for dealing with state authorities and the associations act “between” constituencies and the state. Direct action can involve either unmediated or intermediated problem solving. Individuals can make claims on state officials personally, or they can contact either political parties or influential individuals for the purposes of enlisting or using them as intermediaries. Finally, problem-solving activities are also distinguished in terms of whether they are atomized—undertaken individually or only with family or close friends—or collective. Associational participation and some types of direct action, such as petitioning and protest, are by definition collective forms of problem solving. Other forms of direct action may be undertaken in either an atomized or collective way. This distinction bears on the interests or benefits that are addressed. Atomized direct action primarily concerns personal problems and particularistic, selective benefits (Verba, Nie, and Kim ; Verba, Schlozman, and Brady ).1 On the other hand, collective direct action and 1. Letter writing is a form of individual direct action that, in the United States, for instance, typically addresses broader policy issues; however, this activity is rare in Latin America, and is therefore not part of the survey.
02 Chapter 2.qxd
11/9/2009
11:48 AM
Page 35
,
associational participation are usually undertaken in pursuit of broader group interests, larger policy goals, or public goods. The distinction between particularistic and collective interests is also related to that between state-targeted and society-targeted activities. State-targeted activity can be in pursuit of either particularistic or collective goods, but claim making around the latter constitutes an especially crucial part of democratic representation and policy processes. Society-targeted activity involves only collective benefits, similar to what Rosenstone and Hansen (:) in their study of political participation refer to as “social goods.” The Popular Interest Regime as Organizational Infrastructure The analysis of the organizational component of the interest arena is carried out by examining a delimited set of organizations. Given the present focus on the urban popular sectors, our purview is limited to a set of city-based organizations by, of, or for the popular sectors. Included as popular associations are both what analysts have referred to as grassroots organizations (as associations by and of the popular sectors) and s (as associations for the popular sectors).2 Within the broad set of popular organizations, we further focus the analysis for pragmatic and theoretical reasons. We are especially concerned with those that are “programmatic” in the sense that they have the primary goal of advancing popular-sector interests and undertaking activities to solve collective problems. Although the A-Net includes both labor unions as well as popular associations, the focus is on the latter because they are the most important and most prevalent “base units” of the new urban interest regime. The examination of the A-Net relies on a survey of nearly a thousand associations, approximately in each capital city, implemented through interviews with associational leaders. A “representative” survey based on a random sample was neither feasible nor, for present purposes, desirable. The goal of this analysis is not to generalize about the “universe” of popular associations, which would include many associations that are ineffective and/or ephemeral. Rather, the attempt was to accomplish two goals: to analyze some key associational traits among the great diversity of types of associations and at the same time to capture and explore a “core segment” of the A-Net that is particularly 2. This distinction between grassroots associations and is frequently utilized in the literature on civil society and associationalism to refer to professional versus participatory forms of organization. However, the data suggests that categorizing associations in this way is problematic, as many cases occupy a middle ground.
02 Chapter 2.qxd
11/9/2009
.
11:48 AM
Page 36
visible, well linked, and active. The potential of the A-Net is less a reflection of the “average” traits of associations than it is of the operation—advantages and shortcomings—of such a core segment. The survey was thus based on a chainreferral sample with purposively selected starting points to produce equivalence across national cases on the basis of the importance of types of associations. This selection procedure, partly based on reputation, most likely resulted in an overall sample that particularly tapped the core segment of more active and more linked associations, at the same time that it covered great diversity, including, for example, many nonlinked, isolated associations (see appendix B). In examining the interest regime, we focus on the traits and activities of associations. Parallel to the approach to individual participation, we analyze the strategies associations employ in terms of a basic distinction between statetargeted or claim-making strategies and society-targeted strategies. Associational claim making can take place through institutional or noninstitutional channels, the former consisting of a variety of contacting activities or the pursuit of legal claims and the latter consisting of organizing protests. Two types of society-targeted activities are distinguished: “provisioning” projects that involve providing goods or services to constituencies or organizing local events and “capacity-building” activities, such as training leaders or creating and financing other associations. With respect to all of these strategies, the study explores only whether the association pursues them and their relative importance within the association’s repertoire of action, leaving aside the question of their effectiveness in actually solving problems. Many other factors that may shape the influence of associations are beyond the scope of this analysis. We seek here to take a first step toward examining the organizational infrastructure of the associational world available to the popular sectors. The study conceptualizes multiple forms of scaling within the A-Net. A first is inter-associational coordination or network “embeddedness,” which is observed in the number of working ties that associations have with other associations and gives us purchase on the density of networks. A second is the presence within networks of associations that are particularly oriented toward serving a coordinating function and that thereby may play crucial roles within the A-Net. Two different kinds of coordinating associations, Nodal s and Flexible Fronts, are distinguished in this respect. A third is coordination between associations and unions and the position of these union-linked associations within the A-Net, which may be important because of the skills and resources traditionally possessed by unions, and their consequent possible impact on coordination and strategies in the A-Net more generally.
02 Chapter 2.qxd
11/9/2009
11:48 AM
Page 37
,
Many activities bring associations into contact with the state or, less frequently, political parties. Regarding state-association ties, we distinguish state-targeted claim making from state dependence, such as when associations receive funding from the state, implement state programs, or were partly founded by state actors. Regarding party-association ties, we distinguish between the strategies of associations in problem solving and the strategies of political parties in appealing to associations for electoral support. Three forms of the former are analyzed: issue-based (acting in the interest arena to address some issue), candidate-based (acting in the electoral arena to support a candidate or party), and multifaceted (acting in both interest and electoral arenas through more ongoing collaborative relationships) linkages. Finally, the study examines key associational traits in order to assess both the diverse associational landscape and the relationship between these characteristics and other activities. In this regard, we look at organizational traits concerning how the association is structured and run (e.g., whether it has formal state recognition, permanent leadership, paid staff, formal members) and those related to the financial resources of the association. Analytic Themes Across the Chapters Five major themes are pursued in the analysis of the interest arena: individual problem-solving repertoires, representational distortion, associational strategies, scaling, and state-association ties. These provide the basis for a comparative landscape of the popular interest arena in Latin America in terms of both individual and associational actors, as analyzed in chapters through . At the same time, four of these themes are closely related to aggregate traits of the interest regime, which receive attention in chapters and . Scaling is itself such an aggregate trait. A second, the scope of the interest regime, is derived from an analysis of individual problem solving and the extent to which individuals, in aggregate, choose to participate in associations. Two other aggregate traits, access to the state and autonomy, cannot be easily observed directly. Nevertheless, we derive some purchase on these two aggregate traits by examining, respectively, associational accessing strategies and state-association ties of dependence. Individual Problem-Solving Repertoires. A central theme concerns the extent of individual participation in the interest arena, the types of activities individuals undertake to address both individual and collective problems, and how associational participation fits into the larger problem-solving repertoire of the popular sectors. These questions are taken up by several chapters of the
02 Chapter 2.qxd
11/9/2009
.
11:48 AM
Page 38
volume, from a number of different angles. Chapter presents a comparative analysis of the scope of the interest regime, by developing a theoretical argument concerning collective action problems in forming and participating in associations in the A-Net compared to those for unions in the -Hub. Chapter situates associational participation within the larger individual problem-solving repertoire, examining rates of participation in a variety of direct-action activities, as well as participation in associations, and exploring the correlates of both forms of participation. Chapter expands this analysis, examining rates of participation and the relationship between associational participation and engagement in activities in the party/electoral arena. Representational Distortion. Another theme concerns the class bias in the interest arena. How does popular-sector participation in the interest arena compare with middle-class participation? Several chapters take up this question. Chapter shows that while the middle classes have higher levels of participation, some attributes commonly possessed by popular-sector individuals— such as a history with labor unions or land invasions—are significant predictors of associational participation and to some extent serve to reduce the degree of representational distortion. Chapter approaches the subject of representational distortion from a different angle, showing that in cross-national comparison the extent to which the popular sectors are underrepresented in civil society mirrors the extent to which they are underrepresented in the body of electoral participants: the proliferation of associations has provided important problem-solving vehicles for the popular sectors, but they do not reduce or counterbalance class-based distortion in the electoral/partisan arena. Associational Strategies. Popular associations pursue a variety of strategies in addressing collective problems. This diverse repertoire raises the question of whether some strategies are pursued by particular types of associations. Chapter typologizes strategies and investigates their correlates. The analysis shows that the A-Net is characterized by a preponderance of provisioning, rather than claim-making strategies. Nevertheless, popular associations do pursue a significant amount of claim making within their repertoire of action, and to some degree are able to target the national level. Further, a substantially “dual” pattern exists in which rather than a “division of labor” among associations regarding these strategies, many associations pursue both state-targeted claim-making strategies (through institutional channels and through protest) and a variety of provisioning strategies. Chapters and , respectively, suggest that associations that coordinate networks and those linked to labor unions play special roles in channeling claims toward national-level officials.
02 Chapter 2.qxd
11/9/2009
11:48 AM
Page 39
,
Associational Scaling. The widespread image of fluid networks of associations raises theoretical and empirical questions of inter-organizational coordination. How frequently do associations engage in collective action among themselves? To what extent, in this “flatter” networked model, are there associations that take on the role of coordinating others, and what are the properties of these coordinating associations? Chapter undertakes a comparative analysis of coordination among unions in the -Hub versus associations in the A-Net, moving beyond the Olsonian tradition and incorporating some factors emphasized in the literature on new social movements, to develop a theory explaining why associations may have more difficulty coordinating their activities. Other chapters turn to the empirical investigation of the networks. Chapter assesses the extent to which associations coordinate with others and examines associations oriented toward playing a coordinating role in networks. It contrasts these coordinators with the labor federations of the -Hub and suggests that the A-Net coordinators may be divided into two categories that display markedly different organizational traits and pursue different problemsolving repertoires. Chapter demonstrates that associations with linkages to unions and with some types of linkages to parties are particularly embedded within networks, suggesting that they may wield greater influence than other associations. State-Association Ties. While the new associationalism has been lauded for its autonomy from the state, in fact many associations are linked to state actors through a variety of means. As further discussed in chapter , “autonomy” presents both conceptual and empirical challenges. Hence, our empirical approach is to look at “dependence,” that is, the relations associations have with the state—and also with political leaders and parties—that may challenge associational autonomy. That is, the analysis is focused at a step removed: at the activities and relationships through which associations engage the state or political parties and that may affect or shape associational agendas and decisions. What kinds of state-association relationships have emerged, and with what implications? Chapter compares the relationships that unions and associations have with the state. Chapter explores three types of state-association ties: associations founded in part by state actors, associational dependence on the state for resources, and the associational role in implementing government programs, in a kind of state outsourcing of public services. The role of parties is also important in state-association ties. In the -Hub, political parties intermediated many of the state-union relations. Chapter shows that many associations have relationships with parties, but that these relationships are most likely to remain relatively instrumental and distant.
02 Chapter 2.qxd
11/9/2009
.
11:48 AM
Page 40
Four Countries This volume is based on surveys carried out in the capital cities in four Latin American countries—or, more accurately, their metropolitan areas. The analyses of these cities are not intended to approximate national patterns of the urban A-Net in each country; but rather, these cities are important political sites to understand in their own right. They are extreme forms of “primate” cities, containing – percent of the national population and an even larger proportion of the urban population and of the economy. These megacities have been particularly affected by the recent, dramatic changes in economic policy. They tend to exhibit the most dramatic problems of governability; the biggest changes in the labor market with the rise in unemployed and informal workers; profound social problems with the growth of the poor; and unprecedented levels of violence. Because of their economic and demographic size and their political centrality as capitals, they are uniquely connected to national politics (Grindle a; Myers and Dietz ). Finally, the capital cities have been the primary sites for the development and proliferation of urban associationalism in each country. They constitute spaces in which associational activity has become most visible and where government programs vis-à-vis associations have been most saliently implemented. The four countries in this analysis cover a wide range of socioeconomic variation among Latin American countries that underwent substantial processes of import substitution industrialization () in the twentieth century. Argentina embarked on early industrialization, based primarily on European immigrant labor and supported by an extensive form of temperate agriculture, in ranching and wheat, with a small peasantry. At the beginning of the twentieth century, Argentina was the richest country in Latin America and had developed the strongest labor movement. Chile also had substantial European immigration and an early development of manufacturing, although its economy was dominated by the mining sector and although a larger peasantry made up a pool of surplus labor. Venezuela and Peru were later industrializers, with large peasantries, supplied in Peru by the country’s large indigenous population, which worked in both commercialized export crops on the coast and often large uncommercialized haciendas in the highlands, until both types of large-scale holding were reformed by the military government that came to power in . Venezuela came to be a major oil exporter during the twentieth century, and by midcentury it leaped to very high levels of per capita. Peru remained a relatively poor country by Latin American standards, with the largest peasantbased source of surplus labor of the four countries considered here.
02 Chapter 2.qxd
11/9/2009
11:48 AM
Page 41
,
Through the twentieth century, the patterns of economic growth reordered the relative standing of the four countries to some extent. In Peru and Venezuela had similar levels of per capita, and Argentina, the region’s economic leader, had a more than three times greater, with Chile in between. By , after the major push of primary import substitution, the gap among the countries narrowed. The exception was oil-rich Venezuela, which, according to some measures, became the region’s richest country (see tables . and .). The economic crises and adjustments of the s and particularly the s, when the Latin American region encountered the debt crisis and a period of essentially no growth, severely affected economic levels: by per capita in Venezuela was actually lower than it had been thirty years earlier; post growth in Peru was wiped out by and nearly so in Argentina; only Chile recovered from the downturn enough to achieve levels higher than , increasing by about half again as much. Following the “lost decade” of the s, growth generally resumed in the s until the downturn of –, though the rate was uneven both within and across countries (see table .). The exception was Venezuela, where the economy was quite stagnant, after two years of strong growth (–). The Chilean economy grew robustly, at an average annual rate of . percent – ; and Argentina and Peru were in the middle. Table . Comparisons of Per Capita (Peru = )
Argentina Chile Peru Venezuela
1900
1960
1990
337 230 100 100
184 142 100 322
219 213 100 271
Source: McGuire Development Database. Calculated from Maddison Data in Geary-Khamis dollars.
Table . Long-Term Growth of Per Capita ( = )
Argentina Chile Peru Venezuela
1900
1960
1990
100 100 100 100
202 221 370 1185
239 328 367 995
Source: McGuire Development Database. Calculated from Maddison Data in Geary-Khamis dollars.
02 Chapter 2.qxd
11/9/2009
.
11:48 AM
Page 42
Table . Change in Per Capita, – (percent change, in dollars) Argentina 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003
10.6 9.6 5.7 5.8 -2.8 5.5 8.1 3.9 -3.4 -0.8 -4.4 -10.9 8.8
Chile
Peru
Venezuela
8.0 12.3 7.0 5.7 10.6 7.4 6.6 3.2 -0.8 4.5 3.4 2.2 3.7
2.8 -0.4 4.8 12.8 8.6 2.5 6.8 -0.7 0.9 2.9 0.2 4.9 4.0
9.7 6.1 0.3 -2.3 4.0 -0.2 6.4 0.3 -6.0 3.7 3.4 -8.9 -7.7
Source: , División de Estadística y Proyecciones Económicas, Sección Cuentas Nacionales, http://website.eclac.cl/sisgen/Consulta.asp (accessed April ).
A concomitant of these macroeconomic patterns has been a sharp worsening in the region’s notoriously high level of inequality, a trend that failed to turn around, even once growth resumed in the s. Argentina, a country that was traditionally more equal than others in the region, came to have a high level of inequality, while the Gini index of Chile became one of the highest in the region, as inequality continued to worsen despite the notable improvement in that country’s poverty rates. Politically, the four countries in the present study are also quite diverse. For present purposes, an important aspect of this variation is the nature of the labor movement and the political position of the with which it was historically affiliated. In Chile, as elsewhere, the early labor movement gave support to the Socialist and Communist Parties. Unlike the other countries, this affiliation was never successfully challenged by populist parties. The result was a strong electoral bloc on the Marxist left, which grew through the century and captured the presidency in . In the other countries a multiclass did form and attract the allegiance of the labor movement: the Partido Justicialista () in Argentina, the Alianza Popular Revolucionaria Americana () in Peru, and Acción Democrática () in Venezuela. Though the programs of these three populist parties were more moderate than that of the Marxist Unidad Popular () in Chile, in Argentina and Peru these parties also provoked intense opposition. In three of the four cases, then, the polarization resulted
02 Chapter 2.qxd
11/9/2009
11:48 AM
Page 43
,
Table . Inequality Trends Argentina
Chile
Peru
Venezuela
Income ratio circa 1990: Top/bottom decile (equivalized)
15.9 (1992)
32.5 (1990)
31.4 (1994)
16.6 (1989)
Income ratio circa 2000: Top/bottom decile (equivalized)
32.9 (2001)
37.0 (1999)
41.0 (2000)
24.6 (1998)
Gini circa 2000 (equivalized household income)
50.4 (2001)
56.1 (2000)
47.7 (2000)
45.5 (1998)
Gini change: Circa 1990–circa 2000
7.7
1.4
2.0
3.8
Source: De Ferranti : –.
in military governments in the s and/or s. Only Venezuela remained democratic during this period. All four countries have democratic regimes in the current period, though the regime dynamics and party systems are quite different. The regimes in Chile and Venezuela, which had an extended history of democracy in the twentieth century, are the outcomes of quite different—even opposite—historical trajectories, from polarization to stable and moderate “two-party” (actually two-bloc) competition in the former and the reverse direction of change in Venezuela. In Chile the Marxist parties substantially moderated their positions, and the democratic transition of –, which followed a sixteen-year period of military rule, saw the inauguration of a party system based on competition between two electoral blocs, the center-left Concertación (composed primarily of the Partido Demócrata Cristiano [], Partido Socialista [], and Partido por la Democracia []) and the rightist Alianza (composed of the Unión Demócratica Independiente [] and Renovación Nacional []). The Concertación has held the presidency and the legislative plurality since the return of democracy. While increasing social spending and attacking poverty, it generally maintained the economic model introduced under General Augusto Pinochet as the leading edge of economic reform in the region. The premilitary pattern of political polarization has thus been superseded by stable, centrist two-bloc competition, even after the Concertación presidency passed from the Christian Democrats to the Socialists under Ricardo Lagos in . Electoral dynamics moved in the opposite direction in Venezuela, from
02 Chapter 2.qxd
11/9/2009
.
11:48 AM
Page 44
stable and centrist two-party competition to intense polarization as part of a development that saw the collapse of the parties that had dominated politics for half a century. Starting in the late s, moderated, ejected its left wing, and ruled in cooperation with the social Christian party, , which twice won the presidency, providing the political integration and stability that allowed Venezuela to be one of the few countries that did not succumb to a military rule in the s. By the mid-s the highly bureaucratized and ossified parties had failed to retain the support of the informal sector, particularly in the face of a poor economic record. When in the president entered an agreement with the International Monetary Fund () and adopted the unpopular package of economic reforms prevalent in the region, the country erupted in protest, leading to a bloody military crackdown. These events were followed by two military coup attempts in a country widely considered one of Latin America’s few stable democracies, and, eventually, the collapse of the traditional two parties. After a decade of political fluidity, former army lieutenant colonel Hugo Chávez, the leader of one of the coup attempts, was elected president at the end of . Chávez has been an outspoken firebrand and a leader of a new left pole within Latin America that rejects the more marketoriented economic model. Popular among the lower classes to whom his “Bolivarian revolution” has been more attentive in social policy, Chávez has alarmed the middle and upper classes by his consolidation of power, his restructuring of the judiciary and the legislature, his confrontational rhetoric and symbolism, a lack of attention to certain kinds of public goods like security, and his policy priority on providing the poor with basic needs—education, health care, and better housing. Venezuelan politics has become extraordinarily polarized, and on more than one occasion the opposition has attempted to oust Chávez through legal and illegal means, including a referendum on his recall, a paralyzing work stoppage, and a failed coup attempt. The third wave democracies in Argentina and particularly Peru have been characterized by considerable party instability. Although the and were each the largest party in their respective countries, military opposition had prevented them from holding power from the s until coups in the s put an end to these “impossible games” (O’Donnell ). In the postmilitary democracies, these parties have been normalized and operate without the earlier ban. Neither party emerged from the military period capable of winning the first presidential election of the new democracies; however, both won in the next presidential election. Both presidencies were discredited at the end, although in Peru this occurred more rapidly, where the experience seemed to
02 Chapter 2.qxd
11/9/2009
11:48 AM
Page 45
,
mark the collapse of the party system, as it ushered in an extended period of rule by Alberto Fujimori (–), an outsider, nonparty president who perpetrated an auto-golpe that inaugurated an authoritarian interlude. A second outsider, Alejandro Toledo, was then elected, and outsiders continue to be prominent, although remained a well-organized party, which won the presidency in the second round in . In Argentina the fared better, for perhaps two reasons. First, the came to power in the midst of an economic crisis and carried out deep economic reforms that brought hyperinflation under control and initially produced high growth rates, whereas ’s economic policy was not successful. Second, the successfully transformed itself from a labor-based party () to a clientelistic machine, thereby changing its constituency relations and retaining the support of the lower classes, whereas ’s eclipse reflected its loss of the support of both unionized and informal workers (Levitsky ; Cameron ). The Argentine party system has experienced an unusual combination of stability and change. With the greater adaptability of the , the deep-seated Peronist/anti-Peronist divide continued to structure politics, and Peronism remains the largest political force. In contrast to the continued core support of the , the anti-Peronist forces experienced substantial instability, as the longstanding Unión Cívica Radical () captured the presidency twice but failed to complete either term in the face of economic crisis, both times subsequently experiencing rapid political decline. A much weakened and reduced has become dependent on new alliance partners, which represent the rise and fall of new forces in the party system. At the same time, while remaining dominant, Peronism has fragmented and taken on different forms under two Peronist presidents who sustained a high level of popularity: Carlos Menem (–), in a programmatic about-turn for Peronism, introduced far-reaching market reforms, while Néstor Kirchner (–) partly joined the backlash against the economic reform model in the region. The surveys in each country were carried out in –. The data necessarily reflect a snapshot at that particular moment. Therefore, a description of the conjuncture in each country is in order. The Peruvian fieldwork was conducted in July–September . Growth was just resuming, following a period after the economy had performed well under Fujimori from to and then stagnated. In November Fujimori resigned amid popular protest after a political corruption scandal, and Toledo was elected in spring as the first indigenous president of Peru, promising to rectify the corrupt practices and political abuses of the Fujimori period and to represent the indigenous poor who had traditionally been excluded from
02 Chapter 2.qxd
11/9/2009
.
11:48 AM
Page 46
the political sphere. After his first one hundred days in office (late July–early November ) citizens were frustrated with his already apparent inability to make firm, quick decisions (in stark contrast to Fujimori), but seemed willing to give him the benefit of the doubt. That willingness eroded bit by bit over the next year, so that by fall (around the time of the surveys) Toledo’s approval rating had plummeted. He did manage to stabilize the economy and record modest growth, so that macroeconomic indicators improved. However, that growth failed to trickle down, and Toledo’s indecisiveness and continued political scandals, along with popular protests over issues such as privatization, fostered already simmering discontent with the political system.3 In sum, the surveys were fielded at a time of strong disaffection with politicians and frustration that Peru’s newly installed democratic government was proving to be dysfunctional. In Venezuela, the surveys were fielded in early , during a moment of economic and political crisis. The economy had performed poorly for a decade, following an even longer history of economic stagnation. In the economy experienced a decline of nearly percent.4 The economic crisis itself was directly related to the political crisis. In Chávez increased government control over the petroleum industry, heightening the opposition of both capital and labor, among whom the petroleum workers had historically been a leading group. The year was marked by demonstrations, occasional strikes, and most dramatically a coup in April that briefly ousted Chávez. In midDecember, an opposition movement consisting of the traditional political parties, the union movement, and representatives of business launched a debilitating two-month general strike, which shut down much of the economy, including Venezuela’s state-owned oil company, and which came to an end only weeks before the surveys were fielded. The surveys were thus taken at a particular moment when politics was monopolized by the issue of the ousting or survival of the regime. This period predated Chávez’s most important and most ambitious initiatives to expand social service delivery and sponsor associationalism among the popular sectors, although some associational initiatives, such as the sponsoring of Bolivarian Circles, were already underway. The surveys provide a snapshot of Venezuela at a very particular though interesting time: after the collapse of the long-standing and well-institutionalized party system and immediately following an intense year of regime contestation, but 3. In a September opinion poll, approval of the economic policy and the government, as well as the president, was less than percent ( ). 4. Calculated from ’s online Base de Estadísticas e Indicadores Sociales.
02 Chapter 2.qxd
11/9/2009
11:48 AM
Page 47
,
before the construction of the major chavista organizational structures in popular-sector communities. The fieldwork in Chile was conducted between August and January . Although the – election continued the Concertación’s unbroken string of presidential victories, it brought to power a Socialist rather than a Christian Democrat for the first time since the transition. The election was one of the closest in Chile’s history. President Ricardo Lagos (–) faced the country’s first post-transition economic recession, induced by the international financial crisis of the late s. Further, social tensions surrounding the issue of former dictator Pinochet’s responsibility for the human rights abuses continued, having been reignited by Pinochet’s arrest in London in . In general terms, Lagos continued the market-oriented economic policies that had guided Chile’s economy since the mid-s, making it a model reformer and one of the fastest-growing economies in the region. The basic continuity could be seen in the long-awaited labor reform of , which made only marginal changes and maintained a high level of labor market flexibility; in the dismantling of capital controls completed by ; and in several new free trade agreements. The discovery of a series of corruption scandals in and (which eventually reached the Ministry of Public Works, which Lagos had previously led) prompted efforts to revamp political institutions, and by early , constitutional reform was under way. Thus, despite the election of the first Socialist head of state since Salvador Allende (–), and in contrast to Argentina, Peru, and Venezuela, the moment during which the surveys were fielded in Chile was one of political continuity and relative economic stability, as well as generalized disaffection from the political system and political parties in particular on the part of many Chileans. The Argentine surveys were carried out in July–September . Economic growth was resuming after the dramatic financial collapse of December , which was followed by a devaluation of the currency in January . In the accompanying political crisis, Alianza president Fernando De la Rúa resigned, and after a succession of short-lived interim presidents, Eduardo Duhalde of the was finally appointed as president in January . In a context of high levels of protest by unemployed workers, Duhalde prioritized social spending. He launched a massive workfare program and distributed significant resources to community soup kitchens to cope with the dramatic effects of the crisis and to curb the massive protest. Indeed a movement of unemployed workers and popular associations had formed in the late s and had grown dramatically even before the financial collapse. The instability surrounding the political crisis was overcome with the election of Peronist Néstor Kirchner as president in
02 Chapter 2.qxd
11/9/2009
.
11:48 AM
Page 48
May . Although Argentina continued to witness unprecedented poverty rates and high levels of protest during , the popularity of the new president was high, with an approval rating of about percent. After taking office, Kirchner adopted a confrontational discourse against the neoliberal reforms that had been implemented by president Carlos Menem in the s. At the time the surveys were carried out, popular associations were receiving significant resources from the state. It was a period of resumed economic growth and expanding government programs, under a Peronist president, to ameliorate the effects of the neoliberal economic model.
The Shift in the Popular Interest Regime The historical shift in the popular interest regime from one centered on a union-party hub (-Hub) to one based in urban associational networks (ANet) has occurred in tandem with—indeed, has partly been spurred by—a set of national and international political and economic changes with enormous consequences for the popular sectors. To contextualize the investigations in this book, it is important to elaborate on the historical shift in the interest regime and the factors that brought it about. The Rise and Decline of the -Hub A popular interest regime centered on the -Hub characterized the middleincome countries of Latin America throughout much of the last century. This regime had its roots in the early stages of industrialization, when a new proletarian class was formed—indeed “made”—socially, ideologically, and organizationally (E. Thompson ; Katznelson and Zolberg ; Bartolini and Mair ; Collier and Collier ). The growth of the proletariat toward the end of the nineteenth century gave rise to two new sociopolitical technologies: the labor union and the union-affiliated political party (). In Western Europe, the two emerged in tandem as part of the organizational drive of the socialist movement. Unions with links to socialist and communist parties were also established in Latin America and became the most dynamic form of popularsector organizing from below. However, unions also received a great deal of elite attention, as the dominant classes reacted to a development they found threatening: politically, union discourse was anticapitalist and revolutionary; and economically, unions threatened disruption through strike action and, more generally, limits to employer control of the workplace.
02 Chapter 2.qxd
11/9/2009
11:48 AM
Page 49
,
As a result, political leaders sought to moderate, control, and co-opt the labor movement, and in a large subset of countries multiclass populist s managed to affiliate unions, displacing communist and socialist parties (Collier and Collier ). Among the four countries in the present study, political leaders founded such populist s in the s and s in Argentina (), Peru (), and Venezuela (), but not in Chile, where the more classist Communist and Socialist Parties remained the most important s.5 Whether or not populist parties were formed to win their support, unions were legalized and integrated into the political game, and they became the most important and privileged organizations of the working classes or popular sectors. The label “-Hub” reflects the facts that unions were the base organizations at the core of the popular interest regime and that they were closely integrated into political parties—either populist or leftist—through formal organizational links, interlocking leadership, or close coordination, as well as through a history of electoral mobilization and political socialization. It should also be noted that sooner or later in each of these cases the to which the main unions were affiliated became the largest party (or, in the case of Chile the largest coalition, made up of the two major s) in the country. Of course, only a small part of the lower classes participated in unions. Among wage earners, or proletarians, union density varied substantially, by country and sector. Further, peasants were not included in most countries, although Mexico, Venezuela, and Bolivia were exceptions. More importantly for present purposes, unions did not include the urban informal working class, which grew, particularly starting in the s, with rapid urbanization and the failure of late industrialization to absorb labor at rates comparable to the historical experience of the early developers. Community-based associations had historically been a feature of urban Latin America, including in popular-sector neighborhoods in the cities examined here, with roots that often go back to the mid-nineteenth century. In Buenos Aires, for instance, mutual aid societies and ethnic associations emerged as early as the s, and urban associationalism grew further with early urbanization in the first decades of the twentieth century (Sábato ; ). Later on, in the interwar period, neighborhood associations crossing ethnic lines became relatively common vehicles for representing local interests (de Privitellio ; Gutierez and Romero ; L. Romero ). In general in Latin America, 5. In Peru, dates from the s, when its founder was in exile in Mexico, but appeared as a force inside the country in the s. In Argentina, the Peronist movement took various organizational forms and party labels before finally stabilizing as the Partido Justicialista.
02 Chapter 2.qxd
11/9/2009
.
11:48 AM
Page 50
popular-sector associationalism grew in tandem with twentieth century urbanization, especially in the post–World War II era. Beginning in the s and s, neighborhood associations proliferated with the establishment of shantytowns on the urban peripheries, often through organized land invasions. Neighborhood associations, some of which grew out of invasion organizations, sprung up as a means of demanding urban services as well as self-provisioning for collective neighborhood needs (Dietz ; C. Schneider ; Castells ). To the extent they had political linkages, these were primarily to s: in Argentina, the Peronists; in Chile, the Socialists and Communists, and later the Christian Democrats; in Venezuela, and later ; in Peru, , but also Manuel Odría, who made alliances with neighborhood associations both during and after his presidency.6 While these associations were certainly a part of the postwar interest regime we refer to as the -Hub, the point is that they were nevertheless peripheral when compared to the reach, salience, and political importance of labor unions. Although, as mentioned, some presidents tried to mobilize the support of these associations, which benefited from some limited degree of discretionary distributions, associations rarely became politically important either as a support base or as influential actors, and they never attained the status of the more politically important labor movement. Thus, unionized workers became the most politically privileged, relevant, and mobilized popular-sector actor, benefiting from state attention and relatively favorable social policy, despite the fact that a restrictive labor code established a pattern of state corporatism, through which the state controlled the formation, structure, and activities of unions, and despite the fact that links to populist parties also constrained unions (Collier and Collier ). Although unions leveraged their political strength for particularistic gains for their members, they also won broader concessions, such as subsidies on food and other basic consumption goods whose benefits were important to the popular sectors as a whole. The Political Economy of . The -Hub either emerged or was consolidated under a state-centric economic model oriented toward import substitution. had a number of characteristics that supported this interest regime, among them the promotion and protectionist support of national industry; a large public sector, including state-owned firms as well as a state bureaucracy, that was an important source of employment and unionization; and a kind of 6. See Bamberger (), Pratt (), D. Collier (), Castells (), Coppedge (), Gutiérrez and Romero (), Oxhorn (), C. Schneider (), Dietz (), Crisp and Levine (), Sábato (), Garcés (), L. Romero (), de Privitellio (), and Forment ().
02 Chapter 2.qxd
11/9/2009
11:48 AM
Page 51
,
(peripheral) Fordist regulation of the economy that both promoted production on the supply side and undertook policies that sustained aggregate demand (Lipietz ). The demand-side logic of this inward growth model, which produced industrial goods for a domestic market, provided the structural basis of a “class compromise” by at least making room for a state willing (in line with the political strategies of the incumbents) to adopt policies at the national level that would increase aggregate demand. Pro-union and pro-worker policies—“rigidities” in the labor market that would maintain high levels of employment in the formal sector, minimum wages and rising wages in line with productivity gains, subsidies on basic consumption items, and health care and pensions—all increased aggregate demand for national producers in settings where small markets were often ultimately limiting. Unionization was particularly strong in the state sector, which grew large under . Such policies helped to reduce competition within as well as between industrial classes and solve collective action problems of both workers (competing for jobs in a situation of labor surplus) and employers (reducing competition on the basis of labor cost). The degree to which these policies were put in place varied according to the configuration of political coalitions, their institutionalization in party systems, and the resulting political dynamics. Nevertheless, parallel to the Fordist logic of the Keynesian welfare state, encouraged a form of “organized capitalism” that favored encompassing, peak associations of labor and that focused demand making on policies at the national level. In comparison with other late developers, Latin American countries had unusually deep and lengthy experiences under , leading to particularly pronounced and institutionalized privileges for organized labor and positions of exceptional centrality for s in many countries. Furthermore, in this context, political parties found in unions an available source of organized support with substantial mobilizational capacity. Some unions in Latin America remained affiliated to socialist and communist parties, most notably in Chile. More typically, populist parties, led by middle-sector interests, sought to establish linkages and formally or informally to affiliate unions in order to simultaneously deradicalize, control, and co-opt unions and to mobilize the political and electoral support of workers (Collier and Collier ). If concessions were needed to maintain that support, they did not contravene the demand-side logic of the model. The Decline of and the -Hub. The socioeconomic underpinnings of the union-party hub were disrupted in the s and s, as the model was abandoned in the face of three factors: the limitations of the model itself, an acceleration of economic integration in the global economy, and the
02 Chapter 2.qxd
11/9/2009
.
11:48 AM
Page 52
exigencies of the debt crisis beginning in . The renunciation of the model in favor of a more market-oriented approach challenged deeply entrenched actors and institutions and fundamentally altered precisely those conditions that sustained the -Hub. It weakened unions and altered union-party linkages. The dramatic change from a protected, state-led economy to an internationally open, market-oriented economy changed the social-structural base of politics and ushered in a different political and coalitional logic for incumbent governments. Protectionist measures gave way to free trade policies; demandside regulation was replaced with a concern for efficiency, international competitiveness, and attracting foreign investment; and the state was downsized. Under these circumstances, the demand-side logic for higher levels of formal employment and higher wages is no longer relevant. Powerful national firms began to look to international rather than domestic markets, and imports or transnational corporations rather than national firms increasingly supplied the local market. With more open, internationalized economies that put a premium on efficiency and international competition, labor became no longer both a cost and a source of demand for producers, but only or primarily the former. From the logic of this perspective, unions came to constitute an unwelcome rigidity in the labor market. The effects of these changes in employment hit unions especially hard. State and firm restructuring led to layoffs, especially in those sectors where unionism had been especially strong: state firms and key manufacturing sectors. By the mid-s, union density had dropped almost everywhere, often precipitously. In Argentina, historically the region’s most unionized economy, membership fell from to percent, while in Peru membership plummeted from to percent of the workforce. Only Brazil, which experienced a slight uptick, saw a trend in the opposite direction, and Chile saw some recovery in the s from an earlier steep decline to . percent in (see table . in chapter ).7 Further, ongoing efforts in most countries to flexibilize labor laws regarding employment and collective rights have challenged unions and put unionized workers on the defensive as they face an erosion of the boundaries that had demarcated, even insulated, them from the informal sector (Murillo ; Murillo and Schrank, ; Cook ). The new economic model challenged not only unions, but also their relationship to political parties, so that unions had a harder time maintaining influence within those parties large enough to be serious contenders for power. 7. These figures are taken from a database compiled by Kenneth Roberts and used in his forthcoming book, Party System Change in Latin America’s Neoliberal Era, using data from the and other sources.
02 Chapter 2.qxd
11/9/2009
11:48 AM
Page 53
,
The change in union-party relations was part of a larger pattern of change also occurring in advanced industrial countries (Katz and Mair ), and analysts have widely noted the decline or transformation of s, or more generally of “mass” parties that relied on affiliated social organizations. In Latin America many governing s adopted—indeed initiated—economic reform policies that contravened the interests of unionized workers. This programmatic reorientation has been accompanied by coalitional adjustment, as s have restructured their constituencies by turning increasingly away from unionized workers and toward the popular sectors more generally, a larger lower-class stratum embracing informal workers (R. Collier ; Levitsky ; Burgess and Levitsky ). Roberts’s (forthcoming) analysis demonstrates the extent to which s and the party systems based on them have been particularly vulnerable to decomposition or transformation, experiencing the greatest volatility, discontinuity, and dislocation. The four party systems in the present study are no exceptions. The s in Peru and Venezuela have experienced dramatic collapse;8 during the reform period Argentina’s came to rely more on clientelism based on neighborhood brokers and machine politics than on unions, and Chile’s s have similarly either become distant from unions (the Socialist Party) or significantly declined in electoral strength (the Communist Party). Since approximately the s, then, unions have been weakened and put on the defensive as unionization rates have fallen and labor reform has been pursued—though not always successfully—as a way to flexibilize the labor market and reverse some of the earlier gains of unions. Further, the links between unions and s have weakened, and the day of the classic mass party that drew its core support and mobilizational capacity from affiliated union organizations has been eclipsed. The result has been a disempowering of the unionparty hub within the popular interest regime. The Rise of Urban Associationalism and the A-Net While unions have declined, Latin America has witnessed an unprecedented proliferation of popular associations in urban areas over the last two or three decades.9 Neighborhood organizations, community soup kitchens, microcredit 8. While all existing parties in Peru were dramatically eclipsed starting in , the victory of was evidence of the recovery of what had been the original ; but the traditional left, which had come to represent a greater proportion of the union movement, has effectively disappeared. 9. See, among many others, Chalmers, Martin, and Piester (), Alvarez, Dagnino, and Escobar (), Dietz (), Roitter, List, and Salamon (), Sanborn et al. (), Eckstein (a), and Schönwälder ().
02 Chapter 2.qxd
11/9/2009
.
11:48 AM
Page 54
associations, street vendor unions, community policing and safety groups, women’s groups, s working on environmental or local development issues, and many other forms of associations have come to characterize Latin American metropolises. The heterogeneity of these associations is reflected in the number of labels used by scholars to refer to them, such as civil society organizations, social movement organizations, community organizations, grassroots organizations, s, and advocacy groups. As discussed above, these forms of organizing are not entirely new. Associations among the working classes date back to the early twentieth century in some areas. More generally, as urbanization picked up pace in the postwar era, popular associations often formed in the aftermath of land invasions and the development of new settlements. However, the diversity and sheer number of popular associations distinguish the present period. The flourishing of urban associationalism can be seen as a response to domestic economic and political change, as well as a more supportive international environment, especially in the period since . Economic crisis was perhaps the most immediate stimulus for the new associationalism. For the region as a whole, the international debt crisis of the s was a moment of inflection in economic and social trends, which until then had been generally positive for decades: though it became fashionable to talk of the inefficiencies of the state-led model, Hirschman, looking back from the vantage point of the s, pointed to the strong postwar record of growth and social development and mused that like Europe, Latin America may have had its own version of the “trentes glorieuses—and perhaps a few more.” (Hirschman :). The trend generally continued into the s, despite the oil shocks, and that decade of debt-led growth saw widespread improvement in social conditions: the poverty rate decreased, and even those who remained poor improved their situation. In the s, however, this trend was dramatically thrown into reverse, as the debt crisis and economic adjustment imposed particular economic costs on the popular sectors amid the austerity and recessions of the “lost decade.” Even when these often volatile economies resumed some growth in the s, the distribution of benefits was highly concentrated, and the new economic model largely failed to deliver benefits to the popular sectors, as seen in the stubbornness of poverty rates and inequality after some initial recovery (Londoño and Székely ). The rise in associationalism has also been underwritten by a change in the composition of the working classes, amplified by the new economic model. Proletarianization, the dominant social-structural trend among the working classes when the -Hub was founded, gave way to informalization. While the
02 Chapter 2.qxd
11/9/2009
11:48 AM
Page 55
,
formal sector in Latin America never produced employment at a rate equivalent to that of urban population growth; and while an informal sector grew under , informalization of the workforce accelerated under the new model. Widespread privatization of state firms, the restructuring of private firms, and state reform resulted in large layoffs of the formally employed. The relative growth of formal wage earners that marked the era has thereby been followed by their relative decline or stagnation. According to Portes and Hoffman (:), whereas between and the public sector, where unions were traditionally strong, had accounted for percent of total job growth, more recently public-sector employment has actually shrunk; and whereas large and medium firms had contributed an additional percent of total job growth, that sector’s share of employment creation has been reduced to percent.10 The informal economy has picked up the slack in employment generation. The relative decline in the formal workforce can be observed in Peru and particularly Venezuela, among the four countries considered in this study, with Chile and Argentina escaping from the general trend in the s. Tables . and . display the effects of economic tumult after the debt crisis and the reform period of economic adjustment in the four countries in this study. The very large informal sector, between a third and a half of the workforce, provided the social-structural basis for organizing around territorially based neighborhood organizations. Declines in real wages and the stubbornness of poverty levels provided further impetus for organization around “subsistence” issues as a way of coping in the face of ongoing hardship (Panfichi ; Levine and Romero ; C. Romero ). Even with the general improvement of the s, at the beginning of the new century important socioeconomic indicators were still behind their levels of the s. Chile, which had borne many of the costs of market adjustment much earlier, stood out for making economic and social progress, with a sharp decline in poverty and wage rates surpassing pre–debt crisis levels. Political factors have also contributed to a more supportive environment for associationalism. Democratic transitions unfolded in a way that often involved substantial participation and mobilization among the popular sectors in the antiauthoritarian, pro-democracy movements in the s and s. While this type of activism diminished after the transition, it was an important part of the early stages of the upsurge in associationalism. Under the (generally new) democratic regimes, state policy orientations have also stimulated the 10. Portes and Hoffman (:) show Argentina as an exception, but the data predate the crisis.
02 Chapter 2.qxd
11/9/2009
.
11:48 AM
Page 56
Table . Economic Change and the Popular Sectors, –⁄ Informal Sector (%)
Argentina Chile Peru Venezuela
Real Minimum Wages (1980 = 100)
Average Urban Wages (1982 = 100)
1990
2002
1990
2002
1990
2003
52.0a 37.9 52.7a 38.6
44.5 38.8b 56.2 51.8
40.2 73.3 21.4 55.2
64.2 135.2 33.1 75.9
99 96 29 46
89 141 32 25
Sources: Average Urban Wages calculated from (). Real Minimum Wages: (; ). Informal-sector data from () (on the surveys: Argentine survey conducted in urban areas; Chilean survey conducted at the national level; Peruvian survey conducted in the Lima metropolitan area; Venezuelan surveys conducted nationally). a
.
b
.
Table . Urban Poverty (percent of households in poverty) 1970 Argentina Chile Peru Venezuela
5 12 28 20b
1980 7 32 35 (1979) 18 (1981)
1990 20 33 52 (1991) 33
1999 16a 16a (2000) 31a 45c
Source: Unless indicated otherwise, data are from the dataset employed in de Janvry and Sadoulet (:), which relies on data (de Janvry and Sadoulet : ). a
, Anuario estadístico de América Latina y el Caribe ().
From Tabatabai :–. This source relies on data: though the sources for the individual data are not provided, is listed in the book as a “major source,” and the bibliographic references list numerous studies.
b
c Household data on urban poverty in Venezuela are not available beyond 1994. The estimate for the percent of urban households living in poverty in is calculated based on a . percent increase between and in individual urban poverty rates during those years, according to ’s Anuario estadístico de América Latina y el Caribe ().
rise of associationalism. In response to the dislocations of market adjustment, states have expanded social policies beyond the formal sector, while simultaneously engaging in a pattern of greater public-private partnership in social policy implementation. This new political-economic orientation, far different from the peripheral Fordism of the postwar era, offers substantial incentives for associations to form and to act as providers of services themselves. The new orientation toward social policy has also been accompanied by the decentralization of many policy areas. Local state officials, more proximate
02 Chapter 2.qxd
11/9/2009
11:48 AM
Page 57
,
targets for associational demands than their national-level counterparts, have more control over resources and policies than before (Nickson ; Hiskey and Seligson ; Montero and Samuels a). And, often to mobilize support, politicians at the municipal level across many countries have implemented policies explicitly intended to promote the growth of associationalism, helping found or fund associations, and developing state-sponsored institutional spaces such as mesas de concertación and policy councils in which associations can join together, make claims, and engage in other problem-solving activities (Chavez and Goldfrank ; Dagnino, Olvera, and Panfichi ). Finally, another stimulus for urban associationalism was a supportive post– cold war international environment characterized by a new global convergence on the promotion of civil society. The third wave of democracy has been characterized by the diffusion of norms that champion society-based rather than state-led approaches to many collective problems, and advocates on both the left and the right see civil society as a bulwark against authoritarianism and as a method for deepening democracy (Avritzer ; Goldfrank ). In this environment, new nongovernmental actors such as transnational social movements or advocacy groups (Keck and Sikkink ; Tarrow ) have emerged, while international institutions such as the United Nations, the World Bank, the International Monetary Fund, and the Inter-American Development Bank have integrated support for civil society as an important element of developmental initiatives. While it is impossible to obtain an accurate count of popular associations, much less an accurate accounting over time, there is nearly universal agreement among analysts that their numbers have swelled considerably over the last decades. Foweraker, Landman, and Harvey, for instance, refer to an “explosion in the sheer number of groups and organizations” (:). All four countries in the present analysis experienced this general trend, spurred by the same set of political and economic factors just discussed. In Argentina and Chile severe repression by the military regimes (ruling in – and –, respectively) prompted the formation of human rights organizations in which the popular sectors participated as key actors protesting in the antiauthoritarian movement (Moreno , Oxhorn a). In Peru mobilization against the military regime (–) also played a role, starting in , but in the pre period associational proliferation was also given an exceptional boost by an earlier strategy of the military regime, which sought to transfer political loyalty from to the state by sponsoring associationalism through a new organization, (Sistema Nacional de Apoyo a la Movilización Social). And although Venezuela did not experience military rule, organizing proceeded
02 Chapter 2.qxd
11/9/2009
.
11:48 AM
Page 58
with the goal of reforming political institutions, which had become ossified and unresponsive. In all four cases, economic conditions during the “lost decade” and the period of structural adjustment also provided impulses to associationalism, especially the proliferation of provisioning associations such as mothers’ committees and soup kitchens. Many of these associations were initially independent of the government, but they often became targets of politicians for mobilizing support. In Peru various types of collective kitchens were established: initially promoted by church organizations (with food donations from ) and other s, they were subsequently supported by successive mayors, presidents, and political parties (Barrig ). Government programs, such as milk distribution in Peru and Venezuela and, in the s, workfare in Argentina, both stimulated organizing by enlisting associations in implementation and were often successful in winning political support. For instance, Crisp and Levine (:) report that the Venezuelan vecinos (neighbors) movement, which began in the s with a strong antiparty orientation, experienced “a notable migration of leaders and organized groups into explicitly political vehicles.” The incidence and importance of associations in contemporary urban Latin America can be seen in the proportion of the urban popular sectors that participate in associations. Across the four countries, . percent of the popular sectors participate in just the subset of programmatic associations on which the survey solicited data. Furthermore, for about half of the participants (. percent of the total sample), participation is not merely occasional but takes place at the rate of at least once a week. In addition to a substantial level of participation in associations, it is noteworthy that the popular sectors subjectively consider them to be efficacious vehicles for solving problems and representing interests. As shown in figure ., in every country in the study associations are ranked more highly in this respect than the dominant actors of the -Hub—labor unions and political parties—as well as the legislature, the core elective institution of representation. This gap ranges from being fairly narrow in Chile to strikingly large in Peru, Argentina, and Venezuela. Given the substantial pessimism toward politics that has come to characterize much of the popular sectors in these countries, the absolute proportion of those who expressed confidence in associations as problem-solving vehicles is striking, with an especially high proportion in Peru. The lack of confidence in political parties is one aspect, or arguably one result, of a more general pattern in which parties have become more distant from civil society. Political parties have increasingly adopted plebiscitary, catchall, or clientelistic strategies and make appeals to atomized individuals on the
02 Chapter 2.qxd
11/9/2009
11:48 AM
Page 59
,
70 Argentina Chile Peru Venezuela
60 50 40 30 20 10 0 Popular Associations
Parties
Congress
Unions
Fig. .. Popular-Sector Confidence in Infrastructure of Representation (percent)
basis of the mass-marketing of candidates through the use of images, celebrity, and personality.11 Parties no longer seek organic, institutionalized relationships with popular organizations or, to the same degree, woo electoral support with concessions and benefits to or through these organizations. Reflecting new research on presidential elections in Chile and Peru, as well as in Brazil, Boas (:) reports, “Since the transitions to democracy . . . presidential candidates have increasingly adopted new campaign technologies and strategies that privilege direct connections with voters, and they have moved away from the use of party structures, interest organizations, and other intermediary institutions as ways to reach the electorate during campaigns.”12 This change in mobilization strategies is reflective of a larger shift, also seen in the advanced democracies, in which parties have come to serve less of a “representative” function and to operate less “in society,” and are more restricted to activity in the electoral arena and in government. An aspect of that transformation has been the weakening of ties with social organizations: not only have linkages with unions been challenged, as discussed above, but the linkages with associations have exhibited a pattern that is generally more distant, fluid, and instrumental. Although parties can be important allies and actors in the new interest regime, they do not play a role in the A-Net analogous to their role in the -Hub. 11. See Panebianco (), Kirchheimer (), Roberts (), Weyland (b), Castells () and Lawson and McCann (). 12. See also Novaro (), Hagopian (), Manz and Zuazo (), and Olavarría ().
02 Chapter 2.qxd
11/9/2009
.
11:48 AM
Page 60
Powerful evidence thus suggests that—propelled by economic and political change as well as shifts in the international climate—these once peripheral associations have multiplied and become central structures through which the popular sectors attempt to articulate interests and solve collective problems, forming the base units of a new popular interest regime, with a smaller role for parties. The next chapter undertakes a closer comparison between the A-Net and the -Hub, examining the logic of individual and organizational collective action in each as well as ties between popular organizations and the state.
03 Chapter 3.qxd
11/9/2009
11:49 AM
Page 61
3 , , : - - Ruth Berins Collier and Samuel Handlin
The shift in the popular interest regime from one centered on a union-party hub (-Hub) to one based in urban associational networks (A-Net) has entailed a new organizational infrastructure for popular-sector representation and problem solving. To analyze the dynamics of the new interest regime, it is useful to compare explicitly some important characteristics of the -Hub and the A-Net. This comparison is necessarily drawn in broad strokes, presenting a general picture that overlooks differences among national interest regimes. However, it provides an important context for subsequent chapters that carry out a more focused empirical examination of popular problem solving and the organizational infrastructure of the A-Net. The chapter presents two kinds of comparisons between the popular interest regimes. The first concerns patterns of collective action; the second, relations between states and popular organizations. Collective action is directly relevant to two of the major themes in the volume: the extent of individual participation in the interest regime and the nature of coordination among popular organizations. We develop a theoretical framework for contrasting the logic of collective action in the -Hub and A-Net, one that moves beyond previous conceptions in several respects. We start by making an analytic distinction between collective action among individuals in the formation of and participation in organizations (“individual collective action”) and that among organizations themselves (“organizational collective action”) in coordinating their activities. This distinction, while simple in the abstract, has not been fully
03 Chapter 3.qxd
11/9/2009
.
11:49 AM
Page 62
developed in other studies that have followed in the Olsonian tradition of elaborating a single logic of collective action. The chapter suggests that incentives and disincentives for the two kinds of collective action emerge both from characteristics of organizations themselves as well as from the economic orientation of the state. The analysis suggests that unions, which predominated in the -Hub, tend to be characterized by greater obstacles to individual collective action unless the state steps in to solve these problems, and indeed, during the period of the -Hub the state under did play this role in many ways. Associations, which predominate in the A-Net, do not face the same obstacles to individual collective action for organizational formation and participation. Hence, associations are less reliant on the state in this sense. “Spontaneous emanation” (Berger ) is at least a greater possibility. The relationship is reversed with regard to organizational collective action: coordination among organizational entities is likely to be more difficult in the A-Net than the Hub, and the state, in the two historical periods respectively, reinforces this difference, again providing incentives for organizational collective action among unions but not (or much less so) for associations. These logics of collective action have implications for two aggregate traits of the interest regime: scope and scaling, to which the chapter then turns. Scope refers to the extent to which the interest regime attracts the participation of the popular sectors; it is affected by the logic of individual collective action for organizational formation and participation. Scaling refers to overall trends in organizational coordination within the interest regime; it is affected by the ease of organizational collective action. An advantage of the A-Net is that it is potentially more inclusive; that is, while unions, as the core organizations of the -Hub, are restricted to formal workers, with others able to organize but consigned to a peripheral status, the A-Net does not marginalize associations of the larger popular sectors. In the specific sense of comparing unions in the -Hub to associations in the A-Net, the scope of the A-Net is potentially greater. Even if the state helps solve collective problems for unions, they still have the potential for encompassing a more restricted base of participants, whereas associations are organizational forms that are available to a much larger base, such as amas de casa and informal workers, who would be unable by definition to participate in unions in the -Hub. While empirical comparison is difficult, the data that will be presented below suggest that to date this potential advantage of associations in the A-Net compared to unions in the -Hub has not generally been realized. In terms of scaling, associations have been able to overcome some disincentives for organizational collective action, but the A-Net remains substantially less coordinated than the -Hub.
03 Chapter 3.qxd
11/9/2009
11:49 AM
Page 63
,
The second comparison between the two interest regimes involves the nature of relationships popular organizations have with the state. Unions in the Hub had extensive and complex relations with the state, which often penetrated and co-opted them. Associations in the A-Net are quite different. During the period of democratization, many associations participated in antiauthoritarian opposition to the state; and during times of economic crisis, particularly the s debt crisis, many associations engaged in provisioning or subsistence activities quite independent of the state. But subsequently,“ordinary times” have lent themselves to more routinized patterns of claim making and other forms of associational interaction with the state—as the deliverers of government programs, as participants in institutional fora established by the state, as recipients of state financing, and sometimes as partial creations of the state, in that officials played a role in the founding of the association. This pattern of stateassociation relations in the A-Net has occurred in the context of changes in the state, including, most prominently, shifts in the state’s approach to social policy delivery, political decentralization, and change in state support of associations. State relations are important for two aggregate traits of the interest regime: access and autonomy. Neither trait can be observed directly. Therefore, in the data analysis we examine the accessing strategies of popular organizations, which can be viewed in terms of the institutional channels for voice that popular organizations utilize for presenting claims to the state, the level of the state at which claims are made, the role of political parties as intermediaries, and the kinds of demands that flow through these channels. To explore autonomy, we examine relationships of direct dependence that organizations have with the state or political parties for resources or organizational advantage. Such relations of dependence may provide the incentives for organizations to behave in certain ways, thereby compromising their own autonomy. In this latter vein, autonomy is often seen as potentially jeopardized by the mechanisms that provide access, so that there may be a trade-off between access and autonomy. The examination in this chapter of accessing strategies, state dependence, and party-association relations in the A-Net anticipates the analysis in subsequent chapters, where these issues are addressed in greater detail.
Comparative Logics of Collective Action: Implications for Scope and Scaling The popular interest regime, as an infrastructure constituted by organizations, is shaped by both patterns of collective action among individuals in forming
03 Chapter 3.qxd
11/9/2009
.
11:49 AM
Page 64
and participating in these organizations and by patterns of collective action among the organizations in coordinating their activities.1 In considering differences between the -Hub and the A-Net, then, a natural place to start is a theoretical analysis of these two different “logics of collective action” (Olson ) in the two interest regimes. The present section develops a framework for this analysis, which has implications for scope and scaling, two important aggregate traits of the -Hub and the A-Net. These two traits affect the representational capacity of the interest regime. As Peruzzotti (b) argues, participation is not necessarily required for an organization to represent constituencies, particularly for those organizations that perform monitoring functions. Yet the extent of participation in aggregate is an important indicator of how the infrastructure functions as a vehicle for representation that articulates, aggregates, and otherwise pursues popular-sector interests. Scaling, or engaging in joint activity across organizations, is a powerful option for any set of interest organizations. For the popular sectors, as a group endowed more with numerical than with financial resources, coordination can be particularly important for leveraging strength—exerting influence, undertaking activities on a bigger scale, and reaching higher levels of government. A useful approach to theorizing these dimensions is to view scope and scaling as empirical outcomes of two distinct logics of collective action. The analytic approach developed here elaborates on Olson’s original formulation, which outlined a single and universal logic of collective action, and on subsequent research, which has focused on differentiating distinct collective action challenges faced by different societal groups, particularly capitalists versus the working class in industrial societies. While Olson focused on the size of groups and the problem of free riding, other scholars have stressed both the more onerous problems of collective action faced by labor compared to capital (Offe and Wiesenthal ) and the further asymmetry that the market mechanism provides a degree of coordination that mitigates the need for concerted or organized action among capitalists (Lindblom ). Analysts have since further disaggregated the capitalist side of this dichotomy, positing that differences in sector (Bowman ; Streeck ) and size of enterprise (Shadlen ) affect the logic of collective action. Less work has been done to disaggregate the other half of the dichotomy: the working classes. In pointing to the relative strengths or weaknesses of working-class groups, some analysts have differentiated them according to their structural and associational power (Wright :; see 1. The distinction between individual and organizational collective action, or collective action within and between organizations, is similar to that utilized by B. Schneider ().
03 Chapter 3.qxd
11/9/2009
11:49 AM
Page 65
,
also Silver :), derived from, for example, size of plant, isolated enclaves, competitive versus noncompetitive sectors, skill levels, and craft versus industrial organizations. However, these analyses, rooted primarily in the experiences of advanced industrial countries, are oriented toward formal wage earners and fail to capture an important aspect of Latin American countries with their heterogeneous mixture of informal and formal workers, as well as the unemployed and underemployed, who are more likely to pursue interests through popular associations than through union organizations. For these groups, the analyses of social movement theorists are often insightful. The present approach extends the literature on collective action in several ways. First, as mentioned, we distinguish individual collective action from organizational collective action. Second, the analysis focuses on two sets of factors that shape incentives for individual and organizational collective action— the traits of the predominant base organizational units of the interest regime and aspects of the state. In this section we ask how these other factors, organizational traits and the state, affect the incentives for participation in order to form organizations (unions and popular associations) and, once formed, to sustain participation in them and achieve coordination among them. Many organizational differences distinguish unions and popular associations. Although the upsurge of associationalism in Latin America has attracted great scholarly interest, the specific question of how these distinct organizational forms provide different incentives or disincentives for collective action compared with unions remains underexamined. For this question, the literature on social movements, in its emphasis on framing and mobilizational structures, provides a theoretical touchstone. The other set of factors pertain to the state. Whereas much of the Olsonian tradition is rather “inward-looking” in that it focuses on the characteristics of the group itself, the present approach focuses equally on factors “external” to the organization itself, especially on how the orientation of the state both directly and indirectly shapes incentives for collective action within and among organizations. The indirect effect of state factors bears similarities to arguments in the political process theory of social movements, although the emphasis in that literature is often on more conjunctural political factors that explain episodes or cycles of protest, rather than the more structural factors emphasized here, specifically the orientation of the state toward a particular model of political economy. In this regard, we distinguish the state from the neoliberal state. The analytic framework thus views the logics of both types of collective action as mechanisms by which organizational and state traits shape the scope and scale of the interest regime. Of the organizational traits, three are
03 Chapter 3.qxd
11/9/2009
.
11:49 AM
Page 66
highlighted: (a) resources, that is, constituency participation and finances, (b) interest convergence, and (c) the nature of demands. The state factors derive from the “type” of states, or the political economy of its basic policy orientation. Three dimensions of state policy are particularly relevant: (a) policy regarding the regulation of the base organizational unit, (b) policy regarding state macroeconomic intervention, and (c) social policy. It should be emphasized that these two sets of factors are interrelated, as the organizational factors are themselves influenced by the state. State action may affect the dominance within the overall interest regime of a particular type of organization as well as the three specific organizational traits under examination here. The analysis begins by looking at the logics of collective action that derive from organizational and state traits. The former suggests that unions, which predominated in the -Hub, tend to be characterized by greater obstacles regarding individual collective action; however, the state typically stepped in to solve or ameliorate some of these problems. Associations, which predominate in the A-Net, face less severe individual collective action problems and hence rely less on the state in this sense. With regard to organizational collective action, the problems are reversed: coordination among organizations is likely to be more difficult for associations than for unions. Organizational Traits and Collective Action Labor unions and popular associations, the two predominant base organizations of the -Hub and the A-Net respectively, are quite distinct types of organizations, often with contrasting implications for collective action. The following analysis examines differences between unions and associations in terms of resources, interest convergence, and the nature of the demands typically made. These are traits largely “inherent” to unions and popular associations. In this discussion, it must be remembered that, as we will see in the next section, the state may step in to “solve” collective action problems so that some of the problems described here, particularly those of unions, are often overcome. Resources. Labor unions and popular associations differ in terms of the resources at their command, particularly personnel and material resources, or the type of constituency participation and the type of financing. For unions, resource requirements are very high if they are to have any capacity. Because collective bargaining is their basic activity and their ultimate weapon has historically been the strike—or credible strike threat—they need the constituency participation of a significant number or percentage of relevant
03 Chapter 3.qxd
11/9/2009
11:49 AM
Page 67
,
Table . Organizational Traits and Collective Action Organizational Traits Resources
Unions in -Hub
Associations in A-Net
Personnel: High-cost formal membership
Personnel: Lowcost flexible participation
Material: Significant and self-funded; stable, reliable level
Material: Varies, but rarely both significant and self-funded; uncertain level
Interest Convergence
Low among competing workers but high among union members, high among unions
Demands
Recurrent, often nondisaggregable
Effect on Collective Action Individual Organizational Unions: Assoc: +
Unions: + Assoc: -
High among self-selecting participants, low among associations
Unions: Assoc: +
Unions: + Assoc: -
Often one-shot and disaggregable
Unions: +/Assoc: +/-
Unions: + Assoc: -
workers and, if possible, material resources for a strike fund. Hence the model of constituency participation of unions is one of formal membership, usually drawn from a delimited group of individuals who are employed in similar circumstances. Membership carries costs, which are high, such as dues and the need to conform to organizational choices, like the decision to strike and forego income. Unions thus experience serious obstacles to individual collective action. Although these barriers may be somewhat mitigated by the low cost of participation in terms of time and by the side payments of member services, like health care, pensions, and vacation resorts, the obstacles remain high: unions are difficult to form and, as Olson (:ff.) suggested, unions must often sustain participation through some type of “coercion,” such as the union shop or, as we shall see in the next section, state regulation. Once formed, however, the formal membership model “locks in” both participation and material resources. The stability and ample funding of the resource model help unions coordinate. With formal membership, leaders are able to commit to cross-organizational collective action and pursue long-term strategic goals that may impose short-term costs without fear of defection. Union dues provide predictable funding for day-to-day operations, professional staff, and long-term budgetary and strategic planning. Popular associations rarely have formal members and exhibit considerably
03 Chapter 3.qxd
11/9/2009
.
11:49 AM
Page 68
more variation than unions in the nature of constituency participation. Although exceptions exist, such as some associations of street vendors (Roever ), in most cases it is probably a misnomer to speak of membership at all. Rather, these associations have participants, and the costs to participation are low. Dues are rarely required, and participation is generally voluntary and intermittent and can be tailored to fit one’s changing level of enthusiasm. Further, many associations, particularly s providing services, rely on professional staff and only minimally need or even seek participants. In contrast to unions, then, the more flexible models of participation that characterize most popular associations require lower levels of commitment and tend to facilitate associational formation. On the other hand, associations often have more trouble sustaining individual participation, because, unlike unions, which have substantial barriers to exit (leaving a union not only entails forgoing these benefits, but is also often costly and disruptive, generally requiring a change of jobs) associations are characterized by few barriers to exit. However, because some associations can easily replenish their participants and others have little reliance on them, on balance the associational resource model provides incentives for individual collective action. With respect to collective action among associations, the resource models of associations may, in some ways, provide incentives for coordination. Associations, which are generally resource-poor, sometimes rely for funding or training on other associations that have the raison d’être of coordination and/or receive money from the state or international donors that is explicitly earmarked for such activities. At the same time, popular associations that rely on constituent participation tend to find coordination more difficult precisely because coordination across associations heightens the Olsonian free rider problem of larger numbers and because participants are presented with no barriers to exit. In sum, the union resource model, based on formal membership with dues, makes individual collective action initially problematic for organizational formation, but once unions are formed, it facilitates organizational collective action. For associations, the low resource requirements and low cost resource model is an incentive for individual collective action for organizational formation. At the same time, however, associations have a resource model of substantial uncertainty, which makes individual collective action for ongoing participation and organizational collective action or coordination more difficult. Interest Convergence. A second organizational factor is interest convergence or heterogeneity among individuals and organizations. As is widely recognized, interests are constructed; however, the burden of this task is not everywhere
03 Chapter 3.qxd
11/9/2009
11:49 AM
Page 69
,
the same. This factor makes formation easier for associations than for unions and coordination more difficult. In terms of individual collective action for organizational formation, unions confront a notable difficulty regarding a lack of interest convergence, based on a particularly acute individual coordination problem. Though workers converge on a desire for better wages and working conditions, they directly compete with one another in a flexible, nonunionized labor market, driving wages down and making interest heterogeneity, indeed competition, endemic to proletarian status. The importance of framing and collective identity, like class consciousness, in overcoming this problem for collective action has long been recognized, but the nature of the task is usefully formulated by Offe and Weisenthal, who suggest that forging a collective identity involves a somewhat paradoxical dynamic in which “interests can only be met to the extent that they are partly redefined” (:). Associations, on the other hand, do not face the same kind of competition among would-be participants. They more typically attract participants for projects, issues, or policy goals that are more specific and delimited—for example, neighbors may participate in associations for specific infrastructural projects, or vendors may cooperate to secure space on the streets when its use is threatened. They thus have the advantage that participants self-select, come together around specific common interests, and have less to overcome in forging a convergence of interests for initial formation. The comparison is reversed for organizational collective action: interest convergence is likely to be greater across unions than across associations, making coordination more difficult for associations. The union movement focuses on a limited number of productionist issues, related to wages and work conditions. Even when union interests differ by sector, the most important sectoral unions are generally large and cohesive enough to scale and often to win bargaining rights at the national level. In addition, as we shall see, under , many workers in key sectors had a commonality in the state as an employer and as a target of grievances. Furthermore, the interests of those in the private and public sectors converged on many issues of national policy, such as those that protect jobs, set minimum wages as a benchmark, and regulate individual contracts and collective rights, as well as macroeconomic regulation. Even across sectors, then, enough commonality of interest existed to provide an incentive to establish vertical structures for representing common interests at the national level. Furthermore, the affiliation of unions to s, through institutionalized organizational links, interlocking leadership, or a history of collaboration, has played some role in constructing a convergence of interests. Not only did party
03 Chapter 3.qxd
11/9/2009
.
11:49 AM
Page 70
linkage impart a common party identification, but in addition, political parties could be brokers that helped unions coordinate on specific programmatic priorities or concessions. By comparison, associations in the aggregate cover more heterogeneous interests. The work situations of participants in associations vary widely. While most are in the informal sector, others are formal workers, unemployed, or out of the labor force altogether. Further, even informal workers are a diverse category, sometimes explicitly defined as including wage earners in small or informal enterprises, microentrepreneurs who hire them, domestic workers, and the self-employed. Given this diversity, most associations do not take up productionist demands at all. They focus on a great variety of other, usually consumptionist issues, including most prominently social services, poverty, and neighborhood and local development. With such a great diversity of issues being addressed and interests being advanced, it may be more difficult for associations to coordinate. Of course, interest heterogeneity is not as great a problem for the subset of associations that focus on the same issue. However, as we will see below, a third organizational trait may mean that coordinating around even the same issue may be difficult. There is, of course, some asymmetry in comparing only unions in the Hub to a diverse set of popular associations in the A-Net. However, the point is that in the -Hub, working-class interests were constructed in a homogeneous way and framed around at least a minimal level of common ideological orientation and class identity and around a limited set of productionist issues. By contrast, the A-Net does much less to construct a commonality of interest and identity or to prioritize a subset of issues: its “pluralism” is almost necessarily less homogenizing. Nor do associations have a common partisan affiliation that helps construct common identities or convergent programmatic priorities across these separate issues or that subsumes them under a larger whole. As mentioned above, it has long been suggested that collective identity can help overcome the problem of interest heterogeneity, and the social movements literature similarly has emphasized the importance of common frames for collective action. Some observers have seen a common discourse of rights and grassroots participatory democracy, but an overarching identity may not be readily available when interests are as fragmented as they are in the associational world. Demands. A final factor shaping the logic of collective action is the nature of demands central to each type of organization. Two aspects of demands affect the logic of collective action. The extent to which demands are “deliverable” can affect incentives for sustaining participation and for organizational
03 Chapter 3.qxd
11/9/2009
11:49 AM
Page 71
,
coordination. The degree to which demands can be “disaggregated” can shape incentives for both the formation of organizations and, more directly, coordinating across associations. “Deliverability” is in part affected by the nature of demands, as envisioned on a spectrum running from discrete, one-shot demands to those that are ongoing, and in some cases essentially unbounded, in that they cannot be satisfied at some determined point. It is useful to contrast associational and union demands with those of the new social movements (), which have been important cases for theorizing collective action, and typically have demands that represent one end of the “deliverability” spectrum: transformative, virtually unbounded, postmaterial demands that cannot be “satisfied.” Though s also make more specific deliverable demands, dominating their larger agenda is usually a much broader demand that is ultimately unfeasible within any reasonable time frame. Even a responsive government can, for example, only partially “satisfy” demands for racial equality, fulfill an environmental agenda, or even save the forests. This “insatiability” of demands can be a powerful inducement for continued organization and mobilization; indeed, responsiveness on the part of the government may be as likely to energize as demobilize such movements and sustain them over time. Such transformative and long-term orientations also provide a common, salient goal that subsumes the many immediate, concrete demands made by individual associations, thereby uniting the larger network and facilitating coordination among organizations. A quite different dynamic is more likely to hold both for unions and especially for popular associations, whose demands are more immediate and discrete rather than unbounded. Along a continuum of this dimension, the demands of unions and some popular associations may be seen as intermediate, in the sense that although they tend to be discrete and can be satisfied in a relatively short time frame, they are not one-shot but rather, they recur. For unions, wages as well as many national policies, such as many aspects of macroeconomic policy that unions may be involved in negotiating, are set only for the short term, and then they must be renegotiated. The recurrent nature of these demands gives incentives for sustaining participation and facilitates organizational coordination, as stable coordinating structures are needed for ongoing claim making and concertation. Similarly, associations that administer programs, and those that may be concerned with ongoing issues such as policing, health, or education, may, like unions, have recurring demands. On the opposite end of the spectrum from the s, many associations engage in activities around discrete and immediate demands or problems that
03 Chapter 3.qxd
11/9/2009
.
11:49 AM
Page 72
do not recur, such as urban services, infrastructure investments, or land titles. Once such demands are fulfilled, the activities come to an end. Evidence suggests that having these demands fulfilled may be more likely to demobilize than energize associations (Dosh ). In parallel fashion, some associations engaged in society-targeted rather than state-targeted activities may also address problems that are one-shot in nature, in which participants organize to achieve a particular goal, which can be accomplished and fulfilled. To the extent they engage in one-shot, deliverable demands, organizations may have difficulty sustaining participation, as participants are more likely to exit once an immediate need is fulfilled. Moreover, the one-shot nature of many associational demands makes organizational coordination more difficult, as it is less likely that a critical mass of organizations will be willing to make a commitment to long-term coordination. The second aspect of demands is whether they are directed toward policies that are easy or difficult to disaggregate in terms of constituencies, or the degree to which demands are for targetable or excludable goods. Lowi () insightfully argued that policies that lend themselves to disaggregation display different patterns of group contestation and are played out in different “political arenas,” or through different policy processes, compared to those not easily disaggregated. Distinguishing distributive, redistributive, and regulatory policies, Lowi made the fundamental observation that different types of actors are the “primary political units” in each policy arena—peak associations weigh in on redistributive issues, coalitions of more discrete interest groups tend to contest regulatory policy, and individuals, firms, or small interest groups operate in the distributive arena of disaggregable policy. This correspondence between interest groups and policy type helps us understand the differing logic of collective action of unions and popular associations. While unions and associations may have multifaceted agendas and make a variety of demands that differ in their ease of disaggregation, it is possible to draw some general distinctions and posit some basic implications. Many union demands concern the workplace, and much union activism occurs at the plant level. Nevertheless, a powerful logic operates for unions to present demands at the level of the entire economic sector so that competing firms won’t be disadvantaged by labor concessions. Further, as we will see, particularly under union interests often converged on highly “aggregated” demands at the level of national economic, social, and regulatory policy, including, of course, labor market regulation and legal provisions that regulate unions and collective rights. To the extent unions embrace such aggregated demands or those that involve a larger number of beneficiaries, the free rider problem is greater, and individuals
03 Chapter 3.qxd
11/9/2009
11:49 AM
Page 73
,
may be less willing to pay the cost of participation in collective action. However, organizational incentives for coordination across unions increase when demands cannot be disaggregated. Many popular associations present a different picture. As the above examples of demands for urban services, infrastructure investments, and land titles indicate, popular associations are more likely to focus on distributive demands that can be disaggregated, in the sense that a government response can be targeted to one association and withheld from another. The disaggregability provides an incentive for individual collective action, since success is more likely when the response is cheap (as it is when targeted), and would-be participants are more likely to join an effort that promises to reap quick rewards. However, disaggregability has negative implications for associational coordination. The piecemeal and discretionary response to these demands—a sidewalk here, a health clinic there—is conducive to clientelism and co-optation, and it discourages the establishment of ongoing, institutionalized relations of cooperation among associations. While some analysts see the new associationalism as a bulwark against clientelism, the opposite may be equally likely: fundamental characteristics of associations, their desire for state access and resources, and the nature of their demands make them vulnerable to clientelistic relations. Some associations may, of course, have a larger—and national—political agenda, for example, health and education policies, active labor market policies, poverty relief, and tax and redistributive policies. However, many associations focus on disaggregable demands, and the fact that core demands can be satisfied individually lowers the incentives for cooperation or coordination. Precisely because it is cheaper to dole out targeted benefits and hence disaggregated demands are more easily fulfilled, associations may have an incentive to demand a particular subsidy or distribution for just one neighborhood, rather than to present the demand as a politically more difficult form of “entitlement” for all similar neighborhoods. In sum, for both unions and associations the nature of demands produces countervailing incentives for individual collective action. The recurrent nature of union demands provides an incentive for collective action, but their aggregated nature increases incentives for free riding. The disaggregated nature of associational demands facilitates formation, but their nonrecurrent nature makes sustaining participation more difficult. With respect to collective action between organizations, however, the effect of demands is less ambiguous: they provide incentives for coordination among unions, while discouraging coordination among associations.
03 Chapter 3.qxd
11/9/2009
.
11:49 AM
Page 74
State Characteristics and Collective Action The incentives and disincentives for collective action that derive from state characteristics illuminate key contrasts between the -Hub and A-Net. The two interest regimes developed in distinct historical periods, in the context of very different types of states. The -Hub generally emerged in tandem with an interventionist state fostering industrialization through policies of import substitution, while the A-Net has emerged alongside a leaner neoliberal state that has redefined appropriate state action, retreating from many policy areas and reformulating its intervention in others. The political economy of the and neoliberal states differ regarding three major policy areas that are central in terms of their effects on collective action: state regulation of popular organizations, macroeconomic policy, and social policy. In considering the impact of the state, then, the following discussion will refer specifically to the state of the -Hub and the neoliberal state of A-Net. Regulatory Policy. The first contrast concerns state regulation of the base organizations of the interest regime. The and neoliberal states took very different approaches to regulating, respectively, unions and popular associations. In the current period, neoliberal states have adopted a “pluralist” approach to associations, typically marked only by minimal and voluntary registration requirements, in contrast to the period, when Latin American states extensively regulated unions in a pattern widely analyzed as state corporatism, in which the state legally recognized strike and bargaining rights—necessary for basic union functioning—in exchange for extensive regulation (Schmitter , ; Erickson ; Stepan ; Collier and Collier ). Regarding collective action among individuals, state regulation of unions had somewhat contradictory effects regarding incentives for union formation. Table . State Traits and Collective Action State Traits
State/ -Hub
Neoliberal State/A-Net
Effect on Collective Action Individual
Organizational
Regulations of Popular Organizations
High (unions)
Low (associations)
Unions: + Assoc: 0
Unions: + Assoc: 0
Economic Policy
Interventionist
Marketoriented
Unions: + Assoc: 0
Unions: + Assoc: 0
Social Policy
Employmentbased, centralized
Broader, more decentralized
Unions: + Assoc: +
Unions: + Assoc: +
03 Chapter 3.qxd
11/9/2009
11:49 AM
Page 75
,
Regulations regarding registration, legal recognition, and minimum membership requirements were an impediment to union formation, and, more importantly, monopoly representation presented a barrier to entry that made it virtually impossible to form new unions among an already unionized workforce. At the same time, however, regulatory provisions often took a form that amounted to compulsory membership and had the effect of virtually mandating the formation of and workers’ participation in unions in specific sectors. Taken together, these contradictory effects often created situations in which regulation induced or allowed the formation of a group of unions that were legally recognized and hence legitimized as political actors, but blocked subsequent union formation, especially of dissident or alternative unions competing for the same members. The system of corporatist regulation had unambiguously positive effects on the ability of extant unions to sustain participation. Versions of compulsory membership or closed shop were clearly the most important provision, but various types of incentives also existed. Legal recognition was accompanied by formal participation in an institutionalized system of industrial relations that produced benefits for union members. Further, the monopoly of representation enjoyed by privileged unions sharply reduced any incentives members had for defection, since the benefits of union membership were available through a single source. All in all, state regulation did much to solve the collective action problem of competition among workers, and effectively mandated that people join unions—but not a multiplicity of unions in the same firm or sector. With regard to organizational collective action, regulation often gave strong inducements for coordination among unions, and federations and confederations were formed virtually everywhere and became especially powerful when the state allowed them to bargain collectively. The regulation of this right varied across countries. Chile, for instance, historically allowed only plant-level bargaining except during the Allende period. More typically, however, Latin American countries granted official recognition and bargaining rights to federations at the national level. The state thus provided an incentive for local unions to coordinate and to form peak labor federations and confederations. In comparison, the neoliberal state does little to regulate associations. Registration with the state is fairly common among popular associations but often voluntary. It may be required for associations participating in or serving in the implementation of government social programs, as may additional forms of regulation. In general, however, nonregistration does not affect the basic functioning of associations in a manner parallel to that of unions. Overall, and in comparison with unions, the associational world seems to be relatively
03 Chapter 3.qxd
11/9/2009
.
11:49 AM
Page 76
free of regulation, characterized by a pluralist environment that neither encourages nor impedes individual or organizational collective action. Economic Policy. A second state characteristic concerns the approach to economic intervention and macroeconomic policy. The state was interventionist. It was a major employer both in state firms and in an often bloated bureaucracy, and it undertook responsibility for an extensive range of economic policies, including macroeconomic regulation, an array of subsidies, and wage policy. Since these were policies in which workers had a direct interest, the state itself, and not only private firms, became a target for union claim making, both as an employer and as a policy maker at the national level. The involvement of the state in these roles offered incentives for individual collective action, in both union formation and ongoing participation. Furthermore, as discussed in chapter , the state paralleled the Keynesian welfare state in the advanced industrial countries in that it could make room for a certain degree of class compromise, supporting unionization in many sectors and encouraging a kind of organized capitalism with peak confederations and, in some countries, their participation in negotiations on aspects of economic policy. The economic role of the state and the privileging of union members among the broader popular sectors provided incentives for individuals to remain union members, even if unions intermediated state-society relations in contradictory ways that compromised worker interests. Statist economic policy also offered substantial incentives for organizational collective action, specifically for unions to coordinate in a way that would allow them to scale to the national level and make demands regarding national policy. Under the neoliberal economic model, which is oriented to market mechanisms and deregulation, Latin American states have privatized state firms and retreated from many areas of economic policy intervention. State withdrawal from policy areas—through the privatization of state firms, the elimination of subsidies, or the free rein of the market in areas such as exchange rate policy— can depoliticize these policies, removing them from the arena of political contestation and shifting them to the arena of market competition and provision. Kurtz (b) has demonstrated how the change of economic model (such as the substitution of a land market for state-sponsored land reform) has depoliticized and disorganized the rural popular sectors in this fashion. Further, the neoliberal state relies little on the cooperation or participation of popular associations regarding economic policy, parallel to the way it sometimes used to rely on unions in controlling inflation.2 Economic policy thus does not provide the 2. In some cases, such as Argentina or Mexico, such cooperation has survived or been re-established in the context of the new economic model. See Etchemendy and Collier ().
03 Chapter 3.qxd
11/9/2009
11:49 AM
Page 77
,
incentives for individual or organizational collective action among associations that the state provided for unions. Social Policy. The third state dimension for analysis is social policy. While the and neoliberal states have substantially different orientations toward social provision, both provide some incentives for collective action for popular organizations. Social policy under was primarily employment based and largely restricted to formal workers or to unionized workers, leading to a pattern of truncated welfare states (Garretón ; Mesa-Lago ; Raczynski ). To the extent that unionization helped deliver benefits such as pensions and health care that other workers had more difficulty attaining, this approach to social policy gave strong incentives for individual collective action. The exclusionary character of welfare provision also offered substantial barriers to—or disincentives for—exit, as workers might risk losing important benefits if they left their unions. Finally, the centralized nature of social policy making offered further incentives for organizational collective action or coordination in order to make claims concerning health care, pension systems, and other policies that were almost exclusively handled at the national level under . The neoliberal state has remained active in social policy provision but has been characterized by a different orientation toward making and implementing policy. Many governments have expanded social services to previously excluded sectors and have often decentralized their implementation. Social policy has been especially affected by the larger trend toward decentralization in the region, as governments have often moved to devolve implementation, programmatic initiatives, and budgetary control of various policies to lower levels. Many proponents have anticipated that the result would be greater responsiveness to grassroots demands. Greater responsiveness has also been advanced by a variety of experiments with “deepening democracy,” although their design, implementation, and effectiveness have varied substantially.3 Further, in bringing service delivery to the local level, the neoliberal state has increasingly embraced a partnership model for implementing these policies, based not only on the state bureaucracy but also on private associations. This new orientation of social policy provides a strong impetus to individual collective action in order to form associations that can become service providers and recipients of government programs. The partnership model of implementation, in which associations may become de facto arms of a “privatized” state, is also an incentive for sustaining individual participation in associations, 3. See Roberts (), Abers (), Dagnino, Olvera, and Panfichi (), and Goldfrank ().
03 Chapter 3.qxd
11/9/2009
.
11:49 AM
Page 78
especially for those who draw salaries to run and maintain these programs or whose access to benefits may depend on participation, as is often the case for communal kitchens. In terms of organizational collective action, the decentralization of social policy provision may not provide the same incentives for scaling in the form of national peak organizations as did the state, but may provide substantial incentives for inter-associational coordination on a more local or metropolitan level. Scope and Scaling in the -Hub and the A-Net The analysis above considered several factors, each of which suggested implications for the scope and scaling of the interest regime. How does the theoretical framework just presented bear out empirically? Since the factors did not all point in the same direction, it is not possible to weight and aggregate them into a straightforward prediction. This problem is particularly vexing for scope. The greater ease of individual collective action in associations would suggest greater density in the A-Net; however, the more active role of the state in solving the individual collective action problems of unions may attenuate the potential scope advantage of the A-Net. Thus, despite the fact that associations are potentially open to a greater portion of the popular sectors than just formal workers and may have greater incentives for individual collective action, we have no strong basis for a “prediction” of how scope might differ in the two interest regimes. The outcome might also be strongly shaped by other factors in effect “held constant” in the above discussion, such as the level of industrialization and size of the formal sector. We might expect a greater and more consistent empirical difference with respect to scaling, as the different logics of organizational collective action in the -Hub and the A-Net are less ambivalent—both state and organizational factors seemed to be consistently more supportive of coordination in the former. For these reasons, in addition to the limitation of data that can be marshaled, “testing” the distinct logics of collective action, as suggested above, is not appropriate or feasible. Nevertheless, it may be instructive to examine available data and compare the scope and scale of the -Hub and the A-Net across the four countries examined in the present study. We start with a comparison between the -Hub and A-Net regarding the scope encompassed by the predominant base units, unions and associations, respectively (see table .). One must obviously exercise caution in interpreting these data, in light of a number of issues concerning measurement and comparability. Although these data reflect only the predominant organizational
03 Chapter 3.qxd
11/9/2009
11:49 AM
Page 79
,
Table . Density of Organizational Participation Across Interest Regimes
Union Density in Peak Years Union Density in 1995 Associational Density (programmatic) Associational Density (“political”)
Argentina
Chile
Peru
Venezuela
50.1 (1975) 22.3
25.0 (1976/7) 5.7
26.4 (1988)a 13.5
22.2
35.0 (1973) 12.7 (8.4 in 1985)b 25.2
55.9
17.8
34.2
40.6
68.4
25.9
Sources: Union density from Roberts database and Roberts (forthcoming). Associational density from database. Note: Union density is calculated as the percent of the total labor force that is unionized. Associational density is based on the percent of a random sample of popular-sector adults living in the capital city. Of the two scores of associational density, the first reports those belonging to at least one association specifically dedicated to problem solving; the second indicator also includes those reporting participation in sports, cultural, recreational, and religious organizations if respondents report that this association undertakes political, that is, problem-solving, activities. It should be noted that the data here differs slightly from that presented in figure ., in which the data is not restricted to the popular sectors. a Because of the late adoption of economic reforms in Venezuela, union density peaked later than elsewhere. b Because of the early economic reform in Chile, union density declined earlier than elsewhere.
form of each interest regime, associations of course existed during the -Hub, and unions continue to be important in the A-Net. Further, union density is calculated as a proportion of the total national labor force, whereas associational participation is calculated as a proportion of the popular sectors in the capital city. For these and other reasons, the measures are not directly comparable.4 Nevertheless, some general points can be made. Even though unions, unlike popular associations, have the potential of reaching only a restricted subgroup of formal workers, union density in the -Hub relative to the total workforce (including agricultural workers) was often substantial, aided by the role of the state in solving the collective action problem of workers in forming and participating in unions. Yet associationalism has 4. The populations assessed by the two statistics differ in three ways. As a proportion of the labor force, union density is calculated at the national level, irrespective of the social class of workers (unions include some white-collar workers who are probably middle class), and only from the economically active population. In contrast, our survey-based approach to measuring associational density is calculated only in the capital city, only from the popular sectors, and without regard to whether individuals are economically active. It should also be noted that, in terms of associational participation, the survey asked about a wide range of associations, but the list is not exhaustive.
03 Chapter 3.qxd
11/9/2009
.
11:49 AM
Page 80
also become widespread. Despite the shortcomings in comparing measures, it seems likely that in Peru associational density has outstripped the peak union density of the -Hub, and only in Argentina does it seem clear that associational density might be lagging behind peak union density. The data also demonstrate the degree to which union density declined from the time of the state and the -Hub to the time of the neoliberal state and the A-Net, and, with the exception of Argentina, the degree to which participation in associations appears to outpace union density in the contemporary period. A sharper contrast exists between the -Hub and the A-Net regarding scaling. Unions under the -Hub developed a high degree of cooperation over time, both in terms of coordination among unions in activities and integration into formal coordinating structures. As labor movements became institutionalized, base-level unions were frequently integrated into regional, sectoral, or national labor federations and confederations, which then represented or intermediated their interests in both the private arena of collective bargaining and in interactions with state and party officials. This pattern of hierarchical integration was generally formalized—federations had explicit roles and powers in the representation of local unions, and some type of formal election or delegation from the base organizations to the higher bodies generally provided at least quasi-legitimation. In addition to providing a structure for coordinating and representing base-level unions in making claims through institutional channels, the federations also played a role in organizing coordinated action for protest. Although no data offer a precise measure, the vast majority of unions were integrated into these structures. Our theoretical framework predicts that coordination would be significantly lower in the A-Net, with associations more likely to have few ties and also less likely to have linkages to coordinating bodies. While there are no measures available for explicitly comparing across interest regimes in this respect, this prediction does seem to be generally born out in the data, even with respect to our snowball sample of associations, which, as discussed in chapter , captures a more linked and interconnected “core segment” of the A-Net. Table ., which presents data for the pooled sample, shows that while portions of the A-Net are characterized by relatively dense inter-associational linkage and an active role for coordinators, many other associations, even within a chainreferral sample, remain relatively isolated and without regular activation of linkages to coordinating bodies. This pattern contrasts with that of the Hub, in which unions very rarely operated in complete isolation and were nearly all affiliated with some kind of federation or confederation. Not shown
03 Chapter 3.qxd
11/9/2009
11:49 AM
Page 81
,
Table .. Coordination Among Popular Associations Inter-Associational Linkages (number of linkages) 0 1–4 5–10 11–30 31+ Linked to Coordinating Association
Percent of Associations 19.4 28.9 18.4 19.2 14.1 36.8
in this data, but clearly evident, is also a marked difference in the type of coordinating structures characterizing each interest regime.5 Associations thus exhibit a pattern of, at best, loose coordination in fluid networks, with “federations” or coordinating organizations playing some role. This pattern accords with numerous other findings in the literature on civil society and social movements in Latin America (Chalmers, Martin, and Piester ; Alvarez, Dagnino, and Escobar ; Alvarez ) and cautions against an overstatement of extreme atomization. The implications of this pattern of coordination for the operation of the A-Net as a structure of representation are double-edged. As many analysts have noted, networked patterns of coordination may be less prone to Michelsian oligarchy and threats to democratic accountability: to the subordination of base-level interests and to co-optation of organizational leaders by political elites. However, a pattern of loose, networked coordination may make it more difficult for associations to leverage their strength in national-level claim making and may make associational networks less attractive allies for political parties, although, as will be suggested in chapter , electoral support is often sought among associations.
State Linkages: Implications for Access and Autonomy In addition to providing different logics of collective action, a second contrast between the -Hub and the A-Net regards linkages between the state and popular organizations. These affect the access popular organizations may have 5. For analysis of country variation, further discussion of differences in coordinating associations between the two interest regimes, and a much more detailed investigation of scaling in the A-Net more generally, see chapter .
03 Chapter 3.qxd
11/9/2009
.
11:49 AM
Page 82
to the state as well as their autonomy. Relations with the state are thus fundamental to the way the interest regime is inserted into politics and operates as a structure of representation. As noted in chapter , accessing strategies are so fundamental that many analysts define political activities in these terms. However, as further noted, not all problem solving by popular organizations is state targeted—bargaining with employers is an integral part of the activities of labor unions, while many society-targeted activities are widespread among associations. Yet the state remains the locus of policy-making authority and the target of contestation, so that the accessing capacity of an interest regime is a fundamental trait. Relations with the state also may compromise autonomy, a dynamic that is particularly important given that many elite actors use state resources to shape and control popular organizations or mobilize political support. We can think of two general kinds of relationships that present threats to autonomy. Associations may enter into relationships of dependence on the state, relying on state funding, implementing state programs, or even being partially founded by state actors. More generally, threats to autonomy can emerge from the pursuit of accessing strategies themselves. Access and autonomy are often in tension with each other—autonomy may wane as access increases. Popular organizations may thus face a tradeoff between access and autonomy. Relations that offer access for claim making and opportunities to participate in government programs may provide inducements or incentives that can lead autonomy to be compromised, as when associations modify demands to maintain favor with the state or are induced to become partisan actors. Given this tradeoff, popular organizations that value autonomy may be reluctant to pay the cost of entering into relationships with state and party actors, while others may opt for the access these relationships afford. The relationship to representation is therefore particularly complex. Access The classical form of interest activity is claim making directed toward government officials, and for many popular associations, institutionalized channels of access are central for pressing claims. Thus, the degree to which the interest regime as a whole provides access to officials, and at what level of government, is an important aggregate property of the interest regime. Of course, access to policy makers is not always desired by popular organizations or necessary for them to be effective representative structures. Historically, anarcho-syndicalist unions were unwilling to enter into any relationship with the state (although
03 Chapter 3.qxd
11/9/2009
11:49 AM
Page 83
,
that strategic posture changed once socialist and communist tendencies became dominant within the labor movement). Some popular associations may emphasize society-targeted strategies or may press claims through protest, without seeking institutional channels of access. For many associations, however, accessing strategies are central for representing their constituencies. Some associations target claims primarily toward local-level officials, who are responsible for the decisions and distributions of most relevance, particularly in the context of administrative decentralization; for some claims, however, access at the national level is vital. Therefore, variation in the degree to which a popular interest regime has channels of access at the national, as well as local, level is an important dimension for comparison. The absence of popular-sector organizations from policy-making arenas would leave these constituencies without voice in policy areas central to their well-being. As noted in chapter , it is important to emphasize the distinction between accessing and influence. The latter is notoriously difficult to assess and has posed problems that have often plagued analysts of the interest arena (Baumgartner and Leech :). Accessing refers not to influence but to patterns of contact between popular organizations and elected or appointed decision makers within the state either directly or indirectly through political parties. These patterns can be intermittent or more regularized, and they can take place through channels that may be either formal or informal. Variation in access between the -Hub and the A-Net is influenced by the orientation of the state. With the contrast between the and the neoliberal states, the -Hub afforded more access to unions, particularly at the national level, than the A-Net does to associations. The state established formal systems of industrial relations, involving both a “private” (but highly regulated) space of policy making of primary importance to workers (collective bargaining) and spaces in state institutions, such as conciliation and arbitration boards and labor courts, in accordance with an economic model that could accommodate a certain degree of class compromise at both the micro and macro levels of the firm and national policy respectively. In addition, it often engaged in forms of concertation and negotiation with labor confederations over aspects of national economic policy, though these arrangements were typically informal, ad hoc, and quite asymmetric. As a result, unions in the -Hub had institutionalized channels of access, although the institutional configurations that were implemented varied across cases. The logic of the neoliberal state has provided a different pattern of access under the A-Net. With the withdrawal of the state from many forms of economic intervention and a move toward more insulated, technocratic policy
03 Chapter 3.qxd
11/9/2009
.
11:49 AM
Page 84
making, access regarding macroeconomic policy has been largely closed off. However, other political changes have left the neoliberal state more open to access regarding social policy and other issues. Fiscal and administrative decentralization has brought the locus of policy making and implementation regarding some issues closer to societal actors (Oxhorn ; Montero and Samuels a). Greater emphasis on accountability has arguably put pressure on government officials to be more responsive to societal input. Many localities, and even provinces or states, have begun to implement innovative institutional spaces of access for popular associations. These spaces have primarily been arrangements for participation in policy councils that give associations some voice in social policy or in the distribution of the budget for urban infrastructural improvements, usually at the local level. While this template is becoming more common, it has so far been partially and unevenly implemented, and its ability to deliver real policy-making access to associations, especially beyond the municipal level, is still uncertain. In sum, state channels for access under the A-Net are often substantial but appear to be more confined to the local level and involve a different set of issues and policy areas when compared to patterns under the -Hub. Disparities between the two interest regimes in terms of access also derive from the gatekeeping role of parties, which may act as intermediaries between popular organizations and the state. In the -Hub, parties initially forging ties and mobilizing union support could do so only by offering substantial concessions (Collier and Collier ). The implications and payoffs of party affiliation to unions varied across countries and across time within countries, but in general, the relationship was grounded in a basic exchange.6 As volunteer labor, unions served as electoral “shock troops” that could mobilize the vote and provide large-scale demonstrations of public support. They helped to cement party identification, and, as more general sources of core political support, to legitimate parties among the larger popular sectors. In exchange, unions often gained some degree of access. As members of these s, unionists had influence within the party and often ran for office as party candidates and won elected positions. Populist s were typically the largest party in the country; leftist s, as in Chile, constituted a large, competitive bloc. Since such parties or blocs organized governments or oppositions, they constituted 6. A clear example of the temporal variation in the balance of power between the labor movement and the central party organization can be seen in the contrast between what Collier and Collier () call the incorporation and postincorporation periods, when the political goals and strategies of leaders of populist s were quite different. In the former, unions had substantial clout in the party; in the latter they were often subordinated to party control.
03 Chapter 3.qxd
11/9/2009
11:49 AM
Page 85
Table .
Institutional Channels of Access Across Interest Regimes -Hub
Direct State Access
,
A-Net
Formal Channels
Industrial relations system (mandatory participation)
Participatory policy councils (uneven implementation, limited participation)
Informal Channels
Informal ad-hoc concertation on macroeconomic policy
Irregular patterns of contacting and claim making initiated from below
Level
Predominantly national
Predominantly local
Policy Areas
Macroeconomic policy, national social policy
Social policy, infrastructure, local development and governance
Organic ties; exchange of access for support
More distant relations with possible short-term negotiated exchanges
Party Access
important policy-making structures, and in this way too they provided a channel of access for unions at all levels, from municipal to national, depending, of course, on their own relationship to power. Parties in the current period tend to have more distant relationships with their constituencies and to rely much less on societal organizations for electoral and political support. Consequently, the same basic exchange of support for access has not been reproduced. New parties have often formed around “neopopulist” candidates, who make personalistic appeals that are largely unmediated through organizations.“Audience” (Manin ) parties similarly make appeals to an atomized mass public, so that most parties do not seek to affiliate associations in the same way. Even the old s have adopted new patterns of linkage to voters. These parties have reordered their constituencies and transformed their relationships to unions, which no longer tend to hold such a privileged position as a core support group. While adaptation has typically involved continued targeting of the popular sectors for electoral support, party strategies have generally emphasized contingent, instrumental relations, activated episodically, particularly at election time, rather than more “organic” linkages with popular associations. Even for those parties most oriented toward an associational support base, however, political parties do not institutionally incorporate associations in a way that is comparable, for example, to the ’s labor sector in Mexico, the ’s tercio in Argentina, or ’s labor bureaus in Venezuela, in the earlier interest regime. As Roberts (forthcoming)
03 Chapter 3.qxd
11/9/2009
.
11:49 AM
Page 86
has argued, the neoliberal critical juncture “has undermined . . . mass parties and led to a proliferation of individualized linkages to machine, personalist, or professional-electoral parties. . . . A more fragmented, autonomous, and pluralistic civil society . . . [now has] fluid and tenuous linkages to party organizations.” Parties thus provide less access for associations in the A-Net.7 It should be noted, however, that this pattern is consistent with the strategy of elected leaders to win political loyalty by using associations to implement state programs. The contrasts just described are summarized in table .. While no comparative data directly taps these contrasts, some survey results give more empirical detail about the pattern of accessing in the core segment of the A-Net. The top row of table . compares four indicators of accessing: the proportion of associations for which contacting local and national officials is important, the proportion that reports regular participation in institutional spaces provided by government, and the proportion that reports contacting leaders of political parties as important. The data confirm a pattern of accessing that occurs primarily at the local level. Yet national-level contacting is not insignificant, nor are the institutional spaces provided by the state, at least for the core segment of the interest regime captured in the sample. The data also confirm that political parties are less active as intermediary structures. The table also shows the main issue categories of concern to associations and, for each issue area, the proportion of associations working on that issue that engage in each accessing strategy. Associations are dedicated to a great diversity of issue areas, a broad repertoire that contrasts sharply with the productionist bias of the -Hub. The chain-referral sample makes it difficult to Table . Patterns of Accessing in the A-Net
Incidence (A)
Total (% of A)
Local Contact (% of A)
National Contact (% of A)
Institutional Space (% of A)
Party Contact (% of A)
Social Service Poverty Neighborhood Public Bads Politics Productionist Total
30.5 24.9 21.5 14.6 13.3 13.2 100.0
55.0 64.0 62.3 63.8 69.8 66.4 60.0
29.1 27.1 26.0 28.3 45.2 39.2 29.7
31.7 27.0 35.9 46.2 52.5 31.9 33.3
16.3 17.0 22.0 20.4 22.1 15.0 16.7
7. Since our surveys were fielded, leftist presidents in Argentina and Venezuela have established new forms of linkage with associations. These new developments are discussed briefly in chapter .
03 Chapter 3.qxd
11/9/2009
11:49 AM
Page 87
,
assess the distribution of issue orientations in the associational world, but in these data the emphasis is on social service delivery, poverty, and neighborhood issues, while productionist issues, political issues (such as corruption and human rights), and public bads (such as violence and pollution) take a back seat. To some extent, particular issues are more likely to be pursued through distinct channels or accessing strategies. Associations dedicated to productionist and political themes are considerably more likely to make national-level contacts, compared to associations pursuing other issues. If we infer that these national-level channels are especially vital for the pursuit of these interests, then the overall local orientation of the A-Net may introduce a general bias against the pursuit of these issues. With respect to political problems, this bias might be ameliorated somewhat since these issues—along with public bads— also seem especially conducive to access through institutional spaces and other participatory innovations. However, this is not the case with productionist issues. The A-Net thus may differ from the -Hub not just in terms of its greater diversity of issues addressed, but also because the pattern of accessing seems specifically biased somewhat against productionist and political issues that are more often pursued through national-level accessing, which is harder to secure. The significance of this issue bias, especially regarding productionist issues, can be more fully appreciated if we compare (cautiously, given the nature of the sample) the distribution of issue categories from table . with data on issue salience drawn from our survey of individuals. The categories do not match up perfectly (the responses aggregated into the issue categories in table . were generated through an open-ended question, whereas the responses for salience were selected from a list), but the comparison is nevertheless instructive and suggests a relative inability of the A-Net to address salient, workrelated issues. In response to a question concerning the most important problem in the country, . percent of popular-sector individuals pointed to unemployment or work. In comparison, only . percent chose health, basic needs, housing, or education. Yet, as table . indicates, associations are far more likely to address the latter group of issues than the former. Autonomy As mentioned in chapter , “autonomy” presents both conceptual and empirical challenges. Conceptually, instead of a dichotomous approach, we prefer that of Fox (:; see also Haber :), who defines an organization’s autonomy as the “degree of control over setting its own goals and making its
03 Chapter 3.qxd
11/9/2009
.
11:49 AM
Page 88
own decisions without external intervention, whether by governments, political parties, religious groups or development agencies.” Linkages with the state, either directly or through political parties, can lead to challenges to autonomy. Because the word “representation” invokes autonomous societal actors, Schmitter () appropriately insisted on the word “intermediation” to invoke instead the idea that structures that stood between society and the state to present societal interests were to one degree or another shaped by the state and were channels of “state interest” as well as of societal interest. The kinds of relationships popular organizations have with states and parties and the ways these may potentially challenge autonomy thus become a key dimension for analysis of interest regimes. It is useful to analyze the challenges posed by relationships with external actors, particularly the state, in terms of inducements as well as constraints (Collier and Collier ). Relationships between unions and the state within the -Hub were governed by corporatist labor laws that regulated unions in great detail through a combination of the two mechanisms. Constraints refer to overt controls regulating the activities of unions, their leadership, and their internal governance. Inducements bestow benefits or advantages, which may shape organizational behavior through incentives that derive from, for example, the prospect of continuing government subsidies. The bestowal of benefits may also induce cooperative behavior, even co-optation, providing incentives through discretion in their application, thereby privileging some unions but denying these benefits to other, often dissident elements of the labor movement. Both types of control by the state were extensive and formally embodied in labor law.8 Compared to unions, states generally impose relatively few regulations on popular associations that constitute direct constraints. While most associations register with the government, this formal state recognition seems to be mostly pro forma and largely optional for many kinds of activities. There is no “associational law,” comparable to labor law, that provides either direct regulation or “rights” to carry out basic activities (parallel to the right to bargain collectively), and some associations, those oriented toward certain kinds of self-help activities, may not come into contact with the state at all. More frequently, however, associations are presented with a set of state inducements. Despite the lack of regulation under the A-Net, most associations 8. In addition, other types of inducements and constraints have typically taken place informally in Latin America. These range from many types of co-optive favors to corrupt practices and illegal forms of coercion.
03 Chapter 3.qxd
11/9/2009
11:49 AM
Page 89
,
engage in activities that in various ways lead them to interact with the government. Although state influence tends to be substantially more subtle than for unions, and although associations may indeed maintain greater independence than many unions, interactions with the state may nevertheless have implications for the autonomy of associations. In theorizing how the engagement of the state may create inducements that challenge autonomy, it is useful to think of two kinds of state-association linkage that are relatively widespread and their possible implications. First, the pursuit of accessing strategies and the orientation of state policy may shape associational agendas. Associations that attempt to gain or maintain access to policy makers or party brokers may be induced to enter into cooperative—or perhaps co-optive—relationships, especially with the prospect of repeated interaction. Cooperative relations and some adjustment to the nature of demands may be beneficial to the organization in representing the interests of its base, and an association may, in the process, retain substantial autonomy. At the same time, just as journalists may have an incentive to refrain from critical reporting on valued sources, the effect may be stronger: associations may shape their agendas and political support accordingly, gambling on an unknown return to this behavior. These inducements present potential challenges to autonomy. Similarly, the orientation of state policy and programs may offer strong incentives that guide associational activities and shape the agenda of the larger popular interest regime by signaling the greater “availability” of specific demand-making targets. As a consequence, associations may turn away from or give less priority to other substantive goals. For example, the existence of labor programs in Argentina channeled the energy and demands of both new and existing associations toward a government-initiated program, in a sense diverting attention from other approaches to welfare and income support (Garay ). Thus, through incentives provided by existing programs, the state can shape the substantive orientation of individual associations and even weight or skew the larger, aggregated interest regime toward programs the state finds most acceptable (e.g., Cigler and Loomis ). A second set of inducements stem from more direct dependence on the state, usually for resources. Over the last two decades, states have more frequently become directly involved in funding associations. The growing trend of publicprivate partnerships or state “outsourcing” of service delivery and program implementation makes associations dependent on the government for a major part of their activities. Through the threat of exclusion from a program or denial of funding, the state may implicitly constrain associational behavior, for instance
03 Chapter 3.qxd
11/9/2009
.
11:49 AM
Page 90
Table . Relationships of Dependence (percent of associations) State Dependence
Receives money from the state Distributes government program State actor involved in founding
Party Dependence
Receives money from party Party involved in founding
29.5 23.0 14.1 3.3 6.3
through a tacit understanding between associations and political leaders or state officials that criticisms should be blunted. In addition, the state may effectively exert controls on associations through requirements for funding or participation in programs, such as criteria for constituting a soup kitchen or distributing resources in a particular way. Finally, states have sometimes become involved in the founding of associations and may in that way exert influence, especially on younger associations. Table . shows the proportion of associations in our sample that reported each of these three types of state-association linkages. In addition to these activities involving dependence on states, popular organizations may of course develop similar relations with political parties. In the -Hub, the “exchange” relationship mentioned above often led to substantial restrictions on union autonomy, although these restrictions varied across parties and even across time periods. Subordination of unions to the party occurred, as in Venezuela, when governing populist parties had the responsibility of constructing and sustaining policy and governing coalitions. Other s also faced the classic trade-off between the representation of the core constituency and party or electoral strategies. in Peru conservatized substantially with the goal of escaping the military ban on its electoral victory; while in the face of a more severe ban on the party that stood little chance of reversal, the in Argentina did not subordinate unions or conservatize in the same way. Some subordination of union interests to party goals was also a feature of leftist s, like the Socialist Party in Chile, though, again, these relationships varied through time, with the strategic position and goals of the party. In addition, s could be vehicles of co-optation, so that union leaders became more dependent on favors and career opportunities through the party than they were on the rank and file. In the A-Net, with the change in their strategies for mobilizing electoral support, parties generally do not subordinate popular associations in the way s traditionally subordinated and controlled unions. As described above with regard to access, party-association relationships are less frequent and, even when they exist, tend to be characterized by significantly greater distance. Relationships almost never take the form of formal or informal integration into
03 Chapter 3.qxd
11/9/2009
11:49 AM
Page 91
,
the party, but instead are negotiated on an ad hoc, recurrent basis. As shown in table ., very few associations rely on parties for financial support or have had parties involved in their founding. Just as the greater distancing of associations from parties accentuates a difference between the A-Net and -Hub with regard to access, so it does with the role of parties in their subordination of popular interest organizations. However, this reduced role of parties does not eliminate the risk to autonomy stemming from individual politicians who may seek to form clientelistic relationships with associations. Support mobilization through associations is available to politicians at all levels, from incumbent presidents to mayors and also to those in the legislative branches at both national and local levels.
Conclusion This chapter has compared the -Hub to the A-Net in terms of four dimensions: scope, scaling, access, and autonomy. The first part of the analysis examined the logic of collective action and relations between the state and popular organizations as a means of exploring two aggregate traits of the interest regime. The analysis regarding collective action suggested that the scope of the -Hub and the A-Net might be roughly comparable, despite the much larger group of potential participants under the A-Net, and that scaling might be Table . Comparison of -Hub and A-Net -Hub
A-Net
Scope
State actions solve otherwise greater individual collective action problems
“Natural” ease of individual collective action
Scaling
Unions generally highly coordinated, prominent role of federations and confederations
Associations substantially less scaled, less dominant role of coordinating associations
Access
Accessing predominantly at national level, parties often play intermediary role, mainly productionist and macroeconomic claims
Accessing predominantly at local level, parties more rarely play role as intermediaries, great diversity of claims
Autonomy Detailed state regulation and party affiliation; constraints as well as inducements to affect union activities
Associations more autonomous, but involvement in government programs and state subsidies may create dependence and shape activities
03 Chapter 3.qxd
11/9/2009
.
11:49 AM
Page 92
considerably more extensive in the -Hub than in the A-Net. Although bringing data to bear on these comparisons is difficult, both contrasts seem largely born out. The subsequent analysis of relations with the state suggested that the two interest regimes have been marked by different patterns of access and autonomy. The -Hub offered formal and informal channels of access predominantly at the national level, both directly to the state and through parties as intermediaries, which largely served as conduits for productionist claims. The A-Net is characterized by greater access than has sometimes been assumed, over a much greater diversity of issues; yet the pattern is more fragile, occurring primarily at the local level, rarely involving political parties as intermediaries, and at least to some extent not prioritizing the most salient issues. This divergent pattern of access emerges in tandem with a difference regarding autonomy. The -Hub involved substantial restrictions on union autonomy, marked by the controlling or co-optive influence of both state and party actors. The A-Net is not free of such controls, but relationships that threaten associational autonomy are more indirect, partial, and less pervasive. Table . presents the general contrast that emerges from this analysis, which has focused on sketching the characteristics of the -Hub and A-Net in broad strokes. Of course, the actual operation of the A-Net regarding each of these aggregate traits is considerably more complex, and across countries the emergent interest regime is still taking on different forms and must be a subject of ongoing analysis. Subsequent chapters shed light on each of these aggregate traits from various perspectives with more focused empirical investigation.
04 Chapter 4.qxd
11/10/2009
9:44 AM
Page 95
4 : - Thad Dunning
How do Latin American citizens participate in the interest arena? While the changing character of interest representation in the region, typified by the decline of robust union-party linkages during recent decades, may imply new constraints on political participation, it may also afford opportunities for novel modes of political problem solving. The emergence of new forms of associational life, in particular, has captured recent analytic attention (Chalmers, Martin, and Piester ; Oxhorn b; Korzeniewicz and Smith ; Foweraker ). Yet scholars have largely lacked systematic and comparative data with which to evaluate claims about emergent forms of interest intermediation. Chapter distinguishes the interest arena and the electoral arena; whereas participation in the electoral arena involves voting and other partisan activities, citizens may also undertake “direct action,” that is, make a variety of direct appeals to state officials or engage in other types of state-targeted claim making, and they may participate in associations, for purposes of either state-targeted claim making or society-targeted provisioning. This chapter investigates these latter forms of participation in the interest arena.1 What kinds and patterns of political problem-solving strategies are most prevalent? What types of direct action do citizens employ, and to what extent do they actually participate in 1. This chapter analyzes data from a survey administered to a probability sample of individuals from the capital cities of Argentina, Chile, Peru, and Venezuela. The survey instrument and sampling design are described further in chapter and the appendixes of this volume.
04 Chapter 4.qxd
11/10/2009
.
9:44 AM
Page 96
the associations that have been so widely noted? To what degree do individuals find the interest arena a useful site for solving problems? How do the popular sectors differ from the middle classes in this regard? As two kinds of individual engagement in the interest arena, both direct action and associational participation are important components of interest representation. As outlined in chapter , direct action involves discrete, statetargeted activities, in which individuals typically voice demands regarding single, specific problems or issues that may come up at a particular moment. Given this broad definition, there may be substantial variety in the form of direct action. Citizens may target the state at various levels (from local to national); the character of direct action may be more institutionalized (e.g., signing of petitions) or more contentious (e.g., participating in protests). Furthermore, direct action can be atomized, as in individual, face-to-face interaction with authorities; or it can be collective, involving either face-to-face interaction between collectivities and authorities or larger efforts such as petition drives or protests, in which demands are voiced more distantly. These different forms of direct action may correspond as well to the range of objectives individuals pursue: personal or particularistic problems might be especially well addressed by face-to-face contacts, especially atomized contacting, while participating in protests or signing petitions may correspond to more general demands for policy changes or larger group benefits. The key point is that these various modes of direct action all involve the voicing of specific demands or claims on the state. In contrast to direct action, associations tend to have a multifaceted existence beyond any single particularistic issue or specific claim-making activity, and they tend to work in relation to a set of general concerns or issue areas that may or may not involve state-targeted demands (see chapter ). The associations on which this chapter mainly focuses are “programmatic,” in that they exist in order to address collective problems or advance goals that have been conceptualized as political in chapter . (Associational participation will generally be used to refer to participation in such programmatic associations; when participation in nonprogrammatic associations is analyzed, this is explicitly noted.) Thus, these organizational units of the A-Net may pursue diverse activities, both making claims on the state and providing useful vehicles for collective provisioning of a variety of local public goods that could be, but for one reason or another may not be, supplied by the state. They may also serve as conduits for state program implementation. As a means for political problem solving, associational participation may differ from direct action in various ways. Relative to direct action, participation in associations may be a more sustained kind of problem solving, though individual participation in associations
04 Chapter 4.qxd
11/10/2009
9:44 AM
Page 97
,
may of course wax and wane over time and across different issue domains. In addition, since associations engage general issues and problems of concern to members or participants, they may be less likely to serve as a means for obtaining particularistic benefits for individual members or participants. Who, then, contacts or accesses the state through direct action, and who participates in associations? Past studies of political participation, particularly in the developed world, emphasize that income, education, and other socioeconomic variables explain political participation, from voting to protests (Verba and Nie ; see also Wolfinger and Rosenstone ; Rosenstone and Hansen ). In the civic voluntarism model, Verba, Schlozman, and Brady () argue that socioeconomic status affects participation by providing “resources,” like time, money, and politically relevant skills; by promoting engagement with politics; and by influencing the mobilization of citizens (see also Brady, Verba, and Schlozman ). In Latin America, too, those with higher socioeconomic status have traditionally been seen as having better access to the state, for many kinds of reasons. Whether making personal, particularistic claims or demands for group benefits, those with higher class status may draw, inter alia, on reservoirs of human capital, time, money, and social connections (O’Donnell ; Kurtz b; B. Schneider ); elite groups may also be able to overcome collective action problems by virtue of their relatively smaller numbers (Olson ), engendering disproportionate influence on policy making as well as asymmetric personal access to the state. The popular sectors, on the other hand, may disproportionately lack the resources and contacts for many kinds of direct action, although they may retain an organizational or numerical advantage in others, such as protests (Myers and Dietz ; López Maya and Lander ). The question of a class bias in accessing the state may be particularly relevant in the contemporary period, as the role of traditional vehicles for popular-sector interest representation such as union-party linkages has waned in the wake of rising economic informality and marketization (Portes and Hoffman ). This chapter therefore emphasizes comparisons of patterns of direct action by social class, drawing from this volume’s emphasis on descriptive representation as a useful diagnostic for evaluating bias in the interest arena: asking whether the incidence of specific demands on the state is symmetric or asymmetric across classes may allow us to assess the evenness with which access to the state (if not influence) is distributed. With respect to associations, the resources of time, money, and politically relevant skills associated with socioeconomic status may also play an important role in promoting participation, so that the possibility of a bias against popular-sector participation is present for associations as well. On the other
04 Chapter 4.qxd
11/10/2009
.
9:44 AM
Page 98
hand, various forms of “social capital,” including experience with various modes of collective action, the ability to draw on wide social networks, and ties to communities may be particularly important for associational participation; in this regard, associations working with the popular sectors could enjoy certain organizational and numerical advantages.2 Some recent analyses have noted that the rise of “associative networks” in Latin America may even provide a privileged site for representation of the popular sectors, relative to the middle classes (Chalmers, Martin, and Piester :–, ). Yet these analyses have not been able to exploit systematic and representative data on individual participation in the contemporary interest regime. These hypotheses about who participates in associations and about what kinds of factors facilitate associational participation have therefore gone untested. Related to the issue of participation in associations is the extent to which associations are perceived as useful vehicles for making claims on the state and/or for providing local public goods and services through provisioning activities; presumably individuals would tend to participate more in associations they think are effective. By analyzing individual-level data in terms of the perceived efficacy of associations, this chapter also places a “subjective” lens on the capacity of different parts of the organizational infrastructure to function as representational vehicles. The aim of this chapter is thus to compare distinct modes of participation in the interest arena, in particular, forms of direct action and associational participation. The chapter examines who participates and how, with particular attention to class differences in participation rates across different forms or channels of problem solving in the interest arena. In addition, though the chapter investigates aggregate patterns across the four countries studied, it also focuses on cross-national differences, as different national settings may provide different incentives and structures for individual patterns of participation. Finally, in multivariate analyses, this chapter also investigates individual correlates of participation, above and beyond popular-sector membership, which allows us to explore how involvement in social networks, past experience with organizational activity, and other factors may promote individual problem solving through associational participation and direct action. Together, these steps in the analysis contribute to a portrait of individual participation in the contemporary interest arena. 2. While the resource-based model of Brady, Verba, and Schlozman () emphasizes the role of socialization through organizational involvement, other accounts give an even more central role to social networks in explaining political participation (Lake and Huckfeldt ; see also Putnam ).
04 Chapter 4.qxd
11/10/2009
9:44 AM
Page 99
,
Operationalizing Modes of Political Problem Solving The above distinctions imply a typology of modes of political problem solving based on two nominal dimensions, as depicted in table ., in which direct action and associational participation form the rows, and atomized and collective problem solving form the columns. Note first that the lower-left-hand cell is empty because associational participation is a preeminently collective form of problem solving, one in which citizens by definition cannot engage in an atomized fashion. The table thus identifies three types of political problem solving in the remaining cells: atomized direct action, collective direct action, and (collective) associational participation. Within the categories of atomized and collective direct action, modes of problem solving are further classified by whether they are unmediated, in the sense that the individual herself is engaging state authorities, or intermediated, in that the individual is contacting another actor to solve a problem with the state on her behalf. The table also indicates the ways in which these types of participation have been operationalized through a questionnaire administered in the capital cities of Argentina, Chile, Peru, and Venezuela (see appendix C). With respect to direct action, the analysis focuses on a series of survey questions identifying whether respondents had in the previous five years sought to resolve problems by contacting an official or agency of the national, provincial, or municipal government or by engaging in a legal action; these forms of direct action are considered unmediated. Two other types of intermediated problem solving were also investigated: contacting a political party and contacting an influential intermediary, both of which can be understood as involving actors that intermediate between state and society.3 Each of these activities may be conducted individually, in an atomized fashion, or they may be engaged in as part of a broader social network—for instance, as part of a group from the neighborhood, from work, or from other venues.4 In addition, surveys
3. Respondents were asked whether they had attempted to contact “a person with influence and contacts to help you or to recommend how to resolve problems such as those we have been discussing” (Argentine questionnaire) in the previous five years. A follow-up question, specifying the person the respondent sought to contact with the greatest frequency, asked about the purpose for which the respondent sought to contact this person. “Gaining access to a state entity” was one of the response categories that was coded for this open-ended question. 4. Respondents who had engaged in each of these activities were then asked with whom they had done so on the most recent occasion. Those who did so alone or with family members were coded as having engaged in a direct action in an atomized fashion; those who did so with acquaintances from a neighborhood association, a political party, work, church, or some other source engaged in direct action collectively.
04 Chapter 4.qxd
11/10/2009
.
9:44 AM
Page 100
Table . Operationalizing Problem Solving in the Interest Arena Atomized (Alone or with family members) Direct Action
Unmediated: • Contact the government • Engage in a legal process
Intermediated: Contact political party Contact influential intermediary
• • Associational Participation
Collective (With acquaintances from work, the neighborhood, or other sources) Unmediated: Contact the government Engage in a legal process Sign a petition Participate in protest
• • • •
Intermediated: Contact political party Contact influential intermediary
• •
Participation in programmatic popular associations, including: • Neighborhood • Anticrime (protection) • Food distribution • Parents • Women • Regional/place-of-origin • Cooperatives • Unemployed
asked about two forms of direct action that are considered ipso facto collective acts: participation in a protest and signing a petition.5 Respondents who have engaged in a form of direct action could then be coded as having participated in either an atomized or collective fashion, as described in the cells of the upper row of table .. Associational participation is operationalized as involvement on a regular basis in several kinds of groups and organizations named in the surveys. These programmatic associations are set forth in the lower-right-hand cell of table .. We do not here include unions; as noted in chapter , unions are components of the interest regime, but the present study focuses specifically on associations, the base units of the A-Net. Also, the types of associations included in table . obviously do not exhaust the range of associations in the interest regime, but they are taken to be especially important for popular-sector 5. Signing a petition is conceptualized here as an unmediated form of collective direct action, since individuals signing petitions directly give voice to specific, concrete demands; however, the organization of petition drives obviously does involve intermediation to some extent.
04 Chapter 4.qxd
11/10/2009
9:44 AM
Page 101
,
participation. A host of other associations, such as religion- or church-based and recreational associations, draw the participation of many people but are essentially nonprogrammatic (see chapter ); though we do not include these in table . as instances of associational participation in the interest regime, we will nonetheless examine the incidence of participation in recreational and religious associations at various points below.6
Patterns of Problem Solving: The Aggregate Picture An aggregate analysis of patterns of political problem solving reveals that rates of participation in programmatic associations are quite high, reaching an estimated percent for associations across the four capital cities in our study.7 This figure is impressive, even compared to many developed countries; for instance, Verba, Schlozman, and Brady (:) report that in Europe, estimated rates of voluntary associational participation were percent in Spain, percent in France, and percent in Italy. The same study found that an estimated percent of the population of the United States reported participating in an association (Verba, Schlozman, and Brady :); yet that study analyzed many more types of associations than the set of programmatic associations analyzed in this volume.8 (Indeed, including just church-based and recreational associations would boost the rate of participation to nearly percent in the four Latin American countries.) The level of associational participation in Latin America is therefore quite impressive in comparative terms. The incidence of direct action is perhaps surprisingly high as well. Indeed, while increasing associational participation has been widely noted in the literature, direct action as a mode of representation in the contemporary interest arena has gone largely unremarked. Yet the survey results indicate that this 6. In the analyses below, it is indicated whether both nonprogrammatic and programmatic associations or only programmatic associations are considered. 7. To estimate population parameters, sampling weight ratios are used to adjust for over-sampling of respondents in focus districts, using the inverse of the probability of selection and then setting the effective sample size at the number of interviews. For estimates by country, country-specific weights are used. 8. The way in which they arrive at this estimate is similar to the present study: respondents were asked about involvement with specific types of associations (veterans’, religious, recreational, neighborhood, etc.), some of which took political stands and some of which did not. An aggregate measure of associational participation based on having reported participating in any of these specific types of association was then created. When respondents were asked a general question regarding whether they were members in any association, however, only percent indicated membership (Verba, Schlozman, and Brady : n.).
04 Chapter 4.qxd
11/10/2009
.
9:44 AM
Page 102
form of activity is quite important. Pooling across countries, we observe participation rates of and percent in collective and atomized direct action, respectively; engagement in direct action as a whole approximately matches participation in associations.9 Though we lack fully comparable aggregate data on patterns of direct action in other developing and developed countries, disaggregated estimates presented in the next section suggest that rates of collective and atomized direct action are impressive in comparative terms as well. Behind this aggregate picture, however, lies substantial variation across the four countries (figure .). Peru in particular stands out in terms of associational participation, with percent of citizens participating in programmatic associations; in Peru, unlike the other three countries, the rate of associational participation exceeds the sum of the rates of atomized and collective direct action ( percent).10 In Chile and Venezuela, associational participation exceeds either form of direct action alone, though together the two forms of direct action outpace associational participation. In Argentina, associational participation rates are similar to Chile’s and substantially higher than Venezuela’s; however, associational participation is surpassed by collective direct action, underscoring the high level of direct action in that country. Nonetheless, despite this cross-national variation, the evidence underscores the importance of associational participation: even in Argentina, for instance, where the incidence of direct action outstrips associational participation, the relative frequency of associational participation there is still greater than in Spain, France, or Italy (Verba, Schlozman, and Brady :). There are important cross-national differences in direct action as well, particularly when one compares the incidence of atomized and collective direct action. Comparing across countries, collective direct action is particularly prevalent in Argentina and Peru, while atomized direct action is high in Argentina and Chile. Argentina has the highest rates of both collective and atomized direct action ( and percent, respectively), while Venezuela stands out in terms of 9. An estimated . percent of citizens in the pooled universe have engaged in both atomized and collective claim making, i.e., one form of direct action listed in table . in a collective fashion and another form of direct action in an atomized fashion. Note that for the disaggregated forms of direct action listed in table . and discussed in the next section (contacting the government, engaging in a legal process, and so on), atomized and collective direct action are mutually exclusive, because survey questions asked about the most recent occasion on which respondents engaged in direct action. 10. The Peruvian surveys were fielded shortly after the fall of Alberto Fujimori, widely perceived to have instituted a semi-authoritarian government. While this could account for an upsurge in mobilization in general, and while associational life might have also surged outside of Lima, it is unlikely to account for the much higher participation among limeños in the types of associations we consider here, relative to other capital cities in our study. See the discussion below of the question of mobilization during “normal” versus “crisis” moments.
04 Chapter 4.qxd
11/10/2009
9:44 AM
Page 103
,
60 50
Associational Participation Collective Direct Action Atomized Direct Action
40 30 20 10 0 Argentina
Chile
Peru
Venezuela
Fig. .. Problem Solving in Cross-National Perspective (percent)
low rates of all forms of political problem solving, including both collective and atomized direct action ( and percent, respectively). These important crossnational differences should be borne in mind and will be analyzed further below. Overall, the data suggest that collective problem solving is relatively prevalent. Not only is associational participation robust, but within the category of direct action, collective direct action is relatively important; only in Chile does the incidence of atomized direct action outstrip collective direct action.
Direct Action What kinds of direct action are most prevalent in Latin America, and who engages in direct action? In this section, I examine these two questions in turn, focusing respectively on the incidence of unmediated versus intermediated direct action and on the incidence of associational participation as well as the issue of differential participation. The literature suggests several hypotheses that might guide our expectations. With respect to the incidence of various kinds of direct action, we might expect that, following the literature on patronage politics and clientelism in Latin America, contacting political parties and influential intermediaries would play a particularly important role in problem solving. Relatedly, we might also expect variation across different Latin American countries in this respect; for
04 Chapter 4.qxd
11/10/2009
.
9:44 AM
Page 104
Table . Direct Action in Cross-National Perspective (percent) Argentina
Chile
Peru
Venezuela
Unmediated
Contacted Government Engaged in Legal Process Signed a Petition Participated in Protest
20.33 (1.05) 7.47 (0.68) 26.61 (1.16) 9.92 (0.78)
20.67 (1.05) 7.09 (0.67) 9.44 (0.76) 6.09 (0.62)
16.06 (0.95) 7.47 (0.68) 12.54 (0.86) 17.61 (0.99)
8.70 (0.73) 6.71 (0.65) 11.93 (0.85) 8.86 (0.74)
7.54 (0.69)
2.96 (0.44)
4.43 (0.53)
1.98 (0.36)
Intermediated
Contacted Political Party Contacted Influential Intermediary
1.90 (0.35)
2.61 (0.41)
0.73 (0.22)
0.71 (0.22)
Note: Percent estimates in the cells are calculated using country-specific sampling weights; reported percentages therefore differ from unadjusted sample percentages. Standard errors are in parentheses.
instance, in more “developed” countries that also feature relatively institutionalized party systems (Argentina, Chile), we may expect a relatively important role for party intermediation, while we might also expect a larger role for both parties and influential intermediaries in countries like Argentina where patronage politics appear particularly widespread. With respect to the question of who participates, individuals possessing resources associated with social class may be expected to engage in direct action at higher rates, as elites and members of the middle class are often alleged to have special access to the state in Latin America. We might therefore expect atomized direct action to be especially prevalent among the middle classes relative to the popular sectors. One contrary expectation is that the popular sectors may retain an organizational advantage in protests, which may also act as a substitute for other, more proximate ways of voicing demands on the state. In addition, involvement in social networks or various individual-level indicators of social capital may be important predictors of participation. Table . confirms, across the disaggregated forms of direct action, the crossnational pattern suggested by figure .: rates of atomized participation are strikingly lower in Venezuela, and to some extent in Peru, than in Chile and Argentina.11 In Chile and Argentina, direct contact with government officials 11. Standard errors for estimates in table . and elsewhere in this paper are calculated as for a simple random sample, a common approach that may be reasonable even for the analysis of much
04 Chapter 4.qxd
11/10/2009
9:44 AM
Page 105
,
at the local or national level is a predominant form of direct action, and its incidence dwarfs participation in contentious claim making in those two countries (i.e., protest). (It should be noted that the data reported here do not reflect participation in the intense and “unique” episodes of protest surrounding the – crisis in Argentina or the unsuccessful coup in Venezuela but instead refer to more “regularized” patterns of interest politics through direct action). Commonalities exist across the countries in the relative importance of unmediated problem-solving activities. Contacting, protesting, and petitioning the state are all relatively widespread.12 Of the three unmediated forms of direct action, contacting an office or program might be expected to be the most prevalent, and, indeed, it ranks very high in all the countries. The importance of the others, however, is striking. In Argentina and Venezuela, signing a petition is an even more prevalent form of making state-targeted demands than contacting the government; it is nearly as important in Peru, and still relatively substantial in Chile.13 Petition signing in Latin America is fairly impressive in comparative perspective as well. For instance, more than percent of the Buenos Aires sample signed petitions, compared to an average of percent nationally in the United States from to (Rosenstone and Hansen :). Protest activity is also an important form of direct action. Participation in protest is most prevalent in Peru, where nearly percent of residents of metropolitan Lima had participated in a protest in the previous five years, and least prevalent in Chile. Differences in patterns across countries are thus substantially different with respect to contentious action, relative to institutionalized action: for instance, protest rates in Peru approach the rates at which citizens in Argentina and Chile contact the government. That protest is also quite prevalent in Argentina and Venezuela may be surprising, since, again, survey questions asked separately about protests apart from the unusual episodes more complex sampling designs than the one used here (see, e.g., Verba, Schlozman, and Brady :–; Kish and Frankel, ). 12. Comparing these data to estimates presented by Verba and Nie (:, table –) suggests that government-contacting may also be quite prevalent relative to other developed or developing countries. Their data, for a national sample (which may underestimate the incidence of contacting compared to a sample of the capital city) suggests that percent of Austrians and percent of Nigerians had contacted local officials with others to resolve a community problem. The rates at which citizens engaged in particularized contacts (similar to atomized contacting) at the local level were: percent in Austria, percent in India, seven in Japan, percent in the Netherlands, percent in Nigeria, and percent in the former Yugoslavia (the latter at all levels of government). 13. Note that our surveys in Venezuela preceded the petition drive associated with the attempt of the political opposition to remove President Chávez from office in –, through a recall referendum that eventually took place in August .
04 Chapter 4.qxd
11/10/2009
.
9:44 AM
Page 106
of protests during the crisis periods in each country.14 While it is appropriate to compare countries on a measure of mobilization and protest during relatively normal times, as in table ., separate analysis reveals that . percent of Venezuelans in metropolitan Caracas reported participating in demonstrations related to the conflict over President Hugo Chávez during –, while . percent of Argentines reported participating in a cacerolazo during the unrest associated with the crisis surrounding the devaluation of the peso.15 Clearly, extraordinary circumstances can mobilize people into action. In contrast to unmediated direct action, intermediated direct action such as contacting political parties and influential individuals as intermediaries— those activities most associated with clientelism—are relatively uncommon. Even legal suits generally outpace problem solving through influential individuals or political parties. The relative infrequency with which citizens solve problems through party representatives and influential intermediaries must be considered in light of the emphasis that analysts have placed on clientelism and the increasingly nonprogrammatic orientation of political parties, which, it has been argued, have become more clientelistic and more reliant on selective benefits. With respect to differences across countries, it is perhaps unsurprising that citizens infrequently contact representatives of political parties in both Peru and Venezuela, two countries that underwent near-wholesale party system collapse in the decade prior to the surveys. Peruvians were somewhat more likely to contact parties than Venezuelans at the time of our surveys, a difference that is consistent with analyses that have found a rapid recovery of partisan identification in Peru in the wake of party system collapse (Seawright ).16 Yet the point estimate for party contacting in Chile, a country with a tradition of well-institutionalized parties with particular appeal to the popular sectors, is lower than either Peru or Argentina. And while Argentina scores highest on party contacting—perhaps unsurprisingly, given a history of party ties to the popular sectors and an important role for clientelist appeals by the Peronists (Levitsky ; Stokes )—rates of party contacting in Argentina are 14. In Peru, protests emerged in connection with the end of the Fujimori regime. Many important protests were concentrated in the south and the central Andean region, however, while our surveys were fielded in Lima. Our high estimate of protest rates in Peru may therefore underscore the elevated level of contentious mobilization in the capital during “usual” as well as “crisis” times. 15. In other words, during these moments of crisis, participation in protests in Venezuela and Argentina rose to the level found in Peru—which may again underscore the high level of activation in the latter country. 16. The proportion of Lima residents who contacted and Fujimori’s party (Cambio ) was roughly the same, with other parties being contacted at lower rates.
04 Chapter 4.qxd
11/10/2009
9:44 AM
Page 107
,
statistically on par with engaging in legal processes (and party contacts are less frequent than engagement in legal processes in the other three countries). The aggregate picture is thus one in which party contacting is an infrequent channel for direct action in the interest arena; while overtime comparisons are not possible, the data confirm the present distant relationships between political parties and society suggested by previous analysts. The low rate at which citizens contact influential intermediaries for purposes of accessing a state entity is also surprising. Various studies have emphasized the important role of clientelism in contemporary Latin American democracies (e.g., Roberts ; Weyland ; Stokes ), and clientelism or patronage politics may surely involve asking powerful intermediaries for help with contacting a state entity. Yet in no capital we surveyed have more than percent of citizens contacted an influential intermediary.17 The greatest estimated incidence occurs, unexpectedly, in Chile, despite studies that suggest Chile may be a country in which clientelistic politics play a less important role. It should be noted that instances of unmediated direct action may also take the form of clientelistic relations; local power brokers may in fact be state officials, and citizens may contact them face-to-face for the purposes of receiving selective benefits. The overall conclusion, then, is not necessarily that patronage and clientelism are not widespread, but that these relationships rarely take the specific form of an intermediated linkage between citizens and the state, a form that previous literature would have suggested to be much more common. Another possibility is that societal “brokers,” rather than receiving requests from citizens and then using their power to contact the state, might help citizens themselves engage in unmediated direct action by accompanying them in the activity or recruiting them to other activities such as protests and petitioning. The surveys allow us to explore this question by examining how often direct action activities were conducted in conjunction with local leaders (dirigentes) of neighborhood associations, unions or work organizations, churches, or political parties. Table . suggests the substantial importance of dirigentes in this sort of activity. Pooling across countries, percent of respondents who had engaged in protests reported that dirigentes had played a role. Such dirigentes played much less marked roles in legal processes, but they also helped in contacting both political parties and people of influence (final column of table .). There are notable contrasts across countries. Peru in particular stands out for the role of dirigentes in aiding protests as well as in contacting political 17. Some patronage brokers may be representatives of political parties, of course, yet this does not fully explain the counterintuitive findings—since party contact rates are also so low.
04 Chapter 4.qxd
11/10/2009
.
9:44 AM
Page 108
Table . Role of Brokers in Cross-National Perspective (percent of participants saying a “leader” [dirigente] had helped)
Contacted Government Contacted a Political Party Contacted Influentia Intermediary Engaged in Legal Process Participated in a Protest
Argentina
Chile
Peru
Venezuela
Pooled
19.57 (6.08) 47.76 (9.12)
47.79 (9.02) 33.55 (13.65)
55.35 (4.88) 65.75 (7.57)
54.78 (17.57) 57.00 (11.61)
44.78 (3.30) 56.37 (4.44)
60.54 (10.99) 1.69 (3.37) 39.91 (6.31)
62.91 (11.26) 16.20 (10.78) 38.23 (8.29)
42.81 (9.15) 49.50 (7.57) 71.06 (3.17)
71.20 (9.65) 45.62 (11.71) 39.66 (6.62)
54.86 (5.10) 35.67 (4.79) 60.26 (2.40)
Note: Survey questions asked respondents whether a leader (dirigente) of a neighborhood committee or political party, or one from work or church, had helped them with respect to each form of direct action. Petitions are omitted, as respondents were not asked about the role of a leader (dirigente) in helping them with petitions. Percent estimates in the cells are calculated using country-specific sampling weights; reported percentages therefore differ from unadjusted sample percentages. Standard errors are in parentheses.
parties; over percent of protest participants had done so together with a local leader, an observation that might go some way toward explaining the heightened overall level of protest in Peru discussed above. In the other three countries, the role played by leaders was similar with respect to protests, while local leaders or brokers appear to play a bigger role in the contacting of intermediaries in Argentina, Chile, and Venezuela and a reduced role in Peru. These results, particularly those for individual countries, should be interpreted with some caution as the numbers of respondents who have participated in several forms of direct action are small within countries, and thus standard errors in table . are large,18 and the numbers regarding the activity of particular types of “leaders” is particularly small, rendering it impossible to interpret the degree to which “clientelism” rather than “representation” may be involved. Nevertheless, brokerage is clearly an important element of direct action. Within the broader framework of these regional and national trends in direct action, three further questions regarding descriptive representation based on 18. In addition, there was fairly substantial nonresponse on the question about the role of dirigentes, even for those respondents who had said they had participated in a particular form of direct action. This conceivably could indicate that the role of dirigentes is even more important than it appears, if respondents are reluctant to reveal the role of leaders. The percentages in table . are calculated using only respondents who responded to the question about dirigentes.
04 Chapter 4.qxd
11/10/2009
9:44 AM
Page 109
,
class are important for understanding popular-sector representation in the interest arena. To what extent is there class bias in terms of who participates in the different kinds of direct action? Does social class help predict the tendency of individuals to engage in certain kinds of direct action in an atomized fashion and others in a collective fashion? Finally, how does the perceived efficacy of these different forms of problem solving—in terms of type of activity and in terms of whether it is carried out individually or collectively—vary across social classes? In order to measure the extent of class bias or representational distortion, the popular sectors are understood to comprise those who have not finished high school (educación secundaria).19 This threshold implies that the popular sectors include an estimated percent of the study population across our four countries; among the rest of the population, termed the “middle class,” percent have finished high school but have no college education, and percent have at least some college education. It should be borne in mind that this approach to defining the popular sectors may provide a conservative estimate of the size of this group in each country; however, an alternate threshold, in which those who have finished high school but have no further education are included in the popular sectors, does not alter the conclusions discussed below.20 The distribution on the education variable is also broadly similar across countries, which helps to bolster cross-national comparability. The evidence suggests that the popular sectors are significantly less likely to engage in virtually the entire range of direct action than the middle classes (table .).21 This observation is perhaps unsurprising for institutionalized forms of direct action, given the widely asserted relationship between socioeconomic status and access to the state in Latin America (O’Donnell , Kurtz , B. Schneider ); the representative distortion involved in petition signing, where more than percent of middle-class citizens but less than percent of the popular sectors participate, mirrors patterns in both the developing and developed world.22 Strikingly, however, this class bias in the use of 19. Students finish high school (educación secundaria) after twelve years of education in Chile and eleven years in Peru and Venezuela. Constructing the threshold based on academic degree rather than years of education helps to obviate problems of comparability across countries. 20. Indeed, including those who have finished secondary education but have no further education in the popular sector only sharpens many differences between the popular sector and the middle classes (now, those who have some college education)—presumably because the latter becomes a more homogenous group. 21. See the note to table .. 22. For instance, Rosenstone and Hansen (:) find that percent of blacks and percent of whites had signed petitions, on average, from to in the United States. Rosenstone and Hansen (:) also point out a very large and negative effect of three educational categories (less than nine, nine to twelve, and more than twelve years of schooling).
04 Chapter 4.qxd
11/10/2009
.
9:44 AM
Page 110
Table . Incidence of Direct Action (percent) Atomized Direct Action (A)
Institutionalized
Contentious
Collective Direct Action (B)
Popular Sectors
Middle Class
Popular Sectors
Middle Class
Contacted Government
11.22 (0.62)
12.82 (0.58)
4.03 (0.38)
6.68 (0.43)
Engaged in Legal Process
4.08 (0.39)
5.58 (0.40)
2.01 (0.27)
3.74 (0.33)
Contacted a Political Party
2.91 (0.33)
2.40 (0.27)
1.91 (0.27)
3.28 (0.31)
Contacted an Influential Intermediary
0.92 (0.19)
1.22 (0.19)
0.35 (0.11)
0.55 (0.13)
Signed a Petition
N/A
N/A
12.60 (0.65)
20.46 (0.70)
Participated in a Protest
N/A
N/A
10.21 (0.59)
12.89 (0.58)
Note: Percent estimates are calculated using sampling weights for the pooled sample. Standard errors are in parentheses. Signing a petition and participating in a protest are coded as collective forms of direct action, which is why N/A appears in the columns for atomized direct action (see table . and accompanying text). The middle classes are significantly more likely than the popular sectors to have participated in each form of direct action, with the exception of contacting an influential intermediary (in either atomized or collective forms) and contacting a party in an atomized form; for these latter forms of direct action, the difference is not significant. The estimated standard error for this difference is the square root of the sum of the estimated variances for each percent estimate.
channels for interest articulation also extends to forms of direct action that may be assumed to be more closely associated with the popular sectors— namely, protest. The middle classes engage at statistically higher rates than the popular sectors even in this form of claim-making activity. Further observations may be made about class and atomized or collective forms of direct action. It was noted above that in the aggregate, collective direct action is more prevalent than is atomized direct action (figure .). Table . reveals that this relatively high participation in collective direct action is due to the high frequencies of petition signing and protest; among all modes of direct action that can be engaged in either atomized or collective fashion— that is, all modes except petition signing and protest—atomized participation is more common. The difference between the percent engaged in atomized and collective direct action, for all such activities, is positive and statistically
04 Chapter 4.qxd
11/10/2009
9:44 AM
Page 111
,
significant, and this holds for both the popular sectors and for the middle classes; the only exception is that middle-class (but not popular-sector) citizens are significantly more likely to have contacted a political party collectively.23 On average, among actions that can be undertaken either in atomized or collective fashion, citizens report having engaged in atomized direct action roughly twice as frequently as collective direct action. Up to this point, this section has concentrated on the incidence of different forms of direct action. Probing the perceived efficacy of direct action is also valuable, because it affords a more subjective window on the capacity of different modes of interest intermediation to function as representational vehicles. Efficacy is measured here by exploiting a survey question asking respondents who had engaged in each form of direct action how much that action, when last they undertook it, had helped to resolve the problem in question; those who answered “a lot” or “some” were coded as having engaged in a claimmaking action that was perceived to be effective. The results in table . suggest several observations. In terms of perceived effectiveness, contacting an influential intermediary ranks highest; second comes legal action, which is seen to be especially effective in its collective forms. Protest and petition signing are also perceived as relatively effective, as is contacting a party (but more so in its collective forms). Strikingly, the point estimates suggest that contacting the government is perceived to be the least effective means for advancing a claim through direct action (with the possible exception of claims advanced in an atomized fashion by middle-class citizens). In other words, citizens who have engaged in what might be thought of as political problem solving par excellence—contacting government officials— are the least likely to rate their direct action as useful. In addition, a comparison of tables . and . suggests several disconnects between the incidence of different forms of direct action and their perceived effectiveness. For instance, notwithstanding the greater incidence of atomized direct action (among claims that could be made either collectively or in an atomized fashion), collective claims on the state are seen as significantly more 23. In table ., standard errors for the differences in the percentages across columns A and B (within classes) may be calculated as follows. First, treat subsample groups (e.g., the popular sectors or the middle classes) as a simple random sample of these groups from the population. Then, conditioning on each group, the sample proportions have a trinomial distribution with parameters p (the proportion of the group engaged in individual action), q (the proportion engaged in collective action), and z (the rest of the population, who have done neither), with p + q + z = . The estimate of the difference in rates of individual and collective action for each kind of activity is provided by the sample quantity pˆ – ˆq, and the estimated variance of this difference is: ‒n¹ [pˆ + ˆq – (pˆ – ˆq ) 2]. The reported standard error is the square root of this estimated variance.
04 Chapter 4.qxd
11/10/2009
.
9:44 AM
Page 112
Table . Perceived Efficacy of Direct Action (percent) Atomized Direct Action (A)
Contacted Government Contacted a Political Party Contacted Influential Intermediary Engaged in a Legal Process Participated in a Protest Signed a Petition
Collective Direct Action (B)
Popular Sectors
Middle Class
Popular Sectors
Middle Class
39.28 (2.90) 39.45 (5.77)
40.39 (2.42) 37.49 (5.58)
43.56 (2.81) 53.58 (7.16)
48.36 (3.43) 55.70 (4.86)
55.00 (10.98) 43.05 (4.89)
58.13 (7.89) 50.38 (3.74)
62.32 (17.03) 59.24 (7.06) 52.15 (3.15) 48.67 (2.91)
63.11 (11.62) 60.66 (4.54) 54.49 (2.50) 50.74 (2.09)
N/A
N/A
N/A
N/A
Note: See notes for tables .–..
effective; this trend is evident both among the popular sectors and the middle class. The estimated difference between the perceived effectiveness of atomized and collective claims is substantively large and negative; pooling across classes, it is statistically significant for all categories of direct action.24 For all categories of collective claim making, a majority or near majority of those who had made a claim collectively reported that the action was effective, while for several kinds of atomized action, only a minority reported such satisfaction. The same kind of disconnect between incidence and efficacy emerges across different types of direct action: there is relatively less participation in forms of direct action that citizens report as most effective, and conversely, citizens primarily engage in—or feel able to engage in—forms of action that are considered least effective. For instance, contacting an influential intermediary is viewed as more effective than making other kinds of claims, yet as seen above, it is also the least prevalent.25 We cannot gain further leverage with these data 24. When we disaggregate by class, as in table ., the sample size is reduced, resulting in lowerpower statistical tests; within the middle-class group and within the popular sector, the estimated difference between the perceived effectiveness of atomized and collective action is statistically significant at standard levels only for some of the categories of direct action, though the sign of the difference is negative for all kinds of direct action. 25. The difference is significantly different from zero only when we pool across social classes. Sample sizes are small: since the question involves effectiveness of a given claim conditional on having
04 Chapter 4.qxd
11/10/2009
9:44 AM
Page 113
,
on why this might be the case; it could be, for instance, that only relatively well-connected people are able to contact influential and effective intermediaries. This possibility may suggest that clientelism may be a more pervasive way for patrons—particularly political patrons—to build support than it is for citizens, on average, to benefit. That is, the fundamentally asymmetrical patron-client relationship, especially for the popular sectors on an individual basis, may take place primarily through the initiative of the patron and may not be one the client can easily initiate. The goal here is not to estimate the causal effect of contacting influential intermediaries but rather to describe the rates at which such channels for interest intermediation are actually employed and to probe the way in which participants evaluate their actions in terms of a subjective sense of their effectiveness in the interest arena. Like the finding that atomized action is more frequent but is viewed as less effective than collective action, these results raise intriguing questions for future analysis. Thus far the question of who engages in different forms of direct action has been assessed by cross-tabulating rates of participation by country and by social class. Yet it will also be useful to explore a richer set of individual-level covariates through multivariate analysis, estimating a logistic regression model, with data pooled across countries, in which the dependent variable is participation in any form of atomized or, alternatively, collective direct action. This strategy will also allow us to assess the independent effects of class and other factors. Of particular interest are factors such as social capital, organizational capacity from past involvement in popular organizations, current immersion in social networks, or extensive neighborhood ties.26 Two indicators of past involvement in collective action and organizational life (other than participation in programmatic associations) are particularly germane to popular-sector experience: involvement with a union at any point in the past and having participated in an urban land invasion.27 These forms of organization and collective action were particularly important prior to the recent proliferation of associations and the emergence of the A-Net: unions, of course, were the primary made a claim at all. Table . indicates that the proportions making each kind of claim are small and especially within social classes; for instance, less than one percent of the popular sector has contacted an influential intermediary for purposes of accessing an entity of the state. 26. These variables might, of course, be particularly important for predicting associational participation, in addition to direct action; the next section considers this question and compares results to those presented in this section. 27. The survey question regarding unions asked respondents: “at any moment in your life, have you belonged to a union or an occupational association [asociación gremial]?” The survey questionregarding land invasion asked: “Did you come to live in this place [your current residence] by participating in a land invasion?”
04 Chapter 4.qxd
11/10/2009
.
9:44 AM
Page 114
popular organizations of the -Hub, and starting in the s, land invasions were important means of claiming land for settlement in the popular neighborhoods where much of the population of Latin America’s capital cities now lives. Past union participation is a type of collective action through relatively formal organizations, while land invasions occur through informal networks (though informal invasion groups often subsequently formed the basis of more formal neighborhood associations in the new settlements).28 Current immersion in social networks is measured with a variable tapping involvement (in the past five years) in nonprogrammatic church-based/religious or recreational associations. Finally, neighborhood ties are proxied by the time in current residence. The analyses include controls for gender, age, and the square of age (intended to capture a possible nonmonotonic relationship between age and participation); because of the very different average rates of participation across countries (see figure .), country dummies are also included.29 The results (table .) confirm the above suggestion that class matters for direct action: popular-sector members make direct claims at lower rates than the rest of the population. However, organizational involvement plays a role as well: controlling for class, past union membership significantly increases the probability of direct action, though participation in a land invasion plays no discernible role (the coefficient estimate on the latter variable is negative but not significant). Social networks also appear to matter: involvement in religious/recreational associations is a strong predictor of direct action. However, neighborhood ties, as measured by time in current home, do not play a comparable role: such ties are weakly associated with participation in atomized direct action, and they exhibit a negative relationship to collective direct action, even controlling for age. Age is positively related to participation, while its square is negatively related (people engage in direct action more as they age, and then begin to participate less); women also participate more than men, though, surprisingly, only in the case of atomized direct action.30 The evidence is thus consistent that class, as well as resources beyond those associated with socioeconomic status, appears to be important for participation in direct action: in particular, past union involvement and immersion in social networks. 28. Most land invasions are carried out in groups and in more or less organized but ad hoc form, and, of course, they clearly lack the kind of regulatory relationship to the state, much as in the distinction between formal and informal sectors of the workplace; see Dosh (); Roever (). 29. We thus exploit within-country variation to estimate the coefficients of our logistic regression models. 30. The signs of the coefficients on the country dummies simply capture the rates of atomized and collective direct action in each country relative to the omitted category of Peru; the estimated relationships are consistent with figure ..
04 Chapter 4.qxd
11/10/2009
9:44 AM
Page 115
,
Table . Covariates of Direct Action (logistic regression) Atomized Direct Action Estimated Coefficient (s.e.)
Collective Direct Action Estimated Coefficient (s.e.)
Popular Sector
-0.464* (0.068)
-0.420* (0.086)
Union
0.611* (0.110)
0.658* (0.093)
Land Invasion
-0.195 (0.165)
-0.012 -0.012
Religious/Recreational Participation
0.326* (0.084)
0.720* (0.065)
Time in Current Home
-0.004 (0.003)
-0.006* (0.003)
Age
0.099* (0.013)
0.049* (0.010)
Age Squared
-0.001* (0.000)
-0.001* (0.000)
Female
0.184* (0.083)
0.070 (0.065)
Argentina
0.961* (0.112)
0.389* (0.081)
Chile
0.715* (0.123)
-0.920* (0.110)
Venezuela
-0.208 (0.160)
-0.460* (0.111)
Note: “Programmatic” associations include the types of associations listed in the lower-right-hand cell of table .. “Popular sector”—respondent has not completed high school; “Union”— respondent has participated in a union or work association/guild (gremio) at any time in the past; “Land invasion”—respondent obtained current home through land invasion; “Time in current home”—years respondent has lived in current residence; “Religious/recreational participation”—has participated in a church-based/religious or recreational association. Other variables are self-explanatory. A constant is estimated but not reported. *Significant at the . level.
As often is the case, the empirical findings are open to various interpretations. It may well be that experience in unions or religious/recreational associations plays a causal role in facilitating direct action by, for example, providing members with social or material resources, organizational know-how, or human capital. For instance, it is plausible that a sense of empowerment or a belief in the efficacy of activism might arise from a history of participation in
04 Chapter 4.qxd
11/10/2009
.
9:44 AM
Page 116
unions.31 On the other hand, people who selected into unionized workplaces in the union-party hub might be different from those who did not in (partially unobservable) ways, which may also drive propensities to engage in direct action today. Still, the fact that unionized status tended to be extended to all similarly situated workers as part of their jobs (i.e., union membership was not extended on a selective basis to some workers and not others on the same shop floor) might belie this interpretation. Experience with collective action in the old interest regime may indeed play a role in facilitating claim making in the interest arena. In sum, the evidence in this section shows that the popular sectors engage in direct action at lower rates than the middle classes, and this holds for both atomized and collective direct action, though the differences may not always be substantively large (see table .). Some forms of direct action, such as contacting influential intermediaries, are rare, while other forms, such as contacting state officials, signing petitions, and engaging in protests, appear quite robust. There is also substantial variation across countries in the degree to which citizens engage in direct action and in the kinds of direct action in which they participate. The finding that collective direct action is relatively uncommon among the subset of institutionalized direct actions that may be either atomized or collective suggests that in order to understand how individuals pursue collective activities for the purpose of solving political problems, we must turn from focusing on collective direct action, and instead focus on associational participation. Despite its perceived effectiveness, collective direct action is not routinized in the same way as associational participation: it is by nature a low-incidence activity. Given the vastly greater incidence of associational participation relative to direct action (see figure .), it is plausible that associations provide an important alternative vehicle through which citizens achieve representation. The next section therefore turns to an investigation of associational participation.
Associational Participation As in the analysis of participation in direct action, this section investigates the incidence of associational participation both within and across countries, with 31. See Dosh (), Galiani and Schargrodsky (), and Di Tella, Galiani, and Schargrodsky () concerning the effects of land-titling after land invasions on various political and economic variables. For instance, Di Tella, Galiani, and Schargrodsky () use a natural experiment to investigate the effects of land titling in Argentina, finding that property awards to land invaders influenced beliefs about the rewards to individual effort.
04 Chapter 4.qxd
11/10/2009
9:44 AM
Page 117
,
special attention to patterns of participation by class. Since associations may engage in a wide variety of problem-solving activities—from state-targeted claim making to society-targeted provisioning of goods or services—this section also identifies the kinds of interests that are pursued through associational participation. As throughout the text,“associational participation” refers principally to participation in the programmatic associations listed in table .. However, a host of other associations, such as church-based/religious or recreational groups, may also draw the participation of many people. Although these are nonprogrammatic associations and as such are not included in the interest arena as conceptualized in this project, it is sometimes suggested that such groups can provide social ties or networks that may facilitate participation in the programmatic associations of primary interest here. For that reason, participation in such church-based/religious or recreational associations is also analyzed here. Table . presents a snapshot of associational participation, disaggregated by types of association (programmatic versus all associations) and by country.32 As suggested by figure . above, participation rates are strikingly high in Peru and low in Venezuela, with Argentina and Chile falling somewhere in between. This pattern largely holds for nonprogrammatic associations as well as the programmatic associations considered in figure .. It also holds with respect to the intensity and duration of associational participation: the proportion of citizens who participate in associations more than once a week, and the proportion participating in any association for more than five years.33 For these intensity and duration indicators, as with the aggregate participation rates, Peru scores highest and Venezuela scores lowest, while Argentina and Chile occupy intermediate positions. Though this observation accords with earlier studies of Peru, which have found a substantially high rate of associational participation in that country relative to other Latin American countries (e.g., Dietz : chap. ), the sharp contrast in participation in Peru and Venezuela is nonetheless striking.34 While much attention has focused on the rise of popular associations as a new development potentially giving new expression to popular-sector interests, 32. The data on associational participation come from survey questions that asked respondents whether, in the last five years, they had participated in the specific types of programmatic associations listed in table .. Other questions asking about participation in church-based/religious or recreational associations are also exploited. 33. Here we exploit survey questions that asked respondents how often they participate in associational activities, and also how long they have been involved in an organization. 34. Some of the vitality of associational life in Peru at the time the survey was fielded (in ) might be due to the recent fall of Fujimori, but that factor cannot explain the high percentage of Peruvians who had participated in an association for more than five years.
04 Chapter 4.qxd
11/10/2009
.
9:44 AM
Page 118
Table . Associational Participation by Country (percent) Argentina
Chile
Peru
Venezuela
27.55 (1.16)
28.26 (1.17)
54.73 (1.29)
21.91 (1.08)
> Once a Week
11.89 (0.84)
8.46 (0.72)
26.57 (1.15)
11.75 (0.84)
> Five Years
10.20 (0.91)
12.39 (1.04)
25.87 (1.55)
7.69 (0.87)
-12.18* (2.35)
-5.02* (2.21)
-1.90 (2.70)
-7.10* (2.14)
43.43 (1.29)
51.55 (1.30)
74.13 (1.14)
35.93 (1.25
> Once a Week
24.48 (1.12)
27.11 (1.16)
45.38 (1.29)
22.64 (1.09)
> Five Years
26.80 (1.53)
34.73 (1.70)
47.04 (2.10)
19.04 (1.35)
-16.67* (2.56)
-5.83* (2.64)
-2.35 (2.40)
-12.75* (2.48)
Programmatic Associations
Participation Difference (Pop. Sectors—Middle Class) Programmatic Plus Recreational or Religious Associations
Participation Difference (Pop. Sectors—Middle Class)
Note: Calculated using sampling weights appropriate for country-specific analysis. “Programmatic” associations are nonunion, nonrecreational, and nonreligious. The “Participation difference (pop. sectors—middle class)” is the difference in the percent who have participated in any nonunion association or any programmatic association, across the popular sector and middle class. Estimated standard errors for this difference condition on the group and assume a simple random sample within groups. *Difference is significant at the . level or less, based on a two-tailed z-test.
one must inquire about class-based distortion in associational participations. Do members of the popular sectors participate more or less in associations than the middle classes? On the one hand, as with direct action, resources of time and money may be important facilitators of participation, producing the positive relationship between socioeconomic status and associational participation found in studies of the developed world (Verba, Schlozman, and Brady ). On the other hand, in Latin America associational life in popular neighborhoods has often been seen as especially vibrant; recently, scholars have also analyzed the potential of a “new associationalism” to provide privileged arenas of participation and channels for representation among the popular sectors (Chalmers, Martin, and Piester ). In addition, the kinds of political connections that may favor the incidence of, especially, unmediated direct action among members of the middle class may be less important for explaining
04 Chapter 4.qxd
11/10/2009
9:44 AM
Page 119
,
associational participation, leading to an equalization of associational participation across classes. Despite expectations in the literature that associations may provide a means for especially vibrant political problem solving among the popular sectors, in fact the middle classes participate in associations at significantly higher rates than the popular sectors (fourth and eighth rows of table .). This trend can be observed at the aggregate level for programmatic associations, and it also results when nonprogrammatic associations are considered. In Argentina, for instance, the proportion of popular-sector respondents who participate in associations of any kind is an estimated seventeen percentage points lower than that of middle-class respondents; in Venezuela, the estimated difference is nearly thirteen percentage points, and in Chile, it is nearly six percentage points.35 The only exception to this pattern of class-based distortion is Peru, where the estimated difference between the percentage of popular-sector and middle-class citizens who have participated in associations, while negative, is not significant (table .).36 This initial evidence may therefore belie the suggestion that the activation and invigoration of associational participation among the popular sectors might help close the historical, class-based representational gap. What kinds of functions are served in the interest arena by the associations in which citizens participate? Specifically, do citizens perceive the associations in which they participate as making claims on the state, or are associations instead perceived as engaging in societal provisioning? In what kinds of associations do members of the popular sectors participate—and thus to what extent does the current interest regime make up for the low levels of popular-sector claim making through direct action? To answer these questions, the analysis exploits a series of survey questions that asked, of each respondent who self-identified as a participant in a given association, the extent to which that association engaged in claim-making and/ or provisioning activities. Respondents who said their association engaged “a lot” or “some” in claim-making/provisioning activities are coded as participants in claim-making/provisioning associations.37 Note that the categories are not mutually exclusive: a respondent may deem a given association in which 35. Similar results obtain with a more expansive definition of the popular sectors that includes high school graduates (not reported). 36. As we will see below, this finding also holds when we use an alternate, more expansive threshold, including high school graduates among the popular sectors. This group, including those who have not finished high school and those who have (but have no college), comprises percent of the population. 37. The two relevant questions, asked for each type of association, were worded as follows (author’s translation). Claim making: “How much does or did this [association] work to present issues to the
04 Chapter 4.qxd
11/10/2009
.
9:44 AM
Page 120
he or she participates to be a vehicle for both claim making and provisioning, or he/she might deem one association to be engaged in claim making and another as engaged in provisioning.38 Indeed, the data suggest substantial overlap: an estimated . percent of citizens participate in claim-making and provisioning associations and an estimated . percent participate in neither, while just . percent participate in one type but not the other. This observation is consistent with the finding in chapter that many associations engage in both claim making and provisioning. Notwithstanding this overlap, rates of participation in provisioning associations are markedly higher than participation rates in claim-making associations (figure .). Individuals therefore tend to participate at high rates in associations oriented toward helping the participants themselves than they do in those oriented toward making state-targeted claims. An estimated percent of the study population pooled across countries belongs to a claim-making organization, while nearly percent belongs to a provisioning association; this difference holds up and is statistically significant across all countries, though it is more pronounced in Peru and Chile than in Argentina or Venezuela.39 Other subsample analyses (not reported) show that this pattern holds up in both the popular sectors and middle classes, and also within associations of different types (see table .).40 In the aggregate, then, though associations that individuals choose to participate in are perceived to engage in substantial claim making, they tend to be seen more as vehicles for provisioning. As table . indicates, Latin Americans participate in quite high levels in church-based/religious and recreational associations. In terms of assessing participation in programmatic associations, a comparison to these other two is government that are important for you? Would you say that the [association] works or worked a lot, some, a little, or not at all [for this purpose]?” Provisioning: “And how much does or did this association help its members organize themselves to resolve important problems by their own means? Would you say that the association helps or helped a lot, some, a little, or not at all?” Note that the word “member” (socio, also partner) in this final question can be understood as applying to participants—since it follows a question in which respondents are asked if they have “participated” in the activities of the association in the previous five years (and only those who said yes were asked the follow-up question). 38. In the latter case, the respondent would be coded as a participant in both claim-making and provisioning organizations (though for purposes of calculating associational participation rates, the respondent is coded as participating as long as he or she has participated in at least one association, and otherwise is a nonparticipant). 39. To calculate standard errors and thus perform significance tests, we used an approach similar to the one described in connection with table . above. 40. Interestingly, although food, protection, parents’, women’s, and place-of-origin associations and cooperatives are more likely to be perceived as engaging in self-provisioning, neighborhood and church-based/religious associations are only slightly less likely to be identified as claim making (as opposed to provisioning) groups.
04 Chapter 4.qxd
11/10/2009
9:44 AM
Page 121
,
60 50
Associational Participation Collective Direct Action Atomized Direct Action
40 30 20 10 0 Argentina
Chile
Peru
Venezuela
Fig. .. Participation in Self-Provisioning and Claim-Making Associations (percent)
instructive. Pooling across countries, participation in programmatic associations (as an aggregated category) reaches an estimated percent in total: a rate that far outpaces individual participation in either church-based/religious or recreational associations.41 On the other hand, no single type of programmatic association has estimated participation rates over percent, which suggests that different individuals choose to participate in distinct types of associations (the “breadth” of individual associational participation is analyzed below). In addition, confirming the aggregate finding above, table . shows that the popular sectors participate at substantially lower rates than the rest of the population in nearly all types of associations.42 The lower rates of participation among the popular sectors is all the more striking, given that the surveys 41. It may be instructive to compare the estimates in table . to a set of higher estimates for the United States. Verba, Schlozman, and Brady (: table .) estimate affiliation rates in the United States of percent for unions, percent for recreational (hobby, sports, leisure) associations, percent for neighborhood organizations, and percent for women’s rights groups (see n. regarding this latter category). On the other hand, the estimated rates of participation in food distribution and parents’ associations here— percent in both cases—may be quite high in comparative perspective, as is the estimate of percent involvement in church-based/religious groups in the present analysis (compare Verba, Schlozman, and Brady’s estimate of percent in the United States, from a survey taken in ). 42. Exceptions may be unions and food and women’s associations, where the difference between popular sectors and the middle classes is not significant. It may be worth noting that these women’s groups may or may not make claims on the state; they may be groups of women that engage in provisioning as they approach a whole range of problems from health, education, food, children and child care, and so on. Thus, they may not be strictly comparable to the women’s rights groups discussed in n..
04 Chapter 4.qxd
11/10/2009
9:44 AM
Page 122
Table . Incidence of Associational Participation (percent) Total Sample
Popular Sectors (A)
Middle Classes (B)
Difference (A–B)
TOTAL
35.93 (0.62)
30.92 (0.90)
39.92` (0.85)
-9.00* (1.24)
Neighborhood
13.45 (0.44)
11.80 (0.63)
14.77 (0.62)
-2.97* (0.84)
Protection
8.86 (0.37)
5.81 (0.46)
11.29 (0.55)
-5.48* (0.72)
Food
13.28 (0.44)
12.66 (0.65)
13.78 (0.60)
-1.12 (0.88)
Parents
13.27 (0.44)
11.32 (0.62)
14.83 (0.62)
-3.51* (0.87)
Women
2.11 (0.19)
1.79 (0.26)
2.36 (0.26)
-0.57 (0.37)
Place-of-Origin
2.18 (0.19)
1.59 (0.24)
2.65 (0.28)
1.06* -(0.37)
Cooperative
3.22 (0.23)
1.64 (0.25)
4.48 (0.36)
-2.84* (0.44)
Unemployeda
0.85 (0.17)
1.13 (0.28)
0.60 (0.20)
0.53* (0.25)
Union
12.42 (0.43)
12.00 (0.63)
12.75 (0.58)
-0.75 (0.86)
Church/Religious
21.52 (0.53)
20.36 (0.79)
22.44 (0.73)
-2.08* (1.06)
Recreational
21.17 (0.53)
15.02 (0.70)
26.07 (0.76)
-11.05* (1.88)
54.00 (0.65)
46.48 (0.97)
60.00 (0.85)
-13.52* (1.29)
Programmatic
Nonprogrammatic
Programmatic and Nonprogrammatic Associations, Excluding Unions
Note: Categories are nonexclusive, so sums across the types of associations (i.e., down the columns) may exceed the percent of respondents who have participated in aggregate categories of associations (i.e., the first row and the last). Estimates are calculated using sampling weights for the pooled sample. *Sigificant at the . level. a
This question was asked in Argentina and Chile only. Country-specific sampling weights are used for estimates reported in this row.
04 Chapter 4.qxd
11/10/2009
9:44 AM
Page 123
,
asked about many types of associations that presumably are of particular relevance to the popular sectors (e.g., food distribution groups, cooperatives, associations of the unemployed) and did not include the types of “postmaterialist” associations (e.g., environmental associations) for which a middle-class bias might plausibly be anticipated.43 Again, this evidence challenges the view that associations might play a privileged role in the representation of the lower classes in the contemporary interest regime. I now turn to the individual correlates of associational participation. Is social class in fact the major predictor, or do other variables play an important role? Do other individual-level factors better predict associational participation, controlling for class? As with direct action, this question is pursued through a multivariate analysis, estimating a logistic regression model that includes as covariates variables such as past organizational links, experiences with collective action, immersion in social networks, and neighborhood ties; the proxies for these variables are union involvement at any time in the past, gaining one’s residence through land invasion, participation in religious/recreational associations, and time in current home. The logical relation of these variables is the same for associational participation as for direct action; in addition, however, they may be intimately linked to associational life in other ways. For instance, informal land-invasion groups sometimes formed the basis of more formal neighborhood associations in the new settlements, a type of association included in this study and a primary type that existed in the periphery of the -Hub. Because of the potential importance of recreational and religious associations as vehicles for entry into programmatic associations, we also analyze predictors of participation in an aggregate category that includes both programmatic and nonprogrammatic associations (in which case we obviously drop religious/recreational participation as a covariate). All analyses control for gender, age, the square of age (intended to capture a possible nonmonotonic relationship between age and participation), and country dummies. As anticipated by the cross-tabulations above, the results suggest that members of the popular sectors participate at significantly lower rates than the middle class, even controlling for other variables (table .). Strikingly, however, the apparent effects of previous involvement in forms of collective action (i.e., 43. As above, separate analysis (not shown) confirms that these results persist when employing the alternate, less restrictive measure of the popular sectors. Indeed, when we include high school graduates among the popular sectors some of the results are strengthened, presumably because the higher-education group becomes more homogenous: for instance, the popular sectors participate at substantially and significantly lower rates across every type of association, both programmatic and nonprogrammatic.
04 Chapter 4.qxd
11/10/2009
.
9:44 AM
Page 124
past union participation and/or participation in land invasion) and of participation in nonprogrammatic associations are substantially stronger than the effect of class, as the calculations of marginal effects reported in table . show. For instance, while being a member of the popular sectors decreases the probability of participating in a programmatic association by . percentage points (when other variables are set at their median values), having been a union member is associated with an increase in the probability of associational participation by . percentage points.44 Past participation in a land invasion is associated with an increase in the probability of associational participation of .. Finally, participation in non-programmatic associations (religious or recreational) increases the probability of participation in programmatic associations by . percentage points.45 The empirical findings are again open to various interpretations. Past organizational links (as proxied by past involvement in unions), experiences with collective action, and immersion in social networks through recreational or church-based/religious associations may have a causal effect on programmatic participation in various ways. On the other hand, some people may be “joiners” who have unobserved sources of propensities to participate, which may drive involvement with unions, land invasion groups and nonprogrammatic associations as well as the programmatic associations in which we are interested. Estimating causal effects is not so much the interest here as describing the types of individuals who participate in the interest arena through associational participation. We may want to consider here an additional measure: the “breadth” of associational participation, or the number of different kinds of programmatic associations in which respondents report having participated. The mean of this variable is just over one-half (.), with a minimum of zero and an empirical maximum of six (out of a theoretically possible eight programmatic associations in the surveys). By itself, this evidence does not strongly support the abovementioned possible interpretation of spuriousness—that results in the data 44. In tables . and ., the baseline individual (who has an estimated probability of associational participation of percent) is not a member of the popular sector; has never been a member of a union; did not get her house through a land invasion; is thirty-eight years old; is female; has lived in her current home for years; is Peruvian; and does not participate in a religious or recreational association. (Obviously, for purposes of calculating the “marginal” effect of being female, the baseline individual is instead assumed male, and so forth). 45. Note, however, that not all measures of social networks or ties appear important. For example, the indicator of neighborhood ties, “time in current home,” does not have a statistically significant relationship to associational participation, controlling for the other variables in the specification. Interestingly, this finding persists if we take as the dependent variable participation in neighborhood associations alone (not reported).
04 Chapter 4.qxd
11/10/2009
9:44 AM
Page 125
Table . Covariates of Associational Participation (logistic regression) Programmatic Associations
Programmatic, Religious, and Recreational Associations
Estimated Coefficient (s.e.)
Estimated Marginal Effect (s.e.)
Estimated Coefficient (s.e.)
Estimated Marginal Effect (s.e.)
Popular Sector
-0.244* (0.084)
-.061* (0.021)
-0.290* (0.061)
-0.057* (0.016)
Union
0.774* (0.127)
0.178* (0.026)
0.752* (0.092)
0.112* (0.016)
Land Invasion
0.294* (0.130)
0.072* (0.031)
0.356* (0.106)
0.060* (0.022)
Age
0.105* (0.014)
0.026* (0.003)
0.033* (0.009)
-0.006* (0.002)
Age Squared
-0.001* (0.000)
-0.000* (0.000)
-0.001* (0.000)
N/A
Female
0.344* (0.083)
0.086* (0.021)
0.009 (0.059)
0.002 (0.014)
Time in Current Home
-0.004 (0.003)
-0.001 (0.001)
-0.001 (0.002)
-0.000 (0.001)
Argentina
-0.988* (0.113)
-0.236* (0.026)
-1.132* (0.077)
-0.257* (0.024)
Chile
-1.068* (0.105)
-0.253* (0.024)
-0.849* (0.088)
-0.186* (0.022)
Venezuela
-1.227* (0.105)
-0.284* (0.024)
-1.464* (0.095)
-0.339* (0.024)
Religious/ Recreational Participation
0.831* (0.082)
0.189* (0.018)
N/A
N/A
Note: “Programmatic” associations are nonreligious and nonrecreational associations. “Popular sector”—respondent has not completed high school; “Union”—respondent has participated in a union or work association/guild (gremio) at any point in the past; “Land invasion”—respondent obtained current home through land invasion; “Time in current home”—years respondent has lived in current residence; “Religious/recreational participation”—has participated in a churchbased/religious or recreational association. Other variables are self-explanatory. A constant is estimated but not reported. *Significant at the . level.
04 Chapter 4.qxd
11/10/2009
.
9:44 AM
Page 126
could be driven by the influence of “joiners.” If joiners drive the relationships between the variables in table ., we might expect to observe that important proportions of respondents had joined several different kinds of associations. However, while an estimated percent of citizens have not participated in any programmatic associations and percent have participated in only , just , , and percent of respondents have participated in two, three, and four associations, respectively.46 Using the breadth of associational participation as the dependent variable, the multivariate results are qualitatively similar to those for the incidence of participation in any association (i.e., the dichotomous dependent variable for table .). Table . reports results of estimating Poisson regression models; in the left two columns, the dependent variable is the breadth of participation in programmatic associations, while in the right two columns, church-based/ religious and recreational associations are added. Even controlling for class, some of the variables tapping past collective action and immersion in social networks through religious and recreational participation are robustly associated with the breadth as well as the incidence of associational participation. In sum, the evidence in this section suggests that associations play an important role in individual participation in the interest arena. The incidence of associational participation outstrips collective or atomized direct action. While associations are most often viewed as vehicles for helping members and participants organize to solve collective problems by their own means, they are also seen as engaging in state-targeted claim making. However, the incidence of claim making via associations outstrips the incidence of collective direct action in three of the four countries under consideration, the exception being Argentina. The exception of Argentina points again to the fact that beyond the commonalities suggested in the analysis of pooled data, country-specific factors appear to matter a lot, perhaps indicating the importance of sociotropic factors or broad differences across countries in the character of the popular sectors. Peru, in particular, stands out in terms of high levels of associationalism. This analysis has largely not sought to explain this variation; however, we have been careful to note and control for this cross-national variation where desirable. Some further thoughts are suggested in the concluding chapter, but the current chapter has focused on exploring this cross-national variation with systematic individual-level data, which scholars have lacked in previous analyses. Finally, the evidence suggests that the class bias in descriptive representation 46. An estimated . percent of citizens have joined five associations, while just one respondent has joined six.
04 Chapter 4.qxd
11/10/2009
9:44 AM
Page 127
,
is not less for associational participation than it was for direct action; associations thus do not provide a privileged or offsetting site for popular-sector mobilization and representation. Aside from class, variables tapping organizational experience with collective action and immersion in social networks are important predictors of associational participation. The analysis thus points to the important role of individual resources, both those tied directly to socioeconomic status and those of a more organizational nature, in promoting involvement in the interest regime through associational participation. Table . Breadth of Associational Participation (Poisson regression) Number of Programmatic Associations
Number of Programmatic, Religious, and Recreational Associations
Estimated Coefficient (s.e.)
Estimated Marginal Effect (s.e.)
Estimated Coefficient (s.e.)
Estimated Marginal Effect (s.e.)
Venezuela
-0.204* (0.059) 0.439* (0.075) 0.238* (0.072) 0.079* (0.010) -0.001* (0.000) 0.260* (0.056) -0.003 (0.002) -0.663 (0.009) -0.778* (0.073) -0.784* (0.079)
-0.166* (0.046) 0.495* (0.107) 0.242* (0.078) 0.071* (0.011) -0.001* (0.000) 0.206* (0.044) -0.003 (0.002) -0.436* (0.053) -0.486* (0.050) -0.488* (0.053)
-0.148* (0.040) 0.986* (0.043) 0.197* (0.054) 0.024* (0.006) -0.000* (0.000) 0.023 (0.037) -0.001 (0.001) -0.571* (0.052) -0.489* (0.046) -0.767* (0.055)
-0.211* (0.055) 2.577* (0.183) 0.334* (0.096) 0.037* (0.010) -0.001* (0.000) 0.036 (0.056) -0.001 (0.002) -0.667* (0.059) -0.593* (0.058) -0.821* (0.060)
Religious/ Recreational Participation
0.537* (0.055)
0.639* (0.070)
N/A
N/A
Popular Sector Union Land Invasion Age Age Squared Female Time in Current Home Argentina Chile
Note: Definitions are in table .. A constant is estimated but not reported. *Significant at the . level.
04 Chapter 4.qxd
11/10/2009
.
9:44 AM
Page 128
Gender and Participation in the Interest Arena Before concluding, I turn to the issue of gender, not systematically analyzed thus far in this chapter. Latin America’s large-scale shift from union-party hub to associational networks may have provided new opportunities as well as challenges for the articulation of issues of concern to women and, more generally, for the mobilization of women in political problem-solving activities of various sorts. In the context of the re-democratization of much of the region during the s and s, some analysts saw a replacement of a hierarchical, “masculine” model of interest representation with a decentralized, horizontal, and more “feminine” paradigm; according to this view, the reduced salience of traditional parties and male-dominated formal-sector unions gave rise to new opportunities for grassroots organizations, including those with disproportionately female memberships, to fill the gap (Buvinic and Roza :).47 This “feminization” of participation in the interest arena may have also given an especially heightened role to women from the popular sectors. Jaquette (:), for example, emphasizes that “the mass of women’s movements from the mid-’s to the present has been lower-class urban women (mujeres populares) who organized to demand relief from the state, supply the basic services the state could no longer provide, or feed their families collectively when it was no longer possible to do so individually.” Our data give us unique empirical leverage to investigate the role of women in the contemporary interest arena. Several conclusions stand out. First, women do participate more in the interest arena than men, both via associations and through direct action. Simple cross-tabs (not shown) confirm the relationship between gender and participation; in the multivariate regressions reported above, the “female” dummy variable is a positive and statistically significant predictor of atomized direct action as well as the incidence and breadth of associational participation (tables ., ., and .). Second, the descriptive bias in favor of female participation is most pronounced for programmatic associations; when we include nonprogrammatic (religious and recreational) as well as programmatic associations, we find that men and women do not participate at statistically distinguishable rates (tables . and .). In other words, the disproportionate participation of women in the A-Net is geared toward 47. One could also point to the increasing prominence of formal-sector teachers’ unions, which tend to have disproportionately female membership if not necessarily leadership; according to Education International (), over percent of teachers in Latin America are female, though on average women occupy fewer than percent of leadership posts in teachers’ unions.
04 Chapter 4.qxd
11/10/2009
9:44 AM
Page 129
,
programmatic problem solving.48 Finally, however, despite the greater mobilization of women through programmatic problem solving, the preponderance of women’s participation does not lie in women’s associations; just . percent of respondents, and . percent of women, had participated in a women’s group, making this one of the least common forms of associational participation (table .). By way of comparison, participation in women’s groups is dwarfed by participation in neighborhood, parents’, or food associations. In sum, our evidence is consistent with the idea that women do play a disproportionate role in the contemporary A-Net. The greater participation of women is also especially oriented toward programmatic problem solving. Of course, this greater descriptive representation of women in the interest arena does not guarantee a more effective articulation of “women’s” interests; nor is women’s participation necessarily geared toward the advancement of a problemsolving agenda framed in terms of women’s issues. The evidence does suggest, however, that the disruption of the traditional modes of interest representation associated with the -Hub has heightened the relative salience of female participation in the interest arena.
Conclusion Various observers have commented on the importance of associational networks in contemporary Latin America as well as the apparent transformation of other structures of representation, in the wake of market reforms and the decline of robust party-union linkages (e.g., Chalmers, Martin, and Piester ). However, we have lacked a systematic, comparative investigation of patterns of individual participation in the interest arena. The present chapter has allowed us to probe different modes of individual problem solving and in particular to compare patterns of direct action with associational participation. Several assertions can be made. First, the incidence of associationalism is quite high across the four countries studied; Peru, in particular, stands out for its high levels of associational participation. In three out of four countries, associational participation exceeds either atomized or collective direct action as a mode of political problem solving, suggesting associations play an important role in the interest arena. Citizens choose to participate in associations they see as particularly active in provisioning, those that help members and participants organize to solve problems through their own means. At the same time, individuals do participate in associations that are active in making claims 48. However, note that gender is not a significant predictor of collective direct action (table .).
04 Chapter 4.qxd
11/10/2009
.
9:44 AM
Page 130
on the state; the fact that collective direct action is more infrequent than associational participation may suggest that associations provide citizens an alternate or substitute means of collective claim making. Second, however, direct action constitutes an important dimension of the contemporary interest arena as well. Collective direct action is perceived as particularly effective by those who engage in this kind of problem solving, relative to atomized direct action—though interestingly, respondents were more likely to engage in claim making through atomized than collective means, among the forms that could be engaged in either atomized or collective fashion. Such particularistic contacting tends to take the form of contacting the government directly and, secondarily, contacting parties or engaging in legal processes. Collective direct action is largely confined to protest and petitioning (though in Peru, citizens do contact the state collectively at high rates). Thus, the following general conclusions about direct action find support in the data: () direct action is common and viewed as effective by people who engage in it; () overall, collective direct action is more common than atomized direct action; and () among types of direct action that can take an atomized or collective form, action tends to be atomized. Perhaps surprisingly given Latin America’s reputation for high levels of clientelism, contacting influential intermediaries for purposes of accessing the state is actually quite uncommon (though the relatively small number of respondents who had done so reported that this avenue of problem solving was highly effective). Like the contacting of political parties, the incidence of this form of intermediated direct action is also far outstripped by several forms of unmediated direct action. Of course, this observation does not necessarily contradict analyses that emphasize clientelism’s political role: intermediated direct action may play a crucial role in the strategies of parties for linking with certain types of supporters, even if on average a relatively low proportion of citizens contact the state through influential intermediaries. Moreover, the analysis in this chapter also suggests an important role for dirigentes or influential leaders in abetting forms of direct action we have conceptualized as unmediated, such as participation in protests. Such forms of mobilization by local brokers and influential leaders may be one important channel through which clientelist politics take effect in contemporary Latin America (see, e.g., Szwarcberg on the case of Argentina). It is also important to emphasize that there are large cross-national differences in the incidence of political problem solving. In general, Peru stands out for having the highest participation rates, while Venezuela generally has the lowest; Argentina and Chile are generally positioned in between. The chapter’s
04 Chapter 4.qxd
11/10/2009
9:44 AM
Page 131
,
focus on the individual-level correlates of associational participation and direct action across all four countries should not obscure these basic cross-national differences, which, again, underscore the importance of sociotropic factors in shaping participation. The multivariate analyses of associational participation and direct action suggest that, consistent with a number of hypotheses, several individual-level variables predict participation, including class and other factors. Class distortion characterizes all forms of action in the interest arena, including protest. Beyond that, personal experience with other forms of collective action, specifically land invasions and union participation, appear to play an important role in influencing whether an individual participates in programmatic associations under the current interest regime. While subject to various interpretations, this evidence is consistent with the idea that resources in the form of collective action skills play a particularly important role; it may also suggest the importance of networks and social ties, as indicated by the positive relationship between participation in programmatic associations and participation in other types of civil society organizations, such as those addressing religious and recreational matters. (A related neighborhood measure, time in current residence, does not seem to influence associational participation, however.) It appears that the presence of “joiners” may not provide the best explanation for the relationship between engagement in other forms of collective action or immersion in social networks and associational participation. Perhaps one of the most important findings is that representation within the contemporary interest arena does exhibit a class bias. Members of the popular sectors are less likely to participate in associations or make direct claims than members of the middle class, a representational distortion that holds for inclusive as well as for more restrictive measures of the popular sectors. This finding is consistent with classic studies of the United States that have argued that “resources” associated with socioeconomic status, such as time, money, or skills, are the chief motors of political participation and, with respect to direct action, consistent with perspectives that emphasize the privileged access of elites and middle classes to the state. Yet in the Latin American context, where associational networks have been viewed as particularly important for representing the popular sectors in the wake of the decline of union-party linkages, the finding that the middle classes are more likely than the popular sectors to participate in associations is striking and, perhaps, discouraging. The evidence suggests that associations, while important vehicles for interest intermediation in Latin America, have some distance to travel before they will close a historical, class-based representational gap.
05 Chapter 5:05 Chapter 5.qxd
11/9/2009
11:55 AM
Page 132
5 : Jason Seawright
The political representation of the popular sectors has historically been problematic in Latin America, and much of the optimism about the flourishing of associationalism, among scholars and participants alike, has involved its potential capacity to remedy long-standing representational distortions regarding the popular sectors. Although this volume focuses on popular representation in the interest arena, representational distortions in Latin American polities may characterize both the interest and electoral/partisan arenas. Investigating the issue of popular-sector representation more broadly thus involves assessing distortions in both arenas. This dual examination offers an additional benefit in that it allows for an exploration of the nexus between the two. Literature from the advanced democracies (Verba, Schlozman, and Brady :– ) suggests that interest organizations can play a key role in fostering party/ electoral participation through political cues, recruitment networks, and skill formation. Does the A-Net serve a similar function, and with what consequences for popular-sector representation? This chapter seeks to map and explain patterns of popular-sector inclusion in various aspects of partisan politics within the new, association-centric regime of representation in South America. It asks whether urban associations form a point of linkage between the popular sectors and the world of partisan politics. How important are associations in bringing the popular sectors into politics? Do associations instead enhance the partisan representation of the middle and upper classes? Or are linkages between associations and partisan activity
05 Chapter 5:05 Chapter 5.qxd
11/9/2009
11:55 AM
Page 133
,
weak or largely disarticulated? This chapter pursues these questions by examining patterns of individual participation in both party/electoral activities and in urban associations—the programmatic or problem-solving associations that constitute the A-Net as well as other “civil society” associations, such as churchbased/religious and recreational organizations. The analysis proceeds in three steps. The first question concerns the degree to which representational distortion exists in the electoral/partisan arena. Latin America has been experiencing an extended period of democracy at a moment in world history with substantial convergence on the value of representative democracy and of participation in electoral politics. At the same time, in the context of an economic model that has delivered uneven growth and regressive returns from that growth, a pervasive sense of alienation from politics and distrust in democratic institutions has been widely noted. In this context the question of representational distortion in the electoral/partisan arena arises. To what extent do groups along the social hierarchy differentially participate in the electoral/partisan arena, and to what extent are they generally differently linked to parties? The analysis here suggests that in the electoral/ partisan arena, class-based participatory and representational distortion varies considerably across the four countries: it is very pronounced in Argentina and Chile, and considerably smaller in Peru and Venezuela. That is, the popular sectors in Argentina and Chile participate at considerably lower rates than those of higher class status, but at somewhat even rates in Peru and Venezuela. A second issue concerns representational distortions in the A-Net and other civil society associations. As chapters and note, the proliferation of associations has occurred across all classes and across the socioeconomic hierarchy. Associations have become a more important site of popular-sector participation and representation than in earlier periods. An important issue to examine then is how patterns of associational participation among the popular sectors compare with the middle classes. This chapter builds on chapter ’s analysis of the pattern of class-based representational distortion in the new associationalism by examining that pattern in greater detail through a continuous, rather than dichotomous, conception of socioeconomic stratification, and by including nonprogrammatic civil society associations beyond the A-Net in the analysis. For the most part, however, associations are considered in a disaggregated fashion, so that different types of associations (not only programmatic versus nonprogrammatic, but also different types within each group) may be compared. In general, programmatic and nonprogrammatic associations suffer from similar patterns of class distortion and have comparable patterns of mass-based connections with the world of party politics. Contrary to any
05 Chapter 5:05 Chapter 5.qxd
.
11/9/2009
11:55 AM
Page 134
expectation that the new efflorescence of organizing, the “resurrection of civil society” among the popular sectors, might offset the resource advantages of the middle and upper classes, the patterns of distortion in civil society are quite congruent with representational distortion in the electoral/partisan arena. In Argentina, participation in associational life mirrors the pattern of representational distortion of the popular sectors in party-centric participation. In Peru, the incidence of associational participation is less uneven across the socioeconomic hierarchy, again roughly matching the finding regarding party/electoral participation. Chile and Venezuela occupy a middle ground; representational distortion in associational participation thus does not explain Chile’s relatively higher level of representational distortion in the electoral/partisan arena. Nonetheless, for most possible pairings among these four countries, class patterns in associational involvement correspond with class dynamics in partisan participation. This rough macro-correlation points toward the third avenue of exploration in the chapter, the question of whether cross-national differences in distortion in the electoral/partisan arena might be accounted for by patterns of associational participation. Verba, Nie, and Kim (:–) suggest that the strength of a nation’s organizational system affects the degree to which there is class bias in political participation. Strong organizational systems are those in which political parties and civil society organizations (a) extensively mobilize political participation and (b) monopolize the major channels of participation. Verba, Nie, and Kim effectively argue that countries with strong organizational systems will have little underrepresentation of the popular sectors. This argument rests on two conditions. The first is that lower-class individuals are not underrepresented in the organizations. Otherwise, the effect would be to disproportionately mobilize the relatively higher-class population active within organizations into party-related participation, exacerbating or reinforcing distortions in electoral participation. Yet as we have seen, the correlation between the class basis of associationalism in each country and the pattern of party/electoral distortion is remarkably consistent. A second condition is that there must be evidence that interest organizations actually do mobilize participants into electoral politics. Extant literature offers some reasons to be skeptical about this proposition. As discussed in chapter , party-association relations in the A-Net are weaker and more distant than union-party relations under the -Hub, marked by both organizational integration and an exchange relationship in which unions mobilized party identification and secured party loyalty and votes among their members while parties, in turn, were a source of both benefits and controls for unions and union leaders. With the shift to
05 Chapter 5:05 Chapter 5.qxd
11/9/2009
11:55 AM
Page 135
,
the A-Net, the pattern is generally one of greater disarticulation, as associations are either autonomous or less organically tied to parties. Parties have generally changed as even labor-based parties have loosened their ties to unions and tended to adopt traits as catchall or electoral professional parties. And from the citizen side, a new orientation toward “new social movements” and “civil society” has placed greater emphasis on autonomy from the state and on independence from parties than was true under the -Hub. Yet while parties and associations may not be organizationally integrated and more organic linkages between them may be rare (see chapter ), other kinds of linkages may nevertheless exist. People with a particular political agenda may distinctively sort themselves into activism in parties that would serve that agenda and involvement in associations that are responsive to that agenda. Such self-selection would create a form of linkage based on members who are shared between organizations; these “shared-member linkages” may connect civil society and party politics. The last section of the chapter examines shared-member linkages by examining for each country the relationships between individual participation in different types of associations on the one hand, and party/electoral participation on the other. The analysis suggests that these shared-member linkages are generally robust and provides empirical evidence of a mechanism in which civil society does indeed serve a role in connecting participants with the world of electoral politics. This finding further increases the plausibility that the representational distortions in civil society may, in aggregate, help to sustain the representational distortions in the electoral/partisan arena.
Class and Partisan Engagement This chapter focuses primarily on patterns of descriptive representation in the electoral/partisan arena, exploring whether the popular sectors are included in a specific partisan activity in numbers proportional to the overall population, that is, the degree to which representational distortion exists. In a bivariate analysis, a lack of such class-based representational distortion means that there is no relationship between an indicator of social class and participation in partisan activity. A positive relationship between these two variables indicates popular-sector underrepresentation, while a negative relationship implies overrepresentation. This focus on descriptive representation is useful because it is a measurable form of objective representation, based on behaviors and social roles. Subjective
05 Chapter 5:05 Chapter 5.qxd
.
11/9/2009
11:55 AM
Page 136
forms of representation, that is, those with definitions revolving around perceptions or feelings of being represented within the party system, may be important components of people’s political experience.1 Yet subjective representation is a political attitude, and as such it may have cultural, social, political, or ideological causes outside of the actual functioning of partisan representation. Subjective representation is normatively and analytically important in its own right and is also interesting to view in relation to objective behavior. This chapter focuses centrally on objective representation but also briefly considers subjective representation. Within the realm of objective representation, the literature from advanced industrial democracies provides some reason to expect that the popular sectors will be underrepresented in South American partisan politics. This literature assumes freedom of political mobilization and voluntary engagement, an assumption that is approximately, although not necessarily fully, satisfied in the four countries included in this project. Hence, this scholarship based on advanced industrial democracies may present reasonable hypotheses about popular-sector political participation in the current neopluralist environment in South America. In the literature on advanced industrial democracies, components of social class such as income and educational attainment often emerge as useful predictors of involvement in the electoral/partisan arena. Higher levels of education are robustly associated with higher levels of participation in party politics. Education may encourage party involvement by providing an individual with social status, access to resources, cognitive skills (Nie, Junn, and Stehlik-Barry :–), civic norms, or political motivation. Indeed, the accumulated evidence suggests that, “as an explanatory variable of social and political behavior, formal educational attainment is unrivaled” (Nie, Junn, and Stehlik-Barry :; see also Converse ).2 1. Two examples of recent studies that make use of this same distinction between objective and subjective representation are Uhlaner () and Markowski and Tucker (). Objective representation also comprises less directly measurable concepts regarding whether the substantive outcomes of a decision-making process are in some sense fair or proportionate. Given the obvious problems with operationalizing this broader version of objective representation, the only form of objective representation this chapter will address involves the quality of descriptive representation in participation in civil society or partisan politics. 2. Nonformal education, such as participation in basic community literacy or numeracy courses, may have as much of a relationship with political participation as the kind of formal education emphasized in this chapter (Kuenzi ). However, the survey contains no measures of informal education. Hence, for the present, the relative importance in these four Latin American capital cities of formal education, and its associated class status, as compared with the development of basic skills and personal efficacy through either formal or informal education, must remain a matter of speculation.
05 Chapter 5:05 Chapter 5.qxd
11/9/2009
11:55 AM
Page 137
,
Income is also an established predictor of partisan participation (Verba and Nie :–; Verba, Nie, and Kim :–; Verba, Schlozman, and Brady :–), although the relationship between wealth and participation may be substantially weaker if the linear effects of education and group involvement are controlled for (Parry, Moyser, and Day :–). Verba, Schlozman, and Brady argue that income may affect participation by increasing the number of political conflicts a person is involved in and thereby enhancing the motivation for participation. They also suggest a second effect of income on participation: it increases an individual’s resources for participation, which is particularly important for their study, since their measure of participation includes activities that are based on financial resources, such as monetary contributions (Verba, Scholzman, and Brady :). These findings, mostly drawn from the analysis of North American and Western European democracies, imply that the popular sectors, characterized by lower typical incomes and levels of educational attainment than the middle and upper classes, are objectively underrepresented in the group of partisan participants. The quantitative literature on Latin America is less developed: although there is some evidence against higher socioeconomic status as a predictor of participation in Latin America, it is far from decisive, and indeed the overall picture is complex. For example, Aguilar and Pacek () present indirect evidence that voters at lower socioeconomic levels vote less often, in comparison with wealthier voters, when the economy is performing well. While this result does not decide the question of whether socioeconomic status predicts overall participation, it does at least suggest that decisions about participation are made differently depending on such status. Other studies, with a focus on a single country, city, or even set of neighborhoods, have produced findings that suggest a negative relationship for Latin America between social class and political participation. For a Mexican sample comparing formal- and informal-sector workers engaged in similar occupations, Thornton () finds that informal-sector workers vote and protest more often than do formal-sector workers. Hence, for one very specific classrelated distinction in Mexico, there is some evidence of a connection between (arguably) lower socioeconomic status and political participation. Two excellent studies of participation (Stokes ; Dietz ) within generally lower-socioeconomic neighborhoods of Lima contribute evidence that poorer individuals in Latin America can often participate in politics at fairly high rates. However, because the research strategies in these two projects do not formally include higher-socioeconomic areas, it is difficult to use them as
05 Chapter 5:05 Chapter 5.qxd
.
11/9/2009
11:55 AM
Page 138
the basis for cross-class comparison or generalization about relationships between class and participation. Fornos, Power, and Garand () have shown that, in a comparison of aggregate Latin American turnout levels across countries and over time, there is little relationship between socioeconomic status and electoral turnout—a key indicator of political participation. Yet this finding is in fact compatible with the possibility that there is a very strong relationship between socioeconomic status and activity among voters in any given election (or among potential actors in some other form of political participation). Indeed, such has been the pattern over time in the United States, where education levels have risen as voter turnout levels have fallen—even though education remains a singularly powerful predictor of individual turnout (Converse ; Brody ). Hence, socioeconomic status remains a plausible predictor of political participation within Latin American societies. There remains a need for systematic, individual-level comparison of the political participation of lowerand higher-socioeconomic individuals. To explore the relationship between socioeconomic status and descriptive representation for each of the four countries in this project, this section analyzes six forms of involvement in the electoral/partisan arena: voting, having a party identification, receiving mobilizing visits from party leaders, visiting a party office, attending a party rally, or working for a party.3 The analysis here and throughout the chapter relies primarily on education as an indicator of social class. While income presents an interesting alternative, it has the disadvantages associated with a lower response rate. Missing data on income are apparently random with respect to educational attainment in Argentina and Venezuela, suggesting that missing data may not be a source of bias in those two countries. In Chile and Peru, income data are disproportionately likely to be missing for respondents at the middle levels of educational attainment, suggesting that the omitted cases may lead to bias in those two countries. 3. The lines in the graphs for this chapter are nonparametric regressions. Technical details are available from the author. It should be noted that in the surveys in Argentina, educational attainment was measured in a way that is not directly comparable to the measure used in the capital cities of the other three nations. Among other discrepancies, the Argentine measure had far fewer categories. To enhance visual comparability within the bounds imposed by differences in survey questions, all line graphs that compare relationships between education and an outcome use a version of the Argentine education measure that has been multiplied by two. This results in an education variable with a range and standard deviation that are roughly comparable with those for the Chilean, Peruvian, and Venezuelan measures. However, it is worth bearing in mind that plots involving Argentina, are never directly comparable with those for the other three capital cities. This chapter will try to work around this by emphasizing comparisons in the overall shape of the graph—which should be relatively robust to different measures of educational attainment—rather than values at specific points.
05 Chapter 5:05 Chapter 5.qxd
11/9/2009
11:55 AM
Page 139
,
Voting is the most widespread partisan act, and it requires a relatively minor degree of engagement with the system of partisan representation. A voter may think only about candidates and need only consider the electoral system for the few minutes required to cast a ballot once every four to six years. Furthermore, voting is officially mandatory in all four of the countries under study, although the penalties and degree of enforcement associated with failing to vote differ from country to country. In Argentina any individual who fails to vote must either justify the omission before a judge within sixty days of the election or be fined between about and U.S. dollars.4 In Chile nonvoters must either present a credible explanation to a judge or be fined about U.S.$.5 For Peruvians the punishment for failure to vote is a fine of about U.S.$ and an inability to purchase postage stamps and obtain some other government services until the next electoral cycle.6 Finally, in Venezuela voting is technically mandatory but no fine or other sanction is in place for those who fail to vote.7 These legal differences most likely account for the lower rate of overall turnout among Venezuelan respondents seen in figure .; however, they do not clearly explain the other patterns evident in the data. In any case, the primary issues of concern here are whether class patterns of representational distortion exist in the electoral/partisan arena, and whether civil society has the potential to ameliorate such patterns—not why distortions arise in the first place. The relationship between education and this weakest indicator of partisan engagement, shown in figure ., differs somewhat across the four capital cities. For Chile the pattern approximates the ideal of full turnout, so that there is essentially no representational distortion. In Peru the pattern is similar, although there may be some modest overrepresentation of the middle and upper classes. The high turnout levels in these two cities may be due to mandatory voting laws, although it is worth noting that the penalties for failing to vote are much more severe in Chile and Argentina than in Peru. Thus, the relatively high turnout among Peruvian respondents, as well as the relatively lower turnout among Argentines, may require additional explanation. Argentina and Venezuela are quite different. They display lower overall levels of turnout, particularly among the popular sectors, and present a strong pattern of class-based distortion. In both of these cities the highly educated are substantially overrepresented in the electorate, while the popular sectors are underrepresented. 4. 5. 6. 7.
Decreto No. ., Código Electoral Nacional, Article . Ley No. ., Ley Orgánica Constitucional sobre Votaciones Populares y Escrutinios, Article . Ley No. , Ley Orgánica de Elecciones, Article . Ley Orgánica de Sufragio y Participación Política, Article .
05 Chapter 5:05 Chapter 5.qxd
.
11:55 AM
Page 140
0.8
0.9
1.0
11/9/2009
0.5
0.6
0.7
Argentina Chile Peru Venezuela
0
5
10
15
20
Category of Educational Attainment Fig. .. Turnout by Education
Party identification is a useful window onto the world of electoral representation in general for various reasons. First, party identification is one of the most powerful predictors of vote choice in the United States (Green, Palmquist, and Schickler :–) and in other countries (Dalton, :–). Second, party identification is associated with attention to and participation in electoral campaigns (Dalton, McAllister, and Wattenberg :–). Hence, party identification is a useful way of looking at participation in the broader system of electoral competition. Third, party identification represents a more significant degree of engagement with the electoral/partisan arena than voting. Allowing the system of parties and elections to become a component of self-definition presumably entails a degree of mental and emotional involvement far beyond that required to merely cast a ballot. As a result, it is perhaps unsurprising that rates of party identification are clearly lower than rates of turnout. At the same time, party identification is a less intensive form of involvement in the system of parties and elections than participation in partisan rallies or
05 Chapter 5:05 Chapter 5.qxd
11/9/2009
11:55 AM
Page 141
,
0.5
0.6
0.7
0.2
0.3
0.4
Argentina Chile Peru Venezuela
0
5
10
15
20
Category of Educational Attainment Fig. .. Party Identification by Education
in campaign work. However, in comparison with these more rigorous modes of participation (which will be considered below) party identification has the immense advantage of providing larger sample sizes. As with turnout, the relationship between education and having a party identification varies across the four capital cities (figure .). Chile and Venezuela have the greatest class-based distortion. The strong relationship between educational attainment and partisanship continues up to the completion of the twelfth year of schooling, at which point both curves essentially level off. Thus, for Chilean and Venezuelan respondents, the popular sectors are clearly underrepresented among partisan identifiers. Peru and Argentina have the least distortion. In Argentina the durability of Peronism, which has traditionally found most support among the popular sectors, appears to have attenuated representational distortion. The lack of representational distortion in Peru may be somewhat more surprising, given that the collapse of traditional parties and the rise of new ones (Tanaka ; Seawright ) has dramatically
05 Chapter 5:05 Chapter 5.qxd
.
11/9/2009
11:55 AM
Page 142
weakened the party-organizational apparatuses that might otherwise have been expected to reach out to the popular sectors and promote partisanship. Receiving requests from party personnel for partisan activity is sometimes treated as a cause of participation in the electoral/partisan arena (Rosenstone and Hansen :–; Verba, Schlozman, and Brady :–), but receipt of such mobilization efforts is also itself a form of engagement with the partisan system of representation. Such requests provide a personal connection between an individual and the party, and they also generate opportunities for individuals to communicate their specific concerns to people within the party organization. Receiving an effort at electoral mobilization by a party representative is thus a relatively intensive and information-rich form of engagement with the party system. Accordingly, as seen in figure ., the proportions of respondents in all four capital cities who report receiving mobilizing visits from party leaders are substantially lower than the proportions who vote or have a party identification. This mode of mobilization is particularly unusual in Venezuela, where only . percent of respondents overall report a party leader contact; in the other three capital cities, between about percent and percent of respondents report such mobilizing visits.8 Cross-national differences are apparent regarding not only the overall level of contacts, but also the pattern of class-based differences. In Venezuela, in particular, the proportion receiving mobilization is effectively zero at all levels of educational attainment. This result reflects the specific moment in Venezuelan politics when the survey used here was undertaken: in early , when Venezuela was only about five years away from the near-total collapse of its traditional political party organizations, and new political parties with effective national campaign organizations had not yet been built. Data from an earlier historical moment, when the traditional parties were a major force in Venezuelan society (Coppedge ), or from more recent years, when Hugo Chávez has been building a powerful mobilizing machine, would likely show a different pattern. For the other three capital cities, levels of mobilization are higher, and once again we see evidence that the popular sectors are underrepresented in party 8. Rates of reported party mobilization even in Argentina, Peru, and Chile are toward the low end of the range in comparison with the United States, where for the period from to between . percent and . percent of respondents reported contacts comparable to those discussed in the text (Rosenstone and Hansen :–). Indeed, in every year except , the United States had higher rates of reported party contacts than the highest rate among the South American cities under consideration. In light of the persistent image of the United States as having weak party organizations, this difference highlights the relatively low levels of party mobilization observed in Argentina, Chile, Peru, and Venezuela.
05 Chapter 5:05 Chapter 5.qxd
11/9/2009
11:55 AM
Page 143
,
0.3
0.4
0.5
0.0
0.1
0.2
Argentina Chile Peru Venezuela
0
5
10
15
20
Category of Educational Attainment Fig. .. Party Electoral Mobilization by Education
mobilization efforts in Chile and Argentina. For Chile, levels of mobilization rise from about percent among the least educated to about percent among the more highly educated. For the Argentine data, the relationship is even stronger. The popular sectors are badly underrepresented in the group of people who receive party mobilization. In Peru, by contrast, this type of representational distortion is essentially nonexistent. In comparison with the three modes of engagement with the partisan world discussed up to this point, visiting a party office is a more active and potentially information-rich form of engagement. Unlike voting, it requires a specific commitment to a particular political party. Furthermore, unlike having a party identification or receiving a mobilizing visit, visiting a party office requires individual effort and initiative. The degree of engagement involved in visiting a party office notwithstanding, a reasonable number of respondents in Argentina, Chile, and Peru report having visited a party office, as can be seen in figure .. These individuals, of course, have the opportunity to express
05 Chapter 5:05 Chapter 5.qxd
.
11:55 AM
Page 144
0.20
0.25
0.30
11/9/2009
0.00
0.05
0.10
0.15
Argentina Chile Peru Venezuela
0
5
10
15
20
Category of Educational Attainment Fig. .. Visit a Party Office by Education
specific grievances or issue positions in a way that can provide direct information to, and sometimes pressure on, the political party in question. They also have the opportunity to receive any benefits that may be forthcoming, both material and social. According to this measure, all four countries to different degrees display class-based distortion, with the popular sectors underrepresented among office visitors. In light of the expectation that party office visits in South America are a component of clientelistic exchanges that serve to link parties with the popular sectors (Auyero ; Levitsky ), this result may be somewhat surprising. Nonetheless, a generally positive relationship between education and visiting a party office holds in all four capital cities. Furthermore, this relationship is especially strong in Argentina—the home of Peronism and a prototype for much research on South American clientelism. A reasonably strong relationship also appears for respondents in Chile, although the data suggest that the rate of party visits drops off to some degree among individuals with
05 Chapter 5:05 Chapter 5.qxd
11/9/2009
11:55 AM
Page 145
,
more than fourteen years of formal education. The relationship is weaker and nonlinear in Peru and especially in Venezuela. In both cities, though, there is at least some evidence that the popular sectors may be underrepresented among party office visitors. The decision to attend a party rally involves a level of effort that seems roughly similar to that involved in visiting a party office, although there are generally fewer opportunities to attend a rally (which are held at specific moments during campaigns) than to visit an office (which are often open even between campaign periods). Rallies are designed to generate more drama and emotion than visits to party offices, a fact that may make them more attractive and thereby compensate for the relatively fewer opportunities for involvement. However, participation in a rally is a much less information-rich mode of partisan engagement; rally attendees cannot easily communicate specific ideas or concerns to the people running the event. In general (figure .) overall participation rates in party rallies are at about the same level as office visits.9 For Chile and Argentina, the now familiar patterns are again evident. In Chile the striking positive relationship between social class and participation in party rallies clearly indicates that the popular sectors are underrepresented in the group of rally participants. The same is true, although somewhat less dramatically, in Argentina. In Venezuela, again, there is no relationship between education and rally attendance, which is rather low for all levels of educational attainment. Peru displays its usual pattern of relatively high overall levels of participation, but this time with the difference of a sharp linear relationship between class and rally attendance. Peru, which on most previous measures had relatively minimal representational distortion, shows the second-highest class-wise difference on this type of partisan participation, behind only Chile. A final and most intense form of engagement in partisan politics, working for a political party, is both demanding and information-rich. Individuals need a close enough connection with a specific party to be invited to work for the party organization, and a powerful enough engagement with the party to accept the invitation. Working for a party gives people ample opportunities to express their views to party leaders. Party workers also have chances to shape 9. For Argentina, . percent of the overall sample visited a party office while . percent attended a party rally. In Chile, . percent visited a party office and . percent attended a rally. Among Peruvian respondents, . percent visited an office and . percent were involved in a rally. Of Venezuelans, . percent visited an office and . percent participated in a rally. The Argentine difference is statistically significant (p < .), as is the Chilean (p = .) and the Venezuelan (p = .). The Peruvian difference just misses significance at the conventional level (p = .). With one significant positive difference and two significant negative differences, these two variables may perhaps be seen as having similar overall levels.
05 Chapter 5:05 Chapter 5.qxd
.
11:55 AM
Page 146
0.15
0.20
11/9/2009
0.00
0.05
0.10
Argentina Chile Peru Venezuela
0
5
10
15
20
Category of Educational Attainment Fig. .. Attend Party Rally by Education
the public message of a party by how they present the party to other individuals within society. Because of the intense engagement required to work for a party, we might expect such work to be quite rare in the four capital cities under consideration. Figure . confirms that expectation, with overall levels of party work between about . percent and . percent for each of the four South American cities.10 10. Evidence from other countries suggests that these figures for party work may be somewhat but not exceptionally low in comparative perspective. For example, in the United States, reported party work rates between and ran from a low of . percent to a high of . percent (Rosenstone and Hansen :). A survey of residents of Montevideo, Uruguay, in found that percent reported either working for a party or giving money to a party (Biles :). While it is impossible to determine what share of these individuals worked as opposed to giving money, financial contributions range from about half as common to about twice as common as work for a political party in the United States (Rosenstone and Hansen :–; Verba, Nie, and Kim :–), and the two acts are roughly equally common in India (Verba, Nie, and Kim :–). Hence, it seems likely that the Uruguayan frequency of party work was in the single digits: higher than for the four cities reported here, but not by an order of magnitude. Data from for four different countries show
05 Chapter 5:05 Chapter 5.qxd
11/9/2009
11:55 AM
Page 147
,
0.06
0.08
0.10
0.00
0.02
0.04
Argentina Chile Peru Venezuela
0
5
10
15
20
Category of Educational Attainment Fig. .. Work for a Political Party by Education
The evident relationships between education and party work are more complex, yet to some extent confirm some trends we have already seen. Chile and particularly Argentina again reveal a class-based pattern of participation, but on this indicator the relationship for Argentina is a curvilinear one in which the highest and lowest educational groups do not work for parties, while the middle educational groups do the most party work. In Peru the opposite curvilinear relationship appears: the lowest and highest educational groups are overrepresented in the group of party workers, with the middle educational groups that a much higher proportion of respondents report having worked for a party at some point in their lives than the data for a five-year period reported in the text; in Austria, percent report party work; in India and Japan, the proportion is percent; and in the United States, percent. Data for the elections only in the United States show that percent reported party work that year, for a -to- ratio between lifetime and snapshot measures of party work. If that ratio is constant across nations (an assumption that is not especially likely to hold true), then the rate of party work in these four nations ranges from about percent to about percent, a range that is perhaps somewhat higher than, but generally comparable with, the results for Argentina, Venezuela, Peru, and Chile.
05 Chapter 5:05 Chapter 5.qxd
.
11/9/2009
11:55 AM
Page 148
underrepresented. In Venezuela we see again the lowest level of participation overall, but this time a very slight positive relationship in which less-educated individuals are relatively underrepresented among the group of party workers. That said, however, it may be a mistake to attribute too much to these evident patterns given how low the percentages in question are. Clearly, very few people work for political parties, and so the standard issues of inference using very small samples arise when trying to determine the class composition of that small group of people. Looking across the six modes of partisan engagement considered here, some broader conclusions may be made regarding both the country patterns of representational distortion and the comparisons among them. Table . lays out the different patterns of distortion for each country and mode of engagement with the electoral/partisan arena. Argentina displays a broad pattern of substantial underrepresentation of the popular sectors. In fact, only two indicators fail to show a strong, positive relationship between educational attainment and engagement: having a party identification and working for a political party (and the latter, as mentioned involves very few people). Hence, popular-sector underrepresentation in party politics is clearly dramatic in Argentina. This is the case in spite of the efforts of the Peronist party, generally regarded as an efficient agent for the mobilization of popular-sector voters. Chile also exhibits high representational distortion. There is a clear connection between higher educational attainment and having a party identification, as well as between education and attending a party rally. Hence, for these two forms of partisan engagement, the popular sectors are underrepresented. There is also evidence of underrepresentation of the popular sectors with respect to receiving electoral mobilization and visiting a party office, although the relationships in question seem to be quite nonlinear. However, turnout has no connection to education. On net, therefore, the popular sectors appear to be substantially underrepresented in the electoral-partisan system. Table . Class-Based Distortion in the Electoral/Partisan Arena Argentina
Chile
Electoral Turnout High Low to None Has a Party ID Low to None High Receive Mobilization High High Visit Party Office High Medium Attend Rally Medium High Overall Medium/High Medium/High
Peru
Venezuela
Low to None Low to None Low to None Medium High Low/Medium
Medium High Low to None Low to None Low to None Low/Medium
05 Chapter 5:05 Chapter 5.qxd
11/9/2009
11:55 AM
Page 149
,
Venezuela is a case of moderate representational distortion, with a distinction that can be drawn between “citizen” forms of engagement, made up of voting and having a party identification, and “activist” forms, including receiving party mobilization efforts, visiting a party office, attending a rally, or working for a party. Citizen forms of engagement display high representational distortion in the electoral/partisan arena, adversely affecting the Venezuelan popular sectors. For activist forms of engagement, by contrast, little representational distortion exists, although overall rates of engagement are the lowest in the study. In Peru, by contrast, there is very little evidence of class-based representational distortion for citizen forms of party engagement. The same is true for two forms of activist engagement: receiving party mobilization efforts, and working for a political party, although the second relationship seems to be nonlinear in ways that would not affect popular-sector representation. There is some evidence that the popular sectors are underrepresented among the group of party office visitors, and underrepresentation clearly occurs with respect to attendees at party rallies. Hence, patterns of partisan representation in Peru seem reasonably class-neutral, with two meaningful exceptions. In sum, the Argentine and Chilean popular sectors would seem to face a consistent, and serious, pattern of underrepresentation in the electoral/partisan arena. Conversely, the popular sectors of Peru and Venezuela appear to face comparatively little underrepresentation, although for some aspects of partisan participation in both countries there is evidence of representational distortion. These patterns of objective representation may be juxtaposed with respondents’ assessments of subjective representation. To what extent do the popular sectors, compared to other social strata, feel represented by political parties, the main intermediating structures in the electoral/partisan arena? One approach to measuring subjective representation focuses on the proportion of respondents who feel that involvement in political parties can help fix society’s biggest problems. If a respondent believes that parties are instruments that are capable of doing good work in their capacity as agents for society as a whole, we may conclude that the respondent has at least some minimal optimism regarding the prospect of representation. Hence, this approach involves a relatively low-threshold version of subjective representation. Figure . compares countries on this low-threshold measure of subjective representation by educational attainment.11 11. In this and all other graphs unless otherwise noted, a loess regression with locally optimal varying smoothness parameter (Simonoff :–) was fitted to the data for each capital city.
05 Chapter 5:05 Chapter 5.qxd
.
11:55 AM
Page 150
0.5
0.6
0.7
11/9/2009
0.1
0.2
0.3
0.4
Argentina Chile Peru Venezuela
0
5
10
15
20
Category of Educational Attainment Fig. .. Believe That Party Involvement Helps Society The survey question asks, “Do you think that, if people really got involved in political party activities, that could help solve society’s problems?”
One perhaps surprising conclusion is that at all educational levels Venezuelan respondents report substantial belief that working with political parties can improve society, despite the recent collapse of the party system. By this measure, Caracas residents feel high degrees of subjective representation within the Venezuelan party system, and there is little evidence that popular-sector respondents feel much less represented than middle- or upper-class respondents. For Peru levels of subjective representation are lower than in Venezuela, but there is also little sign of a class trend. However, class-based differences are evident for both Argentina and Chile. Chilean, and, to a somewhat lesser extent Argentine, popular-sector respondents are substantially more likely to be pessimistic about the possibility of improving society through the parties than are middle- or upper-class respondents in those capital cities—or respondents in any class category in Peru or Venezuela.
05 Chapter 5:05 Chapter 5.qxd
11/9/2009
11:55 AM
Page 151
,
Figure . presents another way of looking at social class and subjective representation. For each city, respondents are asked to classify themselves as belonging to the “poor,” the “working class,” the “middle class,” or “the rich,” and to identify the party that best represents each of these groups. In figure ., we see the proportion of people who consider themselves to be in a specific social class and who report that no political party represents that class.12 A more consistent picture of class-based distortion emerges across all four countries, though once again, the distortion is sharpest in Chile and to a somewhat lesser extent Argentina, moderate in Venezuela, and particularly low in Peru. Again, Peru displays little difference in subjective representation across social classes. In Argentina, by contrast, popular-sector respondents disproportionately feel unrepresented. The results for Chile and Venezuela must be interpreted in light of the fact that very few respondents classify themselves as poor, and the popular sectors tended to self-identify as working class. The difference between working-class and middle-class respondents is substantial in Chile, where the representational distortion is greatest. In Venezuela, there is less distortion than in Chile or Argentina, although somewhat more than in Peru. Across all four countries, middle-class respondents are more likely to feel represented than working-class or poor respondents—although the difference is greatest in Chile, substantial in Argentina and Venezuela, and the smallest in Peru. To what extent do these patterns of subjective representation mirror the patterns of objective representation analyzed earlier? For Argentina, Chile, and Peru, the two aspects of representational distortion are aligned. In figure ., we see that the least-educated Buenos Aires residents are quite likely to feel that involvement in political parties cannot help solve society’s problems, a result that fits well with the pattern of partisan underrepresentation facing such respondents. By the same token, figure . shows that the self-identified poor and working classes in Argentina are generally more likely to report that no party represents them than are respondents in the other cities. For Argentina, then, there is a meaningful degree of correspondence between the evidence regarding subjective and objective partisan representation. For Chile the general pattern of objective popular-sector representational distortion in partisan politics is joined with the relatively negative pattern of subjective popular-sector representation noted above. Popular-sector Chileans are the least likely to feel that involvement in parties can help society, and those 12. The rich are excluded from this figure because the four samples include very few respondents who place themselves in that category.
05 Chapter 5:05 Chapter 5.qxd
.
11/9/2009
11:55 AM
Page 152
70 60 50
Poor Working Class Middle Class
40 30 20 10 0 Argentina
Chile
Peru
Venezuela
Fig. .. Feel Unrepresented in Partisan Politics The survey question asks, “Which party responds best to the interests of [the class in question]?” The figure shows the percent of the respondents who classified themselves as belonging to a specific class and who chose the unspoken option that no party represents that class.
who identify as working-class feel that no particular party speaks for them. This pattern of subjective representation combines with a mixed objective pattern regarding citizen forms of partisan engagement, but greater or lesser degrees of popular-sector underrepresentation on the other four measures. Respondents in the Peruvian popular sectors are reasonably likely to report that party involvement can help society, and the popular sectors are relatively less likely to claim that no party represents them. Once again, this finding is reasonably congruent with the objective pattern of relative class neutrality with respect to partisan engagement evident in the data. Venezuelan popular-sector respondents are the most likely to claim that involvement in political parties can help society, and they have about the same score with respect to feeling that some party represents them as do Peruvians. Since Venezuela has a strong pattern of popular-sector underrepresentation with respect to citizen forms of partisan engagement but not with activist forms, it seems plausible that respondents form their sense of subjective representation with respect to non-electoral patterns of party politics. These patterns of objective and subjective representation provide interesting findings. First, as noted previously, it is somewhat surprising that the Peronist partisan machine in Argentina has coexisted with a system of substantial popular-sector underrepresentation. At the same time, the relatively weak
05 Chapter 5:05 Chapter 5.qxd
11/9/2009
11:55 AM
Page 153
,
partisan institutions of Peru are matched with a relative lack of representational distortion. Finally, it is noteworthy that Venezuela shows a mix of popularsector underrepresentation in citizen modes of involvement, which Chávez has conceptualized as hallmarks of the elitist “representative democracy” of the past, with a pattern free of representational distortions in activist modes of involvement, which Chávez characterizes as typical of an emergent “participatory democracy.”13 As a concluding note to the discussion of class and partisan engagement in these four capital cities, it may be worth pointing out that there is no evidence of even a single mode of partisan engagement in a single capital city that overrepresents the popular sectors. Furthermore, in no case do the popular sectors report proportionally higher levels of subjective representation than their middle- and upper-class counterparts. The best-case scenario for popular-sector representation seems to be something like Peru: the popular sectors often achieve equal representation, but occasionally face meaningful underrepresentation.
Class and Civil Society Whereas the previous section examined representational distortion in the electoral/partisan arena, we now turn to class-based distortion in civil society participation. It is worth noting that while much of the rest of this volume focuses on the A-Net, or the subset of civil society organizations with a clear programmatic purpose, this chapter broadens the focus by including other kinds of associations: those focused on religion, sports and recreation, or participants’ places of origin. While associations in this latter category are not explicitly political, they may nonetheless play important parts in connecting citizens with the electoral/partisan arena—as will be demonstrated below. Hence, there is some value in exploring the extent to which civil society organizations, broadly conceived, are representative of the class structure of society. Are the popular sectors over- or underrepresented in civil society as a whole and, more particularly, in certain types of civil society organizations? In the United States (Verba, Schlozman, and Brady :–) and other advanced industrial democracies (e.g., Parry, Moyser, and Day :–),
13. See, for example, Ellner (), Cameron and Major (:, –), Ellner (:, ), and McCoy (:–).
05 Chapter 5:05 Chapter 5.qxd
.
11/9/2009
11:55 AM
Page 154
scholars have found class-based distortion, with those at higher socioeconomic status levels overrepresented among members and participants in civil society organizations. For these countries, civil society has the end result of reinforcing the representational advantages of higher-status groups by increasing these groups’ representation in the interest regime. Is the same true in Latin America? Or, as hypothesized by some researchers, are the popular sectors disproportionately active in civil society, participating in the interest arena in place of the partisan/electoral arena? To analyze the relationship between class and civil society participation, this section examines both educational and income measures of social class and finds that across the four capital cities included in the survey, higher-status individuals are generally—although not universally—overrepresented in civil society associations, both programmatic and nonprogrammatic. In a first pass at this relationship, figure . shows the relationships, for respondents in each of the four capital cities, between educational attainment and involvement in civil society (coded here as a dichotomous variable reflecting whether or not an individual is involved in at least one kind of civil society association, whether or not programmatic). Figure . shows the relationships between civil society participation and household income, as an alternate indicator of social class. While interesting, this indicator suffers from the disadvantage of a lower response rate than the measure of educational attainment.14 In order to provide as much comparability as possible across countries, the graph treats household income as ranks within each capital city, rather than as absolute values. Hence, the comparison at any given point on the horizontal axis is among people with a similar position in a given country’s income hierarchy, rather than among people with equivalent material resources. This approach seems appropriate because the analytic focus is on inequity within nations, rather than on the per-dollar effects of income on civil society involvement. Before analyzing the specific patterns of class-based distortion in civil society participation for each country, some general trends should be noted from figures . and .. First, all countries show some degree of class-based distortion, but Argentina presents a much starker degree of distortion than the other three countries. This is especially evident in figure . but also true for 14. Missing data on income are apparently random with respect to educational attainment in Argentina and Venezuela, suggesting that missing data may not be a source of bias in those two countries. In Chile and Peru, income data are disproportionately likely to be missing for respondents at the middle levels of educational attainment, suggesting that the omitted cases may lead to bias in those two countries.
05 Chapter 5:05 Chapter 5.qxd
11/9/2009
11:55 AM
Page 155
,
Fig. .. Civil Society Participation by Education
figure ..15 Second, for all countries, the middle classes are more represented in civil society organizations than the popular sectors, but a hierarchical relationship may not be universal within the middle classes, since in figure ., the rate of participation approximately levels off above about fourteen years of schooling for two of the four capital cities in this study. Third, overall participation rates vary across countries, with Peru having the highest overall participation rates, Argentina and Chile in the middle, and Venezuela exhibiting the lowest rates of civil society participation. For a more nuanced interpretation of these patterns, we now turn to country-specific analyses of figures . and .. With respect to education, the slopes for Argentina are dramatic and positive. The difference in proportions active in civil society between the lowest and the highest educational categories 15. The closest comparison is between the amount of distortion evident for Argentina and Venezuela in figure .. However, even in this comparison, the Argentine line shows a modestly higher overall increase from the lowest income level to the highest.
05 Chapter 5:05 Chapter 5.qxd
.
11:55 AM
Page 156
0.7
0.8
0.9
11/9/2009
0.3
0.4
0.5
0.6
Argentina Chile Peru Venezuela
0.0
0.2
0.4
0.6
0.8
1.0
Social Class (Income Percentile)
Fig. .. Civil Society Participation by Income Note: Because of the roughly continuous nature of the income variable, this figure presents Nadaraya-Watson kernel regression estimates (Nadaraya ; Watson ; Simonoff :–), using a Gaussian kernel and a bandwidth of ..
is quite large: roughly percent of highly educated Argentines have at least one associational involvement, as compared with about percent of Argentines in the lowest group. Clearly, Argentine civil society is disproportionately composed of the well-educated. The trend for Argentina with respect to income is similar. Individuals with the lowest reported household incomes in the sample have about a percent chance of being involved in some civil society association, while roughly percent of people toward the upper end of the income distribution participate in civil society. Whether considered from the point of view of income or education, the upper classes in Argentina are clearly overrepresented in civil society, compared with the popular sectors. A picture of somewhat more modest underrepresentation in civil society of the popular sectors emerges in Chile. With respect to education, the popular
05 Chapter 5:05 Chapter 5.qxd
11/9/2009
11:55 AM
Page 157
,
sectors in Chile are if anything less well represented in civil society than figure . shows. The Chilean data suggest a substantial, positive relationship between education and civil society involvement, with one anomaly: a relatively high estimated rate of participation for the least educated in Chile. This estimate of about percent involvement in civil society for those with three or fewer years of education may reflect a distinctive feature of Chile’s participatory structure. Alternatively, these high rates may be artifacts of the small sample size at such low educational levels, since only about . percent of Chilean respondents reported having completed three or fewer years of education. Even if the very uneducated are distinctively active in civil society, however, the small numbers of such individuals would make the finding of limited substantive importance: the popular sectors would still primarily comprise the groups for which the Chilean curve shows substantial representational distortion. Aside from those at the very bottom, then, only about percent of less-educated respondents participate in civil society, while nearly percent of better-educated respondents have some associational involvement. The plot of the relationship between household income and civil society for Chile in figure . also points toward underrepresentation of the popular sectors, although the distortion appears to be more modest. Just over percent of the poorest residents of Santiago participate in civil society, as compared with about percent of middle- and upper-income residents. The net result is a body of civil society participants that moderately underrepresents the popular sectors. Civil society in Chile is composed of better-educated, wealthier individuals than the population as a whole, although the class-based distortion is weaker than in Argentina. Paralleling the pattern we observed of participation in the electoral/partisan arena, the pattern of civil society participation by class in Peru is quite different from—and indicates much less representational distortion than—those prevalent in Chile and Argentina. At all levels of both education and income, Peru shows unusually high rates of associational participation. Overall, about percent of Lima residents at any level of education seem to participate in civil society associations. Furthermore, civil society involvement shows comparatively weak evidence of a clear class trend in figure ., with a difference of only about percentage points between the education level with the lowest participation rates (five years) and that with the highest (twenty years). Figure . shows an even more class-neutral result for income, with the middle and upper classes if anything slightly underrepresented. Associational life in Peru probably does not privilege the popular sectors, but neither does it particularly disadvantage them.
05 Chapter 5:05 Chapter 5.qxd
.
11/9/2009
11:55 AM
Page 158
In Venezuela a pattern emerges that is in many ways a kind of midpoint between the finding of high class-based distortion in Argentina and the pattern of low distortion that seems to characterize civil society in Peru. When considered by levels of education, there is a difference in participation rates of about percentage points between the popular sectors and the middle and upper classes, which is much less than the degree of representational distortion in Argentina and basically comparable with the level of distortion in Chile—although there is clearly more distortion than in Peru. The income indicator tells a similar story. Figure . shows a clear positive trend among people in the upper half of the income distribution, but no trend at all among poorer respondents. Of the poorer half of the respondents, a little over percent participate in at least one kind of association, while roughly percent of the wealthier half of the respondents are involved in civil society. This gap of about percentage points is comparable to that found in Chile, but smaller than the distortion evident in Argentina. Hence, the class inclusiveness of Venezuelan civil society—considering either education or income—seems to be comparable to that of Chile, and probably more inclusive of the popular sectors than civil society in Argentina, although less so than in Peru. Summing up the results for relationships between social class and participation in at least one civil society organization, table . shows that in no case is civil society a distinctively popular-sector space. In every country, on both educational and income-level measures, the well-off participate in civil society at least as often as the less well-off. These findings deal a somewhat serious blow to the hope that popular-sector involvement in civil society in general, in conjunction with linkages between civil society and partisan politics, might serve to counterbalance partisan participation and representation advantages, due to personal and economic resources, of the middle and upper classes. Comparing across capital cities, Peruvian civil society is quite inclusive, with high levels of overall participation and relative class equality. Civil society in Argentina includes proportionately fewer members of the popular sectors
Table . Class-Based Distortion in Civil Society Participation Measure of Social Class
Argentina
Chile
Peru
Venezuela
Education Household Income Overall
High High High
Moderate Moderate Moderate
Low Low Low
Moderate Moderate Moderate
05 Chapter 5:05 Chapter 5.qxd
11/9/2009
11:55 AM
Page 159
,
than does the population as a whole. The popular sectors are also underrepresented in Venezuelan and Chilean civil society, although not to the same extent as in Argentina. It may also be worth noting that levels of civil society participation in Venezuela are relatively low across almost all social-class categories. These cross-national differences in terms of the relationship between class structure and civil society participation correspond fairly well to the different patterns of representational distortion in partisan engagement. However, the question of whether civil society involvement serves to redress or exacerbate popular-sector underrepresentation in the electoral/partisan arena is not yet decided. Might there be some specific kinds of associations in which popularsector actors are in fact overrepresented—even if such actors are underrepresented in civil society as a whole? Of particular interest is a set of programmatic associations that are known to be prevalent in lower-class neighborhoods. If distinctively popular-sector kinds of associations have powerful linkages with partisan politics, then civil society may still serve to enhance the partisan representation of the popular sectors, and perhaps account for cross-city differences in class representation. Likewise, are associational categories within which middle- and upper-class actors are overrepresented closely linked with the electoral/partisan arena? If not, then the prominence of social elites in civil society is at least neutral with respect to popular-sector partisan representation. The answer, which involves relationships between civil society participation and partisan participation, is the task of the remainder of this chapter. This section concludes by exploring whether there some kinds of civil society associations in which popular-sector actors predominate. It would be impractical, for a variety of reasons including small sample sizes and the sheer number of types of associations under consideration, to address this issue via the highly disaggregated data displays used earlier in the section. Hence, the analysis will focus on the proportion of popular-sector individuals, defined alternately by levels of educational attainment or by relative household income, among participants in various kinds of associations. If those proportions are lower than the overall sample proportion of popular-sector individuals for a given city, then the association type in question underrepresents popular-sector actors, and vice versa.16 Tables . and . show the proportion of popular-sector respondents, defined by educational attainment, among all respondents in each city in the first 16. Throughout this chapter, all reported proportions are weighted to correct for the deliberate overrepresentation of certain districts in the sample.
05 Chapter 5:05 Chapter 5.qxd
.
11/9/2009
11:55 AM
Page 160
Table . Popular-Sector Participants in Civil Society Associations (percent without high school diploma) Argentina
Chile
Peru
Venezuela
All Respondents
56
41
35
42
Neighborhood Associations Food-Distribution Associations Anticrime Associations Parents’ Associations Producer’s Cooperatives Labor Union Household Place-of-Origin Associations Church/Religious Associations Sports/Recreation Associations
46* 35* 32* 50 36 55 32* 48* 38*
40 24* 22* 41 21* 38 0 47* 32*
37 50* 29 33 13* 27* 36 37 28*
34* 33 30* 34 29* 34* 29 42 27*
Note: Data drawn from the – survey. Each entry reflects the proportion of respondents in a given group who have less than a complete high school education. *In a test of the hypothesis that the proportion for this group is equal to the proportion for nonmembers of this group, p < ..
row, and among individuals who are active in various types of civil society associations, in the subsequent rows.17 Table . codes as popular sector those who have not completed high school. Table . shows the results with a more inclusive cut-point, also including in the popular sectors those who have the national equivalent of a high school education or less. Using this definition, the popular sectors include from to percent of respondents per country—corresponding to the common notion in Latin American discourse that the popular sectors are the great majority of society. Regardless of the measurement threshold used to define popular-sector status, for roughly half of the possible association types, associational members are disproportionately composed of members of the middle and upper classes, in comparison with the sample as a whole. The popular sectors are not overwhelmingly excluded from civil society in any of these four capital cities; almost all of the proportions in table . are over percent, so a majority of civil society participants are drawn from the popular sectors. Yet that majority is
17. Note that the substantive focus here is on whether the popular sectors are adequately represented in associational life. Hence, tables .–. report the proportion of people active in a given association who are from the popular sectors. If the analytic focus were instead on the hypothesis that popular-sector status is causally linked with associational involvement, then the appropriate measure would be the proportion of popular-sector individuals who participate in a given association. This distinction between the quantities appropriate to the analysis of representation and those suited to discussions of causation is important to keep in mind throughout this chapter.
05 Chapter 5:05 Chapter 5.qxd
11/9/2009
11:55 AM
Page 161
,
Table . Popular-Sector Participants in Civil Society Associations (percent without postsecondary education) Argentina
Chile
Peru
Venezuela
All Respondents
77
66
67
70
Neighborhood Associations Food-Distribution Associations Anticrime Associations Parents’ Associations Producer’s Cooperatives Labor Union Household Place-of-Origin Associations Church/Religious Associations Sports/Recreation Associations
72 55* 74 74 44* 73* 64 69* 60*
62 44* 51* 70 59 64 14 74* 62
69 78* 66 66 37* 58* 69 64 63
55* 59* 64 58* 60 60* 57 69 61*
Note: Data drawn from the – survey. Each entry reflects the proportion of respondents in a given group who have no more than a complete high school education. *In a test of the hypothesis that the proportion for this group is equal to the proportion for nonmembers of this group, p < ..
almost always smaller than the popular-sector majority in the population as a whole, often significantly so. A comparison between table . and table . shows that the more restrictive decision to consider only those with less than a high school diploma yields, on average, a higher overall sense of popular-sector underrepresentation. This suggests that the group that has completed high school but has no further education is underrepresented in comparison with those with some post-secondary education, but to a less significant degree than the group that has not received a high school diploma. In Argentina and Venezuela, popular-sector participants are at least minimally underrepresented, compared with the sample as a whole, in almost every category of association for either educational cut-point. In Venezuela the differences are significant in one table or the other for every associational category other than church-based/religious groups and place-of-origin associations. Indeed, many of the differences are of double-digit magnitudes. For Argentina five differences (involving food distribution, religious, and sports and recreation associations, as well as cooperatives and labor unions) are significant using the more inclusive cut-point, and all but three (parents’ associations, cooperative associations, and labor unions) are significant using the restrictive cut-point. Many of the differences in both tables are of double-digit magnitude. In Chile popular-sector respondents are overrepresented among those involved in church-based and religious associations and possibly also among those involved in parents’ associations, and the popular sectors are not dramatically
05 Chapter 5:05 Chapter 5.qxd
.
11/9/2009
11:55 AM
Page 162
underrepresented among participants in neighborhood associations or labor unions. For about half of the other associational categories, however, there is a statistically significant pattern of popular-sector underrepresentation. Finally, in Peru, popular-sector respondents are clearly overrepresented in fooddistribution networks, and possibly also in neighborhood associations. The popular sectors are underrepresented in producers’ cooperatives and, for the more restrictive educational cut-point used in table ., sports and recreation associations. All other associational categories appear to essentially mirror the sample as a whole in terms of class composition. While the popular sectors are generally underrepresented among those active in most kinds of civil society associations, certain categories of associations deserve special attention due to their historical legacy of including the popular sectors. The prototypical case in point is the labor union, which is often seen as distinctively oriented toward the popular sectors. Yet empirically this is not universally the case. Chilean labor unions have developed a group of participants who are roughly as likely to be from the popular sectors as the sample as a whole; the same is true for Argentina if the more restrictive educational cut-point is used. In Venezuela, however, the group of respondents from union households distinctly underrepresents the popular sectors—a finding that is consistent with political alignments in Venezuela during the Chávez period, in which labor unions have allied themselves with representatives of the middle classes and against the Chávez regime. Peruvian labor unions also significantly underrepresent the popular sectors according to either measure. These cross-national results regarding labor unions suggest that, especially for Chile but also for Argentina, unions have the potential to play some role in amplifying the partisan representation of the popular sectors. For Peru and Venezuela, however, labor unions have the surprising potential to exacerbate problems of popular-sector underrepresentation. A somewhat different point can be made regarding those active in churchbased and religious associations—a category that, like unions, had close linkages to partisan politics during the heyday of Latin American populism. These associations do not usually overrepresent the popular sectors, but they are distinctive in that they also do not usually underrepresent them. Hence, if there are strong connections between participation in these associations and partisan involvement, that linkage will not distort overall patterns of representation. Indeed, it may well dilute existing distortions by watering down the existing, disproportionately middle- and upper-class group of participants. In three of the four capital cities, the popular sectors are well represented among people active in these kinds of associations. Even for Argentina, where the popular
05 Chapter 5:05 Chapter 5.qxd
11/9/2009
11:55 AM
Page 163
,
sectors are underrepresented among religious association participants, the gap does not reach double digits. For Peruvian and Venezuelan respondents, the popular sectors are close to being exactly proportionately represented among those active in religious associations. The data for Chile and Venezuela indicate that the category of church-based and religious associations shows with the highest popular-sector participation. Indeed, the Chilean popular sectors are overrepresented among those active in religious associations—one of only two categories in which the popular sectors are significantly overrepresented! Thus, as in the United States (Verba, Schlozman, and Brady :–), religious associations in these South American capital cities have a distinctive potential to enhance the partisan participation of those with lower socioeconomic status, and therefore to improve the political representation of the popular sectors. As will be seen in the next section, however, this potential for enhanced partisan participation of the popular sectors is not realized because there proves to be no meaningful relationship between involvement in religious associations and in party/electoral politics. Some other categories of associations are as inclusive, or even more inclusive, toward the popular sectors than religious associations or unions in at least some of the capital cities in this study. Parents’ associations seem quite classinclusive in Argentina, Chile, and Peru, although less so in Venezuela, where the popular sectors are significantly underrepresented using the more inclusive indicator for popular-sector status. Neighborhood associations proportionately include the popular sectors in Chile and Peru, although not in Argentina or Venezuela. To summarize, in Peru most associational categories include popular-sector actors roughly in proportion to the population as a whole. This is quite consistent with the result, above, that patterns of participation in Peru’s civil society show little evidence of class bias. In Argentina and Chile, there is significant evidence of popular-sector underrepresentation in many, although not all, associational categories—a finding that is also consonant with the overall pattern of popular-sector underrepresentation in civil society discussed above. For Venezuela, there also a broad pattern of popular-sector underrepresentation. These patterns generally correspond to the overall findings for participation in one or more civil society associations discussed in table . above. When considered in terms of household incomes rather than educational attainment, as in table ., the pattern of popular-sector inclusion in the group of civil society activists appears to be generally similar, although different in some particulars. Unfortunately, it is difficult to interpret these differences, both because of the general issues related to the income measure discussed
05 Chapter 5:05 Chapter 5.qxd
.
11/9/2009
11:55 AM
Page 164
above, and also because of more specific evidence for all four countries and most association types that the missing cases are significantly different in terms of associational involvement than the nonmissing cases. As with tables . and ., in which class is operationalized in terms of education, in table . a majority of association types in Argentina, Chile, and Venezuela underrepresent the popular sectors—although few association types show statistically significant differences because of the relatively small samples for this chart. With class measured in terms of income, many more associational categories in these three countries show a pattern of popular-sector overrepresentation, though again these differences are not statistically significant. In Peru, as above, patterns of participation in civil society seem fairly classneutral; only labor unions show evidence of a substantial pattern of popularsector underrepresentation. Church-based/religious groups are once again the category that consistently includes lower-income individuals in approximate proportion to their share of the population across the four capital cities, reinforcing the conclusion above that religious associations have a significant potential for redressing popular-sector underrepresentation in the electoral/partisan arena. For example, in table ., church-based and religious associations are the only category for which the popular sectors are at least proportionately represented in Table . Popular-Sector Participants in Civil Society Associations (percent in bottom three income quartiles)
All Respondents Neighborhood Associations Food-Distribution Associations Anticrime Associations Parents’ Associations Producer’s Cooperatives Labor Union Household Place-of-Origin Associations Church/Religious Associations Sports/Recreation Associations
Argentina
Chile
Peru
Venezuela
74
77
76
77
62 72 48* 84 100 70 91 77 57*
65 50 67 81 77 72 100 85 71
75 93 73 79 68 64 72 77 77
67 59* 64 64 55* 61 59 79 68
Note: Data drawn from the – survey. Each entry reflects the proportion of respondents in a given group who fall below roughly the sample seventy-fifth percentile in terms of household income. The percentile is not exact for each country because the income data is lumpy; multiple individuals often report the same income level. Hence, as the first line of the table shows, percent fall below the threshold in Argentina, while percent do in Chile, and so forth. *In a test of the hypothesis that the proportion for this group is equal to the proportion for nonmembers of this group, p < ..
05 Chapter 5:05 Chapter 5.qxd
11/9/2009
11:55 AM
Page 165
,
all four of the capitol cities under consideration. In contrast to the finding when education levels are used to operationalize class, lower-income individuals may be somewhat underrepresented among those active in labor unions in all four countries, even in Chile where unions are more class-neutral according to the education measure. This difference may be a result of nonresponse bias, or it may be due to the income-enhancing effects of union-mediated collective bargaining. Across the different measurement alternatives considered in this section, the overall implication of the analysis is that the popular sectors are underrepresented in civil society in Argentina, Venezuela, and (perhaps to a somewhat lesser extent overall) Chile—while in Peru civil society achieves a much greater degree of class-neutral participation. This overall finding matches reasonably well with the pattern, noted in the previous section, of popular-sector representation in the electoral/partisan arena across these four cases. Argentina’s relatively severe underrepresentation of the popular sectors in party politics combines with a pattern of substantial underrepresentation in civil society. Peru’s pattern of relative class neutrality in the electoral/partisan arena fits nicely with its relatively class-neutral rates of participation in civil society. Chile and Venezuela, which have somewhat divergent patterns of distortion in popular-sector partisan representation, have moderate patterns of popularsector underrepresentation in civil society. Overall, these results disconfirm the suggestion by some authors that representational distortion in the electoral/partisan arena is counterbalanced by an overrepresentation of the popular sectors in civil society organizations and in the interest arena, more specifically. The following section explores whether participation in civil society is connected with involvement in the electoral/partisan arena in ways compatible with the hypothesis that representational distortion in the electoral world is produced by parallel, but perhaps causally prior, distortions in the interest arena.
Shared-Member Linkages We have seen that the patterns of representational distortion in associational participation show some correspondence to participation in the electoral/partisan arena, with high levels of distortion in Argentina, moderate levels in Chile and Venezuela, and low levels in Peru. This rough correlation between class distortions in the electoral/partisan arena and in civil society participation suggests
05 Chapter 5:05 Chapter 5.qxd
.
11/9/2009
11:55 AM
Page 166
the hypothesis that representational distortions in associational involvement may sustain or even exacerbate problems of popular-sector representation in the electoral/partisan arena if associational involvement mobilizes individuals into direct participation in electoral politics. The data at hand permit a test of one central mechanism involved in this hypothesis: are there distinctive overlaps in terms of individual participation between civil society associations on the one hand and the electoral/partisan arena on the other? If such overlaps do not exist, then it is hard to see how class distortions in associational participation could have any meaningful connection with distortions in the electoral/ partisan arena. By contrast, evidence that association activists are overrepresented in the electoral/partisan arena is compatible with this hypothesis. To assess the plausibility of a relationship between associational participation and electoral politics, this analysis relies on evidence of what may be called shared-member linkages between civil society organizations and the electoral/ partisan arena. A shared-member linkage arises when a political party or movement is disproportionately composed of individuals active in a particular kind of civil society organization. This type of linkage is potentially relevant for mass parties and labor-based parties, as well as the catchall parties that characterize contemporary electoral politics in Latin America. Under the -Hub political parties and labor unions had strong sharedmember linkages, as both unions and political parties shared many of the same members or affiliates. These linkages at the individual level were strongly related to organizational links between unions and parties. For instance, unions in Argentina were historically linked to the not only through strong partisan identities but also through formal organizational connections, such as the tercio system, whereby a third of the party slate was allotted to unionists. Likewise, in Venezuela, ’s internal organization included a formal structure of Labor Bureaus. In one form or another, unions not only had shared-member linkages with parties but also had an organizational presence and influence in the labor-based parties that participated in the -Hub. In contrast, relationships between popular organizations and political parties are less formalized and more distant since the weakening of the -Hub. These different relationships arise partly because of the different nature of the organizations (unlike unions, for instance, most associations do not have fixed members per se but rather intermittent participants) and partly because of parties’ shift to the catch-all model, so that even the historical labor-based parties have loosened their relationship with unions and have more distant relations to other organizations in civil society. Consequently, linkages between
05 Chapter 5:05 Chapter 5.qxd
11/9/2009
11:55 AM
Page 167
,
political parties and individuals are now quite different, more indirect, and at the level of the individual rather than the organization. Lacking the linkage that made the popular organizations a crucial component of electoral representation, currently civil society seems disconnected from the electoral/partisan arena. Nevertheless, shared-member linkages may provide a significant mechanism by which the interest regime connects individuals to the electoral/partisan arena. If civil society organizations and partisans share overlapping members, political parties may develop informal relationships with these associations that have important implications for electoral participation. For example, participating in a neighborhood association may put an individual in closer contact with the state, increasing the likelihood that this individual will vote, form a political identity, or even undertake party activism. While shared-member linkages do not provide a formal relationship between associations and political parties, these linkages may make it easier or more appealing for a civil society participant to engage in various forms of electoral or partisan activity. Through the concept of shared-member linkage one can begin to ask some important questions about the relationship between civil society and the electoral/partisan system. To what extent are there ties between associational participation and the electoral/partisan arena in the sense that participants in particular types of associations are overrepresented among those engaging in party politics, with the possible implication that parties are more apt to be more attentive to the constituencies and interests of those associations? To what extent can the pattern of participation in the associational arena account for popularsector underrepresentation in the electoral/partisan arena? The existence of distinctive links between participants in specific associations and the electoral/partisan arena is a particularly plausible mechanism by which a causal connection could exist between representational distortion in civil society and in partisan/electoral politics. Such participants could carry the interests and perspectives of their associations with them into partisan life—and, of course, vice versa. Hence, representational distortion in either arena may tend to produce distortion in the other. The hypothesis that civil society could reinforce popular-sector underrepresentation in the partisan world relies on a supposition that associational life has a dynamic that is causally prior to party and electoral politics; because of our special interest in that relationship, such a supposition may occasionally be adopted below. Nevertheless, this analysis is fundamentally agnostic about questions of causation, focusing instead on problems of description. Finding descriptive evidence of shared-member linkages obviously does not rule out problems of omitted
05 Chapter 5:05 Chapter 5.qxd
.
11/9/2009
11:55 AM
Page 168
variables, self-selection, and the other well-known problems of causal inference in observational studies. Shared-member linkages are a generalization of the long-standing concerns, in Latin America and elsewhere, with relationships between involvement in political movements on the one hand and, on the other, not only labor unions but also other organizations, particularly churches, which have also been a source of recruitment for certain types of political parties (e.g., Rogowski :–; Stanley, Bianco, and Niemi :–; Carmines and Stanley :; Bartolini :–). The concept of shared-member linkage generalizes from this research agenda: rather than focusing narrowly on one or two kinds of civil society organizations, it raises the possibility that other categories of associations may play a parallel role as informal connections, via a common pool of participants, between civil society and party politics. It should be noted that extant typologies of political linkages do not include the shared-member category. Yet shared-member linkages between partisan engagement and associations are potentially complementary to or interact with other types of linkages. For instance, Kay Lawson has proposed a fourcategory schema of linkage types, with an emphasis on political parties (Lawson :–). Participatory linkage involves bringing citizens into the actual process of government decision-making, as with referenda. Policy-responsive linkage describes the Downsian situation, in which political parties are fully responsive to the aggregate distribution of citizen opinion while implementing a government response to policy issues. Linkage by reward involves the exchange of votes for material benefits. Finally, directive linkage involves a more authoritarian relationship in which political parties help the government maintain control over the population. The analysis of these linkages may be usefully informed by a consideration of shared-member linkages. If a political party has a shared-member linkage with a particular type of civil society association, those associations may provide a useful basis for bringing citizens more directly into contact with government decision-making processes— thereby facilitating the construction of a participatory linkage. Such sharedmember linkages may also facilitate the transmission of information about preferences from citizens involved with the relevant associations to politicians and party leaders—bringing about one of the key preconditions for policyresponsive linkage. Likewise, parties may take advantage of associations’ existing organizational infrastructure to reduce the transaction costs involved in distributing goods, services, or money to voters—generating a situation of linkage by reward. Finally, shared-member linkages between political parties
05 Chapter 5:05 Chapter 5.qxd
11/9/2009
11:55 AM
Page 169
,
and categories of associations may help parties keep track of associational plans and actions—helping maintain the control associated with directive linkage. Hence, the existence of shared-member linkages may serve as a resource for the development of any of these other kinds of linkage relationships. A second influential typology of political-linkage relationships, proposed by Herbert Kitschelt, distinguishes among programmatic linkages (essentially equivalent to Lawson’s policy-responsive linkages), clientelist linkages (parallel to Lawson’s linkage by reward), and charismatic linkages, in which a political coalition is held together by the personal characteristics of the dominant leader (Kitschelt :–). Charismatic linkage may be seen as antithetical to shared-member linkage. After all, shared-member linkage is fundamentally organizational in nature, whereas charismatic linkage is built around a rejection of organization as a political strategy (Kitschelt :). To the extent that shared-member linkage is important in a political relationship characterized by charismatic linkage, then, it may be as the basis for a process of routinization of charisma (Weber :–). Shared-member linkages may thus play an important role in supporting the other modes of linkages that have proved valuable in comparing patterns of representation across countries, as well as across political parties and movements. The existence of a shared-member linkage is neither necessary nor sufficient for any of these modes of linkage; however, shared memberships may serve to increase the efficacy of most linkage types. How can the existence of a shared-member linkage be empirically assessed? Since a shared-member linkage inherently revolves around individual behaviors and attitudes, survey data, with measures of individual involvement in various types of civil society organizations and in the electoral/partisan arena, are essential. Survey data about shared-member linkage between a specific type of association and a particular component of the electoral-political system can usefully be represented via a representation ratio. The proportion of respondents involved in the relevant political party or movement who are active in the relevant kind of association may be compared with the proportion of overall respondents active in that kind of association. If the ratio of these two proportions is significantly greater than , there is credible evidence of a sharedmember linkage between the association category and the party or movement.18 For example, if neighborhood associations had high shared-member 18. The representation ratio has been used by Verba, Schlozman, and Brady (:–). The rationale behind this ratio is relatively intuitive. To decide whether women are underrepresented in the U.S. House of Representatives, for example, most people would proceed by finding out how many women are in the House. At the beginning of , there were . This number would then be divided
05 Chapter 5:05 Chapter 5.qxd
.
11/9/2009
11:55 AM
Page 170
linkages, individuals involved in neighborhood associations would be overrepresented among those engaged in the electoral/partisan arena, in comparison with those not engaged in partisan politics. Extensive research on developed democracies, as well as a small amount of work on developing democracies, establishes a strong baseline expectation that there will be extensive shared-member linkages between various forms of partisan participation and involvement in civil society associations. Specifically, a large body of literature has shown that activists in associations are disproportionately likely to participate in public life, both in the developed world, and in the developing world, and there is no clear reason to expect that the partisan domain should be exempt from this broad finding.19 Hence, if participatory patterns in the four South American capital cities under consideration here are similar to those found in other analyses, it is to be expected that many of the shared-member linkages between civil society and groups of partisans considered below will be substantially higher than . Therefore, the rest of this section will examine the representation ratio of participants in various civil society associations for the group of those engaged in each of the six forms of partisan engagement discussed earlier: voting, party identification, receiving party mobilization efforts, attending party manifestations, visiting a party office, and working for a party. Shared-Member Linkages Between Voters and Civil Society As a starting point in considering shared-member linkages between civil society and the system of partisan and electoral competition, it is useful to analyze the most common act: voting in an election. Table . shows the shared-member linkage ratios between participants in this study’s nine categories of civil society associations and the group of voters in each of the four capital cities. For
by , the total number of representatives in the House (/ = . percent). This percentage is then compared to the percentage of people in the United States as a whole who are women, which is about . percent. The conclusion to be drawn is that women are underrepresented because the percentage of women in Congress is considerably lower than the percentage in society as a whole. To quantify the degree of underrepresentation, it is possible to divide the percentage of women in the House of Representatives by the percentage of women in the population as a whole—yielding the participation ratio used in this chapter 19. For the developed world, see, e.g., Maccoby (), Erbe (), Pollock (), Putnam, Leonardi, and Nanetti (), and Verba, Schlozman, and Brady (). For the developing world, see Booth and Richard (), Bratton (), A. Seligson (), and Klesner ().
05 Chapter 5:05 Chapter 5.qxd
11/9/2009
11:55 AM
Page 171
,
Table . Civil Society Participation, Voting, and the Popular Sectors (representation ratios) Association Type
Argentina
Chile
Peru
Venezuela
Neighborhood Food-Related Anticrime Parents’ Producer’s Cooperatives Labor Union Household Place-of-Origin Church/Religious Sports/Recreation
2.376* (1.106) 1.368 (0.362) 2.575* (1.827) 1.424 (0.575)
1.129 (0.235) 0.580 (0.152) 1.155 (0.381) 1.762*(0.421)
1.979* (0.506) 1.268 (0.259) 1.783* (0.563) 1.974* (0.492)
2.091* (0.431) 1.853* (0.671) 1.724 (0.560) 1.829 (0.706)
NA
1.557 (1.258)
0.717 (0.383)
1.944* (0.655)
1.821* (0.346) 0.579 (1.173) 0.940 (0.193) 1.509* (0.378)
1.661*(0.277) NA 1.191 (0.143) 0.611 (0.076)
1.511* (0.416) 1.206 (1.021) 0.782 (0.092) 0.636 (0.068)
1.838* (0.357) NA 0.877 (0.156) 0.641 (0.093)
Note: Data drawn from the – survey. Shading indicates a category of associations in which individuals without high school degrees are significantly underrepresented. Bold text indicates a category in which popular-sector individuals are significantly overrepresented. *For the hypothesis that the representation ratio is equal to , p < ..
each representation ratio in this and subsequent tables, a standard error and the results of significance testing are reported.20 Looking across the four cases, it appears that individuals from union households are always overrepresented among the group of voters. To some extent, then, links between unions and the party system still exist. Neighborhood associations also quite consistently have shared-member linkages with the group of voters, with the exception of Chile. Members of anticrime associations are overrepresented among voters in Argentina and Peru, while members of parents’ associations are overrepresented in Chile and Peru. Food-related associations are overrepresented only in Venezuela, as are producers’ cooperatives, while sports and recreation associations are only overrepresented in Argentina. Church-based/religious groups are never overrepresented in any city’s group of voters, and indeed such individuals seem somewhat underrepresented in Peru. Half of the associations that have ratios in table . that are significantly greater than are composed disproportionately of non-popular-sector individuals. (The shaded cells in the table are association types for which the popular 20. Because these ratios incorporate sample weights and are probably not normally distributed, significance tests have been conducted and standard errors have been computed using nonparametric bootstrapping (Efron and Tibshirani :–, –). Standard errors involved , bootstrap iterations, while significance tests reflect the results of , iterations. Tests are generated using pivoting, resulting in second-order accuracy and correctness (Hall :–).
05 Chapter 5:05 Chapter 5.qxd
.
11/9/2009
11:55 AM
Page 172
sectors are significantly underrepresented for the more restrictive definition of the popular sectors used in table ..) Of associations whose members are overrepresented in the group of voters, the following also significantly underrepresent the popular sectors: sports, anticrime, and neighborhood associations in Argentina; labor unions in Peru; and neighborhood associations, producers’ cooperatives, and labor unions in Venezuela. None of these linkages is counterbalanced by linkages overrepresenting the popular sectors. Hence, for both Argentina and Venezuela, linkages between civil society and the partisan act of voting involve the overrepresentation of a disproportionately middle- and upper-class group. For Peru there may be some pressure in the same direction, although only via labor unions. In Chile connections between civil society and the act of voting are generally somewhat weak—and the linkages that are significantly different from involve more class-neutral association types. These patterns accord reasonably well with the findings, reported earlier, that turnout in Chile is almost totally class-neutral, that turnout in Peru is nearly class-neutral, and that turnout in Argentina and Venezuela shows a clear pattern of popular-sector underrepresentation. At least with respect to the decision to vote, the Verba, Nie, and Kim () explanation of differences in class representation based on the organizational system thus seems to be plausible. Shared-Member Linkages Between Party Identifiers and Civil Society As a further step in exploring shared-member linkages between the institutions of electoral representation and the world of civil society associations, it is useful to move from voting to identification with each major political party and the overlap between party identification and the various categories of associations. In comparison with other modes of engagement in the electoral/partisan arena, having a party identification is quite common. This allows different political parties to be considered separately—an approach that turns out to be impossible, due to problems of sample size, in the analysis of the more intense modes of partisan participation considered below.21 21. Obviously, such an analysis would have been possible with vote choice, a concept related to voter turnout. However, vote choice is a distinctive enough process that the political behavior literature has largely treated it as a topic separate from the question of participation. Furthermore, it is unclear whether there are normative reasons to want the popular sectors to be equally represented in the electorates of every party. Obviously, it is also possible to analyze party identifiers as a unified group within each society, rather than breaking identifiers down by party. Results for such an analysis are available from the author. They are not reported in the text because such an analysis produces few insights not already available from the party-by-party analysis. Three partial exceptions arise. Sports and recreation associations in Argentina have a significant representation ratio for pooled party
05 Chapter 5:05 Chapter 5.qxd
11/9/2009
11:55 AM
Page 173
,
Table . Civil Society Participation, Party Identification, and the Popular Sectors in Argentina (representation ratios) Association Type Neighborhood Food-Related Anticrime Parents’ Producer’s Cooperatives Labor Union Household Place-of-Origin Church/Religious Sports/Recreation
Peronist 0.713 0.853 0.578 1.134 1.243 1.213 0.860 0.992 0.984
(0.219) (0.186) (0.231) (0.291) (0.926) (0.151) (0.680) (0.191) (0.180)
Radical 1.641 (0.589) 1.067 (0.352) 2.656* (0.956) 1.361 (0.535) NA 1.132 (0.236) NA 1.090 (0.324) 0.952 (0.284)
Note: Estimates of the representation ratios for identifiers of the two major Argentine political parties. Standard errors in parentheses are determined by nonparametric bootstrapping. Shading indicates a category of associations in which individuals without high school degrees are significantly underrepresented. The data in this table indicate that there is no category in which popular-sector individuals are significantly overrepresented. *For the hypothesis that the representation ratio is equal to , p < ..
Tables ., ., ., and . show estimates of the representation ratios for relationships among the political parties of Argentina, Chile, Peru, and Venezuela, respectively, and various categories of civil society organizations. Looking across the four tables, the overwhelming general impression is that sharedmember linkages between party identification groups and civil society associations are quite weak in these countries. None of the countries has more than four representation ratios that are significantly different from one; in almost every case, the proportion of civil society association participants among identifiers of the party is about the same as the proportion among the rest of society. Partisans do not disproportionately represent the points of view of civil society participants, and civil society linkages are probably not responsible for the overrepresentation of middle- and upper-class individuals in the groups of party identifiers in any of the four cases.
identification due to large numbers of participants among identifiers with two minor parties (Afirmación para una República Igualitaria, or , and the Partido Socialista). Argentine unions have a significant representation ratio for pooled party identification because of very large numbers of union members among identifiers with one minor party (). Finally, Peruvian church-based and religious associations have a significant representation ratio for pooled party identification because of large numbers of participants among identifiers with Peru Posible and especially the Christian Democratic smaller party Unidad Nacional. All other significant results in the pooled analysis are present in the party-by-party analysis; party-by-party consideration also reveals several significant ratios that are obscured in the pooled analysis.
05 Chapter 5:05 Chapter 5.qxd
.
11/9/2009
11:55 AM
Page 174
Table . Civil Society Participation, Party Identification, and the Popular Sectors in Chile (representation ratios) Association Type Neighborhood Food-Related Anticrime Parents’ Producer’s Cooperatives Labor Union Household Place-of-Origin Church/Religious Sports/Recreation
0.819 (0.282) 1.031 (0.415) 1.132 (0.500) 1.369 (0.418) 1.897 (1.298) 1.655* (0.326) NA 0.996 (0.177) 0.945 (0.198)
1.643 1.490 1.360 1.222 3.291 0.921
(0.421) (0.508) (0.535) (0.379) (1.844) (0.223) NA
1.238 0.959
(0.192) (0.205)
Socialist 1.610 (0.386) 0.483 (0.240) 1.610 (0.535) 1.423 (0.392) 0.911 (0.705) 1.753* (0.317) NA 0.876 (0.159) 1.258 (0.218)
Note: Estimates of the representation ratios for identifiers of the three largest Chilean political parties. Standard errors in parentheses are determined by nonparametric bootstrapping. Shading indicates a category of associations in which individuals without high school degrees are significantly underrepresented. Bold text indicates a category in which popular-sector individuals are significantly overrepresented. *For the hypothesis that the representation ratio is equal to , p < ..
There are, however, some important exceptions to this general point. In Argentina, people active in anticrime associations are significantly overrepresented among identifiers with the Radical Party. The party might consequently pay distinctive attention to issues of order and security; at the same time, the Radical Party may use anticrime associations to some extent as a political resource. It is also noteworthy that both the Peronists and the Radicals have a representation ratio of about one with respect to labor unions. The surprising weakness of union members among Peronist identifiers may be a consequence of the restructuring of party-union relationships during the s (McGuire :–; Levitsky :–). For Chile, the only ratios that are significantly different from zero involve labor unions. Those from labor union households are significantly overrepresented among the group of identifiers with the Christian Democratic Party ( )—a result that is perhaps surprising because unions were historically affiliated with parties of the left, although the has traditionally had support among white-collar unionists. Less surprising are the relationships between identification with the Socialist Party ( / ) and involvement in labor unions. The shared-member linkage between the party and labor unions suggests that traditional leftist concerns about work relations among the industrial proletariat may still have a voice within the Chilean Socialist Party—its dramatic ideological moderation since the s notwithstanding. It may also be noteworthy that, for two parties ( and the Socialists), neighborhood
05 Chapter 5:05 Chapter 5.qxd
11/9/2009
11:55 AM
Page 175
,
Table . Civil Society Participation, Party Identification, and the Popular Sectors in Peru (representation ratios) Association Type Neighborhood Food-Related Anticrime Parents’ Producer’s Cooperatives Labor Union Household Place-of-Origin Church/Religious Sports/Recreation
Perú Posible 1.172 0.829 1.094 0.887 1.198 1.382 2.578* 1.237 0.913
(0.227) (0.203) (0.281) (0.195) (0.561) (0.267) (1.110) (0.181) (0.172)
0.991 1.218 1.040 1.155 0.811 1.061 1.069 1.103 1.154
(0.145) (0.159) (0.180) (0.146) (0.304) (0.167) (0.415) (0.118) (0.124)
Unidad Nacional 0.928 0.940 1.429 1.207 1.183 1.316 0.739 1.036 1.011
(0.208) (0.219) (0.329) (0.222) (0.628) (0.289) (0.454) (0.175) (0.182)
Note: Estimates of the representation ratios for identifiers of the three largest Peruvian political parties. Standard errors in parentheses are determined by nonparametric bootstrapping. Shading indicates a category of associations in which individuals without high school degrees are significantly underrepresented. Bold text indicates a category in which popular-sector individuals are significantly overrepresented. *For the hypothesis that the representation ratio is equal to , p < ..
associations have representation ratios that approach statistical significance; there may thus be a second connection between civil society and the group of party identifiers. The data for Peru show almost no shared-member linkages between groups of party identifiers and civil society associations. Perú Posible, the governing party at the time of the survey, has a linkage only with place-of-origin associations. This linkage may be due to the mixed European-indigenous identity of Perú Posible’s leader, Alejandro Toledo: Lima migrants who would participate in an association based on origin might tend to identify with Toledo. However, this linkage—while statistically significant—is substantively marginal. Only percent of respondents in Lima report involvement in this type of association. Unidad Nacional and show no signs of any shared-member linkages between the group of its partisans and any category of civil society association. The weakness of these linkages, which is shared across all three Peruvian parties, strongly suggests that, at the mass level, the country’s current system of party identifications lacks meaningful ties to organized society. This lack raises the possibility that other modes of linkage, including policy-responsive and clientelist linkages, may be weak in the party system as currently constituted. For Venezuela the patterns are quite different across the parties. The data show few signs of shared-member linkages between Movimiento Quinta República () or Acción Democrática ( ) identifiers and civil society associations. At the time of the survey, in early , was still in the relatively
05 Chapter 5:05 Chapter 5.qxd
.
11/9/2009
11:55 AM
Page 176
Table . Civil Society Participation, Party Identification, and the Popular Sectors in Venezuela (representation ratios) Association Type Neighborhood Food-Related Anticrime Parents’ Producer’s Cooperatives Labor Union Household Place-of-Origin Church/Religious Sports/Recreation
0.684 0.816 1.168 1.303 1.981* 0.986 0.586 1.027 1.054
(0.149) (0.320) (0.346) (0.457) (0.612) (0.192) (0.603) (0.219) (0.172)
1.043 2.047 1.868 1.599 1.524 1.407
(0.384) (0.969) (0.808) (0.794) (0.944) (0.424)
NA 1.882* (0.546) 0.566 (0.246)
Primero Justicia 1.343 2.347* 0.719 1.447 1.676 1.319 1.763 1.966* 1.538
(0.354) (0.900) (0.367) (0.663) (0.765) (0.343) (1.494) (0.485) (0.327)
Note: Estimates of the representation ratios for identifiers of the three largest Venezuelan political parties. Standard errors in parentheses are determined by nonparametric bootstrapping. Shading indicates a category of associations in which individuals without high school degrees are significantly underrepresented. The data in this table indicate that there is no category in which popular-sector individuals are significantly overrepresented. *For the hypothesis that the representation ratio is equal to , p < ..
early stages of formation as a party, while —one of Venezuela’s two traditional parties—had recently all but collapsed. The more striking result for Venezuela involves the party Primero Justicia. This party, relatively small but well known for its belligerent opposition to Venezuela’s president Hugo Chávez, has more and stronger shared-member linkages to civil society associations than any other party in this study. Considering the party’s recent origins and the weakness of linkages in the other two Venezuelan parties under consideration, Primero Justicia’s linkage profile is quite remarkable and is consistent with accounts that emphasize the early organization of urban property owners and business groups (Levine and Romero :). Summing up, the major finding of this section is that groups of party identifiers in these South American capital cities have few shared-member linkages with civil society associations. The common perception of civil society as a sphere that is relatively autonomous from the electoral/partisan arena is consistent with the finding that there is no shared-member linkage between groups of partisans and civil society participants. Instead, partisans in these four countries are in general no more likely to be active in civil society associations than their nonpartisan fellow citizens. As a consequence, civil society has little or no net effect on representational distortion within the body of party identifiers; the patterns of popular-sector underrepresentation on this variable in Chile and Venezuela must have some other explanation.
05 Chapter 5:05 Chapter 5.qxd
11/9/2009
11:55 AM
Page 177
,
Shared-Member Linkages Between Recipients of Party Mobilization Efforts and Civil Society Patterns of class involvement in civil society and linkages between associational participation and partisan engagement seem to reasonably account for differences across capital cities in the relationship between social class and voting, but not in the relationship between social class and having a party identification. Moving forward to party mobilization, we have seen that the popular sectors are underrepresented in Argentina and Chile, but that there is a pattern of class-neutral representation in Venezuela and Peru. Do these differences correspond with patterns of shared-member linkage between civil society and partisan politics? Unfortunately, too few partisan activists or recipients of party mobilization are available within the survey sample to allow party-by-party analysis as in the previous section on party identification. To solve this problem, the data are pooled for all recipients of party mobilization and for all partisan activists within a given country, regardless of the party within which they are active. Pooling the data in this way may tend to average across and therefore obscure distinctive traits of specific parties. At the same time, however, it is useful in highlighting cross-national differences. Table . shows the estimated representation ratios for being the recipient of mobilizing visits by a party leader. These ratios are generally higher than Table . Civil Society Participation, Party Mobilization, and the Popular Sectors (representation ratios) Association Type
Argentina
Chile
Peru
Venezuela
Neighborhood Food-Related Anticrime Parents’ Producer’s Cooperatives Labor Union Household Place-of-Origin Church/Religious Sports/Recreation
1.814* (0.490) 1.765* (0.371) 2.101* (0.665) 1.829* (0.497)
2.425*(0.496) 2.245*(0.632) 3.364*(0.892) 1.896*(0.453)
1.393* (0.180) 1.407* (0.184) 1.577* (0.250) 1.421* (0.172)
3.005* (0.721) 3.730* (1.343) 2.238 (0.896) 4.290* (1.496)
7.866* (5.875)
1.929 (1.004)
2.033* (0.629)
3.736* (1.316)
1.211 (0.181) 1.182 (0.974) 1.415 (0.279) 1.806* (0.315)
1.517*(0.274) NA 1.346 (0.198) 1.430*(0.230)
1.957* (0.267) 0.744 (0.305) 1.262* (0.130) 1.099 (0.130)
2.304* (0.565) 3.858 (3.222) 2.054* (0.603) 1.580 (0.399)
Note: Data drawn from the – survey. Shading indicates a category of associations in which individuals without high school degrees are significantly underrepresented. Bold text indicates a category in which popular-sector individuals are significantly overrepresented. *For the hypothesis that the representation ratio is equal to , p < ..
05 Chapter 5:05 Chapter 5.qxd
.
11/9/2009
11:55 AM
Page 178
those for voting or having a party identification. In fact, a clear majority of the estimated ratios reach statistical significance. Furthermore, setting aside ratios involving place-of-origin associations, most ratios that are not significant closely approach significance. People with virtually any variety of involvement in civil society are overrepresented among the group of those who have received mobilization from party leaders. Does this pattern of generalized linkage between civil society and party mobilization help explain the underrepresentation, among recipients of party mobilization, of the Chilean and Argentine popular sectors, as well as the essentially class-neutral patterns in Peru and Argentina? These linkages may indeed be part of the explanation, in light of the clear class differentials in civil society participation in Argentina, Chile, and Venezuela, and the more modest differences in Peru, it may in part. The relatively lower extent of class distortion among recipients of party mobilization in Venezuela remains in need of some other explanation, but the other three capital cities fit the pattern. Hence, as with turnout and in contrast to party identification, the strength of the organizational system may help explain many of the cross-national differences in popular-sector representation regarding party mobilization. Shared-Member Linkages Between Party Activists and Civil Society To what extent can characteristics of civil society account for cross-national differences in popular-sector representation in the group of party activists, defined as those who visit party offices, attend party rallies, or work for campaigns? Activists within political parties have more opportunities to make the details of their beliefs and preferences known than mere partisans, and activists can sometimes express themselves in more consequential ways. Hence, if organized civil society is to have a distinctive voice in partisan politics, it is likely to involve shared-membership linkages from the group of party activists. Such linkages, if they exist, may also help explain the underrepresentation of the popular sectors in the group of party activists in Argentina, Chile, and, to a lesser extent, Peru. The data about shared-member linkages between the group of party activists and civil society associations are presented for the four countries in tables . through .. The major finding is that linkages are much stronger for activists than they are for party identifiers or even for voters, and comparable to those for party mobilization. In every country, at least ten of the twenty-seven possible shared-member linkages evidently exist; in Peru and in Venezuela, relatively few of the available linkages are not statistically significant.
05 Chapter 5:05 Chapter 5.qxd
11/9/2009
11:55 AM
Page 179
,
Table . Civil Society Participation, Party Activism, and the Popular Sectors in Argentina (representation ratios) Association Type Neighborhood Food-Related Anticrime Parents’ Producer’s Cooperatives Labor Union Household Place-of-Origin Church/Religious Sports/Recreation
Attends Party Rallies 2.307* 1.158 1.791 1.076 1.378 1.323 1.625 1.466 1.247
(0.679) (0.326) (0.692) (0.417) (1.197) (0.224) (1.475) (0.354) (0.294)
Visits Party Office 1.631 2.153* 1.905 2.339* 3.922 1.890* 1.906 1.695* 1.730*
(0.456) (0.449) (0.656) (0.590) (2.742) (0.230) (1.457) (0.337) (0.302)
Works for Party 3.304* 2.202* 4.340* 1.876 2.253 2.073* 2.213 1.055 1.186
(1.006) (0.664) (1.627) (0.829) (2.648) (0.347) (1.754) (0.427) (0.418)
Note: Estimates of the representation ratios for various kinds of activists in any of the Argentine political parties. Standard errors in parentheses are determined by nonparametric bootstrapping. Shading indicates a category of associations in which individuals without high school degrees are significantly underrepresented. The data in this table indicate that there is no category in which popular-sector individuals are significantly overrepresented. *For the hypothesis that the representation ratio is equal to , p < ..
Across the board, then, there is quite strong evidence of shared-member linkages between the group of party activists and civil society organizations. For each country, shared-member linkages with religious associations are relatively weak. Hence, the church, one of the two traditionally central civil society organizations in discussions of Latin American politics and generally among the most class-neutral of the civil society organizations in this study, seems distinctively lacking in shared-member linkages with the group of party activists. The linkages between the group of partisans and religious associations are also weak, as seen in the previous section. In effect, the church is not well represented among the groups of people most involved in party politics. The potential class-equalizing effects of religious involvement thus seem to be unrealized in each of the four capital cities.22 Labor unions, the other traditional locus of party recruitment, have a much stronger set of connections with party activism than do churches; for virtually 22. Other studies of the relationship between church involvement and political participation in Latin America agree that the connection is, at best, complex and further propose that the relationship may be more evident for Latin America’s Protestant minority than for its Catholic majority (see, for example, Patterson ). This suggests that, if the Protestant community in Latin America continues to expand, the church-related representation ratios considered in this section may become larger over time. Given the relative class neutrality of religious associations, such a development would most likely tend to improve popular-sector representation in the region.
05 Chapter 5:05 Chapter 5.qxd
.
11/9/2009
11:55 AM
Page 180
Table . Civil Society Participation, Party Activism, and the Popular Sectors in Chile (representation ratios) Association Type Neighborhood Food-Related Anticrime Parents’ Producer’s Cooperatives Labor Union Household Place-of-Origin Church/Religious Sports/Recreation
Attends Party Rallies
Visits Party Office
Works for Party
2.821* (0.691) 2.685* (0.941) 2.814* (0.933) 0.946 (0.385) 2.197 (1.584) 2.183* (0.414) NA 0.762 (0.211) 2.346* (0.386)
3.023* (0.824) 3.523* (1.232) 2.825* (1.157) 1.966 (0.695) 4.688* (2.806) 3.199* (0.553) NA 0.687 (0.226) 1.771* (0.421)
1.897 (0.797) 1.152* (0.776) 3.345* (1.460) 0.760 (0.543) 4.819 (3.485) 3.087* (0.634) NA 0.827 (0.320) 1.711 (0.484)
Note: Estimates of the representation ratios for various kinds of activists in any of the Chilean political parties. Standard errors in parentheses are determined by nonparametric bootstrapping. Shading indicates a category of associations in which individuals without high school degrees are significantly underrepresented. Bold text indicates a category in which popular-sector individuals are significantly overrepresented. *For the hypothesis that the representation ratio is equal to 1, p < ..
every form of party activism in each of the countries under study, labor unions have a higher representation ratio than do church-based/religious associations. Indeed, in Chile and Venezuela all three representation ratios between party activists and labor unions are above .. In Peru and Argentina the relevant ratios are lower but still mostly significantly higher than .. Labor union members, of course, are significantly less likely to be from the popular sectors than the average respondent in two of the four countries, and the data for Argentina points in the same direction without reaching significance—so the overrepresentation of union household members among party activists in Argentina and especially in Venezuela (where unions more dramatically overrepresent the middle and upper classes) may further erode the voices of lowerincome individuals in party politics. People in labor union households in Peru are also distinctively less likely to be from the popular sectors, so the overrepresentation of such people among party activists means that the relatively class-neutral patterns in Peru for office visits and party work exist in spite of significant inequitable effects of labor unions. Beyond unions and religious associations, in all four countries neighborhood associations, food-related associations, anticrime associations, and producers’ cooperatives sometimes have higher representation ratios. Therefore, the groups of party activists in these four countries seem to be linked into a wide range of civil society organizations, expanding beyond labor unions and
05 Chapter 5:05 Chapter 5.qxd
11/9/2009
11:55 AM
Page 181
,
Table . Civil Society Participation, Party Activism, and the Popular Sectors in Peru (representation ratios) Association Type Neighborhood Food-Related Anticrime Parents’ Producer’s Cooperatives Labor Union Household Place-of-Origin Church/Religious Sports/Recreation
Attends Party Rallies 1.894* 1.314 2.103* 1.605* 2.440* 1.845* 3.155* 1.072 1.605*
(0.244) (0.211) (0.337) (0.204) (0.811) (0.273) (1.005) (0.140) (0.183)
Visits Party Office 1.648* 1.324 2.299* 1.204 2.761* 1.606* 2.479* 1.473* 1.719*
(0.248) (0.217) (0.372) (0.190) (0.909) (0.273) (0.886) (0.173) (0.199)
Works for Party 1.548 0.976 2.526* 1.668* 3.741* 2.400* 3.052* 1.352 1.488
(0.361) (0.325) (0.560) (0.340) (1.655) (0.504) (1.332) (0.263) (0.292)
Note: Estimates of the representation ratios for various kinds of activists in any of the Peruvian political parties. Standard errors in parentheses are determined by nonparametric bootstrapping. Shading indicates a category of associations in which individuals without high school degrees are significantly underrepresented. Bold text indicates a category in which popular-sector individuals are significantly overrepresented. *For the hypothesis that the representation ratio is equal to , p < ..
church-related associations, which served as the traditional sites of mobilization in the previous interest regime. In Peru, where participants in most of these kinds of associations are roughly representative of the population as a whole, the majority of these additional linkages probably have no effect, one way or the other, on popular-sector representation among party activists. Producer’s cooperatives do underrepresent the popular sectors, so the linkages there in conjunction with the labor-union linkages noted above tend to push toward representational distortion in this domain. Overall, then, observed patterns of civil society participation and party-association linkage are basically consistent with our earlier finding of moderate representational distortion regarding party activism. In Venezuela, several association types underrepresent the popular sectors, but overall participation in civil society involves somewhat less representational distortion. Hence, when linkages are strong across most categories of civil society, as in this section, the end result may be expected to be more classneutral than when only a few association types have strong linkages. This is, of course, consistent with the finding of relatively low representational distortion with respect to these activist modes of engagement in the electoral/ partisan arena. For the other two capital cities, however, most of the linkages found in this section will strengthen middle- and upper-class overrepresentation in the
05 Chapter 5:05 Chapter 5.qxd
.
11/9/2009
11:55 AM
Page 182
Table . Civil Society Participation, Party Activism, and the Popular Sectors in Venezuela (representation ratios) Association Type Neighborhood Food-Related Anticrime Parents’ Producer’s Cooperatives Labor Union Household Place-of-Origin Church/Religious Sports/Recreation
Attends Party Rallies 2.568* 2.128 3.789* 1.204 1.797 2.193* 3.874* 1.013 2.523*
(0.585) (0.882) (1.167) (0.531) (0.703) (0.494) (2.706) (0.354) (0.455)
Visits Party Office 3.080* 3.450* 3.874* 2.058 4.607* 3.107* 2.920 2.133* 2.341*
(0.790) (1.424) (1.372) (1.040) (1.700) (0.683) (2.722) (0.649) (0.526)
Works for Party 3.519* 3.337* 4.441* 3.852* 4.872* 3.636* 4.803 2.710* 2.142*
(0.892) (1.469) (1.552) (1.668) (1.814) (0.778) (4.366) (0.839) (0.554)
Note: Estimates of the representation ratios for various kinds of activists in any of the Venezuelan political parties. Standard errors in parentheses are determined by nonparametric bootstrapping. Shading indicates a category of associations in which individuals without high school degrees are significantly underrepresented. The data in this table indicate that there is no category in which popular-sector individuals are significantly overrepresented. *For the hypothesis that the representation ratio is equal to , p < ..
group of party activists, since people from those classes are generally overrepresented among participants in civil society. Thus, these generalized linkages are roughly consistent with the finding that the popular sectors are underrepresented both among party office visitors and among rally attendees in Argentina and Chile, while they are essentially equitably represented in Venezuela. In summary, while shared-member linkages between groups of party identifiers and civil society associations are generally weak, connections involving voters, recipients of party mobilization, and party activists are far stronger. Civil society organizations may be relatively autonomous from some aspects of mass electoral politics, but they have distinctly closer ties with the more intense partisan politics of activism and campaign participation. At this more intense level, the linkage between labor union involvement and partisan participation remains vibrant, although the tight organizational linkage that once existed between unions and parties has been greatly loosened if not abandoned. However, individuals from other kinds of popular organizations are equally overrepresented. As this observation about civil society in general holds for the specific, programmatic associations in the A-Net, we can further observe that the interest regime organized around the A-Net clearly has a substantial party component—although the lack of strong linkages involving party identifiers suggests that the new interest regime’s party component is quite different from that of the prior regime based on the -Hub. In the new regime,
05 Chapter 5:05 Chapter 5.qxd
11/9/2009
11:55 AM
Page 183
,
associations are less often incorporated wholesale into a political party, and associations play much less of a role in mobilizing or activating a party identification. Instead, linkages are based on patterns of individual behavior, as those more active in one domain also tend to participate in the other. Since such individuals will not check their partisan or associational identities at the door, such shared-member linkages form an important, if perhaps less institutional, component of the A-Net. It is less clear from the data considered here whether political parties play a meaningful structuring role in Latin American civil society, as they did in many countries during earlier political and economic periods. In particular, the individual-level evidence used in this chapter can tell us little about whether parties use civil society organizations as a strategy for building coalitions and mobilizing voters, or whether the linkages documented above arise solely through the choices of autonomous individuals. Direct information about the organization and strategy of Latin American political parties during the period of the A-Net would be needed to cast light on these issues. In any case, the existing patterns of linkages between civil society and party politics, in conjunction with evidence regarding the social classes prevalent in civil society, are generally consistent with the idea that the nature of the organizational system can help account for the degree of popular-sector underrepresentation within the system of partisan and electoral competition. Those cases in which the electoral/partisan arena is relatively class-neutral, such as Peru and, to a lesser extent, Venezuela, have reasonably class-neutral civil societies and/or relatively weak linkages between civil society and political parties. Those characterized by greater representational distortions, such as Chile and Argentina, combine extensive linkages with inegalitarian civil societies. Redressing these representational distortions may first require working within civil society to more adequately include the popular sectors in the city’s organizational life.
05 Chapter 5:05 Chapter 5.qxd
11/9/2009
11:55 AM
Page 184
06 Chapter 6.qxd
11/10/2009
9:59 AM
Page 187
6 : Diana Kapiszewski
This volume asserts that a new interest regime centered on associational networks (the A-Net) is emerging to organize collective problem solving among the popular sectors in urban South America. This chapter and the next two analyze three crucial facets of that interest regime: the strategies that associations adopt to pursue their interests, the ways in which they coordinate in order to do so, and relations to unions and parties. More broadly, the chapters evaluate what the emergence of the A-Net might imply for the representation of South America’s urban popular sectors. As chapter describes, and in line with what Cavarozzi () referred to as the “state-centric matrix,” the “Fordist” state played expansive economic and social roles, and structures of interest intermediation were oriented toward state claim making during Latin America’s era. The region’s transition to a market-based economic model over the last fifteen years entailed the emergence of a new “neoliberal” state focused less on economic and social roles. Further, Latin American countries have engaged in varying degrees of decentralization from the central government to the provincial- or municipal-level state (although the national government obviously remains an important policymaker). Simultaneously, inequality has increased in many countries of the region, and the work force has become more informal.1 * The author would like to thank Wendy Sinek for her comments. 1. See discussion of the shift in interest regime and rise of associationalism in chapter , as well as, among many others, Oxhorn (), Hagopian (), Roberts (; ), and Kurtz ().
06 Chapter 6.qxd
11/10/2009
.
9:59 AM
Page 188
-
These simultaneous transformations raise the crucial question of what strategies the organizations of the popular sectors employ, and to whom they turn, to pursue their interests and address their problems. This chapter uses data from the survey of associations to explore those issues. The first goal of the chapter is to examine problem-solving strategies in the A-Net along two key axes. One axis regards the orientation of popular-sector problem solving—analyzing the distinction between state-targeted (or claimmaking) strategies on the one hand, and society-targeted (or societal) strategies of providing social constituencies with goods, services, or information (provisioning), and building associational capacity on the other.2 The former are oriented toward various levels of government and, when pursued through institutional channels, are what chapter refers to when discussing the idea of accessing strategies. The latter are oriented toward associations’ constituencies or other associations. The analysis reveals that the A-Net is a dual-action interest regime. In the aggregate, popular-sector associations in contemporary urban South America employ a heterogeneous mix of state-targeted and society-targeted strategies. The A-Net is also dual in the individual associational sense: most associations employ both claim-making and societal strategies. Thus, though many popular associations may have their roots in “subsistence” organizing to meet economic needs, this analysis of the core segment of the A-Net suggests that the new interest regime serves both as a structure for coordinating society-targeted provisioning activities, and has some potential to channel claim making on the state—that is, to play a role in people’s efforts at achieving political representation. A second axis then, relates to the balance between institutional claim making—voicing demands through official channels—and contention. The ability of an interest regime’s component organizations to make claims on state officials through multiple channels, offering their constituents various forms of voice, is fundamental to the system’s capacity to serve as an effective vehicle for interest representation. Labor unions in the -Hub enjoyed some degree of institutional access to state policymakers and also engaged in contentious politics (though the balance between the two strategies varied substantially from country to country). This arrangement helped the labor movement win important concessions regarding wage bargaining, labor regulation, and macroeconomic policy. Regarding other types of popular organizations, the contemporary literature examines both forms of state-targeted interest representation: 2. As chapter explains, this study considers both state-targeted and society-targeted strategies to be “political.”
06 Chapter 6.qxd
11/10/2009
9:59 AM
Page 189
,
many studies of interest groups and of civil society organizations focus on institutional claim making on the state; the social movement literature tends to emphasize different forms and cycles of contentious politics, and the literature on social movement organizations discusses both. This chapter demonstrates that while associations make claims on the state via institutional channels more often than they utilize more contentious statetargeted strategies, organizing protest is not uncommon. Further, organizing protest is not a strategy of “last resort” used by associations that are less able or otherwise likely to press claims using institutional channels: most often, organizing protest serves as a complement to associational engagement in other types of strategies. In short, along this axis as well, both the A-Net and its component organizations are marked by duality. The chapter’s second goal is to examine a series of associational qualities: associational ties to the state, associational resources (both material and organizational), and inter-associational linkages (or the degree to which associations are “embedded” in associational networks). These qualities are analyzed both to draw descriptive conclusions about the associations that compose the core of each national interest regime and to elucidate cross-national differences in fundamental A-Net traits. The chapter’s third goal is to explore the relationships that exist between these qualities and particular repertoires of problem-solving strategies. The issue of state-association ties is a key thematic touchstone for exploring the new interest regime. Ties with the state often come in the form of selective benefits or access to state officials, yet they can also make associations directly dependent on the state and thus challenge associational autonomy. Indeed, as noted in chapter , this double-edged relationship was fundamental in the -Hub, in which unions were controlled by the state through “inducements” (i.e., the application of benefits) as well as by constraints (Collier and Collier ). While unions had preferential access to the state, they were vulnerable to co-optation and control by the centralized leadership, particularly in settings where state officials were eager to use them as organizations of control. While we cannot assess the degree to which ties to the state erode associational autonomy, we can examine the incidence of these ties and the relationship between them and strategy choice. This chapter suggests that the component associations of the A-Net are not “disarticulated,” as more than half are tied to the state in some way. Further, two findings emerge regarding the relationship between associations’ ties to the state and their problem-solving strategies. Associations possessing state ties more actively pursued institutional claim making than did associations
06 Chapter 6.qxd
11/10/2009
.
9:59 AM
Page 190
-
without such ties, suggesting an affinity between different kinds of connections with and appeals to the state. Yet, having ties to the state does not appear to make associations notably more cautious about engaging in contention, as associations with and without such ties organized protest at more or less the same rates. In short, the evidence suggests that associations are quite oriented toward the state, and that the state may also shape their actions to some degree. The amount and type of resources associations possess are crucial to their functioning, and thus to the operation of the interest regime. The associations of the A-Net have been characterized as less institutionalized and possessing fewer resources than the heavily bureaucratic and well-funded union organizations that characterized the -Hub. In line with social movement theorists who have focused on organizational and financial resources as key factors affecting cycles of contention, chapter suggests that the lower levels of resources in the A-Net may complicate coordination among associations. This chapter explores a related idea, that resource levels may shape the strategies that associations undertake to solve problems. The analysis confirms that the A-Net’s component associations are poor materially, but implies that they possess more organizational resources than is advanced by some images that emphasize their fluidity and lack of institutionalization. Regarding the connection between resources and strategies, associations with more material resources are in fact more likely to use society-targeted problem-solving strategies than are poorer organizations, while associations with greater organizational resources appear to be better positioned to utilize institutional channels to make claims on the state than are their less-institutionalized counterparts. The number of links that associations have formed within associational networks (i.e., their “embeddedness”) is one aspect of what this volume refers to as “scaling.” While the image of the A-Net entailing fluid networks (in contrast to the structured hierarchy of the -Hub) in which associations may be intermittently connected is pervasive in the literature, how much these associations actually band together to engage in collective action—that is, the “density” of the A-Net’s infrastructure—remains an open question. Although case study analysis is necessary in order to examine thoroughly how networks function and to what extent they are activated for collective action, this chapter and the next advance general statements about scaling and its implications for associational problem solving. This chapter shows that associations are relatively well embedded in associational networks. Regarding the relationship between particular associational qualities and problem-solving repertoires, the strongest and most consistent relationships emerge between associational embeddedness
06 Chapter 6.qxd
11/10/2009
9:59 AM
Page 191
,
and choices regarding strategies: better-linked associations are more likely than are more atomized associations to employ each one of the strategies under study. These findings highlight the importance of scaling within the A-Net, an issue examined in greater depth in the next chapter.
Method The questions of coordination and collective action are central ones in scholarship on civil society and social movements, the two analytic lenses most frequently employed to examine the associational world in Latin America. Often, the empirical referent of this scholarship has been a relatively small number of popular associations. The sample, by contrast, contains almost a thousand associations, affording a broader perspective on problem-solving strategies, associational ties to the state, organizational and material resources, and the incidence of associational coordination. A few additional aspects of the sample bear mentioning before proceeding with the analysis. As appendix B explains in detail, the associations that constitute the sample were selected using a chain-referral or snowball technique. By its nature, a snowball sampling strategy is biased toward associations that work more with others—and, likely, toward more salient and active associations—since the snowball itself depends on referrals from one association to another. Consequently, the sample is neither random nor representative of popular-sector associations in the four capital cities. As further discussed in appendix B, this sample suited our analytic purposes. Our goal was not to derive an “average distribution” of associations on any trait (since seeking to do so would likely have placed in our sample many weak, ephemeral, and relatively inactive associations). Nor was the objective to generalize our findings to the whole “universe” of associations. Instead, the project sought to analyze whether the A-Net in contemporary South America includes a relatively more visible, connected, and active subgroup of associations—a “core segment” with the capacity to act as a structure of interest representation.3 Moreover, a chain-referral technique was in fact one of the only alternatives available. No contemporary study of associationalism in Latin America analyzes a random or representative sample of civil society organizations, as obtaining such samples depends on having 3. The project thus follows the pattern of many studies of the labor movement, with their focus on the more prominent, active unions and confederations.
06 Chapter 6.qxd
11/10/2009
.
9:59 AM
Page 192
-
a sampling frame of all associations, while complete and accurate lists of associations do not exist.4 The fact that associations were sampled using a chain-referral technique does not invalidate cross-national comparison (though caution is of course necessary when making such comparisons), given that the chain-referral procedure was conducted in a similar fashion in each city. Moreover, oversampling betterlinked associations provides additional leverage for understanding associational scaling (i.e., inter-associational linkages and coordination), one of this volume’s key areas of inquiry.5 Finally, the study explores the importance association leaders assign to problem-solving strategies, rather than the outcome when such strategies are employed (i.e., whether associations manage to solve problems or wield influence on policymaking). Presumably the importance with which association leaders regard a given strategy both serves as an indicator of the relative weight of that strategy within a larger repertoire,6 and reflects something about the degree to which the strategy is effective or worthwhile. Nonetheless, the present study has no independent evidence regarding and makes no claims about the effectiveness of any strategy in terms of actually solving problems. Many factors affect associational influence and problem-solving success, and such an investigation is beyond the scope of this analysis. The goal here is to take a first step toward analyzing the organizational infrastructure of the associational world available to the popular sectors, exploring and comparing crossnationally the more active and better linked “core segment” of the A-Net.
Problem-Solving Strategies of Popular Associations It seems inevitable that the transition from an state to one based on neoliberal principles that occurred in many Latin American countries since the s would be accompanied by changes in popular-sector problem-solving strategies 4. Given the nonrandom nature of the sample, it is important to be conservative in assessing “significance.” A critical value of . (which corresponds to a significance level of alpha = .) rather than a critical value of . (which corresponds to a significance level of alpha = .) is utilized for all tests of statistical significance in this chapter. 5. It bears noting that despite the use of chain-referral sampling, each country sample includes many atomized associations. Their presence in the sample offers an opportunity to compare lessconnected and better-linked associations on a variety of characteristics, including their problemsolving repertoires. 6. The chapter thus refers to scales of the “importance” of strategies and of the “frequency” of their use interchangeably.
06 Chapter 6.qxd
11/10/2009
9:59 AM
Page 193
,
and in the form and function of interest representation. In what direction and to what degree the interest regime has shifted, however, remains an open question. While there has been little cross-national inquiry to date,7 excellent case studies of associationalism and associational involvement in problem solving and interest representation in contemporary Latin America have been conducted (e.g., Oxhorn ; C. Schnieder ; Dietz ), some of which have focused specifically on associations of the poor (Assies ; Auyero ; Adams ; Myers and Dietz ). Some scholars have asserted that associations are becoming useful vehicles for popular-sector interest representation (Chalmers, Martin, and Piester ; Foweraker and Landman ; Veltmeyer ),8 while others portray the associational world as largely disarticulated from the state and hence a weak structure for the intermediation of popularsector interests (Roberts ). This section examines the strategy profile of the A-Net in the four countries under investigation in an effort to adjudicate between these two views of the strategy repertoires and representational capacity among associations in Latin America. If associations have become important vehicles for solving popular-sector problems in the contemporary era, what strategies do they choose to carry out that function? Do they most frequently consider societytargeted forms of problem solving important, do they prioritize state-targeted strategies, or do they pursue both strategy genres? If claim making is prevalent, does the A-Net play a role in aggregating and channeling claims to the state through institutional channels, serve as an infrastructure for contentious politics, or both? Strategy Genres: Society-Targeted and State-Targeted In order to assess the strategy repertoires of urban popular associations, we asked associational leaders to indicate how important for their work they considered a range of strategies to be. The strategies about which we inquired can be categorized into two genres, each of which includes two types (outlined in table . below). The first genre of strategies includes society-targeted ones—those that are designed to help another association or deliver a public benefit to an association’s constituency. Much of the literature on civil society in the region emphasizes societal strategies as the main activities of associations. Examples include 7. Roberts and Portes (), based on a large cross-national project, is an exception. 8. Jelin () and Friedman and Hochstetler () share this optimism, though their studies focus only on .
06 Chapter 6.qxd
11/10/2009
9:59 AM
.
Page 194
-
Table . Problem-Solving Strategies Society-Targeted (Societal) Strategies
•
Capacity Building (main target: other associations) • Finance other associations • Train community leaders • Create new organizations
•
Provisioning (main target: members or constituencies) • Distribute goods, services, or information to people or associations • Organize public events
State-Targeted (Claim-Making) Strategies
•
Institutional Claim Making • Contact elected leaders • Contact government agencies • Participate in stated-created fora • Initiate legal actions
•
Organizing Protest
activities aimed at building the capacity of other associations (such as providing financing and training community leaders), and provisioning strategies (offering goods, services, or information—hereafter “goods”—to people or other associations, and organizing public events). The second genre comprises state-targeted strategies. These strategies are often emphasized in the social movement literature and are routinely considered central in the study of interest groups. While they are less-often associated with popular associations in Latin America, they are crucial to the capacity of the A-Net to constitute an organizational infrastructure for political representation. The two types are institutional claim making (contacting elected leaders or governmental agencies, participating in government-sponsored fora, or presenting legal actions), and the more contentious strategy of organizing protest. Analyzing the pooled sample, which includes data from the four capital cities included in the study, allows for the exploration of general tendencies in the A-Net. As the data in table . show,9 the vast majority of associations are active in terms of problem solving. Moreover, in the aggregate, the strategy profile of the A-Net is diverse: claim-making strategies (employed by about percent of associations in the pooled sample) are as important for associations 9. Throughout the chapter, tables present the percent of associations whose leaders indicated that one or more strategies within the genre or type in question was important or very important for their association.
06 Chapter 6.qxd 11/10/2009
Table . Society-Targeted and State-Targeted Strategies (percent of associations) Society-Targeted Capacity Building
73.2 84.6 72.8 67.1 68.3
81.9 97.1* 75.0 74.2 80.8
52.7 49.6 55.7 52.9 52.9
Provisioning
Total StateTargeted
Inst. Claim Making
Org. Protest
72.6 94.6* 68.4 62.1* 65.0
85.5 86.7 90.4 84.6 80.8
82.6 83.3 88.2 82.3 77.9
29.6 32.5 28.8 36.3 20.4
Note: The difference between the percent of associations that employ society-targeted strategies versus state-targeted strategies is statistically significant at approximately the p < . level in Chile. The difference between the percent of associations that engage in capacity building and institutional claim making is statistically significant at approximately the p < . level in each country; the difference between the percent of associations that engage in capacity building and organize protest is statistically significant at approximately the p < . level in each country. The difference between the percent of associations that engage in provisioning and institutional claim making is statistically significant at approximately the p < . level in Chile and Peru; the difference between the percent of associations that engage in provisioning and organize protest is statistically significant at approximately the p < . level in each country. *Difference between the percent of associations in the country indicated in each row versus the percent of associations in all other countries (considered together as one reference group) is statistically significant at approximately the p < . level.
Page 195
(n=948) (n=240) (n=228) (n=240) (n=240)
Total SocietyTargeted
9:59 AM
Total Argentina Chile Peru Venezuela
Both Societyand State-Targeted
State-Targeted
06 Chapter 6.qxd
11/10/2009
.
9:59 AM
Page 196
-
as societal strategies (employed by about percent of associations overall). Of the two strategy types that are prioritized by the most associations, one is a societal strategy type (provisioning), and the other is institutional claim making. Further, as these large percentages suggest, individual associations have a varied repertoire of strategies: almost three-quarters of the associations use both society-oriented and state-targeted strategies. In sum, in contrast to studies that suggest that popular associations most often employ societal strategies, we find that the new interest regime may serve both as a structure for society-targeted activities and as a channel for making claims on the state. Translating these findings into the terms of the first axis outlined above, the A-Net—or at least its more dynamic core segment—is decidedly dual with regard to strategy orientation, both in the aggregate and with regard to individual associations. Regarding cross-national differences, the percentage of associations that use both society-targeted and claim-making strategies exhibits little variation. While national patterns in terms of the relative use of the strategy genres and strategy types considered here are also similar, two exceptions deserve highlighting. First, the difference between the percentage of Chilean associations that prioritizes claim making and the percentage that prioritizes societal strategies is significantly larger than is that difference in any other country. Second, associations in Argentina seem much more oriented to societal strategies (and in particular to provisioning) than are associations in any other country,10 while relatively few Peruvian associations engage in provisioning. Beyond these contrasts, which will be explored in more detail below, the cross-national picture viewed at this level is strikingly homogeneous. Society-Targeted Strategies Table . presents a more disaggregated picture of the two types of societal strategies analyzed here, capacity building and provisioning. Surprisingly— given that popular-sector associations are often portrayed as fighting for survival themselves—more than half of the associations consider building the capacity of other associations to be an important strategy. This finding may bode well for the organizational infrastructure of the A-Net, and it reinforces the assertion regarding the diverse strategy repertoires of its more visible and 10. As we will see below, the large proportion of Argentine associations engaged in provisioning reflects the large percentage of associations that distribute goods (often as part of government programs).
06 Chapter 6.qxd
11/10/2009
9:59 AM
Page 197
,
connected core segment.11 In terms of specific “capacity-building” strategies, associations may fund other associations, train community leaders, or create new associations. As we might expect, very few associations in any country emphasize funding other associations. In Argentina, Chile, and Venezuela roughly equal portions of associations are oriented toward training community leaders and creating new organizations, while in Peru the former strategy is more important for associations than is the latter. Provisioning strategies—including distributing goods to people or associations and organizing public events—are important for almost three-quarters of the sample. Moreover, provisioning strategies are more important for associations in each country than are capacity-building strategies. This finding is consistent with the image of these associations as survival or subsistence organizations, born out of periodic economic crises and encouraged by the shrinking of the state, that have banded together, pooling their resources for common provisioning. The use of provisioning strategies displays greater cross-strategy and crossnational variation than the use of capacity-building strategies. In each country associations are more oriented toward offering goods—a strategy that corresponds closely to our image of the main activity associations carry out in popular neighborhoods—than to organizing public events.12 The differential is particularly stark in Argentina where virtually all associations ( percent) prioritize goods distribution; that percentage is more than double the percentage of associations that emphasize any other single strategy in Argentina and also stands out compared to the corresponding percentages for Peru and Venezuela ( percent and percent respectively). As we will see below, the Argentine pattern reflects the active role of the state in contracting associations to deliver certain services. We would expect organizing public events to be a lower priority for popular associations given the implied financial outlay, and it is consequently somewhat unexpected that on average percent of associations in each country emphasize this strategy. Institutional Claim Making As mentioned above, claim-making strategies are in fact just as important for popular-sector associations as society-targeted strategies: in all four capital 11. While not reported here, on average, just over half of the associations in each country that trained community leaders also worked to create new organizations and vice versa, and in most cases more than half of the associations that engaged in one provisioning strategy also engaged in the other. 12. In fact, with the exception of Peru, more associations were oriented toward distributing goods than toward any other society-targeted strategy (though not all differences are statistically significant).
06 Chapter 6.qxd
Table . Society-Targeted Strategies (percent of associations)
Total Capacity Building
Fund Other Assoc’s
Train Community Leaders
Create New Assoc’s
81.9 97.1* 75.0 74.2 80.8
52.7 49.6 55.7 52.9 52.9
6.0 6.7 5.7 6.7 5.0
37.2 32.5 34.2 43.3 38.8
30.2 30.8 38.2 25.4 26.7
Distribute Total Goods to Provisioning People/ Assoc’s 72.6 94.6* 68.4 62.1* 65.0
60.7 91.7* 58.8 42.1* 50.0*
Organize Public Events 41.4 45.8 42.5 40.4 36.7
Note: The difference between the percent of organizations that fund and the percent that train; the difference between the percent of organizations that fund versus the percent that create new associations; the difference between the percent of organizations that fund versus the percent that distribute goods; and the difference between the percent of organizations that fund and the percent that organize events are all statistically significant at approximately the p < . level in each country. The difference between the percent of associations that train and the percent that create new associations is statistically significant at approximately the p < . level only in Peru. The difference between the percent of organizations that engage in capacity building and the percent that engage in provisioning is statistically significant at approximately the p < . level in Argentina. The difference between the percent of associations that distribute goods versus those that train is statistically significant at approximately the p < . level in Argentina and Chile; the difference between the percent of associations that distribute goods versus create new associations is statistically significant at approximately the p < . level in Argentina, Chile and Venezuela; the difference between the percent of associations that distribute goods versus the percent that organize events is statistically significant at approximately the p < . level in Argentina and Chile. The difference between the percent of groups that organize public events versus train is statistically significant at approximately the p < . level in each country; the difference between the percent of associations that organize events versus create new associations is statistically insignificant at approximately the p < . level in every country. *Difference between the percent of associations in the country indicated in each row versus the percent of associations in all other countries (considered together as one reference group) was statistically significant at approximately the p < . level.
Page 198
Total SocietyTargeted
9:59 AM
Total Argentina Chile Peru Venezuela
Provisioning
11/10/2009
Capacity Building
06 Chapter 6.qxd
11/10/2009
9:59 AM
Page 199
,
cities, more than three-quarters of the associations emphasize institutional claim making. Table . distinguishes three institutional channels that associations might use to target the state. The first refers to a new institutional development in Latin America in the post-transition period and particularly in the wake of the significant decentralization that has occurred in some countries: participation in institutions or “spaces” (“espacios”) sponsored by the government, such as working groups, fora, consultative councils, or other similar entities. A second empirically important but little studied strategy is the initiation of legal actions. The last and most conventional strategy is contacting government officials. Students of civil society and interest representation in Latin America have only begun to study government-sponsored fora such as councils and “spaces.” So far, their analyses have suggested that such government-sponsored entities exhibit a great deal of variation: in some contexts they exist on paper but little more, while in others they operate successfully and draw the participation of associations. Against this background, the finding that more than a third of associations in Argentina, Chile, and Peru participate in such fora is intriguing. Although one cannot draw inferences about how widespread this kind of activity is in the larger associational world given the nature of the sample, the high rate of participation in government-sponsored fora by associations located in a core segment of the A-Net suggests the importance of these entities. Until recently, scholars of Latin American politics also paid little attention to the degree to which individuals and associations employ the second institutional claim-making strategy considered here, initiating legal action to induce policymakers to address their problems (see, however, Peruzzotti and Smulovitz and Gargarella, Domingo, and Roux ). This study offers some evidence that associations have indeed begun to turn to the courts as an alternate channel of interest representation. Initiating legal actions is an important strategy for close to a third of popular-sector associations: more than a quarter of associations in Chile and Peru and a surprising percent of associations in Venezuela engage the region’s notoriously inefficient judiciaries. Contacting (of elected officials and of government agencies and offices, at the national and municipal levels) is by far the most common form of institutional claim making among the popular associations in each of the four countries. Cross-national differences are somewhat greater in relation to contacting than in relation to societal strategies (though few differences are statistically significant), perhaps due to the different character of the national state in the four countries under study. For instance, certain types of government (e.g., more populist) or well-institutionalized parties might make elected
06 Chapter 6.qxd 11/10/2009
Table . Institutional Claim Making (percent of associations) Contacting
Legal Actions
Total Contacting
Elected Leaders
Government Agencies
Municipal
National
82.6 83.3 88.2 82.3 77.9
36.3 37.9 39.9 40.0 27.5
29.3 20.4 28.9 27.5 40.4
75.0 73.8 83.3 76.7 66.7
54.9 50.8 66.7* 54.6 47.9
58.9 50.8 67.1 68.3 49.6
60.0 70.0 77.6* 62.1 31.3*
29.7 37.9 38.6 22.1 20.8
Level of State
Note: The difference between the percent of associations that engage in contacting versus those that participate in government fora is statistically significant at approximately the p < . level in each country; the difference between the percent of associations that engage in contacting as opposed to initiate legal actions is statistically significant at approximately the p < . level in each country. The difference between the percent of associations that contact elected leaders versus contact agencies is statistically significant at approximately the p < . level in Peru. The difference between the percent of associations that contact at the municipal level rather than at the national level is statistically significant at approximately the p < . level in every country except Venezuela. *Difference between the percent of associations in the country indicated in that row versus the percent of associations in all other countries (considered together as one reference group) is statistically significant at approximately the p < . level.
Page 200
State Fora
Target
9:59 AM
Total Argentina Chile Peru Venezuela
Total Institutional Claim Making
06 Chapter 6.qxd
11/10/2009
9:59 AM
Page 201
,
leaders more attractive targets for claim making. Alternatively, the degree to which policy making has been decentralized could encourage (or discourage) claim making at the municipal level. These hypotheses are discussed in more detail below. A slightly higher percentage of associations in Chile than in other countries contact the state (elected leaders in particular). This finding runs counter to analyses of Chilean civil society that observe the gradual demobilization of civil society during the previous authoritarian regime (–) and suggest that civil society has remained relatively demobilized into the democratic period (Oxhorn ; Foweraker ; Roberts ; Adams ). The percentage of Venezuelan associations oriented toward contacting is the lowest of the four countries. This finding is somewhat unexpected given that Venezuela is the sole member of the quartet to have escaped demobilizing authoritarianism. Yet another macro-feature of Venezuelan politics may have served to dampen institutional claim making. As noted in chapter , the period during which the survey was fielded in Caracas followed one in which Venezuela’s typically exceptional level of political institutionalization had broken down. Dramatic institutional changes were being effected by President Chávez, an “outsider” (even a coup leader) who relied on charismatic authority. This period of political upheaval may have been a particularly inauspicious one for problem solving via claim making on the state.13 On the other hand, the survey responses may accurately reflect the crisis of popular-sector representation that formed such fertile ground for Chávez’s Bolivarian movement. Contacting Target: Elected versus Appointed Officials. The specific target of contacting—and especially the question of whether associations are more likely to contact elected (and thus partisan) or appointed officials—is interesting in light of the contrasts among the Argentine, Chilean, Peruvian, and Venezuelan parties. The Peruvian party system virtually collapsed during the s, and the Venezuelan party system followed suit in the s. In Argentina, by contrast, Peronism has been a national electoral force since the mid-twentieth century and has managed to retain considerable strength among the popular sectors. And Chile’s traditionally strong and ideological parties have gradually evolved into more pragmatic parties in catchall blocs that scholars suggest have distanced themselves from close ties with social groups.
13. Venezuelan association leaders were explicitly asked to answer survey questions without reference to events surrounding the period of strikes and mobilizations of –; nonetheless, these dynamics still might have had an impact on responses.
06 Chapter 6.qxd
11/10/2009
.
9:59 AM
Page 202
-
Despite these cross-national differences, the data suggest that elected officials are important intermediaries of state-targeted claim making and that associations also directly target government agencies in each country. Only in Peru do a significantly larger percentage of associations favor contacting government agencies over contacting elected leaders. The unusual party system continuity and institutionalization in Chile is reflected in the degree to which associations emphasize contacting elected leaders: whereas only about half of associations in the other three countries consider contacting elected leaders to be an important strategy, more than two-thirds of Chilean associations do so.14 The four countries are more similar in terms of contacting government agencies: the Chilean and Peruvian associations contact government agencies slightly more frequently than the Argentine and Venezuelan associations, but the differences are not statistically significant. Level of Government Contacted. The question of the level of the state that associations target with their claim making is also important.15 We might hypothesize that associations contact the level of government at which they think that decisions regarding the policies of interest to them are made, and at which they anticipate that their claims will be heard. These expectations, however, predict different outcomes in terms of popular-sector contacting. On the one hand, the four countries under study have all engaged in some decentralization over the last two decades—that is, have devolved authority over fiscal and administrative matters to state and local government—albeit to differing degrees (see Manor ; Stepan ; Montero and Samuels a; Oxhorn ). We might expect decentralization to encourage associations to contact at the local level. More generally, as chapter suggests, associations may be disarticulated from national-level issues, and be more immediately concerned with questions that are primarily local, again making contacting at the municipal level the logical choice. Moreover, if clientelism marks some stateassociation relationships, it may develop more easily between associations and 14. See chapter for a similar finding regarding the prominence of issue-based (or contactingoriented) linkages between associations and Chilean parties. 15. Beyond the local and national levels discussed here, association leaders were also asked about contacting at the level of state government (“regional” level for Chile and “departmental” level for Peru), and at another intermediary level of government. In each country sample, however, the largest percentage of associations engaged in contacting at the municipal and national levels. These results contrast with those of Tendler’s () study of five public policy areas in Ceará, Brazil, in which she analyzes the “paradox of decentralization” (that it was an activist state-level government that empowered civil society organizations to demand public-sector accountability and good performance from the local service providers).
06 Chapter 6.qxd
11/10/2009
9:59 AM
Page 203
,
local-level leaders, again fomenting local-level contacting. Alternatively, associations may simply doubt that national-level contacting will bring results. On the other hand, many policy decisions with important implications for the popular sectors unarguably continue to be made by the national government, suggesting the importance of national-level contacting. Moreover, while decentralization generally devolves control, it does not always devolve resources, and popular associations likely seek to target the levels of government with funds available to finance programs that are important to them. The data reveal that in fact associations are more oriented toward contacting at the municipal level of government than at the national level (with the difference being statistically significant in every country except Venezuela). Cross-national differences loosely track decentralization trends. For instance, larger percentages of associations emphasize contacting at the municipal level in Argentina and Chile than in Venezuela. By the early s Argentina had experienced a great deal of decentralization (though mostly to the provincial level), Chile had carried out little decentralization (mainly to the municipal level), Peru had carried out very little decentralization at all, and Venezuela under Chávez had begun to recentralize.16 Despite the appearance of some connection between the devolution of authority from the national state, and the level of government at which associations choose to voice claims, other factors likely impinge on associational choices regarding the level of the state to contact. Beyond those mentioned above, the nature of the A-Net itself may also be important. National-level contacting may require more “capacity” in terms of human, social, or financial capital than most associations possess (a hypothesis that the next section begins to assess); alternatively, popular associations may simply lack channels for contacting the national state. Moreover, the hierarchical nature of union organization facilitated national-level contacting under the -Hub; by contrast, the A-Net may serve mainly as a structure for distributive claim making at the local level, as conventional wisdom suggests. That is, the A-Net may involve the pursuit of more particularistic benefits for a constituency rather than the making of “advocacy” claims around broader policies or issues. The emphasis on local-level contacting notwithstanding, the amount of contacting that occurs at the national level is striking. This finding challenges the stylized image of an interest regime consisting mainly of local associations 16. It bears noting that despite high hopes placed in local processes such as “participatory budgeting” (Abers ; Goldfrank ), the capacity of municipal governments varies within and across Latin American countries, making it difficult to predict how effective associational contacting at the municipal level may be in terms of actual problem solving, despite its ubiquity.
06 Chapter 6.qxd
11/10/2009
.
9:59 AM
Page 204
-
almost completely disarticulated from the national level. Further, cross-national differences mirror those found with municipal-level contacting: the percentage of Argentine and Chilean associations that contact at the national level is again large relative to the percentage of Peruvian and Venezuelan associations that do so (though cross-national differences are not statistically significant).17 There is thus more contacting overall in Argentina and Chile than in Peru and Venezuela. One possible explanation may be variation in the longevity of the party system. As discussed above, the party systems in the two Andean countries collapsed since the s while those in the two Southern Cone countries have undergone less dramatic reconfigurations; perhaps the Peruvian and Venezuelan collapses led to the disarticulation of association-government links. These and other cross-national differences are examined more closely in the volume’s conclusion. Organizing Protest Despite the high visibility of protest, there is disagreement among scholars regarding how prevalent contentious politics have been in postauthoritarian Latin America. While some anticipated that the transition to democracy and the reinstitutionalization of politics would be a harbinger of greater mobilizational capacity, others expected claim making to be reoriented away from contentious forms and toward institutional channels (O’Donnell and Schmitter ). Others, looking at transitions in retrospect, have suggested that the change in the form and frequency of mobilization that occurred during the course of the political transition was overstated (Bermeo ; R. Collier ). With regard to the neoliberal era, some have portrayed it as marked by “demobilization” (Hellman ; Jelin ; Eckstein b)—perhaps due to a shift in the arena in which battles over economic policy and distributive politics were fought from state to market. Others have warned that social movements must inevitably moderate their strategies and adopt low-risk, institutional forms of action to gain influence with the state (Foweraker ). Yet empirically, the crisis-prone nature of the post-transition period has sparked some spectacular cases of “ protests” (Petras ) and other moments of dramatic mobilization, particularly (among the countries under study here) in
17. In Argentina and Chile, the percentage of associations that engage in contacting at the national level is about half of the percentage that engage the local level; the analogous percentages are about one-third in Peru and two-thirds in Venezuela (the latter figure mainly reflecting the scant amount of contacting at the local level in Venezuela).
06 Chapter 6.qxd
11/10/2009
9:59 AM
Page 205
,
Argentina and Venezuela. In short, the degree to which contentious collective action remains a prominent strategy in the post-transition era is unclear. A related question regards the role that associations play in organizing protest: is doing so simply one more strategy within a menu of activities that associations employ? Even if it is, protest should be recognized as a unique kind of strategy within a repertoire of action, contrasting sharply in form with both societal strategies and with institutional claim making. Societal strategies are often “daily” activities; moreover, as with institutional claim making, their more frequent use may actually increase their effectiveness by establishing channels that can be used regularly, on an ongoing basis. Indeed, associations are often established with the specific purpose of carrying out society-targeted strategies and/or claim making on the state as an ongoing form of action. Organizing protest, by contrast, tends to be a more intermittent activity that may be more effective when it is carried out less frequently.18 From this perspective it is striking that for percent of associations, organizing or helping to organize protest is an important strategy (see table . above). This finding is even more surprising considering the specific nature of the survey. First, the survey explicitly excluded the major protest outbreaks in Argentina and Venezuela in reaction to crisis moments occurring around the time the surveys were fielded (see appendix B) in view of the absence of analogous moments of upheaval in Chile and Peru during the period under study. Second, the data reflect not whether participating in protests is important for associations, but whether organizing or helping to organize protest activities is an important strategy for them, a higher threshold in evaluating the importance of this strategy. In comparative terms, Venezuelan associations are less likely to organize protest than their counterparts in other countries, especially Peru (although the differences are not statistically significant). A further word may be added about this finding regarding Venezuela, given the timing of the survey in early , a moment of high mobilization on the heels of a major national strike and profound political upheaval during which the president was briefly ousted by a coup. First, as just noted, the fact that respondents were asked whether organizing a protest was an important activity for them “apart from any political activity related with the current political situation involving President Chávez,” likely dissuaded association leaders who had helped to organize protests in the early s from indicating that doing so was an important strategy for them. Yet it could also be that associations in countries with a less 18. Argentina’s movements of the unemployed may represent an exception to this general rule.
06 Chapter 6.qxd
11/10/2009
.
9:59 AM
Page 206
-
authoritarian past engage in less protesting. And the form of democracy in Venezuela, the highly institutionalized “partidocracy,” may simply have proved an inauspicious context in which to engage in contentious politics, with the result that the practice of protest never became an important strategy within a broader repertoire of problem-solving activities—even after the collapse of the party system—until the occurrence of the highly mobilizing national strike and the sustained presence of the polarizing figure of Chávez. How does protest fit into the broader strategy repertoire of associations? Is it an activity of last resort, a strategy of the powerless (Lipsky ) undertaken by associations that are unwilling or unable to make claims on the state through institutional channels? Or do associations tend to organize protest in conjunction with institutional claim making (Binderkrantz ; Rubin )? The data offer support for the second view: protest appears to be a wellestablished supplementary activity within a larger repertoire of societal-targeted and state-targeted strategies (table .). In each country under study, associations that employ any of the three noncontentious types of strategies are more likely to organize protest than are associations that do not employ those strategies (though most differences are not statistically significant). In short, evaluating the A-Net along a second axis, the institutional versus contentious nature of claim making, again reveals duality: the A-Net both serves as an infrastructure for protest and facilitates institutional claim making. The positive relationship between institutional claim making and engaging in the relatively confrontational tactic of organizing protest contrasts with more traditional views about how and when protest emerges (Piven and Cloward ). Nonetheless, it confirms the findings of studies by Tarrow () and Table . Organize Protest by Other Strategy Types (percent) Percent of Associations That Organize Protest
Capacity Building Provisioning Institutional Claim Making
Yes No Yes No Yes No
Full Sample
Argentina
Chile
Peru
Venezuela
38.2* 19.9 32.8 20.8 32.2 17.0
42.0 23.1 32.6 30.8 35.0 20.0
38.6 16.8 37.2 11.1 30.3 18.5
40.2 31.9 38.9 31.9 40.5 17.8
32.3 7.1 23.1 15.5 22.5 13.2
*Difference between the percent of associations that did and did not engage in the given strategy is statistically significant at approximately the p < . level.
06 Chapter 6.qxd
11/10/2009
9:59 AM
Page 207
,
Korzeniewicz and Smith (), for example, which suggest that civil society organizations can act as “insiders” (working within the official process to influence the political agenda) or “outsiders” (like social movements, rejecting the process itself). It also offers support for Rubin’s () contention that Latin American associations appear to employ strategies involving both “the streets” and “the institutions” and mirrors a tendency in recent studies of social movements in other regions to see strategic repertoires as involving both collaborative and contentious forms of action (McAdam, Tarrow, and Tilly ; Tilly :–).
Associational Attributes and Strategy Choice The previous section provided an overall picture of the importance of several different forms of problem solving in the A-Net, a crucial first step in analyzing how the new interest regime functions. The analysis revealed that the new structure of interest representation is characterized by substantial duality on two fronts: most associations that employ societal strategies also engage in claim making on the state, and those that engage in any form of noncontentious problem solving are more likely than their less-active counterparts to organize protest as well. It remains important to determine more concretely what distinguishes the more-active and less-active associations. Do associations with particular qualities, resources, or linkages tend to employ certain strategies, while associations with different attributes focus on other strategies? Our analysis (not shown) revealed practically no relationship between the particular issue or issues on which popular associations work and the problem-solving strategies they employ, offering us little basis on which to examine the content of claim making or, more broadly, the goals that associations seek to achieve when they elect particular problem-solving strategies. Stronger relationships did emerge between three other traits of associations and strategy choice: associational ties to the state, associational resources—both material and organizational—and associational “embeddedess” (or the degree to which associations are linked within associational networks). This section offers a picture of these traits and analyzes the relationships between them and associational strategy profiles. State-Association Ties Two conflicting visions have been advanced regarding the connections between civil society and the state in contemporary Latin America, raising the
06 Chapter 6.qxd
11/10/2009
.
9:59 AM
Page 208
-
question of their relative validity. On the one hand, the image of a disarticulated civil society is pervasive in studies of the region; on the other hand, a well-developed literature examines the role of the state in encouraging (both intentionally and unintentionally) the formation of—and even in actively and formally establishing—civil society organizations. An additional and related question concerns whether the ties that associations form with the state in some way compromise their autonomy. In order to begin to address these questions, the present study measures the incidence of three types of stateassociation ties. The first type involves the distribution of goods by associations as part of a government program. In Latin America, ideological currents focus on cutting back the size and role of state bureaucracies. Simultaneous with these currents, however, has been an overall increase in social spending. To some extent, this squaring of the circle has occurred with the outsourcing of services to or other civil society organizations (e.g., Stallings ; P. Evans ; Kessler ; Jones and Hwang ), entailing public-private partnerships in which associations and other private entities provide goods or services in areas such as food, health care, and poverty alleviation. Indeed, acting as subcontractors may be central to an association’s ability to employ some of the society-targeted strategies outlined above; moreover, such outsourcing arrangements may even beget associations that want to take advantage of government programs and act as state subcontractors (Garay ). Associations form another kind of relationship with the state when they receive government funding, either for specific purposes or general operations. As chapter notes, a major difference between associations and unions involves their resource model. Unions draw resources mainly from dues, and thus are largely self-financed. While many associations do receive some funding from members or constituents, they may also depend on funding from outside entities. Governments and private funders often grant money for specific purposes or programs, however, and frequently award funds (and therefore may withhold them) on a discretionary or short-term (though possibly renewable) basis. Becoming dependent on such funds may lead associations to engage in certain activities, focus on specific issues, or adopt particular strategies in the interest of retaining existing funding and of appearing as worthy recipients of additional funds. Engaging in activities that match the programs or preferences of a funder may prevent associations from attending to the issues they would otherwise address, or employing problem-solving strategies that they would otherwise choose. In short, when funders hold the purse strings, associational autonomy may be compromised.
06 Chapter 6.qxd
11/10/2009
9:59 AM
Page 209
,
A final type of relationship between an association and the state is a more direct government role in the formation of the association. While the proliferation of associations is often thought of as an organic development from below mirroring the “spontaneous emanations” from society discussed by early theorists of pluralism (see chapter ), governments have played a key role at times in fostering or actually creating associations for various purposes. For example, such programs have long been hallmarks of Peruvian politics, pursued both by presidents (e.g., Alan Garcia in the s and Alberto Fujimori in the s) and by mayors, at least partly in bids for political support. More recently, the sponsoring and financing of associations and, later, Communal Councils has been a major platform of the chavista movement in Venezuela. The data in table . suggest that the image of a new associational world operating with few ties to the state may be overstated. In Argentina more than percent of associations maintain some sort of ties with the state. This proportion stands out in cross-national terms, but in Chile almost one-half had some ties to the state, and in Peru and Venezuela more than one-third of associations had such ties. While the state does not seem to “penetrate” the A-Net as it did the -Hub through its regulation of unions, it nevertheless plays an important role. With regard to particular types of ties, first, there is substantial crossnational variation in terms of the percentage of associations that participate in government programs. While less than one-tenth of associations in Peru and Venezuela serve as distributors on behalf of the state, more than one-fifth play that role in Chile, and in Argentina a striking percent do so. This crossnational disparity could perhaps be attributed to the breadth in Argentina and Chile of specific programs for associational contracting of this sort and the availability of resources at the national level. Another way to gauge the importance of this subcontracting role is to consider what percentage of the associations that engage in distribution do so as a government subcontractor. When the incidence of subcontracting is evaluated in this way, the percentage of Peruvian and Venezuelan associations rises to approximately one-fifth (that is, doubles) and increases to over one-third in Chile. (In Argentina, the high percentage does not change much, given that almost all associations engage in distribution.) Several insights can be derived from these results. On the one hand, distributing goods is certainly not limited to government programs: even in this core segment of the A-Net, with the exception of Argentina, associations are much more likely to distribute goods independently than as part of a government program. On the other hand, in each country—and again, particularly in
06 Chapter 6.qxd
11/10/2009
9:59 AM
.
Page 210
-
Table . State-Association Ties (percent) One or More Goods Distribution through Government State Ties Government Program Funding
Total Argentina Chile Peru Venezuela
45.8 62.5* 47.8 37.1 35.8
Of Total Sample
Of Those Distributing Goods
23.0 52.5* 20.6 9.6* 9.2*
38.0 57.3* 35.1 22.8 18.3*
28.3 39.6* 28.9 18.3 26.3
GovernmentLinked Founding
14.1 8.3 17.5 16.7 14.2
*Difference between the percent of associations in the country indicated in that row versus the percent of associations in all other countries (considered together as one reference group) is statistically significant at approximately the p < . level.
Argentina—a group of associations involved in an outsourcing relationship with the state through which they deliver goods clearly exists. Moreover, while Venezuela had not developed very far along these lines at the time the survey was fielded, the percentage of associations involved in the distribution of goods there today is probably significantly higher due to programs Chávez has implemented. In short, while this cross-sectional analysis offers little basis for assertions regarding the direction in which this dynamic might be evolving, the data are somewhat suggestive of the possibility that a “new state” that employs associations as subcontractors may be developing—albeit unevenly—across South America. What is more, subcontracting seems to be an emerging feature of the state across a variety of political contexts. A similar picture emerges regarding government funding. Argentina again stands out, with nearly percent of associations receiving government funds. The Chilean and Venezuelan percentages are somewhat lower, and fewer than percent of Peruvian associations receive such funding. While these data suggest substantial ties with the state, in very few cases did the government participate in the founding of an association. State-Association Ties and Problem-Solving Strategies By comparing the strategy repertoires of associations with and without ties to the state, we can begin to identify any relationships that may exist between having such ties and engaging in particular problem-solving activities. For
06 Chapter 6.qxd
11/10/2009
9:59 AM
Page 211
,
instance, one might expect that awarding funding to associations—particularly given that such funding is often short-term (even if renewable)—could open an avenue through which the state could shape associational activities. Similarly, being contracted to deliver goods as part of a government program might come accompanied by strong government influence on the type and content of demand making and could induce a Michelsian dynamic in which association leaders become more interested in fomenting the relationships that help them gain access to the state than in actively pursuing their association’s substantive agenda. To be clear, we can make no more than tentative inferences about associational autonomy; direct observation of that associational trait awaits further research. Nonetheless, the discovery of a relationship between ties to the state and strategy repertoires could suggest that such ties constrain associational autonomy. Identifying such relationships thus represents a first step toward understanding the degree to which the representational capacity of the A-Net may be weakened via connections its component organizations form with the government. In the -Hub, union access to the state often developed in tandem with co-optation: unions were penetrated and regulated by the state (often via political parties), and their demands were shaped and controlled by state actors. Thus, union-state relationships were double edged and have been identified as a defining feature of “state corporatism” (Collier and Collier ). With regard to the A-Net, however, there is less agreement. Some scholars have claimed that associations are able to form beneficial connections with the state while remaining independent from it (Graham ; Grindle b). Others warn that being contracted by the state to distribute services can result in associations becoming subject to the agenda and policy preferences of state actors charged with outsourcing service provision. Alternatively, a “reciprocal” or symbiotic arrangement of convenience may develop, inducing associational orientation toward state programs and resulting in a “captured” interest regime with compromised representational potential (e.g., Streeck and Schmitter ; Piester ; Richards and Roberts ; Edelman ; Foweraker ).19 To offer just one suggestive anecdotal example, the leader of an in Santiago suggested to a researcher in that Chilean often had to prioritize certain issues (not necessarily the ones they would have chosen otherwise) in order to receive state funding. 19. Streeck and Schmitter (), for instance, suggest that civil society organizations were often dependent on the state during the s, and Piester () relates how s in Mexico in the s delivered social services while allowing clientelism and political favoritism to become stronger.
06 Chapter 6.qxd
11/10/2009
.
9:59 AM
Page 212
-
This analysis suggests a relationship between state-association ties and institutional claim making (table .). Associations that have distributed goods through a government program and/or received government funding are more likely than are associations with no such ties to emphasize contacting government officials and participating in government fora (although they are no more likely to prioritize filing legal actions).20 However, associations with ties to the state are neither more nor less likely to pursue capacity-building or provisioning strategies.21 Different hypotheses might be advanced to interpret the connection between state ties and the two claim-making activities to which such ties seem related. In terms of contacting government officials, it could be that acting as government subcontractors affords associations state contacts and knowledge about channels for claim making; likewise, contacting government may have brought these associations to the attention of the state and increased the likelihood that they would be recruited for service delivery. Similarly, with regard to greater participation in government fora, it may be that associations with ties to the state were in fact recruited by the government to participate in such fora. Alternatively, associations that have attended government fora may have become better known to state officials or more familiar with the functioning of government bureaucracy, leading to their receipt of funding or state contracts. Regardless of which of these interpretations is most accurate, the finding of a connection between certain types of state ties and certain types of claim making within the core segment of the A-Net suggests the existence of a subset of associations with multiple and likely interrelated ties to the state. The question of the relationship between state-association ties and associations’ propensity to organize protest is a particularly interesting one. We might expect that forming ties to the state would deter associations from organizing protest, causing them to avoid the strategy completely, or mobilize only when they fail to achieve their objectives through other means. In either scenario, associational ties to the state would have affected associational autonomy.Yet the data suggest little relationship between state ties and associational choices regarding organizing protest: similar percentages of associations with and 20. The weak connection between state-association ties and filing legal actions may be related to the fact that associations activate the courts for different reasons. They may do so to request the state’s help in forcing a third party to accede to their demands or desires; or they may file cases against the state to push it, in a public way, to take action to solve their problems. We might expect a positive relationship between state-association ties and the first alternative, and a negative relationship between state-association ties and the second alternative, hence the less-than-robust finding. 21. Here and throughout the rest of the chapter, country-specific relationships are reported only when they differ from the general trend identified using the pooled sample.
06 Chapter 6.qxd
Society-Targeted Total Institutional Provisioning Claim Making
Contacting
State Fora
Legal Actions
9:59 AM
Capacity Building
Organize Protest
Institutional Claim Making
11/10/2009
Table . Strategy Choice by State Ties (percent of associations engaging in each strategy)
Yes No
58.7 60
xxxa xxxa
93.1* 79.5
86.7* 71.5
49.5* 32.3
29.4 29.3
35.5 27.8
Government Funding
Yes No
58.2 50.6
80.2 70.0
93.7* 78.2
87.3* 70.1
45.5* 32.6
27.2 30.1
29.5 29.6
GovernmentLinked Founding
Yes No
51.5 52.9
67.2 73.5
87.3 81.8
82.1 73.8
46.3 34.6
31.3 29.0
23.9 30.5
*Difference between the percent of associations possessing the type of tie in question versus the percent lacking the type of tie in question is statistically significant at approximately the p < . level. a “Provisioning” is an aggregate variable that includes “distributing goods, services, and information.” Therefore, the percentages in these cells will always equal percent, as every association that “distributed goods, services, and information as part of a government program” would have a positive score on “provisioning.”
Page 213
Distribution Through Government Program
06 Chapter 6.qxd
11/10/2009
.
9:59 AM
Page 214
-
without state ties considered organizing protest to be an important strategy for solving their problems (table .). Combined, these findings reemphasize the variable nature of state inducements (or the application of benefits), as opposed to the more definitive nature of constraints (Collier and Collier ). While government constraints involve direct control and are by definition autonomy-compromising, state inducements—such as funding or subcontracting—may or may not involve cooptation. While the analysis reveals a relationship between state-association ties and institutional claim making, we cannot infer whether this relationship is a co-optive (or autonomy-compromising) one. Similarly, while the findings regarding protest offer some support for the claim that associations in fact can form connections with the state while remaining autonomous from it, one must remain alert to the possibility of associations organizing large-scale “protests” on behalf of the state—which our survey questions would have captured. For instance, in his early study of squatter settlements in Peru, David Collier () shows that neighborhood associations were co-opted by presidents to engage in demonstrations in support of the president. And Venezuela’s Chávez has been instrumental in establishing a wide range of associations to support him, sometimes orchestrating mass demonstrations to allow them to voice loudly that support. Under such circumstances, associations with stronger ties to the state might actually be more likely to organize, or at least to engage, in protest. In short, a more complete analysis of associational autonomy awaits future inquiry. Resources Resources are critical to both the effectiveness and longevity of organizations, and popular associations are no exception. Associations with some organizational and financial base would seem better positioned to attract people to join them, better able to coordinate those people to carry out work effectively, and more likely to have the capacity to sustain that work over time. While the unions that populated the -Hub are generally considered to have been heavily bureaucratic and relatively “wealthy,” as chapter highlights, conventional wisdom holds that the component organizations of the A-Net possess fewer organizational and material resources. The survey provides a new empirical basis for analyzing associational resources. The data reveal substantial heterogeneity in resource levels even among the core segment of presumably more-prominent and better-connected associations captured by the survey. Four different organizational resources serve as measures of associational
06 Chapter 6.qxd
11/10/2009
9:59 AM
Page 215
,
institutionalization: state recognition, a membership base, a permanent leadership structure, and paid staff. State recognition represents a minimal level of organizational formality. States rarely impose requirements to which associations must conform in order to be formally recognized (in contrast to their policies toward unions). We might hypothesize, then, that the usefulness and desirability of achieving state recognition may vary with associational goals, or with the extent to which associations wish to participate in government programs and institutional fora. With regard to a stable associational “membership” as an organizational resource, one notes the contrast to unions (see chapter ); rather than formal members, popular associations are more likely to have “constituents” who participate in their activities. In that context, the existence of “members” may be an indicator of a more formal set of boundaries circumscribing the group, the work commitment, and the level of benefits. Possessing a formal leadership structure could be quite important to associational institutionalization and continuity. As Roever () highlights, permanent and recognized leadership may facilitate inter-associational coordination and may also aid associations in interacting more effectively with the state; as such, leadership may represent an important key to overall associational efficacy. Finally, the existence of a paid staff may indicate an even higher level of organizational formality. Associations that have amassed the financial wherewithal to move beyond volunteer labor may be more likely to sustain themselves and their programs into the future and may be better able to strategize to solve problems. The associations in the sample present a mixed picture with respect to organizational resources.22 As table . outlines, popular associations appear to have adopted the constituency rather than membership model: on average only one-quarter of associations reported having members or associates. Further, only slightly more than one-quarter have paid staff. On the other hand, the vast majority has a permanent leadership structure,23 and a striking percent of associations have state recognition. Together, these findings suggest that associations in the core segment of A-Net are relatively institutionalized 22. Association leaders were asked yes/no questions regarding state recognition, permanent leadership structure, and bank account possession. With regard to members/associates, they were asked to choose among possible ways of describing the people for whom they worked; the percentage presented here is the percentage of respondents who answered “members/associates.” Regarding paid staff, the question was “how many paid staff work for the organization?” and in terms of budget, a monthly budget figure was requested. 23. The figures presented in table . may reflect some overreporting, since interviews were conducted with someone relatively high up in each association, and these respondents may have had an interest in indicating that a leadership structure existed and that he or she formed part of it.
06 Chapter 6.qxd 11/10/2009
Table . Organizational and Material Resources (percent)
State Recognition
Members/ Associates
Permanent Leadership
Paid Staff
Bank Account
80.3 78.8 89.5 73.8 79.6
25.5 17.5 37.3* 25.8 22.1
87.1 77.5 95.6 95.8 80.0
28.3 31.3 23.2 17.1* 41.3*
35.7 29.6 52.2* 20.8* 40.9
Lo
Annual Budgeta Med
Hi
33.3 22.4* 37.6 28.3 57.0*
32.9 27.1 38.7 40.2 16.3*
33.8 50.6* 23.8* 31.5 26.7
*Difference between the percent of associations in the country indicated in that row versus that of associations in all other countries (considered together as one reference group) is statistically significant at approximately the p < . level. a Note relatively high level of missing data, such that total n = . Monthly budgets reported by associations were converted into contemporary U.S. dollars and annualized. Low annual budget = $ through $; medium = $ to $,; high = $, and above (maxing at $,,). These cut-points were set such that a third of the associations in the pooled sample that reported their budget fell in each category.
Page 216
Total Argentina Chile Peru Venezuela
Material Resources
9:59 AM
Organizational Resources
06 Chapter 6.qxd
11/10/2009
9:59 AM
Page 217
,
at the top (despite the fact that those at the top may not be paid), and cater to fairly diffuse groups. An association’s level of material resources may offer even greater insight into its degree of institutionalization, as well as its ability to complete projects in a timely manner and thus, ultimately, its efficacy. Two indicators of the material resources associations possess are the size of their budget and whether they have a bank account. The subset of our sample that provided budget data is not well-off in terms of material resources: associations in the richest third work with an annual budget as small as U.S.$,.24 Moreover, only one-third of associations have a bank account. A few observations regarding cross-national differences merit noting here. While most associations have “associates” or participants, rather than more formal members, a significantly larger proportion of associations in Chile have formal members, and an unusually large percentage possesses a bank account. Further, Peruvian and Venezuelan associations are particularly divergent in terms of paid staff: very few Peruvian associations and an unexpectedly high proportion of Venezuelan associations have a paid staff. Yet almost percent of Venezuelan associations are in the “low” budget category (and very few are in the medium budget category); here the contrast is starkest with Argentina where less than one-quarter of associations fall in the “low” budget category (and where a disproportionate number of the total associations in the highest budget category are located).25 The finding that fewer than one-third of Argentine associations possess a bank account is somewhat unexpected in view of their relative wealth; this phenomenon might represent a legacy of the extensive bank freeze imposed in Argentina in (and of a similar freeze imposed in ); fear of being unable to withdraw needed funds might dissuade associations from keeping their money in formal financial institutions. Resources and Problem-Solving Strategies Western European and American scholars have considered the effects that a series of organizational and financial resources might have on both individual 24. This richest group is quite diverse: about half of the associations in the top third have an annual budget under U.S. $,, another percent has an annual budget between U.S. $, and U.S. $,, and a very small percentage (approximately percent) has an annual budget in excess of U.S. $,. 25. Given that the three budget categories were established such that one-third of the associations in the pooled sample for which we have budget data fall into each category, what is most interesting to evaluate is the degree to which the country samples deviate from these one-third divisions.
06 Chapter 6.qxd
11/10/2009
.
9:59 AM
Page 218
-
and collective action. Resource mobilization theory, for instance, has long posited a relationship between the availability of both organizational and material resources and collective action (Olson ; Gamson ; McCarthy and Zald ). And Verba, Schlozman, and Brady () have observed that in the United States, citizens who are wealthy, highly educated, and skilled in speaking, writing, and organizing are more likely to vote, contact political officials, and work for political organizations; in fact, money and “civic skills” predict political participation even when controlling for political interest. Yet must individuals and groups possess both organizational and financial resources in order to engage in collective action? We might, for instance, expect that material resources would “trump” other types of resources in importance. Some scholars have suggested this is the case: Portes and Landolt () claim that the ability of an organization to achieve its goals is restricted whenever finances are scarce. Yet there is certainly not consensus on this point. Della Porta, Mosca, and Reiter () highlight the success of global grassroots movements comprising some of the world’s poorest people, thus suggesting that capacity for collective action may exist even when material resources are lacking. With respect to organizational resources, scholars have considered how “social capital”—network ties, external alliances, and skilled leadership—can affect individual and collective action. Many studies claim that individuals are more likely to participate in collective action when their social network encourages it, and groups are more successful at staging mobilizations when they have access to larger social networks.26 In terms of collective action, Uquillas and Van Nieuwkoop () find that indigenous groups in Ecuador with high levels of organizational resources were very successful in implementing projects and providing community services; groups with few organizational resources were equally successful only when they allied with an . Contemporary scholarship also points to the importance of skilled leadership. As mentioned above, Roever () has shown the importance of a permanent leadership structure for Peruvian street vendors’ effective interaction with the state. Also, Campero () argues that the shantytown movement in Santiago emerged in large part due to the role of Catholic Church leaders in systematizing common experiences, framing them as a common political struggle, and proposing specific actions. Similarly, Krishna () claims that in India, even strong and cohesive local village community groups need savvy 26. See, for example, Tilly (), Klandermans and Oegema (), Opp and Gern (), C. Schneider (), and Rosenstone and Hansen ().
06 Chapter 6.qxd
11/10/2009
9:59 AM
Page 219
,
leaders to connect the group to political and material resources and engage with the political arena. These ideas are instructive for considering the problem-solving potential of popular associations. Nonetheless, little has been written on the connections between the possession of particular types of resources and the choice of specific associational problem-solving strategies. The data presented in table . suggest that organizational resources are not closely related to associations’ employment of society-targeted strategies: associations with fewer organizational resources are equally likely to work to build the capacity of other associations or engage in provisioning as associations with more organizational resources. Both better- and less-institutionalized associations appear to appreciate the value of strengthening other associations—perhaps motivated by a belief that augmenting the capacity of members of the associational network may in the end “lift all ships.” The fact that associations with fewer organizational resources are as likely as those with more resources to engage in provisioning is somewhat expected, given that most studies have portrayed provisioning activities as the typical strategy of poorly institutionalized, survival-oriented popular associations. A slightly stronger relationship appears between the level of material resources and engaging in societal strategies. While poorer and richer associations are equally likely to work to build the capacity of other associations, associations with bigger budgets are more likely to distribute goods to other associations or people, and to organize public events. This contrast might have been expected: while distributing goods almost by definition implies the possession of some material resources (unless an association is doing so as part of a government program), working to strengthen other associations might instead require knowledge and willingness, which associations can certainly possess independent of their level of material resources. In short, the analysis reinforces the finding in della Porta, Mosca, and Reiter () that financial resources are not required for collective action: society-targeted strategies are not closed off to associations lacking financial wherewithal. The relationships between material and organizational resources and the use of claim-making strategies, however, are the inverse of those between material and organizational resources and society-targeted strategies. Material resources appear to make a rather limited difference in terms of associational choices regarding institutional claim making. The finding that associations with bank accounts are significantly more likely to contact at the national level than are associations without bank accounts is somewhat paradoxical given
06 Chapter 6.qxd
Society-Targeted
Organizational Resources Material Resources
Annual Budgeta
Legal Actions
Yes No Yes No Yes No Yes No
53.0 51.9 51.7 53.1 53.6 46.7 58.6 50.0
73.5 69.0 69.8 73.5 71.5 79.5 81.6 68.3
86.3* 67.4 88.8 80.5 84.5* 69.7 82.2 82.8
79.8* 55.6 81.0 72.9 77.7* 59.6 74.3 75.3
62.8* 48.7 89.5 89.5 62.6* 42.6 46.7* 66.3
32.3 19.3 63.5 55.6 30.8 30.0 37.5 26.1
37.8 29.9 40.5 34.8 38.5 21.3 36.5 36.2
29.7 27.8 30.6 28.9 29.7 27.0 33.9 27.2
28.3 34.8 29.3 29.6 29.2 32.0 26.6 30.9
Yes No Hi Med Lo
59.8 48.9 57.6 49.0 51.2
77.8 69.7 82.9* 70.1 64.3
90.2 78.4 89.5 85.3 76.8
82.2 71.0 82.4 78.9 67.6
57.1 61.6 62.4 72.1 60.9
40.0* 25.7 36.2 30.0 24.6
39.3 34.6 47.6 39.2 32.4
34.3 26.6 28.1 25.0 29.0
25.4 31.8 33.3 34.8 29.0
*Difference between the percent of associations possessing the type of resource in question versus the percent of associations that lacked that type of resource is statistically significant at approximately the p < . level. a
Low annual budget = $0 through $; medium = $ to $,; high = $, and above (maxing at $,,).
Page 220
Bank Account
State Fora
9:59 AM
Total Inst. Contacting Capacity Claim Building Provisioning Making Any Level National Municipal State Recognition Members/ Associates Permanent Leadership Paid Staff
Organize Protest
Institutional Claim Making
11/10/2009
Table . Strategy Choice by Resources (percent of associations engaging in each strategy)
06 Chapter 6.qxd
11/10/2009
9:59 AM
Page 221
,
the lack of a relationship between levels of material resources and nationallevel contacting. A stronger (though inconsistent) relationship appears between organizational resources and engaging in institutional claim making. Associations with state recognition and associations with a permanent leadership structure are more likely than associations lacking these qualities to make claims on the state through institutional channels. In particular, they are more likely to contact the government, especially at the municipal level. Following on the discussion regarding decentralization above, the connection between state recognition and contacting at the municipal level suggests a possible answer to the question of at what level of the state associations seek recognition. And, of course, the finding that associations with a leadership structure are more likely to make claims on the state is consistent with previous findings in the literature, mentioned above, regarding the positive relationship between strong leadership and collective action. Yet other institutional resources appear to make little difference to associational decisions regarding institutional claim making. Associations with a more formalized constituency are no more likely to engage in contacting, participate in fora, or initiate legal actions than associations with a less formalized constituency. Unexpectedly, given the complexity of Latin American legal systems, uninstitutionalized associations are as likely to engage in the burdensome process of filing legal claims as more institutionalized associations. And organizations with paid staff are significantly less likely than associations with no paid employees to contact municipal leaders. In short, to carry out the relatively formal (but comparatively inexpensive) task of contacting state leaders or agencies (the most commonly used claim-making strategies), organizational wherewithal trumps material resources in importance. At the same time, there is substantial scope for less formalized associations to engage in a variety of claim-making strategies. Finally, the existing literature left us with competing expectations regarding the relationship between resources and protest. While resource mobilization theorists posit a positive relationship between organizational and material resources and collective action, others have suggested that it is generally lessinstitutionalized, poorer organizations that organize protest (Offe ; Melucci ). The present study’s findings contradict both sides of the debate: associations possessing more material and organizational resources are as likely to organize protest as associations lacking such resources. Protest is not the particular recourse of poorer, or less-institutionalized associations. To sum up, while some relationships do appear between resources and the
06 Chapter 6.qxd
11/10/2009
.
9:59 AM
Page 222
-
employment of certain associational strategies, they hardly represent a definitive or complete answer to the question of why associations employ particular strategies. The next associational quality examined here, “embeddedness,” goes further in helping us sort out associational choices regarding strategies. Embeddedness The term “embeddedness” refers to the degree to which associations are linked to other associations within associational networks.27 Embeddedness is one facet of scaling, or the extent to which associations coordinate within the A-Net to take action and solve problems.28 Examining embeddedness helps us adjudicate between conflicting claims regarding the incidence of inter-associational linkages in Latin America: while some scholars (e.g., Carroll ; Chalmers, Martin, and Piester ) posit the existence of strong and vibrant associational networks, others paint a picture of greater associational atomization (e.g., Oxhorn a). Moreover, given that scholars often portray Latin American associations as individually weak or ineffective, the number of ties that associations form with other associations may relate to their ability to carry out particular kinds of activities and to their individual effectiveness. Similarly, the extent of inter-associational linkages in the A-Net in the aggregate—that is, the density of its infrastructure—may well be crucial to its potential as a representational structure. An association’s degree of embeddedness was measured by counting its bilateral links to other associations.29 Our data suggest that the organizational infrastructure of the core segment of the A-Net is relatively dense: over percent of associations have some links to other associations, and a third of associations are heavily linked (see table .).30 Yet while the chain-referral sampling technique employed might have produced only linked associations, fully percent are isolates, not cooperating or coordinating their activities 27. To be clear, embeddedness is an attribute of an association that refers to the sum of its bilateral links to other associations (rather than to the overall connectedness of an associational network). 28. The discussion here focuses only on inter-associational linkages. Chapter considers several different aspects of scaling in the A-Net, and chapter further extends the analysis by examining associational linkages with and unions and parties and their implications for scaling in the A-Net. 29. Specifically, association leaders were asked whether they coordinated their work with associations (or aggregating groups such as federations or confederations) of the same type, with religious associations, with neighborhood associations, with foundations, and/or with , and if so, with how many of each particular type of group they coordinated. An association’s total number of bilateral links defines its degree of embeddedness. 30. Specifically, . percent of the total sample has – ties; . percent has – ties; . percent has – ties; and percent has more than ties.
06 Chapter 6.qxd
11/10/2009
9:59 AM
Page 223
,
Table . Associational Embeddedness (percent)
Total Argentina Chile Peru Venezuela
Atomizeda
Linkedb
Heavily Linkedc
19.4 9.2 23.2 10.0 35.4*
47.3 55.0 46.1 48.8 39.2
33.3 35.8 30.7 41.3 25.4
*Difference between the percent of associations in the country indicated in that row versus that in all other countries (considered together as one reference group) is statistically significant at approximately the p < . level. a
"Atomized" associations have no ties.
b
"Linked" associations have – ties.
c
"Heavily linked" associations have or more ties.
with any other association. This finding suggests that a great deal of atomization may exist in the associational world. Moreover, some substantively significant cross-national differences appear. Argentina and especially Peru emerge with higher degrees of associational embeddedness, Chile represents a more intermediary case, and Venezuela displays lower levels of embeddedness (though few of the differences are statistically significant). These contrasts are discussed further in the next chapter. Embeddedness and Problem-Solving Strategies We might expect that embeddedness would facilitate certain associational strategies. Inter-associational linkages and cooperation might make organizing protest easier, for instance, and associations accustomed to interacting with other associations might be more likely to engage in capacity building. Indeed, a substantial subset of the literature on civil society posits a positive relation between associational embeddedness and political participation,31 and argues that associations that are less embedded in associational networks engage in less political activity (e.g., Pásara et al. ; Weyland ). The analysis in fact suggests a strong relationship between associational embeddedness and associational problem-solving efforts: the more associations are linked, the more likely they are to engage in each type of strategy. 31. See, e.g., Campero (), Diani (), Oxhorn (), Veltmeyer (), and della Porta, Mosca, and Reiter ().
06 Chapter 6.qxd
11/10/2009
.
9:59 AM
Page 224
-
With regard to the positive relationships between embeddedness and capacity building and provisioning (table .), while it is not possible to pinpoint causality, one can imagine these qualities working together in a virtuous cycle: linkages with other associations might facilitate the development of a repertoire of society-targeted strategies, which might aid associations in attaining even more links, and so on. Despite the clear relationship, it is important to note that more than one-third of completely atomized associations carry out capacity building,32 and almost percent engage in provisioning. While betterembedded associations are clearly more likely to use societal strategies, such activities are not the exclusive domain of heavily linked associations. Better-embedded associations are also more likely than atomized associations to engage in institutional claim making. The contrasts are significant with respect to contacting at the municipal level and participating in government fora. This finding again suggests that better-embedded associations in the core of the A-Net may be better positioned to engage in a range of activities. Nonetheless, as is the case with societal strategies, a relatively large percentage of atomized associations also engage in institutional claim making (almost three-quarters), and more than percent engage in contacting. In sum (and similar to the relationships that emerged between state ties and resources on the one hand and institutional claim making on the other), while better-embedded associations are more likely to make claims on the state through institutional channels, doing so is not their exclusive domain. Further, just how embeddedness and institutional claim making relate to each other is open to interpretation. On the one hand, embeddedness may encourage institutional claim making: a “strength in numbers” phenomenon may empower better-embedded associations to reach out to the state; further, having more bilateral associational linkages might encourage associations to learn and engage in new strategies, leading them, eventually, to contacting. On the other hand, associations could certainly become familiar with and eventually coordinate with other associations through joint participation in government fora, for instance. Finally, regarding protest, we might think that an important prerequisite 32. While this finding might seem paradoxical, it is important to keep in mind that we evaluated associational linkages via a survey question that asked leaders to identify their association’s ties with other associations; the questions regarding capacity building were in a separate battery that included questions about the sorts of activities associations carried out. In addition, among the indicators of capacity building is training community leaders, an activity that might involve working with individuals but does not require an association to directly link with other associations; further, association leaders might have indicated that they train community leaders with their own associational leadership in mind. It is also possible to imagine associations that funded others not working closely with them.
06 Chapter 6.qxd 11/10/2009
Society-Targeted
(n=184)
State Fora
Legal Actions
34.8*
59.2*
73.9
62.5*
47.3*
22.8
26.1
23.4
15.8*
Linked
1-10 ties (n=448)
50.0
71.9
79.9
74.1
59.2
27.7
29.7
25.9
31.7
Heavily Linked
11+ties (n=316)
67.1*
81.3*
91.5
83.5
68.7*
36.7
51.6*
37.7
34.5
*Difference between the percent of associations in the country indicated in that row versus the percent of associations in all other countries (considered together as one reference group) is statistically significant at approximately the p < . level.
Page 225
Total Inst. Contacting Capacity Claim Building Provisioning Making Any Level National Municipal Atomized 0 ties
Organize Protest
Institutional Claim Making
9:59 AM
Table . Strategy Choice by Embeddedness (percent of associations engaging in each strategy)
06 Chapter 6.qxd
11/10/2009
.
9:59 AM
Page 226
-
for organizing protest is the ability to ensure a high turnout. Poorly attended protests may be less effective in voicing the claim in question or may understate support for the claim, thus undercutting the legitimacy of the organizing association. The data bear out the expectation that well-embedded associations organize protest to a greater degree than their less-linked counterparts. A robust positive relationship appears between embeddedness and protest. While it is not possible to discriminate between associations that do not want to organize protest, and those that are not able to do so, the fact that relatively small percentages of atomized associations organize protests serves as further evidence that better-embedded associations are more active.
Conclusions and Implications With changes engendered by neoliberal reform and democratization, problem solving among the popular sectors and in particular the question of popularsector representation have become newly salient in the literature on developing democracies. This chapter contributes three principal sets of conclusions to that literature. It describes several strategies that associations within the ANet use to pursue their interests, outlines a series of associational attributes and discusses their incidence among associations in the core segment of the A-Net, and analyzes the connections between these attributes and the employment of different problem-solving strategies. The first set of conclusions concerns the strategy profile of the A-Net, focusing on two axes. First, like the -Hub, the A-Net is a dual regime with respect to its component associations’ strategy repertoires. In the -Hub unions engaged in collective bargaining, which is a kind of societal strategy (but one that was regulated by the state and in which the state sometimes participated), and they also directly pressed claims on the state. The A-Net as an organizational infrastructure embraces society-targeted activities and plays a role in interest intermediation between society and the state: while percent of the associations in the core segment employ societal strategies, percent engage in claim making on the state. Further, on the level of individual associations, almost three-quarters of the associations surveyed engage in both societytargeted and state-targeted strategies. The strategy profiles in each country are more similar than different. The minor cross-national contrasts relate mainly to aspects of claim making—in particular, the level of the state that associations target. Those contrasts likely have their roots in significant cross-national differences in the size and capacity of the state and the degree of decentralization
06 Chapter 6.qxd
11/10/2009
9:59 AM
Page 227
,
that has occurred in each. In short, the image of an associational interest regime unable to engage (or little-oriented toward) the state and far more oriented toward societal strategies is overstated. On a second axis of associational activity, the institutional versus contentious nature of claim making, duality again marks the A-Net. Although popular associations seeking to make claims on the state in contemporary urban South America appear to do so primarily through institutional channels, for a nontrivial subset of organizations, organizing protest is an important strategy in their repertoire. Perhaps even more important, associations that actively employ any other strategy are more likely to organize protest than those that do not pursue other strategies actively. Organizing protest is not, then, the particular recourse of associations that lack the capacity to engage in other types of strategies (such as institutional claim making). In this sense the A-Net is again similar to the -Hub, under which unions both enjoyed some degree of institutional access to the state and engaged in contentious politics. The chapter’s second overall set of conclusions addresses a series of attributes that characterize associations within the core segment of the A-Net. An examination of the ties that associations have formed with the state reinforces the notion that the associational interest regime may not be as disconnected from the state as some scholars have suggested. Certainly fewer ties bind the A-Net’s component organizations to the state than linked unions to the state under the -Hub. Nonetheless, it is striking that just under half of the associations in the sample maintain some sort of tie to the state—either distributing items as part of a government program, receiving money from the state, or having had the state involved in their founding (the least common tie). Just as claim making remains a prominent and important strategy for associations in contemporary Latin America, the A-Net appears to have more substantial ties to the state through its component associations than some analyses have posited. With regard to resources, associations are at least minimally institutionalized in the sense that the vast majority has state recognition and a permanent leadership structure. Associations in contemporary Latin America thus may be somewhat less ephemeral and more formal than the more fluid, grassroots, antibureaucratic models that associationalism is often said to follow. At the same time, with the exception of Chile, relatively few of the associations sampled in each country have a discreet membership base, suggesting that associations maintain a relatively informal relationship with their constituency. In terms of material resources, associations are on the whole poor, even in the core segment of the A-Net: two-thirds of the associations in the sample operate
06 Chapter 6.qxd
11/10/2009
.
9:59 AM
Page 228
-
on less than U.S.$, annually, and approximately half of those in the top third have an annual budget of under U.S.$,. The budget constraint is undoubtedly connected to institutional formalization: with the exception of Venezuela, fewer than a third of associations in each country have paid staff. Cross-nationally, associations in the highest budget category are disproportionately located in Argentina, and associations in the lowest budget category in Venezuela. In sum, then, the core segment of the A-Net appears to include somewhat institutionalized, though poor, associations—with some important cross-national differences. Finally, the analysis does not fully support the received wisdom regarding associationalism in Latin America, which portrays an atomized associational world with few internal linkages. While a great deal of heterogeneity exists, almost a third of the associations in the core segment of the A-Net are well embedded, possessing eleven or more bilateral links to other associations. At the same time, despite a sampling procedure heavily biased against finding any isolates, percent of the sample is completely atomized, suggesting a kind of bifurcation between atomized and embedded associations. The findings thus far suggest that we may be witnessing a “shift” toward an interest regime that is less connected to (though certainly not disconnected from) the state and that is composed of organizations that are poorer in organizational and financial resources than unions were under the -Hub. Also, unlike the -Hub, the A-Net includes a substantial number of atomized associations, but has a core segment of quite well-connected associations. Finally, the new regime serves as both a channel for making claims on the state, and an infrastructure for coordinating societal activities. The third set of conclusions regards the relationship between associational attributes and the strategies that associations undertake. In general terms, associations with some connection to the state are just as likely to employ societal strategies as those without state ties. Nonetheless, the fact that in each country a subset of associations that distribute goods (the most common provisioning strategy) do so as part of a government program serves as evidence that a “new state” that employs associations as subcontractors may be slowly developing across some parts of South America, including Argentina and, increasingly, Venezuela. Further, a positive relationship exists between the level of material resources and the employment of provisioning. This relationship suggests that associations that distribute goods are dependent on material resources either from the government or from other sources. Finally, the embeddedness of associations is positively related to their use of various societal strategies.
06 Chapter 6.qxd
11/10/2009
9:59 AM
Page 229
,
The relationships between these associational attributes and the use of state-targeted strategies, however, represent something of an inversion of the relationships just outlined. State-association ties emerge as somewhat important for several forms of institutional claim making, as does the possession of organizational resources. However, the strongest relationship that emerges is, again, between embeddedness and claim making: better-linked associations are far more likely than atomized associations to engage in a range of statetargeted strategies. And while ties to the state and organizational and material resources make little difference to associational choices regarding organizing protest, associational embeddedness is also positively related to organizing protest. That is, protest is not dissuaded by state ties nor is protest simply the recourse of the poor—though it does appear to be the domain of better-linked associations. In sum, for societal strategies as well as for claim making both “through institutions and in the streets,” the associational attribute most closely related to strategy choice is embeddedness within associational networks. This associational characteristic along with other aspects of “scaling” in the A-Net are examined further in the next chapter.
07 Chapter 7.qxd
11/9/2009
11:50 AM
Page 230
7 : , , Samuel Handlin and Diana Kapiszewski
Studies of collective action typically examine how individuals coordinate to pursue their interests, and explore the difficulties of sustaining participation in the face of the free rider problem. Yet equally crucial to the functioning of an interest regime is the coordination carried out among organizational entities, referred to here as “scaling.” Studies of the -Hub highlight the coordination that occurred among unions and the important coordinating role played by hierarchical and bureaucratic labor federations and confederations. Similarly, scholars of Latin America working within the “postcorporatist” civil society and social movement traditions have emphasized the importance of associational coordination. Many studies suggest that the degree to which associations are embedded within networks shapes their ability to engage in problem-solving activities and to act as vehicles of interest representation.1 Other scholarship focusing on the structure of associational networks has called attention to the role that play in coordinating and supporting other associations (Carroll ; Howes ). This empirically rich work offers a wealth of information. Yet the fact that much of it is based on case studies impedes broader generalizations regarding the nature and incidence of associational scaling in the A-Net and complicates theory building regarding the implications of scaling. 1. See Oxhorn (), Fox (), Chalmers, Martin, and Piester (), Korzeniewicz and Smith (), and Roberts and Portes ().
07 Chapter 7.qxd
11/9/2009
11:50 AM
Page 231
,
This chapter leverages the project’s uniquely large sample of popular associations to explore scaling in the A-Net from several angles. The discussion begins with a descriptive portrait of associational networks that identifies three forms of scaling. Linkages among associations in the A-Net constitute the first form. Building on chapter , the present analysis more closely examines associational embeddedness—or the degree to which associations are linked to other associations.2 While some striking cross-national differences emerge, overall the core segment of the A-Net captured in the sample is relatively well linked.3 Scaling may also be achieved through the activities of “coordinating associations”—those for which coordinating other associations is a principal function. Two types of coordinating associations can be distinguished: Nodal and Flexible Fronts. The presence and influence within the A-Net of these two types of coordinators constitute the second and third forms of scaling. These coordinating associations appear central to the operation of the A-Net as a whole: in Argentina, Chile, and Peru, coordinators make up approximately percent of the sample, and more than a quarter of the groups interviewed in each country reported working with a coordinating association. Flexible Fronts and Nodal are quite different from the labor federations that served a coordinating function under the -Hub. While federations were not a homogeneous group, they generally took the form of hierarchical membership-based organizations that stood “above” unions and were formally delegated the power to represent those organizations and their members in public and private arenas of claim making and collective bargaining. Neither kind of coordinator in the A-Net fits this model. Coordinators may direct problem-solving activity within their networks, and may have informal and ad hoc means of applying pressure to other associations. Nonetheless, Nodal and Flexible Fronts are rarely formally empowered to represent a defined set of organizations and their participants. Rather than being “above” other organizations, coordinators in the A-Net are on a more horizontal plane with others whose activities they direct and coordinate within the network. As such, they are less distinct from base organizations than labor federations
2. We use the term “embeddedness” to describe the number of bilateral links an individual association possesses within the associational world, and we treat the concept as one aspect of scaling in the A-Net. It should be noted that chapter , which also examines embeddedness, does not treat it as a form of scaling. 3. Readers are reminded that our surveys captured a specific subsection (what we refer to as a “core segment”) of the A-Net in popular-sector neighborhoods in each country’s capital city, one likely to contain more visible and better-connected associations.
07 Chapter 7.qxd
11/9/2009
.
11:50 AM
Page 232
-
were from base-level unions under the -Hub; indeed, it may be that associations within the A-Net take on or give up roles as coordinators as they develop over time.4 Comparing Nodal s, Flexible Fronts, and base associations along a series of dimensions allows us to explore the infrastructure of the emergent A-Net and how it functions as a vehicle for popular representation. Of particular import are questions concerning associations’ organizational and financial resources, the democratic nature of their internal practices, and their links with external actors such as the state or international donors. The latter are of special theoretical interest as they may generate forms of dependence and complicate the ability of associations to represent constituencies. The analysis shows that Nodal s and Flexible Fronts differ from each other with regard to most of these characteristics. Nodal are more institutionalized than are many other associations, and they tend to be marked by a pattern of “participation without democracy”: they are likely to involve members in activities but unlikely to allow them voice in making key decisions about the ’s leadership or orientation. Further, they are more likely to have organizational and financial ties to both the national state and to international organizations and governments. This overall portrait suggests support for scholarship emphasizing the double-edged nature of these coordinators: power disparities between Nodal and base associations may well frustrate the latter’s aspirations for interest representation via , while ’ multiple external links may threaten their own autonomy, and thus that of associations that are linked to them. Flexible Fronts, on the other hand, bear greater resemblance to base organizations despite serving a coordinating function within the network. Flexible Fronts are no more institutionalized than are base associations, and they are highly participatory and highly democratic. Further, they tend to lack potentially autonomy-compromising ties to external actors: they are no more likely than are noncoordinators to have ties to the national state, and only slightly more likely to be linked to international actors. The findings also imply that while Nodal and Flexible Fronts both serve coordinating functions, they are quite different types of organizations with potentially distinct consequences for their networks. Appreciating the coexistence of Nodal s and Flexible Fronts is central to a more nuanced understanding of the operation of the A-Net. 4. We label all associations that are neither Nodal s nor Flexible Fronts as “base associations.” Base associations thus means “noncoordinators” and does not imply that such groups represent basist constituencies.
07 Chapter 7.qxd
11/9/2009
11:50 AM
Page 233
,
The analysis also explores the relationship between each form of scaling and associations’ use of problem-solving strategies, revealing how different forms of scaling shape problem solving within the A-Net. The first relationship examined is between inter-associational linkages and strategy choice (that is, whether more-embedded groups use certain strategies more than lesslinked groups). A second consideration is whether either Nodal or Flexible Fronts are more likely than are noncoordinators to engage in various problem-solving strategies. If coordinators are more likely than are other groups to use claim-making strategies, for instance, one might conclude that coordinating associations serve as conduits for accessing the state (as did federations and confederations under the -Hub). Alternatively, given their coordinating function within the A-Net, the reluctance of Nodal or Flexible Fronts to use certain strategies might in fact discourage the use of those strategies more broadly. For example, some studies on civil society in Latin America have suggested that the rise of has contributed to a decline in protest. The crosstabular analysis and logistic regression models produce four major findings. First, Nodal and Flexible Fronts are both far more likely than are noncoordinators to have contacts at the national level of government. Given the importance of national-level contacts for pursuing many agendas and interests, the finding supports the idea that coordinating associations provide a crucial channel for accessing the state within the A-Net. Second, while being a coordinator is not an important predictor of contacting at the local level, associational embeddedness is: all associations—both base associations and coordinators—tend to be more likely to engage the state at the local level the more embedded they are within associational networks. Third, contrary to many expectations in the literature, Nodal are not significantly less likely than the rest of the associational world to organize protest; Flexible Fronts, however, appear to be far more likely to do so than either Nodal or noncoordinators. Finally, both being a coordinator and being well embedded in associational networks are significant and powerful predictors of engagement in capacity-building activities. In sum, each form of scaling shapes the A-Net in different ways. Coordinating associations seem to link the A-Net to the national state. Flexible Fronts have a greater propensity to organize protest and may thereby encourage other associations in the network to engage in contentious politics. The segments of the A-Net that are more densely linked are likely the points of the interest regime that are best connected to the local state. And all three forms of scaling seem to be associated with efforts to build the capacity of other associations. Thus, different forms of scaling seem related to the adoption of distinct
07 Chapter 7.qxd
11/9/2009
.
11:50 AM
Page 234
-
problem-solving repertoires, shaping how the A-Net operates in aggregate as a vehicle for popular representation.
Inter-Associational Linkages and Coordinating Associations This section deepens the previous chapter’s examination of associational embeddedness, presenting a cross-national portrait of each of the three forms of scaling just identified. While the four interest regimes under study here diverge on several major points, coordinating associations are numerous in each. On an operational level, the extent of associational scaling within each national A-Net is evaluated using a combination of measures. The analysis investigates embeddedness by evaluating the proportion of associations that report working with ten or more other associations and the proportion that report working with none (which we refer to as atomized).5 The incidence of coordinating associations and the propensity of other associations to be linked to coordinators are also explored. It may bear repeating that the chain-referral technique employed by the project to obtain the sample of associations was designed to capture a “core segment” of more-linked associations and coordinators, rather than to produce a representative sample of the associational world.6 Nonetheless, given that the same sampling technique was used in all four countries, some cautious cross-national comparisons of this more salient core segment, and of the coordinating associations within it, may be undertaken.7 Examining coordinating associations involves making conceptual and operational distinctions between base associations and two types of coordinators, 5. The number of bilateral links an association has within the associational world was determined through a battery of questions that asked associations with how many like associations, religious associations, neighborhood groups, foundations, and (but not parties, unions, or large enterprises) they worked. The number of working relationships was summed across these questions. 6. The chain-referral sampling technique naturally led us to better-linked groups and also prioritized following links to coordinating associations. Nonetheless, the sample does have substantial variation on “associational embeddedeness”: roughly one-fifth of the sample reported working with no other associations and almost one-third reported only minimal linkages (four or fewer). See the previous chapter for a more extensive outline of methodological caveats, and appendix B for a full description of the methodology employed in carrying out this project. 7. Moreover, the goal of the analysis is to examine the properties of coordinators, rather than their prevalence. Consequently, oversampling them offers certain advantages, as long as the snowball strategy was not biased toward certain types of coordinating associations and not others (which we have no reason to believe that it was).
07 Chapter 7.qxd
11/9/2009
11:50 AM
Page 235
,
Nodal and Flexible Fronts. Studies of civil society in developing countries have distinguished between and “grassroots organizations” (Edwards and Hulme ; Streeten ), and posited that sometimes play critical coordinating roles within associational networks (Carroll ; Howes ). The chapter terms these coordinating bodies Nodal and establishes three conditions for categorizing associations as such: self-identification as an , reporting that coordinating other associations is an important activity, and reporting working with at least five other associations. Other studies have posited a different sort of coordinating body that is more likely to be “by” or “of ” the popular sectors and less likely to be an outsider group acting “for” constituencies (the stereotypical profile of ). This set of associations, labeled Flexible Fronts in the present analysis, includes associations that selfidentified as a red (network), federation, or mesa de concertación; reported that coordinating other associations is an important activity; and reported working with at least five other associations.8 The vast majority of associations that do not fall into either category are grouped together as “base associations,” a category mainly used for comparative purposes. This approach to distinguishing Nodal , Flexible Fronts, and base associations represents a departure from common practice in monographic studies, which tend to categorize different types of associations—and especially to distinguish between and “grassroots” or “community-based” associations (Edwards and Hulme ; Streeten ; Jelin ; Friedman and Hochstetler )—using criteria such as their level of professionalization, financial resources, and/or membership model. This approach is eschewed for several reasons. First, the data suggest that these particular characteristics do not coincide in a sufficient number of associations to support categorization along those dimensions. A great many associations appear to occupy a middle ground, possessing some of the attributes commonly associated with and other attributes commonly associated with community-based associations. Second, by adopting a different approach to operationalization, those attributes can become subjects of empirical analysis and their differences studied across the types of coordinators identified. Finally, avoiding building these attributes into the typological definition allows the analysis of the problem-solving strategies adopted by the different types of coordinators to condition on those attributes.9 8. There is thus an overlap between the three forms of scaling: coordinators are, by definition, drawn from associations that are more embedded. 9. We originally intended to make a distinction between “horizontal” and “vertical linkages,” planning to use the latter to refer to connections between base associations and coordinators of federationtype entities. However, doing so proved impossible: despite Tarrow’s () use of terms such as “scale
07 Chapter 7.qxd
11/9/2009
11:50 AM
.
Page 236
-
Table . Scaling in the A-Net (percent) Embeddedness in Associational Networks
Argentina Chile Peru Venezuela
Atomizeda
Heavily Linkedb
9.2 23.2 10.0 35.4
35.8 30.7 41.3 25.4
Linked to a Coordinator
49.8 32.6 38.0 24.5
a
“Atomized” associations have no ties.
b
“Heavily linked” associations have or more ties.
c
In all, the sample contains Nodal s and Flexible Fronts.
Incidence of Coordinatorsc Nodal s
Flexible Fronts
13.0 9.3 6.1 4.8
5.0 8.8 14.4 0.4
Taken together, studies of social movements and civil society in Latin America suggest disparate levels of associational coordination across countries. Many analysts have remarked on the apparent robustness of civil society networks in Brazilian cities, for example, while others have explored the relative atomization of associations in Chile (Oxhorn b; Posner ; Foweraker ). Yet, while highly suggestive of disparate cross-national patterns of associational scaling, most of this scholarship comprises case studies of single countries, rather than broadly comparative assessments. The data presented in table . offer a basis for evaluating and comparing patterns of associational scaling—both in terms of inter-associational linkages and the incidence of “coordinators”—in popular-sector neighborhoods in the Argentine, Chilean, Peruvian, and Venezuelan capitals. Striking cross-national differences emerge. The A-Net in Peru is highly coordinated. More than percent of the associations were heavily embedded within associational networks and (similar to Argentina) Peruvian associations were unlikely to be completely atomized. Peruvian associations were also more likely to be linked to coordinators than were their Chilean and Venezuelan counterparts. And Lima was the only city in which there were more Flexible Fronts than Nodal . The robustness of scaling in Peru accords with studies that have explored the proliferation of community-based organizations
shift,” we found the hierarchical analogy to be inapt for the flatter, more fluid network model of associations. We thus abandoned the idea of characterizing the connections between associations (as horizontal or vertical) in favor of highlighting the similarities and differences among different types of associations—base associations, Nodal s, and Flexible Fronts.
07 Chapter 7.qxd
11/9/2009
11:50 AM
Page 237
,
such as clubes de madres and Comités de Vaso de Leche in the popular-sector districts of Lima (Barrig ; Klarén ), or examined new forms of squatter associationalism that rely heavily on coordination (Dosh ). This finding regarding Peru also complements the observation in chapter that associational participation by individuals is highest among residents of Lima. Perhaps associations tend to coordinate where associational participation is most vibrant: in such contexts, individuals may be more likely to participate in multiple associations or to come into contact with other associational participants in their social milieu in ways conducive to the formation of associational linkages. The Argentine A-Net displays a pattern similar to that of Peru, with a large proportion of heavily linked associations, few atomized groups, and robust linkages to coordinators. Further, Nodal are relatively more prevalent in Argentina than elsewhere. The pattern is consistent with analyses that have documented coordinated action on the part of associations and large-scale mobilization in response to the political and economic crises of the early s, particularly by the unemployed (Svampa and Pereyra ; Garay ). Yet while some of the sampling chains began with organizations of the unemployed, the sample contains relatively few of these associations, suggesting that robust associational coordination and collective action are phenomena that extend well beyond the associations of the unemployed.10 Other recent scholarship on the Argentine case suggests a similar picture, one in which the importance, visibility, and coordination of civil society has expanded rapidly in the last decade (Roitter, List, and Salamon ). Chile displays a pattern of somewhat greater associational atomization, conforming to its demobilized stereotype in some ways but offering some less anticipated findings as well. Many analysts have posited a general demobilization of civil society in Chile during the s (Posner ; Foweraker ; Kurtz a), especially in the popular districts of Santiago. Some have suggested that parties of the left exerted a demobilizing effect during and just after regime transition (Oxhorn ; Roberts ), while others have posited that the extensive inroads made by the clientelistic Union Democrática Independiente party in popular-sector districts have dampened social mobilization (Luna ). The moderate degree of atomization in the Chilean A-Net, especially given the bias in the sample against atomized associations, corroborates the notion that associational world is relatively demobilized. At the same 10. Association leaders interviewed in Argentina and Venezuela were also encouraged to deemphasize activities engaged in and linkages formed in reaction to crisis moments occurring around the time the surveys were fielded.
07 Chapter 7.qxd
11/9/2009
.
11:50 AM
Page 238
-
time, however, more than a quarter of the associations coordinated with many others, and connections to coordinators are approximately as widespread as they are in Peru. The A-Net in Venezuela displays even greater associational atomization than in Chile, and contains far fewer associations that serve a coordinating function. This finding accords with some scholarship that has highlighted fragmented patterns of social organization in the Punto Fijo era (–), and with other work that has demonstrated how the high degree of partisan penetration of civil society in Venezuela works to divide, rather than unite civil society associations (Coppedge ; Ellner ). However, it runs counter to other suggestions that popular association and social movement activity received a significant boost in the late s and early s with the collapse of the traditional political parties and the initial years of the Chávez era. While a finding of associational atomization in Venezuela in the late s and early s may be in part an artifact of survey administration,11 it is nonetheless intriguing: it mirrors the findings in chapter regarding individual participation in associations and contrasts sharply with the highly mobilizational chavista project that would soon follow. To sum up, substantial cross-national variation exists in terms of the aggregate level of embeddedness—or density—of the core segment of the A-Net in each capital city. The A-Net in Peru and in Argentina emerges as more tightly linked than its counterparts in Chile and Venezuela. By contrast, there is greater cross-national homogeneity in terms of the incidence of coordinators and the frequency with which associations are linked to coordinators. In each country except Venezuela, coordinators made up approximately a fifth of the sample, and in each country between a quarter and a half of the base associations interviewed had links to coordinators. This finding suggests that coordinators are an important element of Latin American interest regimes and should constitute a focal point in any analysis of how the A-Net operates as a structure for interest representation. We thus turn to a deeper exploration of these coordinators and their roles in organizing collective problem-solving activities.
Nodal s and Flexible Fronts Nodal and Flexible Fronts appear to be very different types of coordinating associations. This section compares them with each other as well as 11. Again, respondents in Venezuela were asked about activities in “normal times” before the most recent political mobilization in Caracas around the regime conflict of –.
07 Chapter 7.qxd
11/9/2009
11:50 AM
Page 239
,
with base associations with respect to a set of basic characteristics: organizational and financial resources, level of internal democracy, and links to actors external to the A-Net. In examining coordinators’ resources and degree of internal democracy, the chapter makes an implicit comparison with the coordinating organizations of the -Hub. Labor federations and confederations played critical roles in aggregating union demands and channeling them toward the state. They were at least formally “democratic” in the sense of having some procedure for internal elections. Nonetheless, unions were frequently seen as prone to Michelsian oligarchy and a lack of responsiveness to the rank-and-file membership, and their elections were often manipulated. Scholarship on suggests the possibility of a similar set of trade-offs, portraying them as possessing the resources and institutional capacity to aggregate and coordinate other associations (Carroll ), but also suggesting that low levels of internal democracy and discrepancies between the backgrounds of employees and members of base organizations may give partnerships with a “double-edged” quality for grassroots groups (Gideon ). Do the Nodal in the sample embody this trade-off? And how do Flexible Fronts, a very different group of coordinators, compare? A second set of questions regards the links between coordinating associations and actors external to the A-Net. A first important actor is the state. The federations of the -Hub were often penetrated by the state in ways that tended to make them less internally democratic and responsive to the base. As discussed in chapter , while governments rarely control and regulate associations extensively, they do enter into various kinds of relations with them. When such relationships entail the conferral of benefits, they may engender associational dependence on the state and thereby constrain or shape associational behavior.12 Transnational networks are another potential source of dependence—one that is more particular to the A-Net. Studies of associationalism and social movements have begun to document the emergence and influence of such networks (Keck and Sikkink ; della Porta and Tarrow ; Tarrow ). In several strands of literature, a major question has concerned the ways in which these international networks interface with the grass roots. Scholarship generally portrays the link between the international and “the local” in the form of domestic that are plugged into international networks, 12. This study simply acknowledges that coordinators’ ties to the state have the potential to compromise autonomy; it makes no assertions regarding whether these ties do compromise autonomy in any given case.
07 Chapter 7.qxd
11/9/2009
.
11:50 AM
Page 240
-
often receiving funding and resources in ways that place them in relations of dependence (Edwards and Hulme ; Stirrat and Henkel ; Thayer ). To what extent and how do coordinators connect with the state? And how well connected is each type of coordinating association with international actors, both foreign governments and transnational social movements? Nodal are in fact marked by something of a representational tradeoff: they possess more material and organizational resources, but are less internally democratic, than other associations. As table . suggests, such groups are more than twice as likely as other groups in the associational world to have high budgets and much more likely to possess bank accounts and paid staff. Possessing more resources does not necessarily translate into greater efficacy, but such groups at least appear to have more potential to effectively problem solve. Yet with regard to internal democracy, table . shows that Nodal follow a pattern that might be labeled “participation without democracy.” While a relatively high proportion of such groups report frequently involving members in the planning and implementation of activities, it is strikingly rare (in comparison to other associations) for Nodal to allow member input into decisions regarding the leadership and orientation of the organization. Table . Organizational and Financial Resources (percent with given resources) Nodal s State Recognition Bank Account Paid Staff High Budgeta
83.1 65.3*† 69.6*† 63.8*†
Flexible Fronts
Base Associations
79.7 31.3 19.6* 29.5
82.8 35.9 35.1 30.4
* Difference of proportions between that type of coordinator and base organizations is statistically significant at approximately the p < . level. †
Difference of proportions between the two types of coordinators is statistically significant at the p < . level. It is worthwhile to evaluate whether the differences of proportions between different groups in the chain-referral (i.e., nonrandom) sample are likely to be the result of chance. The standard test of the significance of a difference of proportions essentially assumes the use of randomly selected data; here this test is employed as a “thought experiment” with chain-referral data. To be extra conservative in assessing “significance” given the unusual nature of this test, a critical value of ., (which corresponds to a significance level of a=.) rather than a critical value of 1.96 (which corresponds to a significance level of a=.) is utilized. a
Monthly budgets were converted into contemporary USD; low = $–.; medium = $–.; high = $–,. Many associations either failed to respond or replied “don’t know” to the survey question regarding funding; these associations are not included in this analysis. We assume that associations that failed to respond are likely poorly funded, rather than evenly distributed across these three categories.
07 Chapter 7.qxd
11/9/2009
11:50 AM
Page 241
,
Table . Participation and Democracy: Role of Beneficiaries (percent) Nodal s
Flexible Fronts
Base Associations
Participate in Internal Elections
32.9*†
80.0*
61.6
Nearly Always Participate in Planning Activities
67.5*
60.6
51.4
Nearly Always Participate in Implementing Activities
68.4
68.2
55.5
Note: See notes for table ..
Flexible Fronts, by contrast, are among the least institutionalized associational entities. They are less likely to possess various types of resources than are Nodal , and even slightly less likely to possess organizational resources than are base associations. Especially remarkable is the extraordinarily low percentage of Flexible Fronts that have paid staff—relative to both Nodal and base associations. A second contrast regards internal democracy. Both types of coordinators, as well as base associations, involve beneficiaries in activities related to planning and implementation. However, in contrast to Nodal , Flexible Fronts combine participation and democracy: they are slightly more likely than are base associations to have members participate in activities, and much more likely than all other associations to embrace democratic practices in terms of internal elections. These findings suggest that many Flexible Fronts may have been established with the primary goal of uniting other associations, rather than with specific programmatic intentions. If their main function is to serve as a vehicle through which other associations interact, stipulations concerning internally democratic practices may in fact be a principle of their foundation. That is, associations may want these structures simply to serve as a space in which they can come together, rather than as an entity that has “authority” or decisional power “over” them.13 In short, these spaces may be crafted with an emphasis on democracy 13. This interpretation is supported by the fact that roughly a third of Flexible Fronts reported that other associational actors were involved in their foundation. By comparison, only percent of all other associations had another association involved in their formation.
07 Chapter 7.qxd
11/9/2009
11:50 AM
.
Page 242
-
and participatory practices, and with an explicit rejection of the bureaucracy and hierarchy that marked the labor movement pattern of the -Hub. Similarly, entities that focus on facilitating interaction among other groups may have less need for high levels of funding or for many of the hallmarks of institutionalization that characterize Nodal . A second set of questions relates to linkages that associations maintain with actors external to the A-Net—both national and international. Such external linkages may be potential sources of dependence and control. As displayed in table ., there is little variation among the different types of groups with respect to the delivery of government programs: roughly a quarter of Nodal , of Flexible Fronts, and of base associations distributed goods, services, or information as part of a state program. However, large differences emerge regarding state funding. Nodal are much more likely than are either Flexible Fronts or base associations to receive money from the government, which may partially explain their greater aggregate levels of financial resources. Receiving such funding may—or may not—create challenges to autonomy. Many state programs relevant to popular-sector interests have been decentralized. States often target groups with established expertise to help them implement new programs. Thus government contracts and funding may simply enable associations to do what they were already doing more effectively (rather than entailing a forfeiture of associational autonomy or the deprioritization of constituent preferences). On the other hand, if such funds are awarded with certain conditions, receiving them—as are more likely to do—may Table . External Ties (percent)
Ties to the National State
Ties to International Actors
Nodal s
Flexible Fronts
Base Associations
Delivers State Programs
27.3
25.8
22.9
Receives Money from the State
43.4*†
23.4
29.1
Money from Foreign Associations
42.1*†
16.9
8.2
Money from Foreign Governments
21.1*†
7.7
3.5
Note: See notes for table ..
07 Chapter 7.qxd
11/9/2009
11:50 AM
Page 243
,
increase the potential for dependence on the state. Further, even in the absence of explicit conditionality, associations that receive such funds may face incentives to behave in ways that encourage continued funding, for instance, eschewing oppositional activity and carrying out activities that will be funded. Either way, state-association ties have the potential to shape the interest regime, channeling associational activity into programs defined “from above” rather than “from below.”14 Linkages with international actors present another potential source of dependence. Nodal contrast dramatically with both Flexible Fronts and base associations in terms of their international linkages. The largest contrast appears with respect to receiving money from international associations. More than percent of Nodal in the sampled core segment of the A-Net reported receiving money from international groups. This is a surprisingly high figure and attests to the substantial presence of international organizations. While Flexible Fronts are only half as likely as are Nodal to receive money from international associations, they are twice as likely as base associations to have such international ties, although this latter difference is not statistically significant. These more democratic kinds of coordinating associations may also serve something of a “brokerage” function, linking local associations into transnational networks. Examining the receipt of money by coordinators from foreign governments offers additional perspective on the nexus between the international and the local. A similar pattern emerges: Nodal are far more likely than other associations to receive money from foreign governments, Flexible Fronts are considerably less likely to receive such funds, and base associations are highly unlikely to receive money from such sources. While there is an obvious affinity between coordinating associations—especially Nodal s—and international funding, it is impossible to pinpoint the direction of causality, and there is probably a reciprocal relationship. Coordinating associations may be better able than others to leverage those skills, or the contacts and knowledge that such coordination affords them, in a way that makes them more visible and attractive to international donors. At the same time, funding from international networks may give associations special resources to play the role of coordinators: receiving funding may open up opportunities to forge contacts with other associations, make it easier for them to invest time, money, or energy in coordinating others, or facilitate their acquisition 14. For a more detailed discussion of the themes of state-association ties, dependence, and autonomy, see chapter .
07 Chapter 7.qxd
11/9/2009
.
11:50 AM
Page 244
-
of the information and skills necessary to coordinate others. Whatever the specific mechanism, the relationship between coordinating associations, especially Nodal s, and integration into international networks seems clear. This relationship has interesting implications for the study of international networks and advocacy groups (Keck and Sikkink ; Tarrow ), and for considerations of the role of international financial institutions such as the World Bank and the International Monetary Fund. Various types of international actors have become more involved in civil society in developing countries over the last fifteen years, with the funding of local associations and movements becoming increasingly prevalent. As this practice has expanded, analysts have begun to raise serious questions about the efficacy of this sort of external funding, and the potentially negative consequences that could spring from the power asymmetries between local groups and international donors. An interesting debate is thus underway concerning how to measure the effectiveness of aid to civil society groups and to assess its benefits from the vantage point of the local actors. The above analysis presents empirical evidence for two points that are relevant for this larger debate. First, the vast majority of international funding seems to go to coordinating associations, suggesting that such funding may have an important multiplier effect. In short, international funding may contribute to the overall strength of the organizational infrastructure, rather than just aiding immediate recipients.15 Second, international actors as a group— transnational social movements and foreign governments alike—are clearly more likely to connect to the A-Net through Nodal than they are to use Flexible Fronts as brokers. When considered in light of the earlier finding regarding the differences between Nodal and Flexible Fronts in terms of internal democracy, the concerns of the transnational social movements literature regarding the problematic nexus between the international and the local are sustained. However, the fact remains that some Flexible Fronts do have international connections. This suggests that aspects of the international-local problematique are not ineluctable, but rather result from choices on the part of international actors concerning the types of associations through which they channel resources and ideas. Pressuring international actors to choose associations that tend toward the Flexible Front model, rather than , as partners is one step activists could take toward addressing the double-edged nature of the international-local exchange. 15. We do not deny, of course, that it is difficult to assess whether funding through these networks is spurring associations to coordinate others, or otherwise introducing them to new contacts and information sources.
07 Chapter 7.qxd
11/9/2009
11:50 AM
Page 245
,
In summary, the core segment of the A-Net is characterized by two distinct types of coordinating associations with notably different characteristics and practices. Nodal tend to be richer, more institutionalized, but less internally democratic than Flexible Fronts (and base associations). In this sense, despite their very different organizational form, they bear some resemblance to the coordinating institutions of the -Hub and may be vulnerable to parallel pathologies or deficits of responsiveness to the base. Nodal also tend to be more heavily linked to external actors, relationships that could potentially compromise associational independence. They serve as the primary conduits between the associational world and international actors, both in government and transnational social movements. Flexible Fronts, on the other hand, contrast sharply with and with the -Hub’s labor federations and confederations: they tend to be relatively resource-poor and far more democratic. They are far less involved than Nodal in external relationships that may involve forms of dependence (though they are more likely than are noncoordinators to have connections to international actors). The final section of this chapter investigates whether these very different types of coordinating associations engage in equally contrasting strategies for problem solving.
Scaling and Problem Solving The analysis thus far has explored three forms of scaling within the A-Net, examining embeddedness and the incidence and characteristics of coordinating associations. A fundamental goal of this volume is to elucidate how the A-Net operates in aggregate as an organizational infrastructure for popular representation. The final section of this chapter turns to this topic, examining the relationship between scaling and strategy choice. First, the hypothesis advanced in the previous chapter that associations’ level of embeddedness is related to their strategy choices is further explored. The section also analyzes the strategy repertoires of Nodal and Flexible Fronts, as their strategy choices are likely to influence those of the associations that they coordinate, amplifying the importance of their decisions regarding problem-solving strategies. This analysis allows us to explore the effect that inter-associational linkages and the presence of coordinators might have on the A-Net’s capacity to serve as an infrastructure for problem solving. Of particular interest is the relationship between scaling and the use of state-targeted strategies. First, among the various strategies that associations might use to solve their problems, one would expect that contacting and strategies of collective contention would depend to
07 Chapter 7.qxd
11/9/2009
.
11:50 AM
Page 246
-
an important degree on scaling. Contacting at the national level is especially interesting given the weak connections between society and the national level of government that seem to mark the current period; nonetheless, the local state may be the level that associations consider to have most control over the issues that concern them. Moreover, examining the degree to which betterlinked groups and coordinators serve as conduits to the state offers leverage on the question of how the A-Net operates as a structure for popular representation, and such an examination allows for the engagement of theoretical debates regarding political representation in contemporary Latin America. Regarding claim making through institutional channels, several prominent studies of interest representation and the new associationalism in contemporary Latin America have suggested that popular associations face difficulties in gaining access to state officials, and these studies have partially traced these problems to what scholars understand as associations’ atomization or relative disarticulation (Oxhorn b; Oxhorn a; Kurtz a). Similarly, other analysts posit that associations embedded in robust associational networks are better able to make claims on the state (Chalmers, Martin, and Piester ; Roberts and Portes ). Is the use of claim-making strategies shaped by associations’ degree of embeddedness, as these studies suggest, and is the logic of the relationship different with regard to accessing local and national officials? A related question concerns what role coordinating associations play in providing channels for access. Labor federations and confederations of the -Hub were particularly instrumental as conduits to national-level policymakers. Are coordinators more likely than other associations to engage in contacting, a finding that would suggest that they might plausibly serve as conduits to state officials for the larger network of associations? In terms of the relationship between scaling and protest strategies, scholars working in the “postcorporatist” tradition have asserted that a positive relationship exists between the network model of coordination and contentious action. How do the different forms of scaling within the A-Net compare in this regard—that is, are Nodal , Flexible Fronts, or groups deeply embedded in associational networks most likely to be protest organizers? A related question concerns a possible decline in protest activity in Latin America since democratic transitions. Protest continues to be a salient feature of politics in “extraordinary times,” such as the widespread mobilization that has occurred in Venezuela under President Chávez at particular junctures over the last decade, or the immense protests in Buenos Aires during Argentina’s dramatic crisis in –. However, many scholars have suggested that, on the whole, protest has been waning in the region. While the cross-sectional data cannot directly assess
07 Chapter 7.qxd
11/9/2009
11:50 AM
Page 247
,
change over time, they can shed some light on the hypothesis that the decline in protest has its roots in the increasing influence of within the associational world (Foweraker ). Are Nodal less likely to organize protest than are base organizations? And do Flexible Fronts behave more like their coordinator brethren, or more like the base-level groups that they aggregate? Although the analysis is focused on issues concerning claim making on the state, it also briefly investigates the relationship between scaling and capacity building. There is reason to expect that the latter relationship would be strong. Associations that are highly embedded in associational networks may have good reason to work to strengthen their partners. Similarly, capacity building has been seen frequently as a quintessential strategy of coordinators, especially . Nonetheless, since capacity building is such a crucial part of the problem-solving repertoire in the A-Net, it remains important to empirically examine its relationship with scaling. While these questions are framed in terms of aggregate tendencies, or forms of scaling, within the A-Net, they are pursued empirically by examining variables at the level of the individual association. By assessing, for example, the relationship between an association’s embeddedness in networks and engagement in contacting, the findings may shed light on the importance of network density to the A-Net’s capacity to channel claim making on the state. The analytic strategy is thus to assess the correlation between three variables that correspond to three different forms of scaling (one that measures the degree of associational embeddedness and dummy variables that measure whether the association is a Nodal or a Flexible Front, with base associations as the reference category), and four variables that measure associations’ use of four problem-solving strategies: contacting at the national level, contacting at the local level, organizing protest, and capacity building. The analysis is conducted principally through a series of logistic regression models, conditioning on a set of other variables that may also influence strategy choice. A first step is to examine bivariate relationships between the scaling variables and the pursuit of each strategy. Table . displays the problem-solving repertoires of coordinators and re-presents the findings from chapter regarding the relationship between associations’ degree of embeddedness and their use of particular problem-solving strategies. This initial look suggests a positive relationship between scaling and associations’ propensity to use each type of problem-solving strategy: both types of coordinators as well as better-linked groups were far more likely to use accessing strategies and engage in capacity building than were noncoordinators and atomized groups. In particular,
07 Chapter 7.qxd
11/9/2009
11:50 AM
.
Page 248
-
Table . Scaling and Strategy Choice (percent engaging in each strategy)
Coordinators
Embeddedness
Contact at National Level
Contact at Local Level
Organize Protest
Engage in Capacity Building
Nodal
51.9*
74.0*
29.9†
85.7*
Flexible Front
48.5*
74.2*
50.0*
86.4*
Base Associations
26.2
57.6
27.5
46.4
Heavily Linked
36.7*
68.7*
34.5*
67.1*
Atomized
22.8
47.3
15.8
34.8
Note: See notes for table ..
Flexible Fronts were far more likely than were noncoordinators or Nodal to organize protest. However, the coordinator and embeddedness variables are correlated with each other,16 as well as related to other organizational characteristics that may affect strategy choice (such as financial and organizational resources or ties with the state). Multivariate analysis can help unravel these relationships, providing insight into the effect on strategy choice of being well embedded in associational networks and of being a Nodal or Flexible Front, conditional on other variables.17 A few methodological caveats warrant mention. First, it is important to be clear about the goal of the analysis. The logistic models are used as a means of examining some complicated relationships among the attributes and activities of associations in the sample. Given that the sample is not representative of the population of popular associations, the analysis should not be construed as a tool for estimating population parameters.18 Second, in view of the relationship between the sampling technique (chain-referral sampling) and some of the phenomena under study (embeddedness within associational networks 16. Nodal s and Flexible Fronts, by definition, have five or more links, and thus cannot be “atomized”; any coordinator scored on the “embeddedness” variable must fall in either the intermediate or “heavily linked” categories. 17. In this respect, we reiterate that, in contrast to some other studies, organizational and financial resources are not elements of our definition of Nodal or Flexible Fronts. 18. As such, the suggestion is not that the model represents a “true model” in the world that is generating data on the dependent variable, or that the assumption of conditional independence is perfectly satisfied. The logistic model in this context is a helpful tool, but its power should not be exaggerated.
07 Chapter 7.qxd
11/9/2009
11:50 AM
Page 249
,
and being a coordinating association), the observations are obviously not independent, necessitating a degree of caution in interpreting significance levels. A final concern bears on model specification, especially the issue of missing data and its potential influence on results. More than a third of the associations in the dataset were unwilling or incapable of offering specific budget numbers to interviewers (n = ). These missing data present a quandary, since level of financial resources is a theoretically important variable. However, statistical tests indicate that missing budget data were not significantly correlated with any of the other associational attributes of interest.19 Nonetheless, models are estimated that include the resource variable (and thus exclude many associations), as well as models that exclude the resource variable (and for which the number of observations is consequently larger). For each of the dependent variables under examination, three different models are specified, with results presented in tables . through . below. The first column reports results for a reduced form model that examines just the scaling variables: Nodal , Flexible Front, and embeddedness, as well as a set of country controls. Next is a model that adds measures of associational institutionalization and associational ties to the state, allowing for the assessment of two alternative hypotheses regarding strategy choice. Finally, the third model adds a measure of financial resources; as mentioned above, this model includes only associations that reported budget figures. To facilitate interpretation of results, first differences for the statistically significant variables in the second and third models are also reported. Table . analyzes the strategy of contacting officials at the national level. In the reduced form model, all three scaling variables—embeddedness and each coordinator variable—are significant, with the relationships between being a coordinator and contacting local-level officials emerging as somewhat stronger. As other variables are added, however, embeddedness decreases in power. Model II (which adds variables measuring ties to the state and organizational resources) suggests that a change from complete atomization to working with thirty-one or more associations will only result in only a . increase in the predicted probability of an association engaging in national-level contacting, and in model III (which further adds a variable for the size of associational budgets), embeddedness is not statistically significant. This evidence suggests that the density of networks may be weakly related—or even unrelated 19. To make this assessment, we created a dummy variable (“lacks budget data”) that we used as the dependent variable in a logistic regression that included all other independent variables from the full model on the right hand side of the equation. We found no significant relationships between those independent variables and the dependent variable “lacks budget data.”
07 Chapter 7.qxd
11/9/2009
11:50 AM
Page 250
Table . Contacting at the National Level (logistic regression) First Difference
Model I
Model II
Peru
-.819** (.216)
-.725** (.224)
-.697* (.256)
Chile
.046 (.197)
.002 (.208)
-.018 (.241)
Venezuela
-.651** (.214)
-.550* (.220)
-.1.202*** (.372)
.864** (.273)
Ties to the State
.19
Model III
.721* (.343)
First Difference
.15
.024 (.033)
Budget Organizational Resources
.920** (.318)
.18
.388 (.399)
Embeddedness
.529* (.250)
.521* (.252)
.11
.587 (.323)
Nodal
.798** (.265)
.806** (.269)
.18
.635* (.316)
.14
Flexible Front
.869** (.283)
.844** (.285)
.19
.930** (.342)
.22
924
924
N
602
*Significant at the . level; **significant at the . level; ***significant at the . level. Note: First differences show the change in the predicted probability of the dependent variable equaling when the relevant independent variable is set at its maximum versus its minimum, with all other independent variables maintained at their means. Budget is the association’s budget in U.S. dollars, logged. Ties to the state is a nonweighted scale (, ., ., ) constructed from three dichotomous variables (whether someone from the government participated in the founding of the association; whether the association receives money from the government; and whether the association distributes goods, services, or information as part of a government program). Organizational resources is a nonweighted scale (, ., ., ) constructed from three dichotomous variables (whether the association has members, whether the association has state recognition, and whether the association has a permanent leadership structure). Number of ties is the number of ties the association reported within the associational world, divided into categories (, –, –, –, +) and then rescaled –. Flexible Front—dummy variable (yes/no). Nodal —dummy variable (yes/no). For each model, a constant was estimated but is not reported.
07 Chapter 7.qxd
11/9/2009
11:50 AM
Page 251
,
to the A-Net’s capacity to facilitate access to national officials. By contrast, the results strongly suggest that both types of coordinators may serve as access points or conduits to the state in the A-Net. Both coordinator variables are strong predictors of national-level accessing in all three model specifications, and most of the first differences are robust.20 Models II and III suggest that ties to the state and organizational resources have some predictive power as well, thus confirming findings advanced in chapter . While the goal is to analyze separately the effects of each variable on national-level contacting, it is useful to keep in mind that, de facto ties to the state and organizational resources may provide coordinators an additional boost in national-level contacting, as cross-tabular analysis (not shown) indicates that coordinators are slightly more likely to possess both of these attributes than are base associations. Further, in each model, the country dummies are also significant: there appears to be a strong relationship between the country in which associations are operating and the likelihood they will engage in national-level contacting. This finding further confirms the analysis in the previous chapter, which pointed to the existence of important cross-national differences in associational strategy repertoires.21 A different picture emerges for the strategy of contacting at the local level (see table .). In contrast to the finding above of a strong relationship between being a Nodal or a Flexible Front and contacting at the national level, coordinators are no more likely than are base associations to contact at the local level. However, another form of scaling—linkages among associations—appears to be a powerful predictor of engaging the local state. Model III predicts that well-embedded associations (possessing thirty-one or more ties) are percent more likely to contact local-level officials than are completely atomized groups, a substantively significant change. These findings support the analysis in chapter , restated above, regarding the bivariate relationship between embeddedness and local-level contacting: contacting local officials appears to be something of a coordinated activity. As with national-level contacting, organizational resources and ties to the state are also significant predictors 20. In interpreting changes in predicted probabilities, it is important to remember that nationallevel contacting is relatively rare. The prediction (in model III) that Flexible Fronts will be . more likely to contact at the national level than others types of associations, for example, represents nearly a doubling of the predicted probability of national access: the probability that a non–Flexible Front would engage in contacting at the national level was just over percent, while the likelihood that a Flexible Front would do so was almost percent. 21. Since the objective is to explore the relationships between scaling and strategy use (and since the country dummies were simply introduced as controls), the analysis does not dwell on the importance of these variables. Chapters and explore the issue of cross-national variation in greater depth.
07 Chapter 7.qxd
11/9/2009
11:50 AM
.
Page 252
-
Table . Contacting at the Local Level (logistic regression) First Difference
Model I
Model II
Peru
.327 (.201)
-.225 (.211)
-.239 (.252)
Chile
.465* (.216)
.429 (.227)
.318 (.265)
Venezuela
1.577*** (.204)
-.1498*** (.210) 1.004*** (.288)
Ties to the State
Model III
-1.575*** (.317) .21
.785* (.370)
.16
-.075* (.034)
Budget Organizational Resources
1.029*** (.294)
.25
1.458*** (.382)
.33
.16
1.292*** (.330)
.28
Embeddedness
.705** (.243)
.697** (.247)
Nodal
.263 (.302)
.260 (.307)
.238 (.363)
Flexible Front
.164 (.312)
.119 (.315)
.397 (.426)
924
924
602
N
First Difference
Note: See notes for table ..
of local-level contacting: it appears that possessing these attributes, rather than others closely associated with being a coordinator, makes a difference in terms of local-level contacting. As model III also reveals, financial resources were negatively correlated with the pursuit of local contacting strategies: betterfunded associations appeared more likely to eschew contacting at the local level while resource-poor base associations tended to depend on local-level contacts to a greater extent. Taken together, the results thus far suggest an intriguing relationship between scaling and the prevalence of claim-making strategies in the A-Net. More consistently than any other variable considered, the three forms of scaling emerged as strong predictors of contacting at the national and local levels—though they mattered in different ways. Associational embeddedness is a powerful predictor of local-level contacting, suggesting that the denser the associational networks, the greater the A-Net’s ability to serve as a channel between the popular sectors and local state officials. However, network density appears to
07 Chapter 7.qxd
11/9/2009
11:50 AM
Page 253
,
offer limited returns in terms of national-level contacting. Instead, being a coordinator—either a Nodal or a Flexible Front—was the strongest predictor of national-level contacting. This suggests that coordinators—like federations and confederations under the -Hub—may well serve as conduits to national-level officials. The discussion now turns to an examination of contentious claim making (see table .). Like national-level contacting, organizing protest is a relatively uncommon strategy, utilized by only percent of the total sample. Consequently, even relatively small first differences can reflect substantively large changes across subgroups. Two important findings emerge from the analysis. First, embeddedness is a consistent and powerful predictor of organizing protest across all model specifications. This finding accords with the findings of many studies of civil society and social movements. A more surprising set of findings emerges regarding coordinating associations. Nodal are neither significantly more nor less likely to organize Table . Organizing Protest (logistic regression) First Difference
Model I
Model II
Peru
.014 (.201)
.014 (.209)
.163 (.245)
Chile
-.138 (.205)
-.079 (.213)
.134 (.247)
Venezuela
-.573** (.219)
-.558* (.224)
.056 (.310)
.031 (.281)
-.051 (.343)
Ties to the State
Organizational Resources
-.350 (.298)
Embeddedness
.972*** (.245)
.965*** (.246)
Nodal
-.338 (.281)
-.335 (.281)
Flexible Front
.531* (.275)
.545* (.276)
924
924
Note: See notes for table ..
First Difference
.051 (.032)
Budget
N
Model III
-.011 (.374) .20
.673* (.303)
.15
-.383 (.331) .12
.798* (.334) 602
.19
07 Chapter 7.qxd
11/9/2009
11:50 AM
.
Page 254
-
protest than are base associations (the coefficient is negative, but not significant). While it may be the case that Nodal seek to maintain a somewhat lower profile, there is little support for the idea that the presence of Nodal within the associational network suppresses contentious activity. On the other hand, Flexible Fronts are much more likely than are other base organizations or Nodal to organize protest. This finding offers additional support for the notion that coordinators serve as conduits for claim making on the state (here through their organization of contentious activity). Finally, the discussion turns to an alternate to state-targeted strategies: the society-targeted activity of strengthening other associations (see table .). As chapter showed, popular associations in Latin America are just as likely to engage in society-targeted strategies (a genre that includes both capacitybuilding and provisioning activities) as they are to engage in state-targeted strategies. Society-targeted strategies are thus crucial to the overall operation of the A-Net as a problem-solving structure. They are somewhat subordinated Table . Capacity Building (logistic regression) First Difference
Model I
Model II
Peru
.038 (.201)
.128 (.209)
.028 (.245)
Chile
.416* (.202)
.469* (.210)
.481 (.248)
Venezuela
.500* (.198)
.582** (.204)
1.238*** (.314)
.504 (.267)
.606 (.345)
Ties to the State
Model III
First Difference
.025 (.033)
Budget Organizational Resources
.061 (.283)
.630 (.369)
Embeddedness
1.257*** (.233)
1.240*** (.234)
.30
1.108*** (.301)
.27
Nodal
1.511*** (.349)
1.515*** (.349)
.32
1.673*** (.415)
.34
Flexible Front
1.694*** (.378)
1.700*** (.378)
.34
1.561*** (.417)
.32
N Note: See notes for table ..
924
924
602
07 Chapter 7.qxd
11/9/2009
11:50 AM
Page 255
,
in this discussion merely because of the expectation of a robust relationship with scaling. For instance, society-targeted strategies such as training community leaders or providing services to the population are quintessential activities of Nodal . Nonetheless, closely examining the relationship between scaling and societal strategies provides a check on conventional wisdom and offers a reference point for considering the strength of other relationships between scaling and strategy choice. As expected, being well embedded in associational networks, being a Nodal , and being a Flexible Front all emerge as very strong predictors of using capacity-building strategies. Across every specification, each of these variables remains significant at the . level. Further, the first differences for each variable are also substantial. However, in accordance with the bivariate analysis carried out in chapter , associations with more material resources and those with more organizational resources were no more likely than their less-endowed counterparts to engage in capacity building. Table . offers a summary of the findings regarding the relationships between the three forms of scaling and strategy choice. What do these findings, taken together, suggest? In a general sense, the results strongly support the idea that the nature and degree of scaling is a vital characteristic of the A-Net, strongly correlated with the propensity of associations to use a diverse range of problem-solving strategies. An important implication for existing studies of interest associations and social movements is the idea that repertoires of action for any given organization are likely to depend on the activities embraced by other groups to which they are connected through their network. In considering why associations prioritize some problem-solving strategies over others, it is necessary to look beyond external variables implicit in the idea of political opportunity structure, and beyond organization-centric variables such as resources and/or ideology, to focus on the nature of peer organizations and the ways that ideas and practices flow through networks. Moreover, a particular pattern in the relationship between scaling and problem solving can be discerned, one that is central to understanding how the Table . Summary: Forms of Scaling and Strategy Choice National Contacting Embeddedness Nodal Flexible Front
X X
Local Contacting
Organizing Protest
Capacity Building
X
X
X X X
X
07 Chapter 7.qxd
11/9/2009
.
11:50 AM
Page 256
-
A-Net operates. A positive relationship emerges between embeddedness and local-level contacting, organizing protest, and capacity building. One interpretation is that as associations gain strength through collaboration, they are more willing or better able to engage in these strategies. However, being well embedded in associational networks is not related to national contacting, which may be vital for representing the interests of constituencies. Both Nodal and Flexible Fronts are much more likely than is the rest of the associational world to have national contacts. Coordinators, rather than the network as a whole, thus may serve as conduits to the national-level state. Finally, Flexible Fronts prove exceptionally versatile. They serve as conduits for institutionalized claim making on the national-level state, are more likely than any other type of group to organize protest, and they also engage heavily in capacity building.
Conclusions This chapter sought to explore the nature of the organizational infrastructure of the A-Net. It identified three forms of scaling, revealing that the way in which associations are linked and organized in the A-Net is quite different from the way unions were connected and coordinated in the -Hub. One form entails linkages among associations (embeddedness). The other two involve different coordinating entities, either Nodal or Flexible Fronts. Each form of scaling was analyzed from three complementary angles. The examination of the incidence of scaling in the core segment of the A-Net captured in the country samples revealed significant differences. Peru and Argentina exhibited the highest degree of associational scaling, Chile was a more intermediary case, and the Venezuelan A-Net presented the lowest levels of scaling. Further, the A-Net in each country had different emphases in terms of scaling: Peru stood out for its relatively high percentage of heavily linked associations and large number of Flexible Fronts, while the Argentine A-Net had the highest proportion of base associations linked to a coordinator, and the largest percentage of Nodal . The second section explored striking differences between Nodal and Flexible Fronts. Nodal are wealthier than are Flexible Fronts, and they also possess more organizational resources. In terms of constituency relations, both types of coordinators involve beneficiaries in planning and implementation, but Nodal are less likely to allow them to participate in internal elections than are Flexible Fronts. Finally, in terms of external ties, Nodal
07 Chapter 7.qxd
11/9/2009
11:50 AM
Page 257
,
are much more likely to have linkages with both the state and international actors, relationships that other scholarship suggests may entail dependence and controls on autonomy. These findings reinforce common perceptions of as better-institutionalized entities that “float above” base associations, a status that can impede the establishment of effective relationships with associations that are both for and of the poor. In reinforcing this understanding of Nodal while showing that Flexible Fronts operate very differently, the findings are particularly relevant to literature focused on the nature and implications of linkages between transnational actors and social movements, and “the local.” The analysis of the relationship between scaling and the use of problemsolving strategies focused heavily on the degree to which scaling shapes claim making on the state. Associational embeddedness emerged as a key predictor of claim making on local state officials, engaging in capacity building, and, to a lesser extent, organizing protest. Each type of coordinator also had particular strategic tendencies. Flexible Fronts were particularly likely to organize protest. Most strikingly, whereas associations that were well embedded in networks were no more likely to access national-level officials than were more atomized groups, both Nodal and Flexible Fronts emerged as key conduits connecting the network to national-level officials. Together, these findings suggest that, at least within the “core segment” of the A-Net represented by the sample, the more pessimistic vision of an atomized popular interest regime in which associations are incapable of leveraging their numbers to solve problems is overstated. Three broader implications of this analysis for the study of popular-sector interest representation in Latin America are noteworthy. First, it is important to identify the different forms that scaling can take when considering the infrastructure of the A-Net. While some analysts in the social movement and civil society traditions examine the role of within networks, the bulk of attention in this literature has been paid to the importance of what is referred to here as embeddedness, or links between base associations. Indeed, scholars have tended to laud the perceived absence of hierarchy in the emergent associational sphere, explicitly contrasting this “flatter” pattern of coordination with the more hierarchical pattern (dominated by verticalist federations) that characterized the -Hub. Yet this analysis suggests that coordinating associations play a key role within the A-Net: while the topography of the A-Net may not resemble the vertical orientation of the -Hub, it is not entirely flat. The presence and operation of these coordinators deserve significantly more attention.
07 Chapter 7.qxd
11/9/2009
.
11:50 AM
Page 258
-
A related point is that different types of coordinators exist, and they have different qualities, operate in different ways, and may have different capabilities in terms of interest intermediation within the A-Net. Because the example of federations in the -Hub looms so large, one often thinks of coordinating entities as ineluctably marked by characteristics representing a classic tradeoff—institutional capacity and resources, on the one hand, and a lack of internal democracy and a tendency toward co-optation and state control on the other. Nodal , in fact, do appear to bear some similarity to federations under the -Hub in these respects; consequently, partnerships with Nodal may be vulnerable to a similar trade-off in which the associations they coordinate benefit from their resources while suffering in terms of autonomy. Flexible Fronts, however, represent a very different kind of entity, one that operates more democratically, engages in less potentially autonomy-compromising interaction with the state, and is thus more likely to coordinate the interests of popular associations without such trade-offs. Cross-national differences in the types of coordinators that predominate may thus also shape more general differences in how the A-Net functions. For example, although the A-Net is highly linked and characterized by many coordinators in both Peru and Argentina, the orientation in the former toward Flexible Fronts may give it a more “basist” character than the latter, in which Nodal are more present. Finally, the finding that different forms of scaling lend themselves to different problem-solving orientations in the A-Net suggests that the diversity of forms of scaling may constitute a foundation for the two “dualities” of the A-Net outlined in chapter —its inclusion of both state-targeted and society-targeted problem solving, and its facilitation of both contentious and noncontentious claim making. The findings regarding the relationship between scaling and problem solving also suggest that cross-national differences in scaling may strongly shape the aggregate operation of the A-Net as a structure for representation. Despite decentralization, many policies of significant relevance to the popular sectors continue to be formulated and implemented at the national level, especially those that pertain to the popular sectors as a group rather than as disaggregated units. This reality reinforces the importance of closer examination of coordinators (which proved especially likely to contact the state at the national level) in future analysis. Our data do not allow for an evaluation of the outcomes of coordinators’ efforts to represent popular-sector interests. Nodal and Flexible Fronts may not be equally proficient at attaining or wielding influence over policies. Further, their strategic choices might affect associations connected to them through the A-Net in different ways. Further analysis might take a finer-grained
07 Chapter 7.qxd
11/9/2009
11:50 AM
Page 259
,
look at these issues, shifting from a perspective focused on the atomization of associations to one that examines the three forms of scaling analyzed here, considering what each implies for associations’ interactions with their constituents, and probing further into how these different forms of scaling shape popular representation in Latin America.
08 Chapter 8.qxd
11/9/2009
11:50 AM
Page 260
8 Candelaria Garay
The premise of this book is that a major shift has occurred in the institutional structures through which urban popular sectors achieve representation and solve problems in Latin America. In this transformation from the -Hub, a union-party interest regime, to the A-Net, based in associational networks, the central role of labor unions and labor-affiliated parties has waned and popular associations have gained relevance as agents of popular problem solving and interest intermediation. Since labor unions and political parties nonetheless remain important actors in the A-Net, a key question concerns their interaction with the rising associationalism. In exploring associational linkages to labor unions and to political parties, this chapter begins to investigate how such linkages may affect the problem-solving strategies of popular associations in the A-Net. Relationships between associations and labor unions are important for several reasons. First these linkages are an indicator of whether common interests are formed across different segments of the popular sectors—those organized in unions, which typically represent wage-earner and work-related concerns, and those participating in associations, which bring together a more variegated constituency around a variety of issues, such as consumption, social services, and housing. Given the traditional segmentation of interests among the popular sectors in Latin America, this question is of particular relevance in the new interest regime. At the same time, associational linkages to unions are important because they can facilitate the diffusion of collective action repertoires
08 Chapter 8.qxd
11/9/2009
11:50 AM
Page 261
,
and techniques in the A-Net, potentially influencing the problem-solving activity of associations. As described in chapter , the collective action patterns that characterize associations put them at a relative disadvantage compared to labor unions in exerting leverage on behalf of their constituencies. In this light, one might ask whether popular associations linked to unions show distinct patterns of problem solving in which limitations that plague popular associations— such as difficulty in coordination and in making state-targeted demands, especially at the national level—are somewhat reduced.1 Similarly, although associations in the A-Net rarely have the kind of affiliations to political parties that were more typical of unions in the -Hub, it is important to ask what relations they do have to parties. Associational linkages to political parties are fundamental subjects for analysis because they indicate whether associations seek to advance their interests in the electoral arena and what strategies they pursue to do so. Recent studies have often cast such linkages somewhat negatively, pointing to the double-edged nature of party-association relations, in which any advantages may be offset by a loss of autonomy entailed in forging close ties with political parties (see discussion in Oxhorn ). Further, analysts tend to see party-association relations from the angle of political party preferences, which are largely centered on electoral concerns. From this perspective, they generally suggest that political parties either pursue clientelistic relations with popular associations, using them as vehicles for vote maximization, or they ignore them altogether, reaching lowerclass individuals in unmediated ways.2 Equally important is to explore whether associations forge ties with political parties to advance their interests, in what ways they do so, and what roles party-linked associations perform within the broader interest regime.3 Using the surveys of associations this chapter describes some aspects of the linkages of popular associations with labor unions and with political parties, and begins to explore the implications of those linkages for patterns 1. On popular associations and their features, see, among others, the discussion in chapters and , A. Thompson (), Oxhorn (a), Campetella and González Bombal (), and Foweraker (). 2. On clientelistic linkages see Chalmers, Martin, and Piester (:), Auyero (), Levitsky (: chap. ), and Oxhorn (). On parties/candidates avoiding organizational mediations, see Roberts (), Weyland (b), and Boas (forthcoming). On types of linkages between parties and associations, see, among others, Keck (), C. Schneider (), Hipsher (), Oxhorn (), Chalmers, Martin, and Piester (), Roberts (), and Schönwälder (). 3. There are studies that explore how popular associations and parties relate and how association members view political parties. In particular, see C. Schneider (), Levitsky (), Roberts (), and Oxhorn ().
08 Chapter 8.qxd
11/9/2009
.
11:50 AM
Page 262
-
of problem-solving strategies in the A-Net.4 The sample of about associations in each of the capital cities of Argentina,5 Chile, Peru, and Venezuela (n = ) constitutes a unique source of comparative information on popular associations; yet, since surveys were administered through chain-referral sampling, caution should be employed when interpreting such data. The first section of this chapter asks whether union-linked associations employ distinct patterns of coordination and problem-solving strategies and explores their implications for the A-Net. The analysis shows that union-linked associations are much more likely to coordinate with other associations, form part of federated entities, organize protests, access the national state, and sit in policy councils than are associations not linked to unions. Given that unionlinked associations have a larger number of inter-associational linkages, their pattern of problem solving may be consequential for other associations in the A-Net, particularly in Argentina and Chile, where that pattern is stronger than in Peru and Venezuela, where the distinctiveness of union-linked associations is weaker or nonexistent, respectively. Associations linked to unions in Argentina and Chile may thus play central roles in the A-Net, potentially operating as mobilizing structures that aggregate diverse associations in episodes of contention and in pressing their interests through institutional channels.6 The second section of the chapter looks at party-association linkages from two perspectives. First it explores party strategies toward associations, and then it conceptualizes different types of linkages that associations pursue to advance their interests through parties: issue-based, candidate-based, and multifaceted linkages. With respect to party strategy, this analysis indicates that associations are often viewed as sources of electoral or campaign support for candidates running for office, a fact that generally supports the existing literature about party politics in Latin America. At the same time, associations’ strategies toward parties show the prevalence of issue-based linkages, by which associations seek to solve problems by contacting party leaders, although crosscountry variation in associational strategies is evident. In Argentina and Chile issue-based linkages tend to be the dominant strategy, whereas in Peru strategies to support candidates during campaigns are more prevalent, and in Venezuela 4. See appendix B for a description of the procedure. As detailed there, the surveys were administered to about thirty associations in each of eight districts in each capital city between and . 5. The capital city of Argentina, Buenos Aires, here refers to (Greater Buenos Aires), which includes the City of Buenos Aires and the surrounding conurbano (city belt). 6. Mobilizing structures are ties, networks or organizations that facilitate the emergence of contention (Tarrow : chap. ). Mobilizing structures allow the diffusion of ideas, frames, information, and the recruitment of participants, and they offer “structures of solidarity” incentives that facilitate collective action (McAdam :).
08 Chapter 8.qxd
11/9/2009
11:50 AM
Page 263
,
multifaceted linkages, which involve more complex interaction with political parties, are relatively more frequent. Finally, associations with issue-based linkages and those with multifaceted ones appear to be particularly embedded in the A-Net, as they reported a larger number of inter-associational ties than associations with other types of party linkage or without linkages to parties. By contrast, associations with candidate-based linkages are quite isolated, showing a relatively smaller number of connections within the A-Net. Differential levels of embeddedness suggest that associations with issue-based linkages may play a more central role in the A-Net than associations with candidatebased linkages, whose lower levels of connectivity probably circumscribe the spread of such relationships to some extent. The final section of this chapter discusses the general findings and country-level patterns in union-association linkages and party-association linkages observable in the data.
Union-Association Linkages Scholars generally suggest that popular associations are limited in their capacity to advance the interests of their constituencies. In particular, associations are assumed to have difficulty coordinating with other associations, organizing protest activity, and accessing the state through formal institutional channels.7 Associations are often seen as either atomized and resource-poor (Kurtz a) or connected through fluid associational networks (Chalmers, Martin, and Piester ). In either case, they are usually seen as lacking higher-level structures of coordination fundamental for pursuing their interests, particularly beyond the local level (Evans ; Fox ). One might expect union-linked associations to show characteristics closer to those of labor unions—and thus be oriented to productionist issues—and to employ a pattern of collective action more oriented toward protest, state contacting, coordination, and the use of state resources. Such a pattern on the part of union-linked associations might arise in two different ways. On the one hand, popular associations that are generally more active in claim making and protesting and that coordinate with a larger number of associations may pursue union links to advance popular interests more effectively and through broader, encompassing coalitions. Those linkages may in turn encourage and 7. On the difficulties faced by popular collective action in Latin America, and on the atomization of popular-sector interests, see Kurtz (a) and Oxhorn (a). A different view about collective action is presented in Arce and Bellinger ().
08 Chapter 8.qxd
11/9/2009
.
11:50 AM
Page 264
-
reinforce such strategies on the part of associations, providing them with additional resources and experience. On the other hand, linkages with unions may be a source of learning and a mechanism of diffusing repertoires and techniques for collective action from unions to associations that would otherwise be less prone to coordination, protest, and national-level demand making.8 In the first case, union links facilitate and reinforce a particular problem-solving strategy; in the second case, they are a source of change, promoting the adoption of a different problem-solving repertoire on the part of associations.9 This section first presents data concerning the extent to which associations coordinate activities with unions and with which kinds of unions. It then explores whether union-linked associations show distinct problem-solving strategies by examining scaling, claim making, and the use of state resources. Finally, it analyzes the implications of such strategies for the operation of the A-Net by examining the level of embeddedness, understood as the number of inter-associational linkages, of union-linked associations. Higher levels of embeddedness may indicate that these associations are central actors in the A-Net and may transmit strategies to other associations. Union-Linked Associations Associations that are linked to unions constitute . percent of the associations sampled (n = ), with the largest portion corresponding to the Argentine case (. percent) and the smallest to Venezuela (. percent), as summarized in table ..10 One might expect that associations oriented toward productionist issues would be more likely to work with unions. Yet in each of the countries under study, union-linked associations do not differ significantly from the rest of the 8. On the diffusion and adoption of organizational practices, see in particular Di Maggio and Powell (). On the diffusion of information, repertories, and learning through networks in the social movement literature, see, among others, Diani (a). In the case of Latin America, Roberts (:) suggests that labor unions are particularly important in the diffusion of organizational models and methods of collective action to other sectors of civil society. 9. In the literature on social movements, ties among individuals or groups are frequently understood as mobilizing structures (McAdam, MacCarthy and Zald a; Tarrow ) or social networks (Diani a) that facilitate the emergence of contention through the diffusion of ideas, cultural frames, and critical information as well as offering “structures of solidarity incentives” (McAdam, MacCarthy and Zald ) that facilitate the recruitment of movement members or participants. There is a vast literature about how and what type of linkages matter for collective action. See Gould (; ), Ansell (), and Diani (a). 10. The survey question asked: Do you work with labor unions? All associations that responded affirmatively were coded as having union ties.
08 Chapter 8.qxd
11/9/2009
11:50 AM
Page 265
,
Table . Union-Linked Associations
Argentina Chile Peru Venezuela Total
(n=240) (n=208) (n=240) (n=240) (n=928)
Percent
Number
22.5 17.3 12.9 8.8 15.3
54 36 31 21 142
associations sampled in terms of their likelihood of engaging any of the following five issue areas, which are typically pursued by popular associations: consumption (e.g., food distribution, poverty), social services (e.g., education, health), political issues (e.g., advocacy, human rights), production (e.g., employment, job training), and neighborhood improvement and quality of life (e.g., infrastructure and urban services).11 What is of particular importance to the analysis is the kind of unions to which associations reported having linkages. One might identify three types of labor unions in Latin America: traditional corporatist unions, social movement unions, and informal-sector unions.12 In terms of their characteristics, traditional corporatist unions represent formal-sector workers exclusively, are regulated by labor law, and form part of a corporatist structure of interest intermediation. In general, these unions organize skilled or semi-skilled workers with higher incomes, and constitute the centerpieces of the union structure and its peak labor confederation developed under import-substituting industrialization. These unions were thus the dominant actors of the -Hub. Traditional corporatist unions have often, although not always, discouraged internal democracy. Especially during market reforms in the s and s, they have prioritized their organizational survival over the interests of their bases, thus limiting their mobilization and often negotiating organizational payoffs in exchange for moderation.13 Despite the decline of the -Hub, these unions remain important interest organizations (see, e.g., Etchemendy and Collier ). Social movement unions generally reject the role of corporatist unions as “interest organizations” and typically prefer the formation of a broader 11. There are two exceptions. In Peru associations that do not work with unions are significantly more oriented toward consumptionist issues than are union-linked associations, and in Chile they are more likely to engage in neighborhood improvement than union-linked associations. 12. Among others, see Collier and Collier (), Cross () Bensusán (), and Roever (). 13. In particular, see Murillo (), Etchemendy (), Bensusán (), and Cook ().
08 Chapter 8.qxd
11/9/2009
.
11:50 AM
Page 266
-
“workers’ movement.” These unions often seek to develop connections with popular associations and social movements, and they frequently engage in protest around broader political or social issues as well as work-related demands (Seidman ; Bensusán ). Social movement unions usually participate in alternative peak confederations, most of which were formed in the s and s, and sometimes develop higher levels of internal democracy. Publicsector unions, such as those of teachers, health workers, and other state workers, are generally prominent among social movement unions. A final type is the informal-sector union, which is formed by specific categories of informal-sector workers, such as street vendors, artisans, and microproducers. These organizations are generally not considered labor unions because they are not regulated by labor law, neither constrained by it nor receiving the right of collective bargaining. They are typically not formed by “salaried” workers within the informal sector and thus usually do not target their demands to employers. Although they collect membership dues and have some capacity and experience in organizing protests and contacting state agencies,14 informal-sector unions face steeper challenges than formal-sector unions in representing and even putting together a membership base. In fact, informal-sector workers often compete with one another, a dynamic that may prevent effective collective action (see discussion in Cross :). It has been further suggested that informal-sector unions do not tend to coordinate with one another in forming higher-level structures, such as federations, although street vendors are sometimes an exception (Roever ). In recent years some of these unions have begun to participate in alternative peak labor associations, as is the case in Argentina, but they nevertheless remain largely marginal players in the representation of labor issues. Although informal-sector unions are generally considered associations rather than labor unions, they are included in the analysis for two reasons. First, leaders of such popular associations who responded to the survey identified them as labor unions and thus perceived them as organizations that fall into the union category rather than the associational category. Second, informal-sector unions are institutionally closer to the labor movement than to cooperatives or other labor-related organizations, generally understood as associations. Associations have developed linkages to these three types of unions but in 14. For example, in Peru and Mexico, street vendors have been particularly influential (Cross ; Roever ). In Peru, they got leaders elected to the city council in reaction to attempts to evict them from the streets (Cross :).
08 Chapter 8.qxd
11/9/2009
11:50 AM
Page 267
,
different proportions. Associations do not tend to work with traditional corporatist unions in large proportions. Argentine associations reported working predominantly with social movement unions belonging to the alternative peak association (Central de Trabajadores de la Argentina), which includes public-sector workers’ unions, in particular teachers and state workers. In Chile associations reported working with professors’ and health workers’ unions, and with labor confederations, such as (Central Unitaria de Trabajadores) and especially (Central Autónoma de Trabajadores), as well as with informal-sector unions. Chilean public-sector unions have adopted higher levels of mobilization and claim making in the recent years, in particular among health workers and teachers (Frank a:–), which place them closer to social movement unions than to traditional corporatist unions. Peruvian associations reported linkages predominantly with informal-sector unions and to a much lesser extent with public-sector workers’ unions. In sectoral terms, public-sector unions (teachers, health workers, and state workers) are dominant among the formal-sector unions with which associations reported linkages. The comparison of these linkages across countries does not include Venezuela, because there were so few observations in the data set.15 Distinctiveness of Union-Linked Associations Scaling. Scaling is a fundamental aspect of an interest regime, as it captures the extent to which organizations coordinate activities among themselves, and whether they form higher-level structures explicitly devoted to coordination. Labor unions in the -Hub had a high degree of scaling, achieved through a formal, hierarchical organizational structure with a peak national labor confederation at the apex. A large body of literature suggests that popular associations often face difficulty in scaling (Oxhorn a; Roberts and Portes ). While some studies suggest that associations tend to coordinate in fluid, associational networks (Chalmers, Martin, and Piester ; Korzeniewicz and Smith, ) others have suggested the atomization of associations (Kurtz a). A key question thus concerns whether associations linked to unions exhibit higher levels of scaling. To explore this question, different dimensions of scaling are examined. A first aspect concerns the extent to which associations participate in federated
15. Very few Venezuelan associations (only twenty-one) reported union ties of any type. For about percent of these it is unclear what types of unions they work with.
08 Chapter 8.qxd
11/9/2009
.
11:50 AM
Page 268
-
Table . Participation in Federated Entities by Union Linkage (percent) Participation in Federated Entities Linked to Unions Not Linked to Unions
Argentina
Chile
Peru
75.9** 55.9
66.7* 47.7
51.6 56.0
*Difference (linked versus not linked) statistically significant, p < .. **Difference (linked vs. not linked) statistically significant, p < ..
entities, such as federations and redes, which explicitly seek to coordinate or represent other associations as their primary function. In principle, such federated entities can be particularly important for expanding problem-solving activity geographically, reaching other groups, and reaching state agencies beyond the local level (see Evans :; Fox :). One should expect union-linked associations to exhibit a differential pattern in this respect. As shown in table ., associations linked to unions join federated entities at higher rates than do associations unrelated with unions in the cases of Argentina and Chile, while no significant difference is apparent in the Peruvian data. It should be noted that for the purpose of assessing cross-national differences, this section uses the standard test of the significance of a difference in proportions, despite the assumption, in this test, of a randomly selected sample. Given the chain-referral data presented here, we thus use this test as a “thought experiment.” The test simply provides a benchmark that indicates significance if these data had been selected through random sampling. A second aspect of scaling concerns the extent to which associations are connected or coordinate with other popular associations. That question includes two components, whether associations link with (a) the same type of associations (such as neighborhood associations working with other neighborhood associations, communal kitchens with communal kitchens, etc.), and whether they link with (b) different types associations (e.g., neighborhood associations linked with day-care or women’s groups). Linkages with different types of associations can be considered a form “bridging” across different popularsector interests, a strategy that may be important for building encompassing coalitions. One would expect that associations that coordinate with unions would be more likely to work with a larger variety of associations and hence be part of more diverse networks more generally.16 16. Associations working with unions obviously interact with other associations, namely unions. The question is whether the networks they form are especially more diverse or not.
08 Chapter 8.qxd
11/9/2009
11:50 AM
Page 269
,
Table . Inter-Associational Ties by Union Linkage (percent) With Similar Type of Associations Union-Linked Argentina Chile Peru
92.6** 94.4** 87.1
Not Union-Linked 78.5 63.3 79.6
With Different Types of Associations Union-Linked 90.7** 72.2 77.4
Not Union-Linked 72.0 59.3 64.6
**Difference (linked versus not linked) statistically significant, p < ..
As shown in table ., associations that are linked to unions have more ties to both similar and diverse associations compared to those unrelated to unions. A cautious approach to interpretation would suggest confidence in these differences in the case of Argentina for associations working with different types of associations, and in the case of reported ties with similar associations in Chile. By contrast, in Peru, propensity to work with similar and different associations is not significantly related with union linkage. Claim-Making Strategies. Claim making can take the form of both institutional (accessing the state) and contentious strategies (protest).17 Unions have a long tradition of employing contentious strategies, such as protests, strikes, and demonstrations for pursuing worker interests and voicing demands.18 At the same time, as argued in chapter , under the -Hub they stood out relative to other popular organizations in their capacity to advance their interests through institutional channels, negotiating with employers and the state. Popular associations, by contrast, are often assumed to face difficulty in organizing protest activity and in pressing demands on the state, especially at the national level, which is often a target for labor union demands.19 The analysis first turns to an examination of protest. Table . shows that union-linked associations organize protests at significantly higher rates than associations without such linkages in Argentina and Chile. In Peru, by contrast, 17. For a discussion of institutionalized and contentious politics in the social movement literature, see Goldstone () and Tilly (). 18. Associations were asked to respond whether helping organize protest activity was “very important,” “important,” “not that important,” or an activity they “never” did. As I am interested in whether organizations carried out protest activity or not rather than how relevant that activity was for the respondent, I built a dummy variable and coded the first three options and “never,” . 19. See comparative analysis of collective action in Latin America after market reforms in Kurtz (a), which suggests low levels of protest. For an alternative analysis, which shows the opposite argument, see Arce and Bellinger ().
08 Chapter 8.qxd
11/9/2009
.
11:50 AM
Page 270
-
Table . Protest by Union Linkage (percent) Protest Argentina Union-Linked Not Union-Linked
75.9** 28.0
Chile 61.1** 26.7
Peru 51.6 45.0
**Difference (linked versus not linked) statistically significant, p < ..
union-linked associations are not particularly different in their use of contentious strategies. Aside from within-country differences, the propensity of union-linked associations to organize protest is particularly large in the Argentine case compared to both Chile and Peru. For Argentina and Chile, one can further ask if union-linked associations are simply more likely to protest, or if they also actually tend to carry out protest jointly with unions. The latter is indeed prevalent in Argentina and Chile; respectively, and percent of union-linked associations that engage in protest also reported organizing protests with unions. However, protest is not an unusually frequent form of cooperation compared to other strategies that union-linked associations organize jointly with unions: for example, percent in Argentina and percent in Chile reported making demands on the state jointly with labor unions.20 For union-linked associations, joint action is clearly important across different claim-making strategies. Are union-linked associations also more likely to engage in claim making through institutional channels? Accessing the state is a crucial element of an interest regime, as it provides opportunities for representing interests, exerting leverage, and bargaining on behalf of constituents. As suggested in chapter , unions under the -Hub were in many ways better situated to engage in accessing strategies, particularly at the national level, than are popular associations in the A-Net. The question thus arises if union-linked associations are distinctive in the degree to which they engage in accessing strategies. For present purposes, accessing can be examined through (a) contacts with state agencies at any level; (b) contacts with national-level agencies; and (c) participation in statesponsored fora and/or policy-making councils. Contacts with national agencies are crucial for bringing the demands and concerns of popular associations beyond the neighborhood level, the site where associations tend to do most 20. Associations that reported working with unions were asked a battery of questions about the activities associations and unions did together, including “Do you place joint demands on the state?”
08 Chapter 8.qxd
11/9/2009
11:50 AM
Page 271
,
Table . Accessing the State by Union Linkage (percent) Contact State Agencies At Any Level UnionLinked Argentina Chile Peru
79.6 97.2** 58.1
Not UnionLinked 76.9 71.5 56.0
At National Level UnionLinked 58.1* 47.2** 32.3*
Not UnionLinked 32.9 17.5 14.4
Participate in State Fora and/or Councils UnionLinked 42.6 69.4** 61.3**
Not UnionLinked 36.6 35.5 36.8
*Difference (linked versus not linked) statistically significant, p < .. **Difference (linked versus not linked) statistically significant, p < ..
of their activity. At the same time, participation in policy councils and fora entails a regular institutionalized channel for expressing voice in policy making, usually including deliberation, agenda setting and/or oversight (Schattan Cohelo and Nobre ). As presented in table ., union-linked associations are more likely to engage in accessing strategies in all three countries, but not always in the same ways. The differences between union-linked associations and others are greatest and most consistent in Chile, where union-linked associations are more likely to contact state agencies in general, and particularly to do so at the national level and to participate in state councils. In both Argentina and Peru the difference in union linkages is particularly marked for contacting the state at the national level,21 and in Peru it is also strong for participating in policy councils at higher rates than associations not working with unions. To some extent, and in different ways, then, union-linked associations have potentially greater chances to voice their concerns in policy making than associations without such linkages. Use of State Resources. A final issue concerning union linkages is the use of state resources. Unions in the -Hub tended to be well funded through membership dues and sometimes administration of social benefits. Associations, by contrast, tend to find fundraising a tougher activity in general. The use of state resources may be a particularly relevant tool for financing their problem-solving activity. As presented in table ., the use of state resources varies dramatically across countries, with a larger number of Argentine associations 21. Again, a difference with chapter should be noted, in that the analysis in the present chapter is based on the dichotomy of whether or not the association contacts state agencies regardless of how important its leaders consider it to be.
08 Chapter 8.qxd
11/9/2009
.
11:50 AM
Page 272
-
Table . State Resources by Union Linkage (percent) Provides Goods, Services or Information Through State Program
Argentina Chile Peru
State Funding (Program and/or Organizational Funding)
Union-Linked
Not Union-Linked
Union-Linked
Not Union-Linked
57.4 22.2 9.7
52.7 21.5 9.6
63.0 50.0 19.4
60.2 40.7 26.8
receiving state funding than associations in Chile and more especially in Peru, and with a larger portion of associations that receive funds from state programs in Argentina than in Chile and Peru.22 What is noteworthy for present purposes, however, is that union linkage makes little if any difference in the use of state resources. Embeddedness of Union-Linked Associations To some extent, then, union-linked associations are characterized by distinctive problem-solving strategies, in that, depending on the country, they are more likely to participate in coordinating structures or federated entities, to protest, and to present claims at the national level. The question thus arises of how relevant these distinctive profiles might be for the more general operation of the A-Net. We have seen that few associations are linked to unions (about percent of the sample); yet a key question is whether union-linked associations occupy central positions or whether they are marginal players in the A-Net. One way to explore this is to analyze the level of embeddedness of union-linked associations, hereby understood as the number of inter-associational linkages reported by associations.23 Embeddedness offers an important way to assess an association’s reach and the extent to which it can potentially engage in joint 22. Social programs mentioned by associations in Argentina principally include national schemes such as Jefas y Jefes de Hogar Desocupados, an income support program for the unemployed that requires beneficiaries to perform a workfare obligation in associations or state agencies, and Fopar, which involves support for soup kitchens. Both programs were created in . For an analysis of social policy in Argentina and Chile, see Garay (n.d.). 23. The concept of embeddedness and the measure of inter-associational ties is also used in chapter , in examining scaling more generally within the A-Net. Since the focus here is specifically on the centrality of a certain group of associations within networks, this chapter distinguishes embeddedness from other aspects of scaling or coordination.
08 Chapter 8.qxd
11/9/2009
11:50 AM
Page 273
,
Table . Embeddedness by Union Linkage (number of ties) Linked to Unions
Not Linked to Unions
All
Argentina Mean Median
76.1 9
14.2 6
28.1 6
Chile Mean Median
46.3 10
13.6 3
19.1 4
Peru Mean Median
46.6 17
24.1 7
27.0 8
action with or on behalf of other organizations.24 As shown in table . unionlinked associations across all three countries reported a larger number of mean and median ties than do associations not working with unions. This higher level of embeddedness suggests a more central or nodal representational role of union-linked associations within the A-Net, with the potential implication that their strategic choices may carry greater weight within the A-Net more generally. In sum, associations linked to unions display a somewhat distinct profile— including coordination with federated entities, organizing protest, and institutionalized access at the national level—when compared to associations without union links. Those differences are more relevant in Argentina and Chile, and much less pronounced in Peru, where union-linked associations show some distinct accessing strategies, but are broadly speaking similar to other associations in terms of organizing protest and scaling. Associations linked to unions, however, do not differ from those lacking such ties with respect to the use of state resources. Finally, although union-linked associations are numerically few in the A-Net, higher levels of embeddedness imply that union-linked associations can be important vehicles of interest representation reaching relatively large numbers of associations and potentially advancing their interests. Given their ability to access the national state and organize protests, they may be central actors in pressing the interests of organized popular-sector constituencies in national policy and in mobilizing other associations for contention. 24. Analysts of social networks have developed different concepts and measures of embeddedness (Diani a; Gould ) and have used them to explain different collective outcomes, such as social movement strategies (Ansell ) and forms of social movement interactions with the state (Diani b).
08 Chapter 8.qxd
11/9/2009
.
11:50 AM
Page 274
-
In that sense, union-linked associations in Argentina and Chile may be important mobilizing structures in the A-Net, and important bridges articulating common interests and activities across the associational and the union world. Thus, an analysis of contention and interest intermediation in these two country cases should pay particular attention to union-linked associations.
Popular Associations and Political Parties What kind of relationships do popular associations in the A-Net establish with political parties? And how embedded are associations with party linkages within the A-Net? Answering these questions is important for understanding whether political parties are allies for associations, how associations interact with them in solving problems, and how those party linkages affect the representation of popular-sector interests in the A-Net. The -Hub was characterized by labor-based parties, which had organic relationships with unions: they included union leaders in party offices and slates, sometimes participated in union formation, and shared a common ideology or at least some level of programmatic cooperation. To different degrees, parties channeled union demands regarding national policy and provided unions with some resources. Unions, in turn, provided labor-based parties a fundamental linkage to popularsector voters, policy-making support, and mobilization capacity. With respect to party-association linkages, an extensive literature has settled on negative portrayals. Existing studies suggest that Latin American parties generally either prefer short-term, intermittent, and sometimes patronage-based relationships with associations, or neglect them altogether, reaching lower-class individuals in unmediated ways.25 The first part of this section briefly looks at party-association linkages, focusing on party strategy toward associations and specifically exploring whether parties participate in the formation, funding, and electoral mobilization of popular associations. The second part explores party-association linkages from the opposite perspective by focusing on associations’ strategies toward parties. It examines the kinds of linkages that associations establish with parties to advance their interests, builds a typology of party-association linkages, and analyzes its implications for the A-Net. 25. See Oxhorn (), C. Schneider (), Roberts (; ; forthcoming), Weyland (a), and Levitsky ().
08 Chapter 8.qxd
11/9/2009
11:50 AM
Page 275
,
Party Strategy Toward Associations Studies of the subject typically assert that political parties in Latin America have tenuous relationships with popular associations, generally viewing them only as potential vehicles for electoral mobilization or neglecting them altogether. When parties do establish linkages with associations, these linkages tend to be short-term and clientelistic, and just a few parties engage in supporting popular associations more consistently in their problem-solving strategies. Data from the surveys of associations corroborates these arguments about party strategies. As presented in table . the share of associations that were contacted by political candidates soliciting campaign support is notable, especially when compared to the percentage of associations that reported parties had been involved in their formation and in contributing financial support. Indeed, about a third of the associations surveyed reported contacts by candidates soliciting associations’ support, while very few associations reported that parties were involved in their formation (. percent), and fewer reported that parties provide funding (. percent). Cross-national differences in party strategy toward associations are large. At the two extremes, candidate contacts are particularly prevalent in Peru, where half of the associations in the survey reported such contacts, and relatively rare in Venezuela, where such contacts were reported by only percent of the associations at the time of the survey. Chile and Argentina stand in the middle, with about percent of the associations reporting contacts by candidates. These figures contrast sharply with the proportion of associations that reported that political parties participated in their formation and those receiving funding from parties, which is highest in Venezuela and lowest in Peru, although differences are not large. Associations reported that candidates frequently offered something in exchange for associational cooperation, indicating that politicians tend to pursue clientelistic exchanges during campaigns.26 About percent of the associations contacted by parties were offered some type of support, out of which percent were promised pork-barrel benefits for the community, such as public works and infrastructure, or excludable benefits for associations, such as donations. A smaller proportion were offered help in contacting public authorities (. percent), or the promise of passing specific legislation ( percent).27 26. There is a rich literature addressing clientelism in Latin America. See for example, Stokes (), Chalmers, Martin, and Piester (:), Auyero (), Levitsky (), and Kitschelt and Willkinson (). 27. Unspecified support was offered to percent.
08 Chapter 8.qxd
11/9/2009
11:50 AM
.
Page 276
-
Table . Political Party Strategy Toward Associations (percent of associations)
Total Associations Argentina Chile Peru Venezuela Total
Parties Participated in Formation
Parties Provide Funding
Candidates Contacted for Support
4.2 5.8 2.5 12.1 6.1
1.7 1.4 .4 9.2 3.2
35.4 31.7 51.3 15.5 33.4
(n=240) (n=208) (n=240) (n=240) (n=928)
Candidates soliciting campaign support displayed particular traits. In all countries, they belonged to parties that spread across the ideological spectrum, including old and new parties as well as right-wing, labor-based, and left-wing parties.28 Further, most of these candidates soliciting support were running in local elections. Over percent of the associations contacted were asked to support local-level candidates (for mayoral or district-level elections), while about percent of associations were asked to support candidates running for national office (for president or for congress). Local politicians, more than those running for national-level positions or governor, appear to be more prone to seek campaign support from associations. This difference may orient partyassociation exchanges toward local-level political agendas. In short, party strategy toward associations centers on contacts for support during campaigns, and such contacts are extended by parties in exchange for benefits that generally include promises of state infrastructure investments, and excludable benefits offered by parties to associations. Local-level politicians seem to be more involved in such requests than candidates running for national or gubernatorial office, and overall, these practices are pursued by diverse kinds of parties. Associational Linkages to Parties If parties seem primarily interested in electoral exchanges when interacting with associations, how do associations relate to parties seeking to advance their 28. Among others these include, in Argentina the Partido Justicialista () (. percent), (. percent), and Unión Cívica Radical () (. percent); in Chile the parties that form the Concertación ( percent) and those on the right, Unión Demócratica Independiente () and Renovación Nacional () ( percent); in Peru ( percent), Somos Perú ( percent), Cambio (. percent), and Perú Posible ( percent); in Venezuela Movimiento Quinta República () and Patria Para Todos () (. percent), Acción Democrática () ( percent), and (. percent).
08 Chapter 8.qxd
11/9/2009
11:50 AM
Page 277
,
interests? Do they respond to candidates and support them in elections? Do they form more stable relations with parties, engaging jointly in a combination of activities? Do they contact party leaders to solve their problems? The following analysis lays out three types of linkages that associations may pursue with parties to advance their interests and solve problems affecting their communities (see table .). The first linkage, which I call issue-based, is one in which associations contact parties to advance a cause or gain leverage regarding a particular problem. For example, some associations working on may contact party leaders to lobby and educate them on issues and transmit their agenda. These contacts are in principle intermittent and ad hoc, as associations direct messages to the party to solve or garner support for a particular problem or issue of concern. Further, they are also instrumental, as they pertain to an agenda on a particular issue, but not a larger commitment on the part of the association to the party’s ideology, more general programmatic orientation, or electoral fate. A second type of linkage emerges when associations support specific candidates during campaigns by pursuing activities such as mobilizing voters, bringing people to rallies, and promoting the candidate in their respective communities. That type of linkage between associations and political parties can be typified as candidate-based. It also consists of an instrumental tie based on the endorsement of a particular candidate, often in the hopes of obtaining either particularistic or public benefits. Finally, the most complex type of linkage, which I call multifaceted, is characterized by regular and broad interactions between associations and political parties. Multifaceted linkages involve associations collaborating with parties on a variety of issues, such as developing joint community projects, training community leaders, engaging in joint claim-making activities, and supporting candidates during campaigns. Especially when such linkages further involve the participation of party militants in the founding of the association, in contributing to the association’s budget as well as associations supporting party candidates in elections, such relationships may be organic, that is, involving shared members, a common ideology or programmatic orientation, and concern about the party’s electoral fate. Scoring party-association linkages is not simple. Aside from the challenges involved in typologizing the content of particular linkages, relationships between associations and political parties may change over time. For example, associations with multifaceted linkages to parties may narrow their interactions to only electoral concerns and specific exchanges. Likewise, associations created by party militants may become disengaged from those parties and
08 Chapter 8.qxd
11/9/2009
.
11:50 AM
Page 278
-
Table . Associational Linkages to Political Parties Issue-Based
Candidate-Based
Multifaceted
Associations contact party leaders for problem solving
Associations support candidates
Associations and parties work jointly in different domains
Goal/Arena of Interaction
Interest
Electoral
Interest and Electoral
Characteristics of Linkage
Instrumental, occasional
Instrumental, occasional
Complex, stable
Subject of Linkage
from electoral concerns. With those caveats in mind, it was possible to create measures of linkages between associations and parties based on responses to three questions concerning whether associations contact parties to solve problems, have supported candidates, and, more generally, collaborate with parties.29 Those associations that reported only contacting party leaders to solve problems and that did not report collaborating with parties or supporting candidates are considered to be characterized by issue-based linkages. Associations that reported supporting candidates but did not report working with a party are considered to have candidate-based linkages. Associations with candidate-based linkages may also contact parties for problem solving, but the distinctive feature is their support for particular candidates. It seems likely that such support was provided in exchange for programmatic issues, for interestgroup concerns, or for excludable benefits, but the data do not provide accurate motivations for such support. Associations that reported collaborating with political parties are considered to have a multifaceted linkage. In a pooled analysis, about a quarter of the associations sampled reported issue-based, candidate-based, or multifaceted linkages with political parties ( percent). The share of associations with party linkages is remarkably similar across countries, ranging from . in Peru and . in Chile, to . and . percent in Venezuela and Argentina respectively. Table . shows that the most prevalent type is issue-based, which represents percent of all reported linkages, although these are not that much more prevalent than other linkages. 29. The three survey questions are the following: () Does your organization work with political parties? followed by What party? and by a battery of close-ended questions about different activities associations may pursue with parties, such as protest, placing demands on the state, sharing information, joint community projects, training, and receiving goods from political parties. () Has your organization supported a political candidate in the past five years? Respondents were further asked to identify the party of the candidate and what position she or he was running for. () Do you contact party leaders for problem solving?
08 Chapter 8.qxd
11/9/2009
11:50 AM
Page 279
,
Table . Distribution of Party-Association Linkages (percent)
IssueBased
CandidateBased
Multifaceted
Total Associations with Party Linkages
41.3 (100)
29.1 (70)
29.7 (72)
100 (242)
Candidate-based and multifaceted linkages each represent about percent of total linkages. The pooled data mask some country variation regarding the types of linkages that prevail (see table .). Issue-based linkages are more frequent among associations in Argentina and Chile, and they are less prevalent in Peru and Venezuela. In Peru candidate-based linkages are most common, whereas in Venezuela multifaceted linkages predominate. Multifaceted linkages are particularly infrequent in Chile and comparatively more frequent in Argentina and Venezuela. Associations with multifaceted linkages reported collaborating with parties on multiple issues: exchanging ideas or information with parties ( percent), placing joint demands on the state ( percent), carrying out community projects together ( percent), and, to a lesser extent, protesting with political parties ( percent). Over half ( percent) of the associations with multifaceted linkages further reported that they had supported candidates in elections. In each capital city, multifaceted linkages are concentrated rather than dispersed across multiple parties.30 In Argentina, such linkages exist primarily with the Peronist Party (), the traditional labor-based party.31 By contrast, Venezuelan associations have developed multifaceted linkages with new parties, primarily with President Hugo Chávez’s Movimiento Quinta República ().32 As mentioned, these linkages are less frequent in Peru and, particularly, in Chile. Candidate-based linkages, which are infrequent with the exception of the Peruvian case, develop around local-level candidates at higher rates in Argentina and Chile, whereas in Peru and Venezuela associations reported 30. Data collected do not allow me to establish which parties are involved in issue-based linkages. 31. For a discussion of labor mobilizing and elite parties, see Roberts (). In Argentina, associations tend to have multifaceted linkages with the . Of the associations that work with parties, percent reported to work with the , percent with the , and percent with other smaller parties. 32. Information about the party/parties associations work with is available for percent of the linkages. About percent reported working with or with and allied (at the time) with , percent with , and percent only with .
08 Chapter 8.qxd
11/9/2009
.
11:50 AM
Page 280
-
Table . Party-Association Linkages by Country (percent)
Total Associations Argentina Chile Peru Venezuela
(n=69) (n=51) (n=56) (n=66)
Multifaceted
Candidatebased
Issue– Based
Total PartyAssoc. Linkages
36.2 11.7 23.2 42.4
14.5 17.6 57.1 28.8
49.2 70.5 19.6 28.8
100 100 100 100
supporting candidates running for national office—president and congress— at significantly higher rates.33 In the latter cases, candidates from new parties, such as Cambio /Nueva Mayoría (former President Alberto Fujimori’s parties) in Peru and in Venezuela reaped the most support.34 Embeddedness of Party-Linked Associations As with union linkages, we can explore how these different types of linkages matter for the A-Net more generally. An analysis of embeddedness, again understood as the number of inter-associational linkages, shows that associations with issue-based linkages and to a lesser extent those with multifaceted linkages are significantly more connected than associations with candidate-based linkages or without party links. Associations with candidate-based linkages, which probably serve as vehicles of electoral support, appear to be less embedded in associational networks, even compared to associations without party linkages (see table .). Two points might be suggested. First, the activity pursued through issuebased and multifaceted linkages may be relevant for a larger segment of associations in the A-Net, as it is carried out by associations that exhibit high levels of embeddedness. By contrast, associations with candidate-based linkages are the least involved with other associations, a fact that would seem to circumscribe the spread of their strategic choices and interactions with political parties 33. In Argentina and Chile associations with candidate-based linkages reported supporting locallevel candidates at higher rates, percent and . percent, than national-level candidates, percent and percent respectively. By contrast, percent of the associations in Peru and . percent in Venezuela reported supporting national-level candidates, and only . percent in Peru and . percent in Venezuela supported local-level candidates. 34. Of the associations with candidate-based linkages, . percent reported supporting in Venezuela and percent Cambio /Nueva Mayoría in Peru. Caution, however, should be employed when analyzing these data.
08 Chapter 8.qxd
11/9/2009
11:50 AM
Page 281
,
Table . Embeddedness of Associations by Party Linkage IssueBased (100) Mean Ties Median Ties
54.3 7
CandidateBased Multifaceted (70) (72) 14.1 4
32.4 6
Without Party Linkage (686) 21.0 5
within the A-Net. Second, each type of party linkage may affect associations in different ways. Whereas associations that pursue issue-based and multifaceted linkages also tend to pursue linkages with other associations in the A-Net, associations with candidate-based linkages are relatively cut off from the associational world, perhaps engaging in more particularistic and go-it-alone strategies, without expanding their ties across diverse associational networks. This point is relevant because parties have sometimes been found to disarticulate associational networks or generate divisiveness in the associational world when linking with associations (see Oxhorn ). The effects of party linkages on associations thus seem to depend on the types of linkages that are established. Candidate-based linkages are associated with lower levels of embeddedness and more atomized associations. Associations with candidate-based linkages seem further isolated with respect to other types of linkages, such as union linkages. While about percent of associations linked with political parties are also linked to labor unions, union linkages vary by type of party linkage. Only percent of associations with candidate-based linkages are linked to unions, while about percent of associations with multifaceted and percent with issue-based linkages are connected to unions. This further shows that associations with multifaceted and issue-based linkages are more likely to be engaged in more diverse networks than associations with candidate-based linkages. In sum, attempts at establishing clientelistic exchanges with associations are an important aspect of party strategy toward the associational world. Candidates running for local-level elections appear to be particularly invested in such attempts, which involve soliciting occasional campaign support from associations, with excludable rather than programmatic goods offered in exchange for cooperation. However, this strategy by parties may not be widely successful. Despite some differences across countries, it is important to emphasize that associations’ strategies toward parties were not predominantly centered on electoral issues. Associations with issue-based linkages, which contact parties for problem solving, and/or associations with multifaceted, complex
08 Chapter 8.qxd
11/9/2009
.
11:50 AM
Page 282
-
linkages with parties, were more prevalent than associations with candidatebased links in three of the four countries under study. That associations do not succumb to parties’ requests for support may indicate that they have other sources of power or influence or that they perceive better prospects in terms of problem solving by establishing other types of linkages with parties, or even establishing none at all. The data thus suggest that a view of the associational world as vulnerable to clientelism and easily manipulated by a candidate or party strategy may be overstated.
Cross-National Comparisons The chapter has revealed variation across countries in associational linkages to unions and parties, both in terms of the degree to which union-linked associations are distinctive and in terms of the predominant types of linkages that associations form with parties. This section explores some of the factors that potentially account for this variation and then lays out overall country patterns. The Distinctiveness of Union-Linked Associations: Argentina and Chile Versus Peru and Venezuela Compared to other popular associations, the problem-solving repertoires of union-linked associations are distinctive in Argentina and Chile, but not in Peru and Venezuela. In the former two countries union-linked associations are more active in scaling (especially participating in federated entities such as federations or redes), in organizing protest, and in accessing the state, particularly at the national level, compared to other popular-sector associations. The fact that union-linked associations also tend to be significantly more embedded in broader associational networks further indicates that their distinctive pattern of activity may be particularly important for the operation of the A-Net more broadly. By contrast, no distinctive pattern of activity characterizes union-linked associations in Peru, except for some tendency toward more frequent state contacting at the national level and participation in policy councils. Any differences in Venezuela are less important, since few associations have union linkages (a fact that precluded the inclusion of Venezuela in this part of the analysis because of insufficient observations). The distinctive pattern of activity of union-linked associations in Argentina and Chile arises in a particular union context with more mobilized publicsector unions and social movement unions, which have been available for
08 Chapter 8.qxd
11/9/2009
11:50 AM
Page 283
,
broader popular-sector alliances. The organization of some of those unions, moreover, takes the institutional form of a national labor central, which serves a broad coordinating function. In Peru and Venezuela no such pattern of active public-sector unions and social movement unions has emerged. In Peru, most associations reported contacts with informal-sector unions, and in Venezuela the few union-linked associations had a higher tendency to be linked to traditional corporatist unions. These two types of unions, informal-sector and traditional-corporatist unions, seem to be less conducive to the pattern of problem solving observed in Argentina and Chile. Thus, the different developments within the union movement in each country seem to affect union-association linkages and the strategies of union-linked associations. Specifically, it seems that ties to social movement unions and/or public-sector unions, in particular teachers’ unions, are more likely to be associated with a pattern of problem solving involving contestation and coordination than ties to informal-sector unions or to traditional corporatist ones, especially industrial unions. This argument is consistent with the literature on labor associations that distinguishes the mobilization strategies of traditional corporatist unions from those that have formed an alternative social movement unionism (Seidman ; Bensusán ), as well as public-sector unions (Frank b; Murillo and Ronconi ). Public-service workers’ unions, in particular those of teachers, became the most contentious labor unions in countries such as Mexico (Cook ) and Argentina in the s (Murillo and Ronconi ). Recent studies have further found public-sector unions to be particularly prone to engaging in institutionalized demand making and in contention in the case of Chile (Frank b:–). Indeed, public-sector unions have had a greater propensity to engage in contention than industrial unions, as they are relatively less exposed to competition and protected by job stability clauses.35 Further, some of these unions have left-leaning leadership, with a broader, oppositionist orientation. In the early s some of these unions formed an alternative central or peak association, such as in Argentina, which is frequently led by teachers or public-sector workers’ unions, and in Chile, which is relatively smaller and led by unions in the education sector. To a greater or lesser extent these labor centrals began to link up with community organizations, and in turn, studies of specific social movements indicate that popular associations have found social movement unions particularly good coalition partners for contestation and protest. A case in point is the unemployed movement 35. See discussion in Murillo and Ronconi (). The authors find causes other than lack of competition and job stability to explain variation in teachers’ militancy in Argentina.
08 Chapter 8.qxd
11/9/2009
.
11:50 AM
Page 284
-
in Argentina, which comprises community associations and has forged important ties with social movement unions, especially those in the (Svampa and Pereyra ; Garay ). Cross-National Variation in Types of Party-Association Linkage With respect to party-association linkages, across the national cases there is remarkable stability in the number of associations with party linkages, which constitute about percent of the associations sampled in each case. Yet, countries differ in terms of what type of linkage predominates. Both Argentina and Chile show a predominance of issue-based linkages, although that is more pronounced in Chile than in Argentina, where fewer associations have such linkages and where multifaceted linkages are also important. In both countries, issue-based linkages are formed at higher rates than in Peru and Venezuela. Moreover, associations with issue-based linkages to parties are more embedded than those with other types of party linkage, so the activities of associations with these linkages may be particularly influential within the A-Net as a whole. Peru is characterized by the dominance of candidate-based linkages, which are much more prevalent than in the other countries. At the same time, Peruvian associations establish very few issue-based linkages. Candidate-based linkages are generally formed by associations with lower levels of embeddedness. Thus, candidate-based linkages are probably not a key strategy of problem solving that diffuses through inter-associational linkages in the broader A-Net. Finally, in Venezuela, multifaceted linkages predominate compared to other types, but the percentage of associations with such linkages, while higher than in Chile and Peru, is similar to that in Argentina, which has an even higher level of issue-based links. Multifaceted linkages tend to be developed by associations with relatively higher levels of embeddedness, which may thus play some role in the operation of the A-Net more broadly. In accounting for the presence of these diverse party-association linkages, a number of factors may be suggested. The presence of issue-based linkages appears to be related to the institutionalization of the party system. More stable party systems seem to facilitate regular patterns of interaction between associations and political parties. If issue-based linkages require some level of predictability and regularity in the interaction between associations and parties, such contacts can more easily develop in the context of stable party structures. More stable party systems are characterized by greater continuity in party organization and leadership with which associations may establish issue-based
08 Chapter 8.qxd
11/9/2009
11:50 AM
Page 285
,
linkages, even through informal ties. These relations may be difficult to form if parties are disrupted and new party organizations fail to consolidate following party system collapse. Among our cases, Chile and Argentina show higher levels of party system institutionalization than Peru and Venezuela, both of which experienced party system collapse in the s.36 Indeed, Chile has longstanding parties that reemerged after the democratic transition as key political actors. Likewise, Argentina has higher levels of party stability, even after the recent dismemberment of the Radical Party, which largely occurred after the financial and political crisis.37 The development of candidate-based linkages appears to be related to the strategy of the party leadership and associational responses to that strategy. When party leadership is not involved in organizational building and instead seeks instrumental support from associations, then the likelihood of finding candidate-based linkages is higher. In Peru party system collapse, which discouraged issue-based linkages, and a party strategy of extending clientelistic interactions with associations may help explain why in that country candidatebased linkages predominate vis-à-vis other forms of party linkage. Regarding the responses to clientelistic party initiatives, recent studies of Lima show that associations appear to adopt a pragmatic strategy of supporting parties that tend to their needs and withdrawing support from parties that do not deliver on their promises (Schönwälder :). The absence of stronger allies, such as social movement unions, may also generate further incentives for associations in Peru to engage in candidate-based linkages. It should be further noted that Peru has a history of top-down attempts at mobilizing popular-sector associations for political projects (D. Collier ; Stokes ). Such supportmobilization projects did not lend themselves to the development of multifaceted linkages in the past, and the more recent attempts of interaction with associations launched by the United Left in the City of Lima, which sought to respect the autonomy of popular associations (Schönwälder ), vanished with the disappearance of United Left from the partisan arena in the s (Roberts ; Schönwälder ). By contrast, when party leadership is involved in developing broader linkages with associations, multifaceted linkages may be more prevalent. In the case of Venezuela, party system collapse and the emphasis placed by the electorally dominant chavista movement in building ties with associations—and 36. Cameron and Levitsky (), Burgess and Levitsky (), Roberts (), and Seawright (). 37. On the crisis, see Levitsky and Murillo () and Woodruff ().
08 Chapter 8.qxd
11/9/2009
.
11:50 AM
Page 286
-
even fostering their formation—may help explain lower levels of issue-based and candidate-based linkages, and the prevalence of multifaceted linkages. In Argentina a larger share of multifaceted linkages than in Chile may reflect the strategy, incipient at the time of the surveys, of parties seeking to extend linkages to the organizations of the unemployed and some community-based movements. By contrast, the major Chilean parties are more oriented toward unmediated linkages to society and this more general orientation toward circumventing organized actors may help explain the low levels of multifaceted linkages in that case. Despite the importance of party strategy, we should not overlook the role of associations themselves in shaping their ties to political parties. Oxhorn () has shown that during the democratic transition in Chile, some associations, fearful that political parties would generate divisiveness in the associational world, intentionally chose a more distant relationship with them. These associations were also concerned about the potentially negative effects of party ties, like the high levels of divisiveness in the context of polarizing partisan competition in the late s and early s.38 Associations in all cases do not readily succumb to party requests for campaign support either. As we saw above, although a large percent of associations received requests for support extended by parties often in exchange for some material benefit, only a few reported— or admitted to—embarking on such exchanges. Overall Country Profiles Three country patterns can be discerned regarding associational linkages to unions and parties. Argentina and Chile display a similar profile. More associations have linkages with public-sector or social movement unions, and undertake more mobilizational patterns of problem solving. In addition, there is a greater tendency of associations to form issue-based linkages to parties. In Peru and Venezuela, by contrast, associations do not tend to link specifically to social movement unions, and those that do have union linkages do not engage in distinctive problem-solving strategies. Yet associations in these two countries differ in terms of the nature of party linkages, with ad hoc candidatebased linkages predominating in Peru and multifaceted linkages predominating in Venezuela. What do these patterns tell us about the operation of the A-Net? In Chile 38. On the relationship of political parties and associations in Chile see C. Schneider () and Hipsher ().
08 Chapter 8.qxd
11/9/2009
11:50 AM
Page 287
,
and Argentina linkages to parties and to unions may affect a significant number of associations in the A-Net given the embeddedness of union-linked associations and those with issue-based party linkages. Unions seem to be good allies for mobilization and demand making, while parties appear to be more distant, contacted for solving specific problems. In Peru, by contrast, although union-linked associations are well embedded, they do not have a distinct pattern of problem solving, and associations working with parties tend to be quite isolated in the associational world. Thus union and party linkages are not particularly relevant for the activities of a broader network of associations. In the case of Venezuela, parties appear to be a more important institutional actor than labor unions for associations’ problem-solving activity. The high embeddedness of associations with multifaceted linkages to parties may make them particularly important actors in the A-Net. The similarity between Argentina and Chile may come as a surprise for two reasons. First, based on the existing literature we would not have expected such a pattern of union-association ties in Chile. The decline of the union movement during the Pinochet dictatorship (–) and its structural weakness thereafter, coupled with the atomization that has been usually understood to be a consequence of sweeping market reforms implemented by the military (Kurtz a), would seemingly make union-association ties improbable. Even though the number of associational ties to unions is not great, the embeddedness of these associations may make their activity consequential for other associations in the A-Net. In this way, union-linked associations in Chile may operate as mobilizing structures for protest and state contacting and as bridges among disparate associations. Second, the similarity in party-association ties is also surprising given that Argentina has a marked reputation for patronagebased organizations and clientelism. This feature of Argentine politics is rooted in the local offices of the Peronist Party, called basic units (unidades básicas), which have generally been the focus of most research on clientelism in Argentina. Patterns of partisan engagement of the associational world, however, do not uniformly fit this pattern.39
Conclusion The new interest regime, the A-Net, is more diverse than the -Hub, involving a broader range of popular organizations. Understanding how different 39. On the patronage activity of basic units and militants, see in particular Auyero () and Levitsky ().
08 Chapter 8.qxd
11/9/2009
.
11:50 AM
Page 288
-
associations engage in problem solving and how relevant they are for the A-Net is an important issue for future research. Union-association and partyassociation linkages may provide good starting points for identifying different patterns of problem solving in the A-Net. With cross-national differences in incidence, linkages between unions and associations are an important feature of the A-Net. In Argentina and Chile they indicate a degree of continuity between the old and emergent interest regimes, as union-linked associations are more likely to adopt strategies of protest and national-level state contacting. Higher levels of protest and scaling on the part of union-linked associations can facilitate episodes of large-scale mobilization, such as the ones spearheaded in Argentina by the associations of the unemployed movement, which emerged largely out of community-based associations but gained further capacity though its connections with unions. National-level accessing strategies of union-linked associations can also be important for policy-making input. Though further research is needed to reveal how influential union-linked associations are, one can observe that although they are not dominant in quantitative terms, they are more embedded in networks of associations, thereby potentially having greater capacity to operate as mobilizing structures through the A-Net more generally. Finally, union-linked associations are important because they may act as bridge organizations, representing structurally different segments of the popular sectors: labor unions, which are focused on work-related concerns and typically represent the wage earner, and popular associations, which focus on a broader array of issues such as social services, consumption, and infrastructure, and represent a more variegated constituency frequently defined territorially, at the neighborhood or community level. This question is particularly interesting given that “bridging” across different segments of the popular sectors in the -Hub was generally the task of political parties in the electoral arena rather than of organizations in the interest arena.40 With respect to party-association interactions, different types of linkages result from strategies of both parties and associations. The greatest prospects for coalition formation between associations and parties occur through multifaceted and issue-based links, the former involving a broader common agenda and concern about the electoral fate of the party, and the latter involving alliances surrounding particular issues. Associations with issue-based and multifaceted 40. Exceptions to this claim are the social movement unions; see in particular Seidman () and Keck ().
08 Chapter 8.qxd
11/9/2009
11:50 AM
Page 289
,
links tend to be characterized by higher levels of embeddedness. Candidatebased links, by contrast, reveal a different strategy on the part of associations. They represent a pragmatic, probably more atomized approach to solving problems, involving the exchange of cooperation for (usually) material benefits. Associations with candidate-based linkages tend to have lower levels of interassociational linkages, and consequently have less potential impact on the activity of other associations in the A-Net. It seems likely that these associations perceive candidate-based linkages as the only viable avenue for problem solving, one that seems more promising or, to them, feasible, than a collective problem-solving strategy. Overall, the relative importance of issue-based and multifaceted linkages, even in the face of extensive efforts of parties to obtain support from associations in exchange for benefits, appears to indicate that associations do not succumb to those efforts en masse. It should be noted, however, that party-association ties are continually negotiated in a fluid ongoing process. Since the surveys were carried out, two of the four countries have witnessed important changes in party politics. In Argentina, multifaceted linkages between parties and associations may have been growing since . Indeed, President Néstor Kirchner (–) governed through a broad coalition in which organizational linkages were critical, including alliances with unemployed groups, community-based social movements, as well as traditional labor unions. Preexisting popular movements were mobilized strategically in attempts to pursue national policy goals, and some associational leaders were appointed to public office and, to a lesser extent, included in party slates. In Venezuela, the formation of popular-sector associations and their mobilization behind his project is one of President Chávez’s most salient political strategies and may entail growing levels of multifaceted linkages in the Venezuelan A-Net. Whether these ongoing changes favor the formation of broader coalitions that include associations in participatory projects, or whether they reinforce or generate fragmentation in the associational world, remains an open question. Based on the analysis here presented, multifaceted linkages do not necessarily fragment the associations involved. The new ties forged by Chávez’s and the Kirchner faction in the resemble multifaceted linkages. In both cases, particularly in Venezuela, the risk exists that the broader orientation of the ANet may be asymmetrical, primarily reflective of the top-down goals and policies of state leaders. Yet, in Argentina, the fact that associations were already connected and active before the extension of party linkages from above may result in the formation of party-association links that afford associations more
08 Chapter 8.qxd
11/9/2009
.
11:50 AM
Page 290
-
influence, with more symmetrical forms of exchange or negotiation between the associations and Kirchner’s faction in the . Nonetheless, multifaceted linkages might evolve into candidate-based linkages, with the possible effect of marginalizing some associations. The evolution of party-association linkages in these two cases is still in flux as party systems undergo significant change.
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 293
9 Samuel Handlin and Ruth Berins Collier
This study has undertaken a comparative analysis of patterns of popular-sector participation and representation in the interest arena in four Latin American countries, examining both individual activities and those of popular associations. Popular representation has been historically problematic in the region, with polities often characterized by the exclusion of popular-sector actors or limitations on their political influence. The issue is particularly crucial in the contemporary period, at a time when norms concerning good governance and societal accountability have spread and seemingly gained global priority, a context potentially more favorable for reversing these historical patterns of economic and political exclusion. In considering popular representation, this study focused on the interest arena, a particular site of representation and state-society intermediation in the polity. This is the site in which both individuals and interest organizations seek to address specific interests and problems, in contrast to the electoral arena, in which representatives are chosen but preferences only bluntly signalled. In exploring popular-sector representation in the interest arena, then, the study focused both on the participation of individuals and on the nature and activities of popular associations. To assess the ways individuals participate in the interest arena, the study examined both “direct action,” or in-person forms of claim making on state officials with regard to specific issues or policies, and participation in popular associations, which, by contrast, can provide organizational channels for sustained activity. With respect to associations, the study
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 294
.
examined them, both individually and collectively, as the popular interest regime or organizational infrastructure through which the popular sectors address collective problems and pursue collective interests. The analysis sought to explore the implications of a shift in the popular interest regime from the -Hub, characterized by the predominance and “privileged” influence of partyaffiliated labor unions, to the A-Net, characterized by the role of a more diffuse and diverse set of associations. One reason for this focus on the interest arena is the suggestion that, amid vast changes to both electoral and interest arenas, their relative importance as sites of representation may be changing, with the interest arena waxing as the electoral/partisan arena wanes. This is not to suggest that voting itself is a less important political act or that the electoral arena does not remain the crucial site for contestation over recruitment to power. Rather, the issue is that political parties themselves may be becoming less central as intermediaries and representational structures, with interest organizations potentially becoming more important. In the context of the advanced countries, analysts of both political parties and interest organizations have suggested that interest organizations are taking over this kind of main representative function (Schmitter ; Mair ). Scholars have also advanced this idea more specifically regarding Latin America, suggesting that parties are increasingly retreating from representational roles and functions “in society” (Weyland ; Roberts forthcoming; Luna ). Some empirical support for this idea was presented in chapter , which showed that confidence in associations is far greater than in parties as agents for problem solving (see figure .). Our attention to these questions arises in the context of new economic models and political decentralization in the region, which also make an exploration of the interest arena particularly pertinent. Many perspectives on these vast political economic changes have seen them as largely unidirectional, entailing a retreat of the state, decentralization, and a reduced policy space for political contestation. Yet, some scholars of advanced countries suggest a more nuanced “bifrontal” process of state reform, in which the state is retreating in some areas in which it has traditionally intervened but assuming new or expanded roles in others. For example, Levy (:) has argued that in the advanced countries the state role has not so much decreased as “migrated” from industrial policy to social and labor market policy, and that centralization of power has occurred in important policy areas even as some functions are decentralized. Similar bifrontal dynamics may hold in Latin America. While many macroeconomic policies have been removed from political contestation, social policies have been expanded dramatically under market liberalization, especially
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 295
,
programs for informal workers (Garay n.d.). And although a general trend toward political, fiscal, and administrative decentralization is evident, many policy areas do remain contested at the national level, including those pertaining to many of the “second generation” social policies that have come to constitute focal points for political contestation in the last decade.1 These more complicated bifrontal dynamics call for an analysis focused on how the state orientation invites or discourages contestation regarding social and other policies, about where these policies are contested, and about how collective actors, especially the popular sectors, coordinate to pursue their interests with regard to these policies. An exploration of popular representation in the interest arena is thus partly a question of how the shift in the popular interest regime intersects this other shift in the orientation of the state.
Popular Representation in the Interest Arena In chapter we highlighted five analytic themes for exploring participation and organization in the interest arena, and we now return to them. Individual Problem-Solving Repertoires. Popular-sector individuals adopt a wide repertoire of strategies to solve problems in the interest arena. Associational participation is quite high, roughly comparable to rates in advanced industrial countries. The incidence of direct action, including various forms of contacting bureaucratic, elected, or party officials, is also quite robust, and some types are more frequently employed than might have been expected. For instance, contrary to the widespread perception that the judicial branch of government is a largely irrelevant site of problem solving, particularly for the lower classes, popular-sector individuals engage in legal strategies to a perhaps surprising degree. Nevertheless, associational participation emerged as occupying the preeminent place within the overall problem-solving repertoire of individuals. In three out of four countries, associational participation exceeds both atomized and collective direct action as a mode of political participation and problem solving. In direct-action strategies, individuals undertake demand-making activities around specific issues, which may be personal and particularistic (as in a personal problem or request having to do with a state agency) or more collective 1. In some instances, the expansion of these programs has involved a process of centralization, as programs become too large or too politically important to be entrusted to provincial authorities. See Garay (:) for a discussion of this dynamic in the case of workfare programs in Argentina.
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 296
.
(specific issues that may range from a neighborhood problem to a preference regarding a national policy). The popular sectors engage in direct action around particularistic problems more frequently than around collective issues. Yet, collective strategies are vital for popular-sector representation, not just because of the group interests pursued, but also because individuals consistently rate them as more effective than individual strategies. While any form of direct action may be collective, petitioning and protest account for, by far, the highest incidence of collective direct action. Petitioning, an activity that has not received much analytic attention, is not primarily a middle-class activity, but is also salient among the popular sectors. More widely recognized in contemporary Latin America are dramatic episodes of contention, especially in Argentina during the economic crisis of – and in Venezuela during the regime contention of –. The surveys sought also to capture the role of protest aside from these unusual episodes and found protest to be an important, intermittent but quite “regular,” component of the larger repertoire of popular-sector strategies. Nevertheless, even more than petitioning and protest, associational participation emerged as the most widespread strategy for the pursuit of collective interests and group problems. In particular, since collective direct action to contact officials through institutional channels is so low, associations—which engage in such contacting relatively frequently—allow participants to engage in this vital strategy in mediated form. It is interesting that the use of intermediated direct action is generally very limited. In a region known for high levels of clientelism, citizens rarely seek out influential individuals or political parties to intervene on their behalf, despite the fact that it is considered an effective strategy. It does not follow from the relative infrequency of these approaches that clientelism is not an important part of Latin American politics. First, other kinds of problem solving can also be marked by clientelism. Second, what may be an important strategy for a political party may not necessarily reach a high percentage of individuals, who, given the asymmetries involved, may not be in a good position to initiate such relationships. Nevertheless, the relatively low incidence of intermediated strategies and the fact that the incidence is highest in Chile, a country widely considered to be less clientelistic than others in the region, challenge analysts to be systematic and comparative in assessing the role of clientelism in Latin American politics. Within these general trends, some intriguing cross-national variation emerged. In Venezuela, at the time of the surveys, the interest arena was one in which all forms of popular-sector participation were low relative to the other cases: not only associational participation but also both atomized and collective
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 297
,
direct action. Argentina presents the opposite picture of active participation in the interest arena. Argentines took relatively frequent opportunity to engage in both atomized and collective forms of direct action; and though their level of associational participation did not reach that in Peru, they did maintain an unusually high level of union participation, so that overall, the level of participation in interest organizations was high. In Peru and Chile the forms of participation were more differentiated. In Peru atomized direct action was low, suggesting that Peruvians were unlikely to attempt to address particularlistic problems through the interest arena. On the other hand, they participated in associations at notably high rates and also frequently engaged in collective direct action, especially protest. Chile presents the opposite picture. Individuals were relatively prone to address particularistic problems through atomized direct action, but made the least use of collective direct action, although they particpated in associations at an intermediate rate. Representational Distortion. We approached the question of representation in the interest arena through the lens of descriptive representation. With some interesting variation, the analysis suggested a general pattern of class-based representational distortion. Virtually all rates of participation in direct action, including rates of protest, varied across social class in a pooled analysis, with the popular sectors nearly always participating less than the middle classes. Most forms of direct action are shaped by “human capital” and personal resources that often come with education and higher social status, including a sense of efficacy. Therefore, class-based distortion in forms of direct action that involve contacting officials or initiating judicial proceedings is not surprising. More unexpected is that a pattern of class-based distortion extends to protest, often thought of as a “weapon of the weak” in Latin American societies. It is also surprising that representational distortion was found, with important cross-national variation, with respect to participation in a set of associations selected in this study precisely because of their presumed greater relevance to the popular sectors. Associational participation was free of representational distortion only in Peru, and levels of distortion were highest in Argentina, an unanticipated finding given the country’s history of popular-sector mobilization and organization in unions. The role of class, and presumably the resources that it brings, in disadvantaging the popular sectors is ameliorated slightly by the prior experience of some popular-sector individuals with other forms of problem solving and collective action. Particularly important in predisposing people to participate in both direct action and associations is a history of union membership; prior experience in a squatter invasion is also related to a greater likelihood of participating
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 298
.
in associations, but not to undertaking direct action. Thus, prior experience in these forms of collective action is predictive of subsequent problem-solving activity. However, with the decline in union density and possibly the diminution in the incidence of land invasions with the exhaustion of peripheral urban land, it is unlikely that future generations will gain such skills through these formative experiences. In any case, in sheer numerical terms, the presence of those coming from such backgrounds did not make up for the discrepancy between classes in rates of participation. Thus, with the exception of Peru, the interest arena is marked by significant representational distortion. The analysis also found a pattern of representational distortion in the electoral/partisan arena, where the popular sectors are generally underrepresented across a set of measures of political engagement. Intriguingly, at a cross-national level, patterns of representational distortion in the electoral/partisan and interest arenas roughly mirror one another. In both arenas, distortion is highest in Argentina and lowest in Peru. In general, then, the “new associationalism” does little to compensate for representational distortion in the electoral/partisan arena; rather, it seems to more nearly replicate it. There is some evidence that distortion in the two arenas may be related. Participants in specific types of associations are generally no more likely to be partisans than are nonparticipants, and associations are not as active in mobilizing members into partisan and electoral politics as were unions in the -Hub. However, participants in associations are disproportionately likely to be party “activists”—those who participate in other ways in party/electoral activity—forming a pattern of “shared-member linkage” between the two arenas. These shared-member linkages may constitute a source of informal partisan influence in the interest regime. At the same time, because the pool of associational participants reflects class-based distortion, as discussed above, if associations serve to mobilize participants into party/electoral activism, they may do so, in the aggregate, in a manner that affirms or accentuates patterns of class-based distortion in the electoral/partisan arena. Associational Strategies. A primary goal of the study was to explore the repertoire of strategies associations pursue in trying to represent popular-sector interests. Based on data from a “core segment” of the A-Net, the volume examined state-targeted strategies through institutional and noninstitutional channels and society-targeted strategies of provisioning and capacity building. As many studies have suggested, the most utilized strategies of associations are society-targeted. However, a majority of associations report “dual” repertoires, which include state-targeted strategies as well as society-targeted strategies. Most claim-making activities are targeted to the local level, no doubt furthered
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 299
,
by the decentralization of the state and, as suggested in chapter , most likely related to the difficulties with scaling faced by associations lacking resources, characterized by high interest heterogeneity, and tending to make discrete and disaggregable demands. However, the repertoire of some associations in the A-Net includes national-level claim making, particularly on the part of those that play coordinating roles within networks and those with ties to labor unions. A final point is that claim-making activity occurs through both institutional and noninstitutional channels. Rather than a division of labor between routinized and contentious forms of claim making, there seems to be an affinity between the two. Furthermore, contention has become a rather common component within the claim-making repertoire, mirroring a point made by Meyer and Tarrow (:) that “social protest has moved from being a sporadic, if recurring feature of democratic politics, to become a perpetual element in modern life.” Many of these strategies bring associations into contact with the state. For example, provisioning activities often take the form of implementing government programs in a form akin to state “outsourcing” and cannot be reduced to “self-help” activities in a sphere of action separate from the state. The overall picture, then, is one in which associations routinely engage the state, through both state-targeted activities and as an integral part of many society-targeted activities. However, it is important to remember that such engagement should not be confused with effectiveness or influence. Channels for voice are in place, but the effects of voice remain unexplored here. Associations with certain traits tend to employ particular strategies more often. Material resources seem particularly related to the utilization of provisioning activities, and state ties and organizational resources have some importance for claim making. In addition, associational scaling emerged as a major factor shaping the propensity to engage in particular strategies, including national-level contacting. Associational Scaling. While some studies have suggested a largely disorganized and atomized associational world, the data suggest that, within a core segment, many associations have regular working relationships with one another. Even within our chain-referral sample, however, some associations do remain isolated, particularly in Venezuela at the time of the surveys and, to a lesser extent, in Chile. The forms by which coordination occurs in the A-Net are far from uniform and appear to have different implications, especially regarding state-targeted strategies. The number of inter-associational linkages possessed is predictive of the propensity to engage in local-level contacting, but not nationallevel contacting. In contrast, two distinct kinds of “coordinating associations,”
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 300
.
Nodal and Flexible Fronts, are more likely to engage in national-level contacting but not local contacting. These entities thus appear to play a role as conduits to national-level officials and may be vital to the operation of the A-Net more generally as a vehicle for popular representation. Evidence also suggests that unions—particularly newer forms of social movement unions and publicsector unions—may play a similar role in shaping the strategies of associations within the interest regime: in Argentina and Chile union-linked associations are more likely to engage in both noninstitutional claim-making strategies such as protest and institutional strategies such as contacting, especially at the national level, and further, to have more linkages to other associations, thereby potentially holding nodal positions of greater influence in the A-Net. State-Association Ties. Aside from engaging the state through claim-making activities, many associations are to some degree dependent on the state for resources and may call on political parties to act as intermediating actors, posing potential threats to associational autonomy. While associations in the A-Net are not marked by the severe state controls that characterized unions in the -Hub, a large number of associations are recipients of government financing or implementers of government programs, and some had state officials involved in their initial formation. Associations may also have relationships with parties that allow for a more intermediated form of access to the state. While ties between associations and political parties may not approximate the organic organizational linkages that typically characterized unions and s, the idea that associations are largely disarticulated from the world of partisan politics cannot be sustained. About a quarter of the associations in the sample do have linkages with parties, although many of these contacts constitute little more than instrumental relations in the context of electoral politics. These looser patterns of interaction with the state and political parties lend themselves to substantially greater autonomy than unions had in the -Hub.
National Models of the A-Net While the volume as a whole explored these five analytic themes regarding the broader interest arena, chapter sought to conceptualize and examine four aggregate traits of the interest regime, using them as a basis for comparing the -Hub and the A-Net. Here we return to these four dimensions in order to examine emergent national models of the A-Net. The shift in the popular interest regime is general to all the countries in the study, as well as to others in Latin America, but has taken different forms. Each
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 301
,
country experienced the decline of the -Hub in its own fashion and had its own experiences with a transition to marketized economies. Each too has been characterized by its own set of political entrepreneurs in the state and social actors on the ground that have contributed to the political construction of the A-Net. Understanding the diversity of national models is fundamental to a consideration of the nature of the shift in the popular interest regime and the contours of the organizational infrastructure as a potential vehicle for popular-sector voice and problem-solving activities. We note again that, unlike the survey of individuals, the survey of associations did not employ a random sample but rather a chain-referral procedure, implemented in similar fashion across countries, but nevertheless making cross-national comparison of frequencies somewhat problematic. Yet the survey yielded a unique comparative dataset across a large number of associations, and on this basis we cautiously advance some ideas about these emergent national models, followed by a discussion of possible explanations for this diversity. We begin with an assessment of scope, defined in chapter as the extent to which the popular sectors participate in the A-Net, an issue explored in different ways in chapters and . While many associations act to represent the interests of a relatively broad group of constituents, not just participants, the question of how much the infrastructure of the A-Net is utilized by individuals to solve problems is nevertheless fundamental. Table . presents two measures (from table .): the extent of participation in programmatic associations in the A-Net, that is, those associations whose raison d’être is problem-solving activity, and the extent of participation in a larger group of associations that also includes nonprogrammatic associations that nevertheless undertake problem-solving activities and to that extent may be here considered “political.” While unions are considered part of the A-Net and many have been integrated into associational networks in some way, this study primarily focused on associations. Given that associations, unlike unions, are not in principle restricted in coverage, an important question in terms of associational representation is whether or not they are in fact significantly more encompassing than unions under the -Hub. Table . Scope (percent)
Associational Participation “Programmatic” Associational Participation “Political”
Argentina
Chile
Peru
Venezuela
22.2
25.2
55.9
17.8
34.2
40.6
68.4
25.9
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 302
.
Scope offers one of the most vivid cross-national contrasts in the volume, with the A-Net in Peru emerging as strikingly extensive. Popular-sector individuals in Peru are more than twice as likely to belong to a programmatic association as those in any other country. When we add other civil society organizations that report engaging in problem solving, the gap narrows somewhat but remains large. This second measure also suggests a clear difference between Argentina and Chile, on the one hand, and Venezuela, on the other. An overall ordinal picture thus emerges with Peru having high scope, Argentina and Chile medium, and Venezuela low.2 A second dimension, scaling, captures the extent of cooperation and coordination among organizational entities. Chapters and approached this from different angles. The former explored three forms of scaling by analyzing inter-associational linkages and two kinds of coordinating associations, Nodal and Flexible Fronts. The three types lend themselves to different problem-solving strategies, with coordinating associations of both kinds crucial for national-level accessing, and inter-associational linkages more important for local contacting. Chapter explored the role of unions in the A-Net by examining whether union-linked associations displayed different patterns of problem solving. Union-linked associations are more likely to organize protests, sit on policy councils, and contact state officials, and this more active profile may be especially important because they tend to play more central roles within the larger associational network, working and coordinating more with other associations. Table . reproduces the cross-national findings from these chapters regarding a variety of measures of scaling. Peru and Argentina stand out as having the most extensive scaling among the four countries. Each is characterized by high embeddedness with far fewer isolated associations and a relatively high proportion of heavily linked associations. Yet they display somewhat different scaling patterns in that Argentina is oriented more toward coordinating with new unions and Nodal ,3 organizations that tend to play particular coordinating roles, while Peru is oriented toward a pattern in which associations coordinate through the formation of Flexible Fronts among associations. Chile emerges in an intermediate position. As in Argentina and Peru, both unions and coordinating associations 2. While associational scope is slightly higher in Chile than Argentina, union density is significantly higher in the latter. A measure of scope that combines participation in associations and unions puts Argentina somewhat ahead of Chile. Both easily surpass that of Venezuela and continue to lag far behind that of Peru. 3. As chapter shows, most of these unions fall into the category of new “social movement” unions rather than the older major industrial unions.
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 303
,
Table . Scaling (percent) Argentina
Chile
Peru
Venezuela
9.2
23.2
10.0
35.4
35.8
30.7
41.3
25.4
Works with Coordinator
49.8
32.6
38.0
24.5
Works with Union
22.5
14.4
12.5
8.9
13.0 5.0
9.3 8.8
6.1 14.4
4.8 0.4
Embeddedness in Networks
Incidence of Coordinating Associations
Isolated Heavily Linked
Nodal s Flexible Fronts
have a significant presence in the A-Net, but the proportion of isolated associations is significantly higher, and the proportion of heavily linked associations slightly lower. Venezuela displays the least associational scaling along every measure. There are hardly any coordinating associations in the Venezuelan A-Net, and over one-third of associations are completely isolated. Finally, we turn to the dimensions of access and autonomy. Unlike unions, which are highly constrained by state regulation, state influence over associations tends to be exercised more by inducements or the application of benefits (Gamson :–). As discussed in chapter , both the pursuit of access and other forms of interaction with the state can present such inducements that may challenge autonomy. Autonomy is a salient issue given the history of corporatist unions, which tended to be regulated by and dependent on the state as well as co-opted by political parties under the -Hub. By contrast, much literature on popular associations has signaled autonomy as one of the main advantages of associations. Because of the difficulties of measuring autonomy, which would require more fine-grained observations than are possible with the type of data analyzed here, this study has approached this important trait of the A-Net at one step removed. Just as we look at the A-Net as an organizational infrastructure or channel of representation but stop short of making claims about its actual effectiveness, so too we note aspects of associational relationships with states and parties that entail potential threats to autonomy through inducements, but we stop short of directly evaluating degree of autonomy in the A-Net. Similarly, we do not directly observe access as such, which is a complex concept, but instead examine the prominence of accessing strategies and the
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 304
.
Table . Accessing and State Dependence (percent)
Accessing
State Dependence
Contacted Municipal Authority Contacted National Authority Accessing Linkage to Party Delivers Government Program Receives Money from Government
Argentina
Chile
Peru
Venezuela
70.0
77.6
62.1
31.3
37.9
38.6
22.1
20.8
24.6
20.2
10.0
19.6
52.5
20.6
9.6
9.2
39.6
28.9
18.3
26.3
state level at which they take place. In table . we reproduce data from chapter regarding the incidence of accessing strategies and add data regarding the combined incidence of issue-based and multifaceted party-association linkages, those that involve the party serving as an intermediary in associational accessing strategies (see chapter ). These indicators provide a measure of accessing. Also presented are the proportion of associations that are dependent on the state. Both accessing and dependence may function as inducements to further cooperation with the state in order to maintain funding or program participation, though associations maintain discretion about the implications of these relations and the effect of the inducements; that is, the degree to which the desire to maintain access, funding, and program participation shapes associational behavior remains an empirical question. Chapter suggested that those associations dependent on the state more frequently pursued accessing strategies. Table . indicates that a similar relationship between accessing and state dependence exists as well at a cross-national level. Argentina and Chile possess the highest levels of accessing overall. State dependence is also higher in these two countries, although Argentina clearly outpaces Chile in this respect, with the proportion of associations involved in distributing government programs particularly high. Peru and Venezuela, by contrast, both have lower levels of accessing and a lesser incidence of relationships of dependence on the state. While the two are similar in many respects, however, it is important to note that local accessing is substantially higher in Peru than Venezuela, while contacts with parties are more frequent in the latter. Putting these dimensions together, we can discern some strikingly different national variations of the A-Net. Figure . presents rough scores on scope and scaling and figure . on associational engagement with the state, specifically
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 305
,
state dependence and accessing strategies, both of which have implications for associational autonomy. The Argentine A-Net reflects a pattern that might be called a Statal Web. It is most oriented to interacting with the state, has a high level of scaling and an intermediate scope, with the composition of participants being the most balanced among associations and unions. The Peruvian pattern, by contrast, can be labeled an Extensive Autonomous Web. Associational relationships with the state are low, in terms of both accessing and dependence. Scope is extraordinarily high, but substantially restricted to associations rather than unions, and scaling is also very high, with dense networks and coordinating associations that take the nonhierarchical, nonbureaucratized form of
High Argentina
Scaling
Peru
Chile
Venezuela Low Low
Scope
High
Fig. .. Scope and Scaling in Comparative Perspective
High
Chile
Argentina
Accessing Peru Venezuela Low Low
Dependence
Fig. .. Accessing and State Dependence in Comparative Perspective
High
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 306
.
Flexible Fronts. The Chilean A-Net occupies an intermediate ground in some respects, and can be labeled a Liberal Web. It is characterized by high levels of accessing but less state dependence than that of Argentina. At the same time, the Chilean A-Net displays considerably higher levels of state linkage than that of Peru. The scope is of intermediate level, and a significant number of associations are relatively atomized, so that overall scaling is lower. Finally, the Venezuelan case represents a pattern of Narrow Neopluralism. Scope was extremely low at the time the surveys were fielded, with scaling weak and interactions with the state relatively rare. This finding regarding Venezuela, of course, reflects the fielding of the surveys after the breakdown of a very strong -Hub but before the most ambitious chavista initiatives to spur associationalism had really begun. As discussed in more detail below, these subsequent events, especially the massive initiative to sponsor the formation of Communal Councils across the country, have undoubtedly completely reshaped the Venezuelan A-Net. While comparable data for the most recent period in Venezuela is not available, it would seem that the chavista initiatives have dramatically increased the scope of the A-Net while also providing greater access and facilitating more scaling, albeit to degrees that are difficult to ascertain, but at the cost of compromising autonomy through very extensive relationships of state dependence. It is possible, then, that a pattern that might be labeled Extensive Partisan Neocorporatism has emerged since the surveys were fielded. Figures . and . attempt to graphically display these radical changes and indicate a direction of movement. As the Venezuelan case indicates, national configurations of the A-Net are not fixed, but rather still evolving. And like the -Hub, they are subject to concerted efforts by political entrepreneurs to reshape the world of popular organizations. Nevertheless, it is worth considering why those variations picked up by the surveys in had emerged, as well as how each model might have changed in subsequent years. The -Hub emerged in a “founding moment” that often took a long time to work itself out, involving actions and reactions. So too, the emergence of the A-Net should be seen in terms of an extended period of actions and reactions. The analysis in this book is thus a snapshot of one phase in a process. In the following sections, we provide some preliminary thoughts, in a first step toward theory generation, about why the different national patterns had emerged at the time of the snapshot and also provide some brief updating. Four explanatory factors are highlighted—one is economic, the others are political. The former is the nature and timing of economic crisis during the decades before the surveys were fielded. Economic crisis was important as an
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 307
,
impetus for associationalism around subsistence and provisioning activities and hence has implications for scope. One political factor is the timing of the shift: the decline of the -Hub and the rise of the A-Net. Despite the relative synchronicity of the debt crisis, economic marketization unfolded across a quite extended period of time, so that some countries have had a substantially longer period for the A-Net to develop and consolidate. Two other political explanatory factors are the continuity of party systems and the nature of political projects launched “from above” by elite actors. More stable party systems may allow for greater regularization of relations between associations and parties or associations and state officials. The specific initiatives of elite actors, either mobilizational projects or social policy initiatives, can offer incentives or disincentives for associationalism and claim-making activity. Argentina—Statal Web The analysis above suggested that the A-Net in Argentina can be considered a Statal Web. The case is characterized by an intermediate level of scope combined with comparatively high union density. It further has a high degree of scaling, distinguished by the key role of Nodal as coordinating entities and by the centrality of union-linked associations in the network. These patterns are accompanied by extensive state-association relationships, both in the degree to which associations engage in accessing strategies and in the form of state dependence. In particular, the proportion of associations delivering government programs is extraordinarily high. In terms of scope, the Argentine A-Net has a particular character in that associationalism is intermediate but is augmented by substantial union participation. This “mixed” nature may be a product, on the one hand, of a relatively late and more partial shift in the interest regime and late onset of associational proliferation, combined with, on the other hand, post-shift social and political dynamics that have provided strong spurs to associationalism. The decline of the -Hub occurred more partially and only about ten years before the surveys were fielded, with the Menemist reforms of the early s (Etchemendy ; Murillo ; Levitsky ), and associationalism did not really gain steam until relatively late (Campetella and González Bombal ). This point is corroborated by the data on the age of associations, which show that, tied with Venezuela, the Argentine A-Net has the highest proportion of associations (roughly percent) founded in the ten years before the survey. The initial period of economic reform saw recovery from the crisis of –, but unemployment and informalization in this traditionally “full employment”
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 308
.
economy began to accelerate in the second half of the decade, producing, relatively late, the social-structural shift on which popular associationalism has been based. The time period for associational participation to expand in earnest has thus been relatively narrow, more akin to Venezuela than Chile or Peru. However, despite the relatively late development of associatonalism, other factors seem to have combined to spur participation and mobilization. Because of the unusually high scope of the Argentine -Hub and the swiftness and extent of market reform in the s, the Argentine popular sectors had an unusually large number of potential associational participants with mobilizational skills and experiences from past union organization. Associationalism was further encouraged by the deterioration of the economic position of the popular sectors during the latter half of the s—which saw a rise in informality, unemployment, and poverty rates—and then, more dramatically, by the economic crisis that began in , arguably the most profound in Argentine history. A final factor concerns the interaction between state policy and the emergence of organizations of the unemployed, which began to engage in dramatic and extended mobilizations in to expand the initially small and limited government workfare program. This program, Plan Trabajar, provided a target for the demands of those in need of work as well as for associations in need of workers and was thereby a stimulus for further participation in these organizations (Garay ). Despite the late timing, the Argentine A-Net quickly became characterized by a high level of scaling, marked by high network density and significant numbers of both union-linked associations and Nodal . In an explanation for this extensive scaling, two factors seem potentially relevant. Many associations have developed linkages to a new labor movement that is more oriented toward social movement unionism—one that continues the tradition of collective action within the framework of a highly scaled labor movement, as seen in the formation of a new confederation. That union-linked associations occupy nodal roles within the A-Net and also scale more frequently provides some evidence that repertoires for coordination and collective action are diffusing through those linkages, perhaps enhancing the overall level of coordination within the A-Net. A second plausible factor is the role of the state and, specifically, the orientation of social policy. As Garay () further argues, Plan Trabajar not only spurred participation, but also created incentives for coordination among associations of the unemployed, and, by having workfare participants provide labor to community-based associations as part of the program, it created stakeholders across the associational world and thus further incentives for coordination around the program.
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 309
,
The orientation in the Argentine A-Net toward accessing strategies, especially at the national level, may partly result from these high levels of associational coordination. Accessing strategies would also seem to be facilitated by policy continuity and institutionalization of the party system, also present in Chile but notably absent in Peru and Venezuela. Continuity in the social policy of Carlos Menem, president for a decade from to , was briefly disrupted by his successor Fernando de la Rúa, particularly regarding the Plan Trabajar. However, his government lasted for little more than two years before he was forced to resign, and his successor, Eduardo Duhalde, expanded workfare again. Substantial continuity also marked the party system. Although the Unión Cívica Radical was a shadow of its former self after its failures to tame hyperinflation during the s and largely fell apart at the national level after the economic crisis that began in early , the Partido Justicialista remained the mainstay of the party system, particularly for the popular sectors, and was seen as a paradigmatic example of adaptation and continuity within the region (Levitsky ; Seawright ; Roberts forthcoming), especially prior to its factionalization in . It seems plausible that this impressive level of political continuity facilitated the high degree of accessing strategies undertaken by associations in the A-Net, with channels for access solidifying as relationships between associations and officials in the state and parties become more regularized. The orientation of state policy, especially regarding social policy, also encouraged accessing strategies and increased the prevalence of certain forms of state-association relationships of dependence. As noted above, a major aspect of the state response to economic hardship was a series of workfare programs, which stimulated high levels of organizing around the issue of unemployment. A feature of Plan Trabajar and the subsequent Programa Jefes de Hogar Desocupados was that associations participated in program implementation either as beneficiaries, becoming in effect dispensers of workfare programs, as in the case of Plan Trabajar, or by providing the institutional space where beneficiaries could do their workfare obligation, as in the case of Programa Jefes de Hogar Desocupados. Other programs such as food distribution also bring associations into service delivery or make associations reliant on state resources. These programs encouraged accessing strategies in several ways. Most basically, they created targets for demand making to government officials, including national officials. Further, since part of the state response to demands involved the establishment of limited policy councils incorporating some associational participants, the programs further spawned institutionalized channels for accessing. Yet the design of the programs and the state-association relationships
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 310
.
also carried co-optive potential through the incentives of benefits—either program participation, informal access, or institutionalized inclusion. Indeed, particularly in a context of great social and political unrest, that dynamic became particularly salient regarding the unemployed groups, many of whom eventually became Néstor Kirchner allies (Garay ). Changes Since . The surveys were fielded in the summer of , as the economic crisis was just starting to recede and immediately after the election of Kirchner, who subsequently led the country through an economic recovery and built a formidable multiclass coalition. An important aspect of his strategy was to cement support from both the traditional unions (Etchemendy and Collier ) and key unemployed groups, which came to constitute an important feature of the Argentine A-Net more generally Under Menem, the unemployed associations had emerged as among the largest and most contentious associational actors in the interest regime, mobilizing for expansion of the government’s workfare programs. In response to their dramatic demonstrations, Kirchner took active steps, not only to negotiate, but to create linkages with them, resulting in exchange relations in which access was traded for support, with the president offering significant inducements, including seats on policy councils and important positions in his government. This strategy of linkage also offered numerous advantages to Kirchner. Garay () suggests that because the unemployed organizations emerged outside traditional machine networks, linking with them was a means for Kirchner and his Frente Para la Victoria () to attract popular-sector constituencies while circumventing other factions of the party with which he was struggling for power. As organizational allies, the unemployed activists were also selectively mobilized for other purposes. For example, numerous observers pointed to the speed with which several allied unemployed groups began to demonstrate outside gas stations after Kirchner called for a national boycott of Royal Dutch/Shell in . And with a commitment to economic stability and a coalition that embraces wide segments of the middle class, Kirchner and his also had a substantial incentive to limit and control the more destabilizing forms of mobilization of the unemployed. While some of this control was achieved through co-optation via inducements, he also took stronger measures to crack down on the activities of associations that remained intransigent. Kirchner’s approach to the rest of the A-Net was considerably more hands off, suggesting likely continuities or the entrenchment of the Argentine pattern found in this study. Despite postcrisis rhetoric and some evidence of a move to the left, Kirchner generally maintained but did not greatly expand social programs, drawing fire from some quarters for relying too heavily on growth
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 311
,
to lift people out of economic hardship (Gerchunoff and Aguirre ; Grugel and Riggirozzi ). Further, general continuities in social programs suggest that the pattern of inducements and, specifically, a top-down shaping of the interest regime through state-association “partnership” in the implementation of government programs, are likely to have solidified for those associations that have been brought into the state’s system of delivery and distribution. The Argentine A-Net stands out in several respects. Overall, it is particularly statist in terms of both its accessing orientation and its level of dependence. The unemployed associations in particular came to acquire a privileged and central position within the A-Net, a degree of integration into policy making, and an exchange of access and concessions for political support. In notable ways, the state shapes the interest regime to make certain organizations more dominant and to influence their demands. It does this through both its policies and the strategies of political leaders, in interaction with mobilizational forces from below. Within the associational world, associations of the unemployed and demands for jobs, rather than other forms of welfare or redistribution, have become prominent in response to the existence of a government workfare program; at the same time, traditional unions have seen a resurgence, in part related to Kirchner’s strategy of garnering support while remaining within inflation targets. The emergent interest regime may be marked by a kind of “neocorporatist” pattern of negotiated cooperation regarding the unemployed associations as well as traditional unions, which have retained relative strength, compared to those in other Latin American countries. Both of these central actors in the A-Net entered into some type of relationship with the state, consisting of concertation and inclusion into policy-making discussions. In exchange, they have generally supported the government. Yet compared to historical patterns of state corporatism, both are characterized by greater distance from the state, with more ad hoc exchange relations. With many other popular-sector organizations important and operating within the A-Net, the interest regime is quite novel, in terms of both the nature of the “neocorporatist” component and its internal variety. Chile—Liberal Web The analysis in this volume has revealed an associational interest regime in Chile that may be surprising given the past assertions regarding weak, disarticulated popular associationalism in that country. The A-Net took on the characteristics of a Liberal Web, with intermediate levels of participation and
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 312
.
coordination and a much stronger pattern of state linkages than that found in Peru or Venezuela, but with considerably less state dependence than in Argentina. Some of the most prominent statements suggesting an image of Latin American associationalism that is atomized and disarticulated from the state have emerged from empirical work within Chile (Oxhorn b; Kurtz b). Yet, the intermediate position of the Chilean case regarding the extent of scope and scaling, as well as the relatively high level of accessing, represent something of a departure from that literature and underline the importance of comparative perspective. The intermediate level of scope in the Chilean A-Net may result from the emergence of an extensive pattern of popular organization and activity in the s, followed by a degree of demobilization in the s. The decline in the Hub began exceptionally early in Chile. The Pinochet regime’s combination of outright repression and early experiments in market reform, targeted at the power of both unions and political parties, decimated the -Hub. Leaders of s were largely driven underground or into exile. The lifting of tariffs and the Plan Laboral, a major labor reform implemented in , also had devastating consequences for the organizational power of the labor movement (Angell ; Frank a). The destruction of the -Hub, however, was not immediately attended by a flourishing of associationalism, as the harshly repressive, demobilizing, and overtly atomizing policies of the Pinochet government put a damper on all forms of popular-sector organization during the s. The s witnessed a kind of opening, with the eruption of protest in the context of one of the region’s most profound debt crises, followed by the approach of the long-awaited democratic transition. In this context, the decade witnessed a surge of popular associationalism, especially in the poblaciones of Santiago. Chile had great difficulties weathering the debt crisis, and the economic downturn hit the popular sectors particularly hard, leading to a sharp rise of subsistence organization around basic needs. The recovery of the economy in the mid-s coincided with increasing opportunities for contestation around the regime issue. Consequently, rather than dissipating with economic recovery, the wave of organization and mobilization was channeled in a new direction, and popular associations became key actors in the regime transition. Other factors underwrote this upsurge in mobilization. Most important were the strong legacy of organization and protest and the extensive experience of the popular sectors with forms of collective action, especially during the mobilizational Allende years. In addition, associations could draw on the participation of many former unionists who had joined the ranks of the unemployed. While earlier measures are not available, it seems likely that
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 313
,
the scope and coordination of the emergent Chilean A-Net were high during the s. Many observers have suggested that popular-sector organization in posttransition Chile (and in Latin America more generally) dwindled. This trend toward demobilization provided the empirical basis for numerous studies suggesting the great difficulties facing popular-sector collective action in Latin America under the neoliberal model. We do not dispute that such a decline took place. However, the data suggest that the subsequent decline left the Chilean A-Net at intermediate levels of collective action, rather than an atomized and minimal A-Net, as has sometimes been portrayed. A further point to consider is that fourteen years separated the fielding of the surveys from the transition, nearly as long as the military regime itself lasted. A more responsive democratic government has had time to become entrenched and to launch a more extensive set of policies addressing popular-sector interests. While the policy orientation of the state does not seem so conducive to engaging associations as in Argentina, incentives for associational formation and coordination for more routinized claim-making purposes may have risen. Indeed, the Chilean A-Net is characterized by a relatively high orientation toward accessing strategies, but only a moderate degree of associational dependence on the state. The high degree of accessing may be underwritten by the unique stability of the Chilean government and party system during the posttransition period. The sustained dominance of the Concertación represents the longest uninterrupted rule by a single party or bloc in third wave South America. And while the party systems of the -Hub era have undergone substantial transformation in every other country in the study, the Concertación parties have been stable in Chile, though all major parties have moved toward a catchall orientation and transformed their linkages to society. The data suggests that party contacting in Chile does not lag behind when compared, for example, to Argentina, which has seen Peronist continuity. In comparative perspective, contacting in Chile thus appears relatively robust, at least more so than portrayed in other studies (Posner ; Luna ). Despite this comparatively high orientation toward accessing, threats to associational autonomy are lower than in Argentina, given the greater distance between associations and the state. The Chilean state does less to fund associations or to bring them into the delivery of social services. In the democratic period in Chile, this different approach to state-association relations follows the military’s hostility to collective actors and is a continuation of a more technocratic approach to policy making in the context of sustained economic success and center-left political hegemony. The Concertación has neither relied
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 314
.
heavily on associations for mobilizing electoral support nor has it been confronted with an associational world heavily engaged in contentious activity. Changes Since . The period since the surveys were fielded has been characterized by greater continuity in Chile than in any of the other three countries in the study. The economy has been relatively stable and successful, and in Michele Bachelet succeeded Ricardo Lagos with a mandate to continue on the same course. While her administration has seen a rise in mobilization, these protests have largely consisted of students contesting higher education policies or a multiclass coalition of citizens unhappy with the progress of transportation reform. Neither issue seems to be highly related to the activities of popular associations per se. And while other left-of-center governments may be searching for innovative ways to use the A-Net to mobilize popular-sector political support, such as in Argentina and Venezuela, neither Bachelet nor Lagos seems to have had the inclination or the need to do so. Boas () reports that Bachelet’s presidential campaigns created and/or enlisted the support of “citizen networks” but did not reach out to associations. In sum, then, two somewhat unexpected features of the Chilean A-Net are its intermediate level of scope and its orientation toward accessing strategies, particularly at the national level. While these features characterized the prePinochet interest regime, they now recur in a nonmobilizational, nonideological, and attenuated form, and in the context of a neoliberal rather than an interventionist state. It is widely understood that a goal of the Pinochet government was to disarticulate popular-sector organization and to impede the ability of the popular sectors to engage in collective claim making, especially at the national level. While this goal was partly accomplished through repression during the military period, studies have suggested that market reforms have had a lasting political effect into the democratic period, constituting a legacy of Pinochet’s war against the organized popular sectors. The analysis suggests that long-term effects on popular organization may be partly overstated. The regime seems to have effected a significant and enduring demobilization when compared to the pre-Pinochet Frei and Allende years. However, fourteen years into the new democratic period, the Chilean A-Net, within this comparison set, can be characterized as a Liberal Web with intermediate scope and scaling, and an orientation toward claim-making strategies. Peru—Extensive Autonomous Web One of the more intriguing findings of the study is the distinctiveness of the Peruvian A-Net. This contrast emerged most strongly along the scope
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 315
,
dimension, with the popular sectors in Peru more than twice as likely to participate in associations as those in the other countries. The Peruvian A-Net is also marked by very high scaling of a particularly basist form, characterized by high inter-associational linkages and many Flexible Fronts, combined with relatively low state linkages. If we take these attributes together, the A-Net in Peru approximates the stereotype of an extensive, interconnected, and largely autonomous sphere of associational life that can be found in some of the rosier statements concerning civil society more generally. At the same time, it exhibits traits that some analysts see as a trade-off in terms of disarticulation from the state: the Peruvian A-Net is an organizational infrastructure that functions less than the others as a channel for accessing strategies, especially at the national level. The unusually extensive scope of the Peruvian A-Net in part derived from an early shift in the interest regime. While unionization and union activism remained high during the s, the military regime that ruled between and embraced an anti-Aprista agenda, catalyzing an early shift in the balance between unions and new popular associations within the interest regime. Although within the Peruvian context unionization remained relatively high until the mid-s, the military moved against , to the point of even favoring the Communist Party, which became a more important ally of the labor movement. More central for the present discussion, the military government attempted to foster support among nonunionized segments of the popular sectors by promoting the formation of popular associations in the newly settled urban areas. In the first few years of the military period, this attempt to mobilize support was carried out to a significant degree through (the Sistema Nacional de Apoyo a la Movilización Social). This initiative had a particular social-structural basis in Peru, which is the poorest country of the four in this study, with comparatively high postwar levels of informality and unemployment and with an exceptionally high number of squatter settlements, which each year extended further into the desert surrounding Lima. These settlements often developed community-based associations for the purpose of bringing urban services. In the democratic period, both political and economic trends have supplied further incentives for associationalism. The country experienced vast economic turmoil during the s, catalyzing a wave of subsistence organization and protest activity. In response to popular mobilization, the government of Fernando Belaúnde of Acción Popular laid the groundwork for a nationwide food distribution initiative in (Copestake ). Yet it was the leftist Izquierda Unida (), under the mayorship of Alfonso Barrantes, that implemented the Vaso de Leche program, reaching out to grassroots associations that had been
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 316
.
formed and integrating them as Comités de Vaso de Leche to be distributors of the government program (Barrig ; Klarén ; Blondet, a, b). , newly ascendant in the early s, also created ties with associations in many of the poorest squatter settlements in Lima, which historically had been linked either to or to the military government. For its part, as the labor movement weakened and many elements defected to the left, also made a concerted effort to target nonunionized portions of the popular sectors, establishing ties with clubes de madres that had been established by women in many of Lima’s lower-class neighborhoods. The high level of associational participation found in the survey may also have been further spurred by the fall of Alberto Fujimori, ushering in a freer political environment, which was marked by increases in protest and popular mobilization. It should be noted, however, that the data suggest that much of the Peruvian A-Net was established relatively early: percent of associations were formed in or earlier, far more than in any other country in the study. The Peruvian A-Net is also characterized by a high degree of scaling, with a basist orientation in which inter-associational networks are dense and Flexible Fronts are the most prevalent coordinating entities. This extensive scaling may, again, derive partly from the early emergence of the A-Net, as associations have generally had much more time to develop working relationships. It also may be the case that extensive scope is simply conducive to higher scaling. The data show that Peruvians have a greater tendency to participate in multiple associations, and thus they may serve as potential bridges between them. The high level of associational participation and coordination in the Peruvian A-Net is accompanied by a disarticulation from the state, in which associations are less oriented toward accessing strategies and have relatively low levels of state linkages. The relative autonomy of the associational world is in some ways surprising given that, as noted above, Peru has perhaps the longest history of projects from above seeking to mobilize and create linkages with the associational world. One plausible explanation underlying the low levels of accessing strategies and other state-association ties may center on the high level of volatility in the party system. The decade and a half before the surveys witnessed the collapse of the three most important political parties in the country, followed by the semi-authoritarian presidency of Alberto Fujimori, who founded a succession of weakly organized political parties. The collapse of the party system can be seen as having two kinds of effects. First, the collapse of the traditional parties and subsequently Fujimori’s Cambio seems to have weakened them as intermediaries in the interest arena, although as chapter suggests, Peruvian associations have continued to be the recipients
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 317
,
of campaign appeals from political parties.4 Second, the collapse of the party system and the ensuing volatility also created extensive turnover among state officials—not just those who are elected but, as patronage and appointment powers change, also those in the bureaucracy. Since patterns of access are facilitated by informally routinized contacts between associations and officials, extensive turnover on one end of the relationship is likely to lead to a disruption in linkages. Such political dislocation may also cause changes in the policy orientation of the state, especially if new actors are not committed to a policy repertoire conducive to associational demand making and program implementation or to otherwise providing resources to associations. While Fujimori, and then Toledo after him, kept the Vaso de Leche, with some changes that decentralized the program, overall the data suggest that associational involvement in state programs and receipt of state funding was the lowest among the four countries. Changes Since . As the surveys were being fielded, there were some signs of tightening linkages between the A-Net and the state. A national law passed in established Mesas de Concertación para la Lucha contra la Pobreza (Roundtable for the Fight Against Poverty) at the national, regional, and local levels. These mesas bring together state and associational actors to develop shared antipoverty programs and coordinate activities. Furthermore, the decentralization reform established formal institutions to incorporate the participation of associations in policy making. This national law mandated the establishment of participatory governance councils for every regional and municipal government in the areas of infrastructural improvements and land use. With the election of ’s Alan Garcia in , the question arose concerning additional changes in state-association relations, given the centrality of associations to aprista mobilizational efforts during his first presidency in the s. Garcia initially took actions to increase state control more generally over the A-Net. A new law required all s to register with the state and declare how funds from international donors would be used, and it included a provision allowing the ministry to outlaw s whose activities were deemed to conflict with state development goals. The surveys captured an extensive and widely interconnected popular interest regime characterized by substantial disarticulation from the state, with associations generally more oriented toward society-targeted strategies and rarely 4. was beginning to experience an electoral resurgence at the time the surveys were fielded. The data suggest that this manifested itself in occasional instrumental attempts to seek the electoral support of associations but very rarely entailed acting as an intermediary between associations and the state.
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 318
.
involved in government programs and other relationships of dependence. Yet the recent developments just discussed, especially the possibility of new forms of state linkage, both in terms of increased access through participatory mechanisms and through various forms of regulation, raise the question of whether the picture of the A-Net captured in this snapshot is again changing. Venezuela—Narrow Neopluralism to Extensive Partisan Neocorporatism? More than any other case in the present analysis, Venezuela has been in a state of constant flux. The analysis captured the Venezuelan A-Net at a particular moment in this fast-changing transition. At the time of the snapshot when the data were collected, the Venezuelan A-Net was marked by low participation in and low coordination among associations that were generally the most disarticulated from the state among the four cases considered here. In this sense, the A-Net closely mirrored more pessimistic assessments of the new associationalism, combining both difficulties with collective action and dislocation from the state and the policy arena. This finding highlights the weakness of the popular interest regime Hugo Chávez inherited and puts his subsequent project of mobilizing popular-sector support through associations in an interesting light. The surveys were fielded before the most ambitious of the chavista initiatives were launched or reached fruition, and the fragmented A-Net they reveal points to a void that had emerged out of the dynamics of the Punto Fijo years and the subsequent breakdown of the party system. While the A-Net has undoubtedly been transformed since the surveys were fielded, understanding this demobilized and fractured starting point is crucial for thinking about how the chavista project has unfolded. The exceptionally low scope of the Venezuelan A-Net at the time of the surveys is consistent with several explanatory factors explored in the other cases, starting with a relatively late shift in the interest regime. The -Hub in Venezuela remained strong into the early s, and popular associations remained relatively peripheral. While the neighborhood association movement was fairly robust by regional standards, it had a middle-class orientation (Ellner ), and popular associations in the aggregate remained relatively marginal in the interest regime, despite their longstanding importance in some popularsector neighborhoods with particular histories of political activism, such as Enero. Political and economic dynamics may not have been conducive to a sharp increase in associational life. The Venezuelan economy was stagnant during much of the last two decades of the century, and the economic position of the popular sectors deteriorated, contributing to an uptick in social movement
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 319
,
activity. Yet this deterioration occurred gradually, and the country did not experience a precipitous economic crisis, as in Chile and Argentina, that might have provided a sharper spur to subsistence organizing. Nor were political dynamics in the previous decade especially conducive to associationalism. The long-standing constitutional and party system imploded during the s, leading to a period of instability and transition. Small parties of the left, such as the Movimiento al Socialismo and La Causa R, won some positions in local and state government during the s, and they used these platforms to experiment with participatory projects that spurred associationalism in the context of a larger tendency toward decentralization (López Maya ). Yet these initiatives were implemented on a relatively small scale. The Chávez government subsequently built on or took inspiration from these programs; as the surveys went into the field, however, the government had only recently embarked on more ambitious projects to sponsor the development of new forms of associationalism. Finally, alone among the four countries in the study, Venezuela did not undergo a regime change in the s, so that this other spur to popular organization and activation was not present. Other high profile political events, such as the caracazo of and the political conflict of –, may have catalyzed popular protest more than regularized forms of associational participation (López Maya and Lander ). It is difficult to determine the factors underlying the extraordinarily low scaling of the Venezuelan A-Net, which was the least dense among the four cases and characterized by the fewest coordinating associations. One important element may have been the dominance and subsequent collapse of “partyarchy,” in which the Punto Fijo political parties penetrated and co-opted civil society organizations (Coppedge ). To the extent that the parties acted as coordinators within the associational world, the collapse of and may have had a fragmenting effect on scaling. Further, partyarchy may have led to the disarticulation of associational networks in another way, even before the collapse: Ellner () suggests that the neighborhood movement, arguably the most coordinated set of associations in Venezuela during much of the Punto Fijo era, became so wary of encroachment that it came to eschew coordinating structures during the s, leading to increasing fragmentation. Along with these low levels of scope and scaling, the Venezuelan A-Net captured in was also highly disarticulated from the state, in a fashion similar to the Peruvian case. Once again, a lack of political and institutional continuity—specifically, the collapse of the Punto Fijo party system and the rapid changes that marked the early years of the Chávez government—may help to account for this outcome. In these circumstances, it may be that the abrupt
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 320
.
arrival of a new set of elected and appointed officials disrupted many informally routinized patterns of contacting. These discontinuities may have been exacerbated by changes in the distribution of power among levels of government, including some tendencies toward (re)centralization. It is in this context that accessing emerged as so low in the Venezuelan A-Net, with the scant degree of local accessing particularly exceptional. The initial phase of the chavista period was characterized by a less ambitious set of strategies regarding social policy and associationalism, compared to those that were eventually to follow. Although the new Bolivarian constitution of included innovative provisions for increasing citizen participation and encouraging civil society (Renaud et al. ), the major programmatic initiatives to follow through on these provisions were not launched during the initial years of the Chávez era. The antipoverty Plan Bolívar, the major but short-lived social policy targeted toward the popular sectors prior to the time of the surveys, utilized the military, rather than popular associations, as a vehicle for implementation. As a consequence, popular associations in Venezuela were among the least likely to serve in the delivery of government programs compared to other cases. More directly, the drive to promote popular associationalism was still in its infancy when the surveys were fielded. The Bolivarian Circles, associations primarily consisting of chavista partisans and predominantly located in popularsector communities, had proliferated by early when the surveys were fielded.5 Programs to establish other kinds of associations, such as Technical Water Roundtables, Self-Managing Community Organizations, and Urban Land Committees, were not yet fully developed, and other participatory initiatives launched by chavista mayors in various municipalities remained on a relatively small scale (Harnecker ). Even so, the data do seem to reflect some effects of these fledgling initiatives. The principal area where the new chavista orientation toward the A-Net was perhaps already apparent is the direct supply of state resources to associations. Changes Since . The growth of these initiatives and many more soon to follow dramatically changed the Venezuelan A-Net. When the surveys were fielded, Chávez and the had only just recently embarked on a major strategic reorientation, one that would eventually entail the most ambitious attempts 5. The number of actual Bolivarian Circles and participants is hard to know reliably, with the best estimate putting the national number of active members between , and , circa July (Hawkins and Hansen ). Participation may not have been fully picked up by the survey instrument, since they represent a unique form of explicitly partisan local organization that may not have fit respondents’ conception of the types of associations being asked about.
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 321
,
by any contemporary Latin American government to support and shape the A-Net, and to utilize it as a tool for the political mobilization of the popular sectors. In addition to further pushing the formation of associations such as the Technical Water Roundtables (see López Maya ), Health Committees, and Urban Land Committees (see García-Guadilla ), the government launched two much larger initiatives: the first a series of social programs called the Bolivarian Missions, and the second a drive to form Communal Councils in neighborhoods across the country. In combination, these initiatives were intended to build and consolidate popular-sector support for chavismo and to realize more fully the promise of a politics of inclusion and popular participation that had been written into the Bolivarian constitution of . They do so both by delivering resources to the popular sectors in a highly “branded” manner (Handlin ) and by establishing a set of chavista-affiliated associations that can receive resources directly and can be given the task of cooperating with state entities to implement social programs. While this process is still developing and varies greatly from neighborhood to neighborhood, the overall picture is one in which the government is shaping the emergence of a politicized associational infrastructure while giving these associations a privileged place within the A-Net through their access to state resources and to state entities, especially to organs of social policy delivery. Below we briefly describe both the Bolivarian Missions and the Communal Councils, and then consider how these initiatives may have reshaped the A-Net in terms of scope, scaling, access, and autonomy. The initiative of the Bolivarian Missions was launched before the Communal Councils, and its effect in reshaping the A-Net was more indirect. Early saw implementation in Caracas of a health program called Barrio Adentro, which established local health clinics staffed by Cuban doctors in low-income communities. As the opposition began the process of calling for a referendum to remove Chávez from office, the president expanded Barrio Adentro to other areas of the country and launched a series of other social programs that would be grouped under the rubric of the social “missions.” Between early and August , when Chávez survived the referendum, a total of eleven missions were launched, with eighteen more subsequently added.6 The most expansive and politically important of the Bolivarian Missions have been Barrio Adentro (technically four missions, Barrio Adentro I, II, III, and IV), a subsidized 6. Some of the later missions, like Misión Che Guevara, have replaced earlier ones, so a simple addition would be deceptive. For a useful overview of the timeline of the development of the missions, see D’Elia ().
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 322
.
food program (Misión Mercal), a program to register and give proper identification to all Venezuelans (Misión Identidad), a housing program (Misión Habitat), and a series of educational missions providing literacy training (Misión Robinson I), ongoing education at a very remedial level (Misión Robinson II), high school degree equivalency (Misión Ribas), and higher education for those unable to find places in the standard university system (Misión Sucre). The missions have primarily served the function of delivering resources to popular constituencies and were critical to Chávez’s victory in the referendum of and the rise in his levels of support afterward (Penfold-Becerra ). More pertinently for this discussion, the missions also seem to have played a role in reshaping the associational landscape. On the one hand, they firmly put the delivery of important programs at the local level under the auspices of the state, rather than outsourcing delivery to associations. On the other hand, the Chávez government has encouraged partnerships between certain missions and popular associations that are at least informally affiliated with chavismo. The degree to which these partnerships are actualized appears to vary substantially among different communities. But at least in some situations Health Committees work with Misión Barrio Adentro to make sure that programs meet local needs, Urban Land Committees work with Misión Habitat regarding the implementation of housing and urban development projects, and Communal Councils work with whatever missions are located within their territorial sphere. While in practice, longer standing neighborhood associations also may form some kind of working relationship with the missions, in general these linkages are formed overwhelmingly with associations that have been sponsored by the chavista movement. The overall effect, then, is that social policy sets up incentives for chavista-sponsored associations to form and to consolidate their roles within communities while, in aggregate, disincentivizing the formation and endurance of nonaffiliated associations that otherwise might play a similar role. An even more dramatic reshaping of the Venezuelan A-Net has occurred through the ambitious Communal Council program. This program is the centerpiece of Chavez’s attempt to create an “Explosion of Popular Power,” which he describes as one of the five “motors” of his larger project of “Socialism in the st Century.” Covering groups of two to four hundred families in urban neighborhoods or twenty to fifty families in rural areas, the councils are formed voluntarily and run by unpaid community members, but in order to be registered and recognized they must adhere to an extensive set of formal rules established in the Ley de Consejos Comunales. Their principal activities include liaising with the Bolivarian Missions, coordinating other chavista-sponsored
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 323
,
associations in the area, and designing and implementing local development projects. These projects are financed by a variety of state entities, including , a national agency with responsibility for overseeing the council program. Reliable data on expenditures are not available, but the broad consensus among analysts is that state funding is extraordinarily high. The council program has rapidly expanded due to massive publicity, incentives to form councils in order to receive resources, and substantial efforts by various national, state, and municipal entities through workshops and mesas de trabajo. Government data suggest that by early over , councils had been officially registered with the national government, with many others presumably in the process of forming. While open to all, the program has been adopted somewhat unevenly, because of the greater incentives for popularsector communities to form councils (resources are distributed roughly on the basis of need) and because of the high degree of politicization (Handlin ), so that participants are overwhelmingly likely to be chavistas (Handlin and Collier forthcoming). In sum, this program to spur associationalism is unprecedented in its scope and has characteristics reminiscent of certain dynamics of the -Hub. First is the degree to which societal and statal spheres have been blurred, with a pattern of state regulation over and penetration of these organizations of interest intermediation, somewhat suggestive of state corporatism.7 Second, whereas the general pattern is that the A-Net is relatively “flat,” in that specific categories of organizations are not privileged, the Communal Council initiative has created some semblance of a “hub” in Venezuela. The councils have become privileged actors within the Venezuelan A-Net, due to their access to resources, their roles in liaising with the Bolivarian Missions, and the emphasis placed on them by entities across levels of government. The effect of this privileged position is that many other associations—especially neighborhood associations—have found themselves marginalized or have been forced to convert themselves into Communal Councils to remain relevant actors. Finally, the council initiative is reminiscent of the -Hub in that overall these privileged actors are affiliated with Chavez’s movement and his political party, the Partido Socialista Unido de Venezuela (), with evidence suggesting that working relationships between councils and the are common. This massive program, then, along with other smaller initiatives already mentioned, has dramatically reshaped the Venezuelan A-Net. It seems clear that the scope of the Venezuelan A-Net has greatly increased. 7. Hawkins et al. () have conceptualized the councils as “quasi-governmental” entities.
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 324
.
While data strictly comparable to the surveys are not available, a question from the Latin American Public Opinion Project’s America’s Barometer on participation in community-based associations suggests that popular-sector participation in urban Venezuela is now about as high as in Peru and significantly higher than in Chile or Argentina. The scaling of the changing A-Net in Venezuela is more difficult to assess. The councils frequently work together on projects that have an impact on multiple communities, and the state is experimenting with arrangements that might facilitate coordination, from roundtables set up by local officials to broader initiatives such as the recently inaugurated Misión de Abril, which set forward a plan to form Comunas Socialistas that may be akin to federations of Communal Councils in a given parish. It remains unclear whether this plan will be realized. As discussed above, Communal Councils seem to have substantial access at the local level, as there are formal channels for requesting resources (although there remains much dissatisfaction concerning the response) and councils appear to have a privileged position in making claims more generally (Machado ). Anecdotally, some channels for access seem to exist also at the national level through and other agencies involved in conducting projects like housing that the councils oversee, but it is uncertain whether other channels are available. The apparent pattern of state dependence, given the reliance on the state for resources and the involvement of the state in helping councils form, necessarily raises questions of autonomy, and some studies have concluded that many councils have become clientelistic vehicles for local chavista power brokers (García-Guadilla ). There is thus little doubt that the character of the Venezuelan A-Net has changed dramatically. By understanding the nature of the Narrow Neopluralism captured in the present analysis, however, one can understand the space that Chávez could occupy and the dynamics of the chavista project. These initiatives are characterized by the simultaneous goals of turning the A-Net into an organizational infrastructure for the chavista movement while also spurring a more participatory democracy that reverses patterns of representational distortion and can provide greater accountability for government. There are obvious innate tensions in this combination of politicizing and participatory goals. This more “basist” A-Net is largely emerging through top-down processes and on a scale and with a degree of overt politicization that is unprecedented in Latin America. We tentatively call the new pattern one of Extensive Partisan Neocorporatism, to capture what appears to be a broad scope, a high degree of politicization, and state dependence.
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 325
,
However the chavista associational project develops, its long-term impact is unpredictable. Peru in the s and s offers an example of initiatives— not nearly as ambitious, but nonetheless significant—to build associational support from above. On the one hand, these efforts seem to have been long lasting, contributing to the extensive scope of the Peruvian interest regime. On the other hand, because of changing political dynamics, the top-down project has not reproduced a pattern of state dependence in the long run. Of the four, the Peruvian A-Net is now the most disconnected from the state. Mobilizational projects from above can thus have lasting ramifications on the nature of the A-Net but these effects are subject to political dynamics, especially in a country like Venezuela that has been characterized by so much recent political turmoil. This book has explored the nature of popular-sector problem solving in the interest arena and the characteristics of the emergent associational interest regime. Individuals participate at quite high rates in the interest arena. They engage in a variety of “direct action” strategies to address both particularistic and collective issues. For the latter, however, protest and petitioning are the most frequent approaches and are not reserved for moments of crisis. Rather, they have become quite regular components within a repertoire of action. Even more striking are the rates of associational participation, underlining the importance of associations as channels for pursuing collective interests. Although this level of activity itself is noteworthy, all types of activity—including protest—reveal a marked pattern of class distortion in which the popular sectors participate at lower rates. Popular associations are the base units of a new organizational infrastructure, conceptualized in the aggregate as the popular interest regime. We neither assert nor imply that, with the shift from the -Hub to the A-Net, these associations have become more influential than unions. Unions have lost their former place of privilege in the interest regime, and within the more heterogeneous A-Net popular associations are more numerous, attract more participation, and subjectively are more highly appraised as vehicles for representation. But unions continue to be important actors, and, more than in the -Hub, some types of unions may be networked with associations. The interrelations of these different actors within the A-Net remains a crucial question. These popular associations in the A-Net resemble neither a vibrant and autonomous civil society oriented almost exclusively toward “self-help” activities, nor a sphere of uniformly weak, ephemeral, ineffective associations unable to give voice to popular-sector interests, disarticulated from the state, and
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 326
.
unable to press demands. Rather than these two images often portrayed in the literature, the present empirical analysis along a set of dimensions demonstrates a more nuanced terrain. Generally, the A-Net, as well as individual associations within it, is marked by a diverse strategic repertoire, including both state-targeted and society-targeted strategies. Likewise, with cross-national variation, the A-Net is characterized by relatively frequent state-association interaction, not only in terms of state-targeted claim making but also in terms of society-targeted strategies, which often involve participation in government social programs as Latin American states adopt a kind of outsourcing approach to service provision. While most of these interactions appear to occur at the local level, evidence suggests that the A-Net as a whole has at least some capacity to direct demands toward the national level. In this respect, the presence in associational networks of labor unions and coordinating associations such as Nodal and Flexible Fronts appears especially critical. While the associational infrastructure may have more capacity for representation than has been suggested in more pessimistic studies, the new associationalism appears to be characterized by significant debilities. Though the A-Net has the potential for greater scope than the -Hub, that advantage generally has not been realized. Despite the existence of some degree of interassociational coordination, many associations remain atomized and quite isolated. And while associations are not completely disarticulated from the state, the looser form of scaling, lower level of resources, and contingent party ties generally afford less access at the national level. As a result, issues of critical import to the popular sectors, such as macroeconomic and labor-market policy and the broad outlines of social policy, are beyond the reach of the A-Net. Evaluating effectiveness is beyond the scope of the present study, which has been limited to mapping the organizational infrastructure of popular-sector representation and examining the existence and operation of channels for voice. Further research must address the way in which this activity may or may not translate into popular-sector influence in policy making, as such effectiveness should not be inferred from a relatively high level of individual or associational activity or any other findings in this study. Although the surveys do not allow for an evaluation of associational autonomy per se, a few preliminary comments may be made. We saw that, compared to unions, associations do not face the same problems of individual collective action and hence do not generally rely on the state to solve these problems— a reliance that was an underpinning of state corporatism in the -Hub. Nevertheless, associations may be tied to the state in other ways. Like unions, they may be the object of support-mobilizing politicians, who may promote their
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 327
,
formation. In addition, whether for top-down support mobilization or in response to bottom-up demands, governments may develop social programs that not only influence the direction of demands but may also provide an incentive for cooperative relations when these programs are implemented through popular associations. The frequency of state-association linkages, through diverse forms of dependence as well as through accessing strategies, suggests the pervasiveness of incentives that may more subtly compromise autonomy, affording the state opportunities to shape agendas and blunt political opposition. But associations do not advance an agenda in a policy or political vacuum. Rather, they respond to the policy and political situation in which they exist. State interactions are an essential part of representation, and the state is necessarily involved in the influencing the associational agenda. While it is beyond the scope of this study to observe degrees of autonomy, it has attempted to point to state-association linkages that raise these questions, which should be pursued in further research. The study has attempted to map out these general patterns in the emergence of the A-Net and to point to cross-national differences in the interest regime and its insertion into politics. An underlying concern of this project was the idea that common shifts in the economic orientation of the state, in political regimes, and in the popular interest regime might together augur a new critical juncture in Latin American politics. The -Hub was constructed in what has been described as the critical juncture of “labor incorporation” (Collier and Collier ). During that period, the projects of actors in state and society combined to create a formal and unionized working class that was given privileged representation. The -Hub was thus an institutional response to the growth of proletarianism, on terms that varied across the cases but nevertheless provided structures for the political inclusion of the formal working class. The way in which this institutional response unfolded varied substantially across countries, but it was generally characterized by a heavy state hand, often in connection with the political project of a reformist president with a transformative, modernizing agenda. The basic structures of the -Hub were thus often put into place in a relatively short period, but sometimes were constructed over a longer time and multiple presidencies. Yet once in place, they set the parameters of the political game, strongly shaping subsequent patterns of political change. A key question is whether the rise of the A-Net is part of a new critical juncture, an enduring and consequential restructuring of mass politics in response to both economic and social-structural change. While one can expect the shift in interest regime to have lasting consequences, it would seem that by its nature
09 Chapter 9.qxd
11/9/2009
11:51 AM
Page 328
.
the A-Net is more fluid than the -Hub in both its institutional structure and patterns of collective action. The A-Net, as an organizational infrastructure, is more institutionalized than the social movement model with its ebbs and flows of activity described by theorists who analyze waves of contention. Yet a number of features make it less structured and more mutable than the -Hub. Associations often have difficulty sustaining participation as well as funding, and their networked form of coordination is far less institutionalized than the hierarchies of the -Hub. They are also less regulated by the state, which does not legalize them to perform recurrent functions as it does unions (e.g., collective bargaining); when the government does partner with associations to implement programs, these relationships are not permanent, but temporary or renewable and dependent on program continuity. Party-association relationships are less common; and when they do exist, they take the form of instrumental and tenuous interactions rather than organic, institutionalized linkages. In general, the A-Net appears to be emerging gradually and more fluidly than did most instances of the -Hub. Its activities are less focused and it is less the result of initiatives from above. As we have seen, however, in many instances political elites have engaged the A-Net and have done so in different ways, in terms of support mobilization, program implementation, and/or participatory input. Indeed, in Argentina, Venezuela, and Peru, sitting presidents have in different ways pursued coalitional strategies that engage or foster popular associations and/or participatory institutions in popular-sector neighborhoods. In the former two, alliances between political parties or movements of the “left” and popular associations have emerged as integral aspects of national political dynamics, although the nature of these still emergent alliances has differed, reflecting contrasts in both leadership strategies and the organizational and mobilizational capacity of the societies. In all four countries under study, the institutional structures emerging out of this more gradual process are still incipient; the process is still ongoing. Yet at the very least, we can already discern significant differences in the contemporary orientation of the A-Net, which both reflect political realities and appear to be shaping politics in their own right. As this book went to press, the global economic crisis erupted, with unknown effects on political economic trends in Latin America. It may be that we have captured a reorganization of popular politics at the end of a neoliberal interlude, a trente inglorieuses, now coming to a close. Whatever the effects of the crisis and its aftermath on future patterns of popular participation and structures of representaton, this volume has provided a framework for further analysis of the interest arena and political dynamics more generally.
10 Appendix A.qxd
11/9/2009
11:51 AM
Page 329
APPENDIX A: SELECTION OF FOCUS DISTRICTS
.
In each capital city, eight districts were selected to focus the analysis. Selecting these “focus districts” served two functions: first, given the particular interest in popular-sector political participation, focus districts were identified to produce an oversample of individuals in nonwealthy yet diverse economic and political settings within each city; and second, these districts delimited specific geographic areas in which to concentrate the analysis of associational activity. In each country, the focus districts are territorially delimited units that, to the extent possible, also serve a political function, either as an electoral unit for legislative elections or as the seat of municipal government. Since the research project focuses on participation among the popular sectors, the first step in selecting focus districts entailed stratifying districts within each capital city based on poverty levels and eliminating the wealthiest districts from consideration. Typically, a notable discontinuity exists between the wealthiest districts and the rest, making it easy to identify and eliminate the former. The resulting “universe” of nonwealthy districts was then stratified on two variables, one economic and one political: the extent of poverty in the district and the strength of the district’s support for left or labor-based parties (s). Cut-points on each of these variables were then established to dichotomize the districts into high and low categories on each, yielding four cells in a x table (see table A.). Two districts were then selected from each cell, yielding a total of eight focus districts in each country. Within each cell, focus districts were selected based on two criteria: historical and political importance, and prior research and information about the district. The following sections describe the selection of focus districts in each country.). Table A. Schema for Selecting Focus Districts Extent of Poverty Strength of Ties with Left/
Less poor, stronger Left/
More poor, stronger Left/
Less poor, weaker Left/
More poor, weaker Left/
10 Appendix A.qxd
11/9/2009
11:51 AM
Page 330
.
Argentina The metropolitan area of Buenos Aires (Area Metropolitana de Buenos Aires, ) consists of two geographic areas: the city of Buenos Aires, the federal capital, which is an autonomous city with its own elected authorities; and Greater Buenos Aires, also known as the Conurbano or city belt, which comprises twenty-four municipalities belonging to the province of Buenos Aires. The city of Buenos Aires historically has had a higher concentration of middleclass occupants and fewer members of the popular sectors, whereas Greater Buenos Aires has had a more popular-sector demographic. To capture this variation in the metropolitan area, the Argentina research team selected four districts from the city of Buenos Aires and four districts from Greater Buenos Aires, choosing one district from the city and one district from Greater Buenos Aires for each of the four cells of stratification. While the research team utilized similar measures to assess the level of poverty and ties with the left/ in each district, the specific cut-points used to eliminate wealthy districts and stratify popular-sector districts vary for the two regions of metropolitan Buenos Aires in order to ensure diversity of focus districts within each region, as the city is wealthier and less Peronist than Greater Buenos Aires. In the city of Buenos Aires, socioeconomic data are not available for electoral districts, complicating the selection of focus districts. Instead, the research team used school districts as the unit of selection because they roughly correspond to electoral districts and socioeconomic data are available for these units. The four focus districts from the city of Buenos Aires actually consist of five school districts, as two neighboring school districts were combined to form a single focus district, because together these two school districts correspond closely to a single electoral district. Unlike the city of Buenos Aires, Greater Buenos Aires has clearly identifiable municipalities that served as the unit of selection for choosing focus districts. To measure the level of poverty in each district, the team leaders used the Unsatisfied Basic Needs index (Necesidades Básicas Insatisfechas, or ) calculated for each school district using census data.1 In the city of Buenos Aires, eleven districts were eliminated because they had poverty rates below percent as judged by the , leaving eleven districts for sampling. In Greater 1. For the city of Buenos Aires, the team used measures from the census. For Greater Buenos Aires, however, the team used census data, since the data was not available for all districts at the start of the project.
10 Appendix A.qxd
11/9/2009
11:51 AM
Page 331
,
Buenos Aires, the team eliminated districts with poverty rates below percent. This excluded two districts, leaving twenty-two districts to constitute the universe of districts in Greater Buenos Aires. To assess ties with the left/s, the team leaders utilized percent voting for the Partido Justicialista () in recent elections, using the electoral district that roughly corresponds with the school district(s) selected by the research team. For districts in the city of Buenos Aires, this measure consisted of the percent of the district voting for the in the legislative elections. In Greater Buenos Aires, ties with the left/ were measured by averaging vote for the in the legislative elections and vote for the in the presidential elections. The following cut-points were used for the city of Buenos Aires:
• •
Extent of poverty: NBI score of percent Ties to the left/: percent vote share for the
After stratifying popular-sector districts, the team leaders chose the four focus districts for the city of Buenos Aires. Figure A. displays the data. In the “more poverty, weaker left/” cell, School District XIX was selected. This school district maps onto two neighborhoods, Villa Soldati and Pompeya, and includes electoral district . Selected in the “less poverty, weaker left/” cell were School Districts XX and XIII, which map onto neighborhoods Mataderos and Liniers and electoral district . As mentioned above, these neighboring school districts were merged into one focus district for analysis. In the “more poverty, stronger left/” cell, School District IV was selected. This district more or less maps onto the neighborhood La Boca, electoral district . Finally, in the “lower poverty, stronger left/” cell, School District XXI was selected, which coincides with the neighborhood Villa Lugano and electoral district . In Greater Buenos Aires, the research team set the following cut-points for extent of poverty and ties with the left:
• •
Extent of poverty: NBI score of percent Ties to the left/: percent average vote share for the
As figure A. shows, four districts were selected among the twenty-two popular-sector districts in Greater Buenos Aires. In the “less poverty, weaker left/” cell, the district of Quilmes was selected. Levitsky’s () study of Peronist party adaptation analyzed party-organizational linkages in Quilmes, increasing the research team’s qualitative knowledge about this district. In the “more poverty, weaker left/” cell, the research team selected the district of San Miguel.
10 Appendix A.qxd
11/9/2009
11:51 AM
Page 332
.
35
School District XX1 (Villa Lugano)
% Vote for PJ
30
School District IV (La Boca) 25
School District XIV (Villa Soldati/Pompeya) School Districts XIII + XX (Mataderos/Liniers) 20
15 0
5
10
15
20
25
% Households NBI
Fig. A.. Selection of Focus Districts in City of Buenos Aires
In the “more poverty, stronger left/” cell, the district of José C. Paz was selected. José C. Paz borders San Miguel in the northern/northwestern section of the Conurbano. Until both San Miguel and José C. Paz formed part of a single district, General Sarmiento. San Miguel contained the industrialized section of General Sarmiento, and consequently industries and working-class neighborhoods were historically concentrated in this area. In contrast, the informal sectors were based in what became José C. Paz, with the result that the has historically been stronger in San Miguel than in José C. Paz. Finally, in the “less poverty, stronger left/” cell, La Matanza was chosen. Like Quilmes, La Matanza was analyzed by Levitsky (). These four focus districts reflect the geographical spread of Greater Buenos Aires, with two districts (Quilmes and La Matzanza) in the southern/southwestern part of Greater Buenos Aires and two districts (San Miguel and José C. Paz) in the northern/northwestern area.
10 Appendix A.qxd
11/9/2009
11:51 AM
Page 333
,
60
Jose C. Paz
55
% Vote for PJ
50
La Matanza
45
40
San Miguel Quilmes
35
30
25 5
10
15
20
25
30
% Households NBI
Fig. A.. Selection of Focus Districts in Greater Buenos Aires
Chile In Chile, the comuna is the unit of municipal government and was used for selecting focus districts; the eight focus districts were selected from a total of thirty-four comunas in Gran Santiago.2 To eliminate wealthier districts and stratify comunas based on the extent of poverty, the team leaders used the average household income measured through the Caracterización Socioeconómica Nacional (CASEN) household economic survey carried out in by the Instituto Nacional de Estadísticas (INE). The spatial distribution of poverty within Gran Santiago created a natural breakpoint between wealthier districts and poorer districts: twenty-eight comunas are clustered between , and , pesos in average monthly household income; the remaining six comunas range from ,, to ,, and were therefore excluded. 2. Gran Santiago comprises the thirty-two comunas of the Province of Santiago plus the adjacent comunas of Puente Alto and San Bernardo, which are bedroom communities for urban workers.
10 Appendix A.qxd
11/9/2009
11:51 AM
Page 334
.
District ties to the left/ in Santiago were measured using comuna-level electoral data for municipal elections from to , averaging the combined vote for the Communist Party and the Socialist Party in the three municipal elections that had taken place since the transition at the time of survey administration (, , ). The following cut-points were used to assign popular-sector districts to the four cells:
• •
Extent of poverty: average monthly household income of , pesos Ties to the left/: percent average Partido Socialista () and Partido Comunista Chilena h) vote share
Figure A. shows the selection of focus districts in Santiago. The cases in the “more poverty, weaker left/” cells are La Pintana and Recoleta. Under the military regime, poor people living in shantytowns in the rich district of Las Condes were relocated by the army to new housing that was created for them in the marginal southern district of La Pintana, which today is the poorest comuna in Gran Santiago. Recoleta is a district with a history of small manufacturers, which presents interesting variation in terms of its industrial profile relative to the other districts. The two are also relatively dispersed within the quadrant in terms of poverty. In the “less poverty, weaker left/” cell, the districts of Peñalolén and La Florida were selected. Peñalolén combines sectors of new middle-class housing with traditional neighborhoods that feature historically high levels of activism by the poor. Located within this district are both one of the oldest land invasion neighborhoods and one of the newest. La Florida is a bedroom community that features a growing middle class. In the “more poverty, stronger left/” cell, the districts of Pudahuel and Pedro Aguirre Cerda were selected. Pudahuel is notable for the high level of electoral support it has given to the Socialist Party since the transition. In this respect, it forms an interesting contrast with Pedro Aguirre Cerda, another district that scores high on the left-vote variable, but which concentrates most of this support on the Communist Party. Pedro Aguirre Cerda is a newer district, created in from parts of San Miguel, La Cisterna, and Santiago Centro; the district experienced a percent population decline between and . In the “less poverty, stronger left/” cell, the districts of Huechuraba and Cerrillos were chosen. Huechuraba is a northern district that is undergoing significant change; it includes new sectors of significant wealth (quasi-gated communities), more traditional popular-sector areas, and sectors of small business and production. Cerrillos is a more traditional district and is one of the few remaining comunas in Santiago with an active industrial presence.
10 Appendix A.qxd
11/9/2009
11:51 AM
Page 335
40
,
Pudahuel
35
% Vote for PCCh+PS
30
Pedro Aguirre Cerda
25
Huechuraba 20
Cerrillos 15
Recoleta
La Florida
La Pintana Peñalolén
10
5 200
250
300
350
400
450
500
550
600
650
Average Monthly Household Income (000s Chilean pesos)
Fig. A.. Selection of Focus Districts in Santiago
In addition to displaying variation on variables of secondary interest such as industrial profile, this sample is evenly dispersed geographically throughout Gran Santiago. Among the eight focus districts, two are in the north (Huechuraba and Recoleta), one is in the west (Pudahuel), two are in the southwest (Cerrillos and Pedro Aguirre Cerda), one is in the south (La Pintana), and two are in the southeast (Peñalolén and La Florida). Peru In Peru the distrito is the unit of municipal government,3 and eight focus districts were selected from thirty-seven total distritos.4 To measure the extent of 3. In addition to the municipal level of government, Lima also has a metropolitan government above the municipalities. 4. The sampling universe also excludes small outlying districts, most of which are very small beach towns in Lima’s southern cone.
10 Appendix A.qxd
11/9/2009
11:51 AM
Page 336
.
poverty, researchers in Peru often use a socioeconomic scale ranging from A through E (with A being the richest and E being the poorest) to classify individuals, households, neighborhood blocks, and districts. Based on data supplied by the public opinion firm Apoyo S.A., the Peru research team constructed a measure of the extent of poverty for each district by adding the percentage of blocks in each district that correspond to socioeconomic levels D and E. Figure A. shows a natural break point between richer districts and poorer districts.5 The sampling universe includes only those fourteen districts with higher percentages of blocks at the lowest socioeconomic levels—that is, those with percent or more. District ties to the left/ in Lima are measured using electoral data between and . After the transition to democracy, the political left in Peru gained considerable electoral momentum and consistently participated in elections until the end of the decade, when the party system collapsed. Therefore, the years in which electoral support for the left can be measured are between and (the period between the transition to democracy and Alberto Fujimori’s more authoritarian regime).6 The measure used for this project is the average vote for the main leftist political bloc, Izquierda Unida (), in municipal elections in , , , and . The following cut-points were used to divide the popular-sector districts into four cells:
• •
Extent of poverty: percent of the blocks in the district categorized at socioeconomic levels D and E Ties with the left/: percent average vote share for in the four municipal elections
In the “more poverty, weaker left/” cells, Chorrillos and San Juan de Miraflores were selected as the only two cases in that cell. In the “less poverty, weaker left/” cell, the districts of Lima-Cercado and La Victoria were selected. Lima-Cercado was included in Henry Dietz’s () study of political participation in Lima, which also used survey research. Both Lima-Cercado and La Victoria are important industrial and commercial centers for the metropolitan area. 5. The richer districts cluster between percent (Miraflores, San Isidro) and percent (La Molina) of blocks at socioeconomic levels D and E. The popular class districts range between percent (Lima-Cercado) and percent (Carabayllo) of blocks at socioeconomic levels D and E. 6. Municipal-level electoral data is not available for the s because Fujimori abolished districtlevel elections during his presidency.
10 Appendix A.qxd
11/9/2009
11:51 AM
Page 337
,
50
Villa El Salvador Independencia
45
El Agustino
% Vo Vote te for IU
40
San Martín de Porres
35
San Juan de Miraflores 30
Lima-Cercado
25
Chorrillos
La Victoria
20 30
35
40
45
50
55
60
65
70
75
80
% Blocks at Socioeconomic Level D+E
Fig. A.. Selection of Focus Districts in Lima
The districts of San Martín de Porres and Independencia were chosen for the “less poverty, stronger left/” cell. Like Lima-Cercado, San Martín de Porres is an important industrial and commercial center of the city and also houses the largest number of informal-sector workers in the city. Independencia was chosen because other scholars interested in similar themes have studied it (e.g., Stokes ). In the “more poverty, stronger left/” cell, the districts of Villa El Salvador and El Agustino were selected. Villa El Salvador is perhaps one of the most studied districts of Lima and has a close history with the left. El Agustino was selected because it has an interesting background as a strong leftist district in the s while also being a strong Fujimorista district in the s, and because of the many existing studies of its political and social characteristics.
10 Appendix A.qxd
11/9/2009
11:51 AM
Page 338
.
Venezuela In Caracas, the eight focus districts were selected from a total of thirty-one municipal districts, known as parroquias.7 The research team constructed an index to assess the extent of poverty within each district, generated from three census variables: illiteracy rate, unemployment rate, and “rancho” rate, which refers to the proportion of all residences at a slum or squatter level of shelter quality.8 The three variables were compiled into an equally weighted, additive index using the Z-scores for each of the three variables. Consequently, the index ranged from - to +. A parroquia with a score of is approximately average for Caracas on the three variables; one with a score of - is relatively wealthy; and a parroquia with a score of + is relatively poor. The threshold for popular-sector districts was set at -. on the poverty index, eliminating four districts from the potential universe of popular-sector districts and leaving twenty-seven popular-sector districts. The measure of ties with the left/ is based on the percentage of votes for La Causa R in the municipal elections. Since Venezuela has a mixed proportional representation system, the relevant measure is the percentage of the population voting for the individual La Causa R candidates in these elections.9 The research team decided to use the vote for La Causa R during this single election as a measure of ties with the left/ due in part to the historical weakness of the left in Venezuela. Under the previous party system, there was no viable leftist party active at the local level. However, during the late s and the s, La Causa R arose as a leftist electoral challenger to the traditional parties, particularly at the municipal level. Consequently, La Causa R represents the pre-Chávez face of the left in Venezuela. The research team established the 7. Two parroquias were excluded from the analysis due to their semirural character. While these parroquias are technically part of Caracas, their overall economic and social profile are sufficiently dissimilar to the other parroquias to merit exclusion. Furthermore, these districts have very small populations and do not reflect a significant portion of the population of Caracas. There was a change in parroquia boundaries in the late s. The research team adjusted for this by calculating weighted average scores for the new parroquias using the scores from the previous parroquias multiplied by the percentage of their area that is composed by each of the old parroquias. While this weighted average results in measurement error, this was the most systematic and reasonable way to calculate values for the new parroquias. 8. The “rancho” rate refers to the quality of housing within the district but does not make reference to legal status or ownership of the housing. 9. There were about seventeen congressional districts in metropolitan Caracas for that election, and results were available only for the level of the district. Districts often include more than one parroquia, and sometimes split parroquias between them. For the cases in which a parroquia was split among more than one electoral district, the research team assigned the parroquia the average of the different districts to which it belonged.
10 Appendix A.qxd
11/9/2009
11:51 AM
Page 339
,
following cut-points to divide the twenty-seven popular-sector districts into the four quadrants:
• •
Extent of poverty: . on the poverty index Historic ties with the left/: percent vote share for La Causa R
In the “more poverty, weaker left/” cell, the parroquias of La Vega and San Agustín were selected as the only two districts in this cell. La Vega consists of primarily self-constructed squatter houses on relatively steep hillside terrain. The neighborhood is known for its dependence on government antipoverty programs, high levels of unemployment, an almost entirely informal economy, and high levels of crime. In contrast, San Agustín is a more-established area of privately built apartment buildings, and private homes. While this district once housed working-class participants in Caracas’s formal economy, it is now characterized by relatively high reliance on the informal sector. The districts of Candelaria and Petare were selected for the “less poverty, weaker left/” cell. Candelaria has a traditionally working-class and lowermiddle-class character, although the overall state of the district has decayed noticeably in the last decade. It is one of the few parroquias in Caracas with a substantial population of workers who used to be involved in the formal economy and therefore served as an important district for this study. Petare was selected since it has a population of over one million people, or around onequarter of Caracas’s total population, making it a politically, economically, and socially important district. In the “more poverty, stronger left/” cell, the districts of de Enero and Sucre were selected. Many of the residents of de Enero live in a large complex of apartment buildings constructed during the Pérez Jiménez dictatorship in the s. Like most of the popular-sector districts in Caracas, de Enero is dominated by informal-sector employment. In recent years, this parroquia has also emerged as an exceptionally strong bastion of chavismo. Like Petare, Sucre is home to over a million residents and houses one-quarter of the city’s population and therefore was selected due to its size and importance. Finally, in the “less poverty, stronger left/” cell, the districts of Antímano and San Juan were selected. Residences in Antímano primarily consist of selfconstructed squatter homes on the hillside. However, this parroquia also has a formal, industrial sector in the section of the district nearest to the downtown. Like La Vega, Antímano is known for its poverty, its reliance on government funds, and its high levels of crime. The parroquia of San Juan is home to mix
10 Appendix A.qxd
11/9/2009
11:51 AM
Page 340
.
Sucre
43
23 de Enero
Antímano 38
%V Vote ote for L LCR CR
San Juan
San Agustín
33
La Vega 28
Candelaria 23
Petare 18 -2.5
-2
-1.5
-1
-0.5
0
0.5
1
1.5
2
2.5
Weighted Poverty Index
Fig. A.. Selection of Focus Districts in Caracas
of poorer, working-class, and middle-class inhabitants living in large, privately owned buildings. While much of San Juan’s population work in the informal sector, this parroquia also contains part of Caracas’s shopping and business center. Compared with the other focus districts in Caracas, San Juan has less of a reputation for high crime rates or for heavy reliance on government funds. In addition to attaining variation on degree of formal employment and other variables, this sample also covers a diversity of geographic locations within the city of Caracas, with one parroquia in each cell falling relatively close to downtown Caracas ( de Enero, San Agustín, San Juan, Candelaria) and one farther out from the city center (La Vega, Petare, Sucre, Antímano). This geographic
10 Appendix A.qxd
11/9/2009
11:51 AM
Page 341
,
distribution is particularly important in Caracas, since many problems facing poor communities depend on proximity to the central city. In the s and s the state spent lavishly on infrastructure in the central city, so infrastructural issues such as transportation or public services should be much less relevant for centrally located parroquias than for those in more peripheral zones.
10 Appendix A.qxd
11/9/2009
11:51 AM
Page 342
11 Appendix B.qxd
11/9/2009
11:52 AM
Page 343
APPENDIX B: SURVEY OF ASSOCIATIONS
. The survey of associations was carried out between and in Lima, Peru; Santiago, Chile; Buenos Aires, Argentina; and Caracas, Venezuela by local research teams headed and closely supervised by country coordinators from the University of California, Berkeley.1 Thirty associations were interviewed in each of the eight focus districts, yielding a sample of approximately associations in each country and a total sample of approximately associations.2 Selection of associations was based on a chain-referral procedure designed to yield a theoretical minimum of starting points (and snowballs) in each city, but in fact consisting of many more snowballs due to a lack of linkages and referrals in some chains. The target population for the survey of associations consisted of leaders of neighborhood, district, regional, or national associations with offices or activities in the focus districts. Associational leaders were interviewed as respondents. For associations with official leadership positions, the research team attempted to interview the association president, but when the president was unavailable the team interviewed another official (typically a vice president or member of the board of directors) with experience in and substantial knowledge of the association. A popular association was defined as a group with a collective name that undertakes regular or semiregular activities that either directly involve or address popular-sector issues and concerns. Associations did not have to be formally registered with the state in order to be included in the study; instead, the level of formalization was included as part of the survey. Two points about the sampling procedure should be noted. First, a chainreferral sample is likely to produce a selection of associations that are more networked compared to a random sample. Second, the starting points of each chain were to some degree selected based on reputation, thus potentially producing a sample of relatively more active, visible, and/or effective associations. Thus the sample departed from a representative sample of all associations based on random selection. Drawing such a sample was not possible since no reliable registries existed from which to determine a population and sample 1. Pre-testing of the survey instrument was carried out in each country during the year prior to the final survey administration. 2. While the goal was to collect interviews in each country, in Chile some interviews were duplicates or otherwise unusable, and hence the Chilean sample consisted of associations.
11 Appendix B.qxd
11/9/2009
11:52 AM
Page 344
.
randomly. More importantly, obtaining a representative sample was not our goal and does not serve our analytic aims. Our analytic goal is not to generalize about the larger universe of associations (thereby including a representative proportion that are small, ephemeral, and/or ineffective) but to draw some conclusions about the organizational infrastructure that may be available for problem solving and advancing interests. Just as an inquiry about the -Hub might place greater weight on the more active unions and confederations rather than seek to generalize about all unions, so an inquiry about the A-Net places greater weight on the more active and influential associations. Hence, the probable biases in the sample relative to a random sample capture the orientation and analytical goals of this research. The sample, then, is not designed to include the entire diversity of associations, nor is it designed to capture complete networks. Rather, it seeks to steer a middle ground between the two goals of exploring the more effective and linked core segment of the A-Net and capturing the substantial heterogeneity of the popular interest regime. Given the heterogeneous nature of the referrals and the fact that the resulting samples yielded many more than the minimum of twenty-four snowballs in each country, this sample displays great diversity along many dimensions and even captured a large number of isolated associations that do not coordinate their activities with others. Nevertheless, the sample is predisposed toward what we may call the “core segment” of the A-Net: those associations that are more prominent, active, and linked.
Sample Design Three associations in each district were selected to serve as entry points for the chain-referral sample, each reflecting a different type of association. Since this project aims to capture the face of the A-Net in each country, the types of associations that served as starting points in the referral chains differ somewhat across the four countries. In each country, the starting points for each district included one neighborhood association and one community organization, such as a soup kitchen or a mothers’ group. For the third starting point, the country coordinators used their qualitative knowledge to select one additional type of association to capture the most significant forms of associational activity in that district. This sampling choice for initiating the snowballs rests on a context-specific reason for establishing some equivalence across national cases on the basis of the importance of the type of association. It renders inappropriate cross-national
11 Appendix B.qxd
11/9/2009
11:52 AM
Page 345
,
comparisons of the distribution of associations with a particular substantive focus (e.g., communal kitchens). Nevertheless, as described below, the same guidelines were used for the chain-referral procedure across the four countries, so that the sample provides a reasonable basis (indeed a uniquely rich, if imperfect basis) for analyzing and comparing the core segment of the A-Net, including coordination and scaling within the associational world. The initial associations in each chain provided names and contact information for other associations with which they worked. From the list of contacts provided by each association, the research teams selected the associations to interview next. The research team similarly asked all successive associations for lists of contacts, and chose among them for the associations to interview in the next stage of the referral chain. This process continued, resulting in a maximum of ten associations interviewed in each chain, or three ten-link chains in each district. It also produced a much longer list of linked associations that were not interviewed. If the research team could not complete ten interviews for that referral chain, they began a new chain in the same district. In pursuing the chain-referral procedure, the following guidelines were used:
•
•
•
Respondents were asked, across several questions, to list associations with which their association cooperated in a variety of ways. Associations were chosen from among those listed as associations that do the same kind of work, religiously based associations other than churches, neighborhood associations, professional associations other than traditional unions, associations of the unemployed, , and federations. The research teams did not interview: political parties, firms/businesses, foundations, charities, unions (with the exception of informal unions or those not registered under the labor law), universities, government agencies, foreign governments, churches (though church-based associations active in community issues were interviewed), organizations for which contact information was lacking, or organizations that were no longer active. In most cases, fora or “spaces” (the word used to describe inter-associational groups like “mesas de concertación”) were interviewed only if they had an organizational component.
The project team developed a systematic procedure for following the chainreferral procedure in each focus district:
•
Scaling is a main analytic concern of the project. Consequently, links to federations or coordinating associations were followed if available. These
11 Appendix B.qxd
11/9/2009
11:52 AM
•
•
Page 346
.
included links to metropolitan or national federations and whose offices were outside the district in which the contact was produced. When interviews in different districts led to the same federation, the contact was pursued by only one interviewer. With the exception of coordinating associations mentioned above, contacts within the district were prioritized. Once a snowball left the district, only links to coordinators were followed. If these ran out before ten interviews were obtained, another snowball was started. The research teams prioritized interviewing diverse types of organizations generated by the same contact within any one snowball. The research teams prioritized against interviewing organizations with which the original organization coordinated only marginally.
Argentina In Buenos Aires the survey of associations was fielded between June and October . The survey was coordinated by Candelaria Garay and implemented by a local project coordinator and a team of eight university students (most of them graduate students) and two interviewers of state-run household surveys. For each of the eight focus districts, the research team identified the following start points for the survey chain:
•
•
•
One neighborhood association (sociedad vecinal or sociedad de fomento). In Buenos Aires, neighborhood associations are the “prototypical” territorial group and often tend to be as old as the neighborhood. These associations emerged to solve infrastructure and service provision challenges at the time when the neighborhood was settled. One community association, such as a day-care co-op, a soup kitchen, or a community center. Community associations tend to be newer phenomena than neighborhood associations and come in two types: those that emerged largely to solve consumption issues and meet service delivery needs for the area; and recreational associations that today tend to meet both recreational and other community needs, such as providing daycare or school support. One unemployed workers organization. Unemployed workers associations were chosen because they constitute a fairly new and important phenomenon that arose with the implementation of free market reforms in the s. If no unemployed workers association could be found in the district, an was selected as the starting point.
11 Appendix B.qxd
11/9/2009
11:52 AM
Page 347
,
To identify the three initial starting points for each survey chain, the research team used two primary methods. First, the team identified some associations by examining lists or databases of civil society associations in Buenos Aires. When existing lists or databases failed to provide three starting points for each district, the research team relied on personal contacts with individuals or organizations that operate in the focus districts. Alternately, the country coordinator contacted , universities, or coordinating bodies (such as a network of day-care centers) to ask for references for potential associations that could serve as the starting point for the survey chain. The survey of associations yielded interviews in Buenos Aires. Chile In Santiago the survey of associations was fielded between July and December . The survey was coordinated by Christopher Cardona and Diana Kapiszewski and administered by a local coordinator and fifteen Chilean university students. In each of the eight focus districts, the research team identified three start points for the survey chain:
• • •
One neighborhood organization (junta de vecinos). Juntas de vecinos are registered with the municipal government, making them relatively easy to access. One community organization dealing with economic development and/ or work issues (such as an employment training center, cooperative, or association promoting infrastructural improvement). One community organization working on some other issue of particular importance in the comuna in question. The particular profile of the organization chosen as the entry point of this type differed among comunas.
To identify these three initial starting points for the survey chain, the research team used a variety of methods. The initial associations were selected through personal contacts of a member of the research team, through contacts from local governments or coordinating bodies of other associations, or through randomly selecting an entity from an existing directory. In Santiago the referral chains resulted in a total of interviews. Peru In Lima the survey of associations was fielded between June and August . The survey was coordinated by Sally Roever, supplemented by a local advisor,
11 Appendix B.qxd
11/9/2009
11:52 AM
Page 348
.
a field coordinator, and eight interviewers who were upper-level students or recent graduates of the Pontífica Universidad Católica del Perú. In each of the eight focus districts, the research team identified three starting points for the survey chain:
• • •
One neighborhood association (associación de vecinos). One community organization (organización social de base), typically a comedor popular (communal kitchen) or club de madres (mothers’ club.) One . Peru has a large sector, making investigation of these associations particularly pertinent.
These associations were identified based on the research team’s knowledge of the districts, or on the basis of personal contacts of members of the research team who were asked to suggest the names of associations that could serve as entry points. The survey of associations yielded a total of interviews. Venezuela In Caracas the survey of associations was fielded between January and June . The survey was coordinated by Jason Seawright and implemented by a local field coordinator and fourteen university students from the Universidad Central de Venezuela and the Universidad Simon Bolívar. In each of the eight focus districts, the research team identified three starting points for the survey chain:
• • •
One neighborhood association (associación de vecinos). One community organization, such as a women’s group or an association mobilized around a specific local issue. One additional community organization, preferably one working with a Bolivarian Circle, intended to highlight the associations that were either created after the rise of Hugo Chávez or that were integrated into the chavista movement.
In contrast to the experience conducting the survey of associations in the other three capital cities, in Caracas the research team faced significant obstacles in extending the survey chain much beyond the initial starting points. This difficulty suggests the overall fragmentation and lack of scaling of the A-Net in Caracas. To address this difficulty, the research team continued generating new chains
11 Appendix B.qxd
11/9/2009
11:52 AM
Page 349
,
when the initial chain was exhausted. The team accomplished this by returning to the original list of all referred associations identified by the initial association, interviewing other associations, and, when necessary, beginning new starting points if the original chains were completely exhausted. While the research team used the general system to select linkages from those recommended by the initial association, the team adapted to the aforementioned difficulties by interviewing each association that they could reach in the snowball. Although the research team was not able to extend twenty-four chains of ten linkages, they did collect interviews through the strategy of reviving old clusters and starting new referral chains when necessary.
Instrument for Survey of Associations The survey instrument consisted of questions. Given that many were followup questions contingent on previous answers, no respondent answered each question. In the first section, the respondents were asked to identify his or her type of association and what sorts of problems it addresses. Next, the survey inquired about the strategies the association employs to solve problems, and what political or governmental entities it contacts to do so. The third section requested that the leader describe the type of people for whom this organization works (or those who benefit from its work), and how those individuals are integrated into the work of the organization. In the following section, leaders were asked to indicate the other entities or associations with which they coordinate in order to solve problems. In the fifth section, leaders were asked to identify any umbrella groups in which the organization participates, along with other external linkages. The final two sections included questions to determine the organizational and financial resources of the organization and information regarding its founding. The survey of associations was implemented in Venezuela during the first six months of , which followed the coup attempt against Chávez’s government in April and overlapped with the latter portion of the general strike conducted from December to February . To ensure that responses did not reflect mobilization around this conflict, the survey asked respondents about their participation in the events surrounding these crises, and then asked respondents to answer regarding their other activities during the previous five years, to capture patterns during “normal” rather than crisis times.
11 Appendix B.qxd
11/9/2009
11:52 AM
Page 350
12 Appendix C.qxd
11/9/2009
11:52 AM
Page 351
APPENDIX C: SURVEY OF INDIVIDUALS
. An original survey of individuals was carried out between and by local survey firms supervised by the research team leaders in Buenos Aires, Argentina; Caracas, Venezuela; Lima, Peru; and Santiago, Chile.1
Sample Design The survey was designed in two parts: a random sample of individuals throughout the entire metropolitan area, and an additional sample of individuals in each of eight focus districts, or an oversample of in the focus districts, thus yielding a total of , per city. In the analysis of this data, sampling weights were utilized to adjust for oversampling of respondents within focus districts, allowing for unbiased estimates of population parameters. In Argentina and Chile both the citywide sample and focus district oversamples were selected using equality probability sampling. In Argentina the sampling firm randomly selected small census units known as radios censales. There was no clustering or stratification in the selection of these radios censales, which were selected based on simple chance proportional to population size. The sampling firm selected , radios censales to produce a final sample of ,. Within each radio censal, one block was randomly selected, then one corner within this block, and finally one house in relation to the corner. In Chile the sampling firm selected , square city blocks to yield the final sample of ,. These square blocks were randomly selected from all blocks in the city after weighting these blocks based on population size. Within each square block, one house was randomly selected based on a registry produced from census data. In Peru the sampling firm randomly selected , city blocks from its computer database, as well as a direction (north, south, east, or west), corner, and door for each block to yield a final sample of , respondents. In Venezuela the sampling firms utilized cluster sampling and then randomly sampled within the cluster. An initial random selection of voting centers was made from the 1. A pre-test of the survey instrument, slightly adapted in each country to account for local idioms, was implemented in each of the four cities a year before carrying out the final survey of individuals.
12 Appendix C.qxd
11/9/2009
11:52 AM
Page 352
.
National Electoral Council (Consejo Nacional Electoral) database for the metropolitan area of Caracas, starting with eighty voting centers for the citywide survey and nine additional voting centers for each focus district (yielding a total of seventy-two voting centers for the oversample). For each voting center, the sampling firm then selected twenty residences randomly from the four square city blocks surrounding the voting center, with a response rate yielding a final sample of ten respondents per voting center. In each of the four countries, after selecting the sampled house the interviewer enumerated all residents of that household and selected the adult with the most recent birthday to be the respondent. Survey nonresponse may be one possible source of bias, as is the use of replacements for nonrespondents. In view of the pattern of low response rates in Latin American surveys, interviewers in Argentina, Chile, and Peru revisited each potential respondent who was not available to complete the survey using a combination of conversion letters and in-person visits. In Argentina and Chile the interviewers attempted to contact respondents at least five times before substituting another respondent; in Peru the interviewers made at least ten attempts. Response rates varied in these three countries, with somewhat low response rates in Peru (. percent) and rather high response rates in Argentina (. percent) and Chile (. percent). This process differed in Venezuela, where the interviewers revisited the potential respondents repeatedly until they achieved a percent response rate in each voting center and thus a total response rate of percent.
Construction of Sampling Weights In each capital city the sampling weights were constructed by dividing the population of each district by the number of respondents sampled from each district (see the tables below for district-level data and weights for each of the four capital cities). For purposes of data analysis, however, sampling weight ratios, rather than the sampling weights themselves, were used, in order to give an effective sample size that is the same as the size of the original sample itself. Two sets of ratios were utilized: one for use with the pooled cross-national dataset, and one for analysis within each country. With respect to analysis of the whole dataset (across countries), the sampling weight ratio was constructed by dividing each sampling weight by the average weight (of ,.) in the whole dataset. Thus, respondents with this average weight received a score of on the sampling weight ratio variable; respondents with above-average weights
12 Appendix C.qxd
11/9/2009
11:52 AM
Page 353
,
(that is, those from districts that were undersampled due to the sample design) have weight ratios greater than ; and respondents with below-average weights (that is, those from districts that were oversampled due to the sample design) have weight ratios less than . For purposes of country-specific analysis, an analogous procedure was followed, but for each country, sampling weights were divided by the withincountry average weight to create a country-specific weight ratio for each respondent. Sampling weight ratios were implemented using the weighting routine in SPSS. Tables C.– contain data on the total adult population, number of respondents sampled, and sampling weight assigned to each focus district in Buenos Aires, Santiago, Lima, and Caracas. In Buenos Aires the survey of individuals was implemented from August to September by the survey firm Gallup. The sampling weight is (B)/ (C). So, for instance, each respondent in Distrito Escolar IV (first row) “represents” . people in the population, whereas a respondent from a nonfocus district in Buenos Aires (penultimate row) “represents” ,. people in the population. The survey of individuals in Santiago was implemented from August to October by the survey firm Statcom Estadísticos Consultores Limitada. The Peruvian survey firm, Opinión y Mercado implemented the survey of individuals in Lima from August to November . In Caracas the survey of individuals was implemented from February to May, by the Venezuelan survey firm , of Alfredo Torres Uribe and Associates. Table C. Construction of Sampling Weights for Buenos Aires (A) District Distrito Escolar IV Distrito Escolar XIII + XX Distrito Escolar XIX Distrito Escolar XXI José C. Paz Matanza Quilmes San Miguel Rest of Metropolitan BA Total, Metropolitan BA
(B) Population (18+) 63,286 151,560 53,204 68,516 138,146 815,886 349,270 160,547 6,489,383 8,289,798
(C) Number Sampled
(D) Sampling Weight
92 103 92 95 91 182 126 92 609 1482
688.891:1 1,471.456:1 578.304:1 721.221:1 1,518.088:1 4,482.890:1 2,771.984:1 1,745.076:1 10,655.801:1 —
12 Appendix C.qxd
11/9/2009
11:52 AM
Page 354
.
Table C. Construction of Sampling Weights for Santiago (A) District Cerrillos Huechuraba La Florida La Pintana Pedro Aguirre Cerda Peñalolén Pudahuel Recoleta Rest of Greater Santiago Total, Greater Santiago
(B) Population (18+)
(C) Number Sampled
(D) Sampling Weight
96 95 139 111 102 116 112 108 601 1480
529.625:1 522.842:1 1,859.856:1 1,109.090:1 810.422:1 1,259.862:1 1,140.446:1 995.176:1 4,750.784:1 —
(C) Number Sampled
(D) Sampling Weight
99 98 101 96 113 119 119 127 612 1484
3,317.758:1 2,742.418:1 1,663.574:1 2,201.573:1 1,980.540:1 3,020.361:1 4,095.495:1 2,555.811 8,622.904 —
(C) Number Sampled
(D) Sampling Weight
115 95 115 285 95 115 185 105 370 1480
1,187.365:1 608.589:1 1,150.2:1 2,796.246:1 409.305:1 960.052:1 2124.73:1 780.914:1 3,986.316:1 —
50,844 49,670 258,520 123,109 82,663 146,144 127,730 107,479 2,855,221 3,801,380
Table C. Construction of Sampling Weights for Lima (A) District Lima Cercado Chorrillos El Agustino La Victoria Independencia San Juan de Miraflores San Martín de Porres Villa El Salvador Rest of Metropolitan Lima Total, Metropolitan Lima
(B) Population (18+) 328,458 268,757 168,021 211,351 223,801 359,423 487,364 324,588 5,277,217 7,648,980
Table C. Construction of Sampling Weights for Caracas (A) District Antímano Candelaria La Vega Petare San Agustín San Juan Sucre Veintitrés de Enero Rest of Metropolitan Caracas Total, Metropolitan Caracas
(B) Population (18+) 45,322 136,547 132,273 796,930 38,884 110,406 393,075 81,996 1,474,937 3,222,864
12 Appendix C.qxd
11/9/2009
11:52 AM
Page 355
,
Survey of Individuals Instrument Survey questions focused, inter alia, on participation in activities of popular associations; the kind and nature of “direct action” participatory activities (see chapters and ); experiences with work and employment as well as workrelated political participation; participation in partisan and electoral politics; and sociodemographics. While the survey of individuals contained the same core set of questions in the four countries, in two countries adjustments were made to the questionnaires to adjust for recent episodes of political turmoil. At the moment the surveys were fielded, these countries had experienced important political crises in the very recent past: Venezuela, with the attempted coup against Hugo Chávez in April and the general strike from December to February ; and Argentina, with the devaluation of the peso, a presidential succession crisis, and the freezing of bank deposits (corralito) in December and January . It was therefore important to distinguish patterns of protests, petitions, or other forms of participation related to these political crises from patterns during more “normal,” noncrisis times. The survey asked respondents about their participation in the events surrounding these crises, and then asked respondents to describe their other political participation, exclusive of these actions, during the five-year period prior to the interview.
12 Appendix C.qxd
11/9/2009
11:52 AM
Page 356
13 Contributors.qxd
11/9/2009
11:52 AM
Page 357
CONTRIBUTORS
. Ruth Berins Collier is Heller Professor of Political Science at the University of California, Berkeley. Her research has focused on the interplay of regime change and forms of popular participation and has included comparative analyses of Latin America, Africa, and Europe. Earlier books are Regimes in Tropical Africa: Changing Forms of Supremacy, -; Shaping the Political Arena: Critical Junctures, the Labor Movement, and Regime Dynamics; The Contradictory Alliance: State-Labor Relations and Regime Change in Mexico; and Paths Toward Democracy: The Working Class and Elites in Western Europe and South America. Thad Dunning is Associate Professor of Political Science at Yale University and a Research Fellow at Yale’s Whitney and Betty MacMillan Center for International and Area Studies. His book, Crude Democracy: Natural Resource Wealth and Political Regimes, was published in by Cambridge University Press; other work has appeared in International Organization, The Journal of Conflict Resolution, Political Analysis, Studies in Comparative International Development, and other journals. Dunning’s research interests lie in comparative politics, political economy, Latin American politics, and methodology. Candelaria Garay is a PhD candidate in the Political Science Department at the University of California, Berkeley. Her research interests include comparative social policy, collective action, and party politics. She is writing a dissertation that seeks to explain the different configurations of social policy expansion to those outside the formal sector in Latin America. Her work has appeared in Politics & Society. Samuel Handlin is a PhD candidate in the Department of Political Science at the University of California, Berkeley. He is writing a dissertation about the use of social policy and linkages with popular associations to win the votes of the poor and transform electoral coalitions during South America’s “left turn.” Diana Kapiszewski, who holds a PhD in political science from the University of California, Berkeley (), is Assistant Professor of Political Science at the University of California, Irvine. Her research interests include the role of courts in politics in Latin America and comparative constitutional politics. She recently co-authored “Doing Courts Justice: Studying Judicial Politics in Latin
13 Contributors.qxd
11/9/2009
11:52 AM
.
Page 358
America” (Perspectives on Politics, ), and edited an Encyclopedia of Latin American Politics (Greenwood, ). Jason Seawright is Assistant Professor in the Department of Political Science at Northwestern University. His research interests include political behavior, parties and party systems, and political methodology. His work has been published in journals including Political Analysis, Studies in Comparative International Development, and Political Research Quarterly.
14 References.qxd
11/9/2009
11:53 AM
Page 359
REFERENCES
. Abers, Rebecca. . Inventing Local Democracy: Grassroots Politics in Brazil. Boulder, Colo.: Lynne Rienner. Abers, Rebecca, and Margaret Keck. . “Muddy Waters: The Political Construction of Deliberative River Basin Governance in Brazil.” International Journal of Urban and Regional Research (): –. Adams, Jacqueline. . “Art in Social Movements: Shantytown Women’s Protest in Pinochet’s Chile.” Sociological Forum (): –. ———. . “The Imagination and Social Life.” Qualitative Sociology (): –. Aguilar, Edwin E., and Alexander C. Pacek. . “Macroeconomic Conditions, Voter Turnout, and the Working-Class/Economically Disadvantaged Party Vote in Developing Countries.” Comparative Political Studies (): –. Alvarez, Sonia. . “‘Deepening Democracy’: Social Networks, Constitutional Reform, and Radical Urban Regimes in Contemporary Brazil.” In Mobilizing the Community: Local Politics in a Global Era, edited by Robert Fischer and Joe Kling. Beverly Hills, Calif.: Sage. ———. . “Latin American Feminisms ‘Go Global’: Trends of the s and Challenges for the New Millenium.” In Cultures of Politics, Politics of Culture: Re-Visioning Latin American Social Movements, edited by Sonia Alvarez et al. Boulder, Colo.: Westview. Alvarez, Sonia, Evelina Dagnino, and Arturo Escobar. . “Introduction: The Cultural and the Political in Latin American Social Movements Research.” In Culture of Politics, Politics of Culture: Revisioning Latin American Social Movements, edited by Sonia Alvarez et al. Boulder, Colo.: Westview. Angell, Alan. . “Unions and Workers in Chile During the s.” In The Struggle for Democracy in Chile, edited by Paul W. Drake and Iván Jaksic. Lincoln: University of Nebraska Press. Anheier, Helmut K., and Lester M. Salamon, eds. . The Non-Profit Sector in the Developing World: A Comparative Analysis. Baltimore: Johns Hopkins Center for Civil Society Studies. Ansell, Christopher. . “Symbolic Networks: The Realignment of the French Working Class, –.” American Journal of Sociology (): –. ———. .“Community Embeddedness and Collaborative Governance in the San Francisco Bay Area Environmental Movement.” In Social Movements and Networks: Relational Approaches to Collective Action, edited by Mario Diani and Doug McAdam. Oxford: Oxford University Press. . . Opinión : (September ). Arce, Moisés. . “The Politics of Pension Reform in Peru.” Studies in Comparative International Development (): –. Arce, Moisés, and Paul T. Bellinger Jr. .“Low Intensity Democracy Revisited: The Effects of Economic Liberalization on Political Activity in Latin America.” World Politics (): –.
14 References.qxd
11/9/2009
11:53 AM
Page 360
.
Assies, William. . “Urban Social Movements in Brazil: A Debate and Its Dynamics.” Latin American Perspectives (): –. Auyero, Javier. . Poor People’s Politics: Peronist Survival and the Legacy of Evita. Durham: Duke University Press. Avritzer, Leonardo. . Democracy and the Public Space in Latin America. Princeton: Princeton University Press. Axford, Barrie, and Richard Huggins, eds. . New Media and Politics. Thousand Oaks, Calif.: Sage. Baiocchi, Gianpaolo. . Militants and Citizens: The Politics of Participatory Democracy in Porto Alegre. Stanford: Stanford University Press. Ballón Echegaray, Eduardo, Manuel Castillo, et al. . Movimientos sociales y crisis: El caso peruano. Lima: Centro de Estudios y Promoción del Desarrollo. Bamberger, Michael. . “A Problem of Political Integration in Latin America: The Barrios of Venezuela.” International Affairs (): –. Barrig, Maruja. . “Women, Collective Kitchens, and the Crisis of the State in Peru.” In Emergences: Women’s Struggles for Livelihood in Latin America, edited by John Friedmann et al. Los Angeles: UCLA Latin American Center Publications. Bartolini, Stefano. . The Political Mobilization of the European Left, -: The Class Cleavage. New York: Cambridge University Press. Bartolini, Stefano, and Peter Mair. . Identity, Competition, and Electoral Availability: The Stabilisation of European Electorates, –. New York: Cambridge University Press. Baumgartner, Frank, and Beth Leech. . Basic Interests: The Importance of Groups in Politics and in Political Science. Princeton: Princeton University Press. BBC News. . “Peru Tightens Control over .” November . Bensusán, Graciela. . “El impacto de la reestructuración neoliberal: comparación de las estrategias sindicales en Argentina, Brasil, México, Canadá y Estados Unidos.” Paper presented at the III Latin American Congress of Labor Sociology, Buenos Aires. Bensusán, Graciela, and Maria Lorena Cook. . “Political Transition and Labor Revitalization in Mexico.” In Labor Revitalization: Global Perspectives and New Initiatives, edited by Daniel B. Cornfield and Holly J. McCammon. Vol. of Research in the Sociology of Work. Stamford, Conn.: JAI Press. Bentley, Arthur Fisher. . The Process of Government: A Study of Social Pressures. Chicago: University of Chicago Press. ———. . Behavior, Knowledge, Fact. Bloomington, Ind.: Principia Press. Berelson, Bernard, Paul F. Lazarsfeld, and William N. McPhee. . Voting: A Study of Opinion Formation in a Presidential Campaign. Chicago: University of Chicago Press. Berger, Suzanne. . Organizing Interests in Western Europe: Pluralism, Corporatism, and the Transformation of Politics. New York: Cambridge University Press. Bermeo, Nancy. . “Rethinking Regime Change.” Comparative Politics (): –. Biles, Robert E. . “Political Participation in Urban Uruguay: Mixing Public and Private Ends.” In Political Participation in Latin America: Volume I, Citizen and State, edited by John A. Booth and Mitchell A. Seligson. New York: Holmes and Meier. Binderkrantz, Anne. . “Interest Group Strategies: Navigating Between Privileged Access Strategies of Pressure.” Political Studies :–. Bird, Richard M., and François Vaillancourt, eds. . Fiscal Decentralization in Developing Countries. New York: Cambridge University Press.
14 References.qxd
11/9/2009
11:53 AM
Page 361
,
Birdsall, Nancy, Carol Graham, and Stefano Pettinato. . “Stuck in the Tunnel: Is Globalization Muddling the Middle Class?” Working Paper No. , Center on Social and Economic Dynamics, Washington D.C. Blondet, Cecilia. a. Cucharas en alto: Los comedores populares, años de historia. Lima: Instituto de Estudios Peruanos. ———. b. “Los comedores populares: años de historia.” In Cucharas en alto: Del asistencialismo al desarrollo local: Fortaleciendo la participación de las mujeres, edited by Cecilia Blondet and Carolina Trivelli. Lima: IEP Ediciones. Blumler, Jay G., and Dennis Kavanaugh. . “The Third Age of Political Communication: Influences and Features.” Political Communication : –. Boas, Taylor. . “Varieties of Electioneering: Empathetic Personalism in Chile’s Presidential Election Campaigns.” Paper presented at the American Political Science Association, Chicago. ———. . “Varieties of Electioneering: Transforming Presidential Election Campaigns in Latin America.” PhD diss., University of California, Berkeley. Booth, John A., and Patricia B. Richard. . “Civil Society, Political Capital, and Democratization in Central America.” Journal of Politics (): –. Bowman, John R. . Capitalist Collective Action: Competition, Cooperation, and Conflict in the Coal Industry. New York: Cambridge University Press. Brady, Henry E., Sidney Verba, and Kay L. Schlozman. . “Beyond SES: A Resource Model of Political Participation.” American Political Science Review (): –. Bratton, Michael. . “Political Participation in a New Democracy: Institutional Considerations from Zambia.” Comparative Political Studies (): . Brody, Richard. . “The Puzzle of Political Participation in America.” In The New American Political System, edited by Anthony King. Washington, D.C.: American Enterprise Institute. Brysk, Allison. . “Democratizing Civil Society in Latin America.” Journal of Democracy (): –. Burgess, Katrina. . “Loyalty Dilemmas and Market Reform: Party-Union Alliances Under Stress in Mexico, Spain, and Venezuela.” World Politics (): –. Burgess, Katrina, and Steven Levitsky. . “Explaining Populist Party Adaptation in Latin America: Environmental and Organizational Determinants of Party Change in Argentina, Mexico, Peru and Venezuela.” Comparative Political Studies (): –. Buvinic, Mayra, and Vivian Roza. . “Women, Politics, and Democratic Prospects in Latin America.” Technical Papers Series. Sustainable Development Department. InterAmerican Development Bank. Washington, D.C. Cameron, Maxwell A. . Democracy and Authoritarianism in Peru: Political Coalitions and Social Change. New York: St. Martin’s Press. Cameron, Maxwell A., and Steven Levitsky. . “Democracy Without Parties? Political Parties and Regime Change in Fujimori’s Peru.” Latin American Politics and Society (): –. Cameron, Maxwell A., and Flavie Major. . “Venezuela’s Hugo Chávez: Savior or Threat to Democracy.” Latin American Research Review (): –. Campbell, Angus, et al. . The American Voter. New York: Wiley. Campero, Guillermo. . Entre la sobrevivencia y la acción política: Las organizaciones de pobladores en Santiago. Santiago: . Campetella, Andrea, and Inés González Bombal. . “Historia del sector sin fines de
14 References.qxd
11/9/2009
11:53 AM
Page 362
.
lucro en Argentina.” In Estudios sobre el sector sin fines de lucro en Argentina, edited by Mario Roitter and Inés González Bombal. Buenos Aires: . Carmines, Edward G., and Harold W. Stanley. . “The Transformation of the New Deal Party System: Social Groups, Political Ideology, and Changing Partisanship Among Northern Whites, –.” Political Behavior (): –. Carroll, Thomas F. . Intermediary : The Supporting Link in Grassroots Development. West Hartford, Conn.: Kumarian Press. Castells, Manuel. . The City and the Grassroots: A Cross-Cultural Theory of Urban Social Movements. London: E. Arnold. ———. . Power of Identity. Vol. of The Information Age: Economy, Society, and Culture. Cambridge, Mass.: Blackwell Publishers. Cavarozzi, Marcelo. . “Beyond Transitions to Democracy in Latin America.” Journal of Latin American Studies (): –. Cavarozzi, Marcelo, and Vicente Palmero. . “State, Civil Society, and Popular Neighborhood Organizations in Buenos Aires: Key Players in Argentina’s Transition.” In New Paths to Democratic Development in Latin America: The Rise of the -Municipal Collaboration, edited by Charles A. Reilly. London: Lynne Rienner. Centeno, Miguel. . “The New Leviathan: The Dynamics and Limits of Technocracy.” Theory and Society (): –. . . Anuario estadístico de América Latina y el Caribe . Santiago. ———. . Anuario estadístico de América Latina y el Caribe . Santiago. ———. . : Base de estadísticas e indicadores sociales. Santiago. Chalmers, Douglas A., Scott B. Martin, and Kerianne Piester. . “Associative Networks: New Structures of Representation for the Popular Sectors?” In The New Politics of Inequality in Latin America: Rethinking Participation and Representation, edited by Douglas A. Chalmers, Carlos M. Vilas, Katherine Hite, Scott B. Martin, Kerianne Piester, and Monique Segarra Oxford: Oxford University Press. Chavez, Daniel, and Benjamin Goldfrank, eds. . The Left in the City: Participatory Local Governments in Latin America. London: Latin America Bureau and Transnational Institute. Chhibber, Pradeep, and Mariano Torcal. . “Electoral Strategies, Social Cleavages, and Party Systems in a New Democracy: Spain.” Comparative Political Studies (): –. Cigler, Allan J., and Burdett A. Loomis. . Interest Group Politics. rd ed. Washington, D.C.: CQ Press. ———. . “Always Involved, Rarely Central: Organized Interests in American Politics.” In Interest Group Politics, edited by Allan J. Cigler and Burdett A. Loomis. Washington D.C.: CQ Press. Clark, Terry Nichols, and Seymour Martin Lipset, eds. . The Breakdown of Class Politics: A Debate on Post-Industrial Stratification. Baltimore: Johns Hopkins University Press. Clark, Terry Nichols, Seymour Martin Lipset, and Michael Rempel. . “The Declining Political Significance of Social Class.” International Sociology (): –. Coelho, Vera Schattan. . “Democratization of Brazilian Health Councils: The Paradox of Bringing the Other Side into the Tent.” International Journal of Urban and Regional Research (): –.
14 References.qxd
11/9/2009
11:53 AM
Page 363
,
Collier, David. . Squatters and Oligarchs: Authoritarian Rule and Policy Change in Peru. Baltimore: Johns Hopkins University Press. Collier, Ruth Berins. . The Contradictory Alliance: State-Labor Relations and Regime Change in Mexico. Berkeley: International and Area Studies, University of California, Berkeley. ———. . Paths Toward Democracy: The Working Class and Elites in Western Europe and South America. New York: Cambridge University Press. Collier, Ruth Berins, and David Collier. . “Inducements Versus Constraints: Disaggregating ‘Corporatism.’” American Political Science Review (): –. ———. . Shaping the Political Arena: Critical Junctures, the Labor Movement, and Regime Dynamics in Latin America. Princeton: Princeton University Press. Conaghan, Catherine, and James Malloy. . Unsettling Statecraft: Democracy and Neoliberalism in the Central Andes. Pittsburgh: University of Pittsburgh Press. Converse, Philip E. . “Change in the American Electorate.” In The Human Meaning of Social Change, edited by Angus Campbell and Philip E. Converse. New York: Russell Sage Foundation. Cook, Maria Lorena. . Organizing Dissent: Unions, the State, and the Democratic Teachers’ Movement in Mexico. University Park: Pennsylvania State University Press. ———. . “Labor Reform and Dual Transitions in the Southern Cone and Brazil.” Latin American Politics and Society (Spring ): –. ———. . Politics of Labor Reform in Latin America: Between Flexibility and Rights. University Park: Pennsylvania State University Press. Copestake, James. . “Multiple Dimensions of Social Assistance: The Case of Peru’s ‘Glass of Milk’ Programme.” Journal of Development Studies :. Coppedge, Michael. . Strong Parties and Lame Ducks: Presidential Partyarchy and Factionalism in Venezuela. Stanford: Stanford University Press. Corrales, Javier. . “Market Reforms.” In Constructing Democratic Governance in Latin America, edited by Jorge Domínguez. Baltimore: Johns Hopkins University Press. Crisp, Brian, and Daniel H. Levine. .“Democratizing the Democracy? Crisis and Reform in Venezuela.” Journal of Interamerican Studies and World Affairs (): –. Cross, John C. . Informal Politics: Street Vendors and the State in Mexico City. Stanford: Stanford University Press. Dagnino, Evelina. . Sociedade civil e espaços públicos no Brasil. São Paulo: Paz e Terra. Dagnino, Evelina, Alberto Olvera, and Aldo Panfichi, eds. . A disputa pela construção democrática na América Latina. São Paulo: Paz e Terra. Dahl, Robert Alan. . Who Governs? Democracy and Power in an American City. New Haven: Yale University Press. Dalton, Russell J. a. “Democracy and Its Citizens: Patterns of Political Change.” Unpublished ms., Center for the Study of Democracy, University of California, Irvine. ———. b. “Political Cleavages, Issues and Electoral Change.” In Comparing Democracies: Elections and Voting in Global Perspective, edited by Lawrence LeDuc, Richard Niemi, and Pippa Norris. Thousand Oaks, Calif.: Sage. ———. . Citizen Politics: Public Opinion and Political Parties in Advanced Industrial Democracies. rd ed. New York: Chatham House. Dalton, Russell J., and Manfred Kuechler. . Challenging the Political Order: New Social and Political Movements in Western Democracies. New York: Oxford University Press. Dalton, Russell J., Ian McAllister, and Martin P. Wattenberg. . “The Consequences of
14 References.qxd
11/9/2009
11:53 AM
Page 364
.
Partisan Dealignment.” In Parties Without Partisans: Political Change in Advanced Industrial Democracies, edited by Russell J. Dalton and Martin P. Wattenberg. Oxford: Oxford University Press. De Ferranti, David M. . Inequality in Latin America: Breaking with History. Washington, D.C.: World Bank. de Janvry, Alain, and Elisabeth Sadoulet. . “Growth, Poverty, and Inequality in Latin America: A Causal Analysis, –.” Review of Income and Wealth (): –. D’Elia, Yolanda. . Las Misiones Sociales en Venezuela. Caracas: . D’Elia, Yolanda, et al. . Las Misiones Sociales en Venezuela: Una Aproximación a su Comprensión y Análisis. Caracas: . della Porta, Donatella, and Mario Diani. . Social Movements: An Introduction. Oxford: Blackwell. della Porta, Donatella, Lorenzo Mosca, and Herbert Reiter. . Globalization from Below: Transnational Activists and Protest Networks. Minneapolis: University of Minnesota Press. della Porta, Donatella, and Sidney G. Tarrow. . Transnational Protest and Global Activism. Lanham, Md.: Rowman Littlefield. de Privitellio, Luciano. . Vecinos y ciudadanos: Política y sociedad en la Buenos Aires de entreguerras. Buenos Aires: Siglo Veintiuno. Diani, Mario. . Green Networks: A Structural Analysis of the Italian Environmental Movement. Edinburgh: Edinburgh University Press. ———. a. “Introduction: Social Movements, Contentious Actions, and Social Networks: ‘From Metaphor to Substance?’” In Social Movements and Networks: Relational Approaches to Collective Action, edited by Mario Diani and Doug McAdam. Oxford: Oxford University Press. ———. b.“Leaders or Brokers? Positions and Influence in Social Movement Networks.” In Social Movements and Networks: Relational Approaches to Collective Action, edited by Mario Diani and Doug McAdam. Oxford: Oxford University Press. ———. c. “Networks and Social Movements: A Research Program.” In Social Movements and Networks: Relational Approaches to Collective Action, edited by Mario Diani and Doug McAdam. Oxford: Oxford University Press. Dietz, Henry A. . Urban Poverty, Political Participation, and the State: Lima, –. Pittsburgh: University of Pittsburgh Press. Di Maggio, Paul J., and Walter W. Powell. . Introduction to The New Institutionalism in Organization Analysis. Chicago: University of Chicago Press. Di Tella, Rafael, Sebastián Galiani, and Edgardo Schargrodsky. . “The Formation of Beliefs: Evidence from the Allocation of Land Titles to Squatters.” Quarterly Journal of Economics (): –. Dosh, Paul G. J. . “Demanding the Land: Urban Social Movements and Local Politics in Peru and Ecuador.” PhD diss., University of California, Berkeley. ———. . “Surprising Trends in Land Invasions in Metropolitan Lima and Quito.” Latin American Perspectives (): –. Eckstein, Susan E. a. “Power and Popular Protest in Latin America.” In Power and Popular Protest: Latin American Social Movements, edited by Susan E. Eckstein. Berkeley and Los Angeles: University of California Press. ———. b.“Where Have All the Movements Gone? Latin American Social Movements at
14 References.qxd
11/9/2009
11:53 AM
Page 365
,
the New Millenium.” In Power and Popular Protest: Latin American Social Movements, edited by Susan Eckstein. Berkeley and Los Angeles: University of California Press. Edelman, Marc. . “The Persistence of the Peasantry.” Report on the Americas (): –. Eder, Klaus. . The New Politics of Class: Social Movements and Cultural Dynamics in Advanced Societies. London: Sage. ———. . “Identity Mobilization and Democracy: An Ambivalent Relationship.” In Social Movements and Democracy, edited by Pedro Ibarra Güell. New York: Palgrave Macmillan. Education International. . Equality with a Gender Perspective in Latin America. Available at http://www.ei-ie.org/en/news/show.php?id=&theme=development& country=global (accessed March ). Edwards, Michael, and David Hulme. . Beyond the Magic Bullet: Performance and Accountability. London: Earthscan/James & James. ———. . “Too Close for Comfort? The Impact of Official Aid on Nongovernmental Organizations.” World Development (): –. Efron, Bradley, and Robert J. Tibshirani. . An Introduction to the Bootstrap. Boca Raton, Fla.: Chapman and Hall/CRC. Ellner, Steve. . “Obstacles to the Consolidation of the Venezuelan Neighbourhood Movement: National and Local Cleavages.” Journal of Latin American Studies (): –. ———. . “The Radical Potential of Chavismo in Venezuela: The First Year and a Half in Power.” Latin American Perspectives (): –. ———. . “The Contrasting Variants of the Populism of Hugo Chávez and Alberto Fujimori.” Journal of Latin American Studies (): –. Erbe, William. . “Social Involvement and Political Activity: A Replication and Elaboration.” American Sociological Review (): –. Erickson, Kenneth P. . The Brazilian Corporative State and Working-Class Politics. Berkeley and Los Angeles: University of California Press. Escobar, Arturo. . “Culture, Economics, and Politics in Latin American Social Movements Theory and Research.” In The Making of Social Movements in Latin America: Identity, Strategy, and Democracy, edited by Arturo Escobar and Sonia Alvarez. Boulder, Colo.: Westview. Espinoza, Vicente. . Para una historia de los pobres de la ciudad. Santiago: Ediciones Sur. Etchemendy, Sebastián. . “Constructing Reform Coalitions: The Politics of Compensations in Argentina’s Economic Liberalization.” Latin American Politics and Society (): –. ———. .“Repression, Exclusion and Inclusion: Government-Union Relations and Patterns of Labor Reform in Liberalizing Economies.” Comparative Politics (): –. Etchemendy, Sebastián, and Ruth Berins Collier. . “Down but Not Out: Union Resurgence and Segmented Neocorporatism in Argentina (–).” Politics and Society (): –. Evans, Geoffrey. . “Class Voting: From Premature Obituary to Reasoned Appraisal.” In The End of Class Politics? Class Voting in Comparative Context, edited by Geoffrey Evans. Oxford: Oxford University Press. Evans, Peter B. . Dependent Development: The Alliance of Multinational, State, and Local Capital in Brazil. Princeton: Princeton University Press.
14 References.qxd
11/9/2009
11:53 AM
Page 366
.
———. . Embedded Autonomy: States and Industrial Transformation. Princeton: Princeton University Press. ———. . “Government Action, Social Capital, and Development: Reviewing the Evidence on Synergy.” World Development (): –. Ferradas, Pedro, and Rebeca Sevilla. . Ciudad y pobladores de Lima metropolitana, –. Lima: . Fisher, William F. . “Doing Good? The Politics and Anti-Politics of Practices.” Annual Review of Anthropology (): –. Forment, Carlos A. . Democracy in Latin America, –. Chicago: University of Chicago Press. Fornos, Carolina A., Timothy J. Power, and James C. Garand. . “Explaining Voter Turnout in Latin America, to .” Comparative Political Studies (): –. Foweraker, Joe. . “Popular Movements and Political Change in Mexico.” In Popular Movements and Political Change in Mexico, edited by Joe Foweraker and Ann L. Craig. Boulder, Colo.: Lynne Rienner. ———. . Theorizing Social Movements. London: Pluto Press. ———. . “Grassroots Movements and Political Activism in Latin America: A Critical Comparison of Chile and Brazil.” Journal of Latin American Studies (): –. Foweraker, Joe, and Ann L. Craig, eds. . Popular Movements and Political Change in Mexico. Boulder, Colo.: Lynne Rienner. Foweraker, Joe, and Todd Landman. . Citizenship Rights and Social Movements: A Comparative and Statistical Analysis. Oxford: Oxford University Press. Foweraker, Joe, Todd Landman, and Neil Harvey. . Governing Latin America. Cambridge: Polity. Fox, Jonathan. . “Democratic Rural Development: Leadership Accountability in Regional Peasant Organizations.” Development and Change (): –. ———. . “The Difficult Transition from Clientelism to Citizenship: Lessons from Mexico.” World Politics (): –. ———. .“How Does Civil Society Thicken? The Political Construction of Social Capital in Rural Mexico.” World Development (): –. Frank, Volker. a. “The Elusive Goal in Democratic Chile: Reforming the Pinochet Labor Legislation.” Latin American Politics and Society (): –. ———. b. “The Labor Movement in Democratic Chile, –.” Working Paper , University of Notre Dame. Friedman, Elisabeth Jay, and Kathryn Hochstetler. . “Assessing the Third Transition in Latin American Democratization: Representational Regimes and Civil Society in Argentina and Brazil.” Comparative Politics (): –. Galiani, Sebastián, and Ernesto Schargrodsky. . “Property Rights for the Poor: Effects of Land Titling.” Working Paper, Washington University in St. Louis and Universidad Torcuato Di Tella Business School, Buenos Aires. Gamson, William A. . Power and Discontent. Homewood, Ill.: Dorsey Press. ———. . The Strategy of Social Protest. Homewood, Ill.: Dorsey Press. Garay, Candelaria. . “Targeting Dissent: Unemployed Workers, Community Associations and Social Policy in Argentina –.” Paper presented at the Annual Meetings of the Latin American Studies Association, Las Vegas. ———. .“Social Policy and Collective Action: Unemployed Workers, Community Associations, and Protest in Argentina.” Politics and Society (): –.
14 References.qxd
11/9/2009
11:53 AM
Page 367
,
———. n.d. “Labor-Market Outsiders and State Expansion: Bridging the Social Policy Divide in Latin America.” Unpublished m.s., University of California, Berkeley. Garcés, Mario. . Tomando su sitio: el movimiento de pobladores de Santiago, –. Santiago: . García-Guadilla, María Pilar. . “Ciudadanía y autonomía en las organizaciones sociales bolivarianas: Los Comités de Tierra Urbana como movimientos sociales.” Cuadernos del :. ———. . “El poder popular y la democracia participativa en Venezuela: Los Consejos Comunales.” Paper presented at the International Congress of the Latin American Studies Association, Montreal. Gargarella, Roberto, Pilar Domingo, and Theunis Roux, eds. . Courts and Social Transformation in New Democracies: An Institutional Voice for the Poor? Aldershot, UK: Ashgate Publishing. Garner, Roberta Ash, and Mayer N. Zald. . “The Political Economy of Social Movement Sectors.” In The Challenge of Social Control: Citizenship and Institution Building in Modern Society. Essays in Honor of Morris Janowitz, edited by Gerald Suttles and Mayer N. Zald, –. Norwood, N.J.: Ablex. Garretón, Manuel Antonio. . “The Political Opposition and the Party System.” In The Struggle for Democracy in Chile, –, edited by Paul Drake and Iván Jaksic. Lincoln: University of Nebraska Press. ———. . “New State-Society Relations in Latin America.” In Redefining the State in Latin America, edited by Colin I. Bradford. Paris: . Gay, Robert. . “Popular Incorporation and Prospects for Democracy.” Theory and Society (): –. Geddes, Barbara. . Politician’s Dilemma: Building State Capacity in Latin America. Berkeley: University of California Press. Gerchunoff, Pablo, and Horacio Aguirre. . “La política económica de Kirchner en la Argentina: Varios estilos, una sola agenda.” Boletín del Real Instituto Elcano de Estudios Internacionales y Estratégicos :–. Gideon, Jasmine. . “The Politics of Social Service Provision Through : A Study of Latin America.” Bulletin of Latin American Research (): –. Goldfrank, Benjamin. . “Urban Experiments in Citizen Participation: Deepening Democracy in Latin America.” PhD diss., University of California, Berkeley. ———. . “Making Participation Work in Porto Alegre.” In Radicals in Power: The Workers’ Party and Experiments in Urban Democracy in Brazil, edited by Gianpaolo Baiocchi. London: Zed Books. ———. . “Los procesos de ‘presupuesto participativo’ en América Latina: Éxito, fracaso y cambio.” Revista de Ciencia Política (): –. ———. . “The Politics of Deepening Local Democracy: Decentralization, Party Institutionalization, and Participation.” Comparative Politics (): –. Goldstone, Jack A. . “Bridging Institutionalized and Noninstitutionalized Politics.” In States, Parties, and Social Movements, edited by Jack A. Goldstone. Cambridge: Cambridge University Press. Gould, Roger V. . “Multiple Networks and Mobilization in the Paris Commune, .” American Sociological Review (): –. ———. . “Uses of Network Tools in Comparative Historical Research.” In Comparative
14 References.qxd
11/9/2009
11:53 AM
Page 368
.
Historical Analysis in the Social Sciences, edited by James Mahoney and Dietrich Rueschemeyer. New York: Cambridge University Press. Graham, Carol. . Safety Nets, Politics, and the Poor: Transitions to Market Economies. Washington, D.C.: Brookings Institution. Green, Donald P., Bradley Palmquist, and Eric Schickler. . Partisan Hearts and Minds: Political Parties and the Social Identities of Voters. New Haven: Yale University Press. Grindle, Merilee Serrill. . “The Social Agenda and the Politics of Reform in Latin America.” Working Paper /-s, David Rockefeller Center for Latin American Studies, Cambridge, Mass. ———. a. Audacious Reforms: Institutional Invention and Democracy in Latin America. Baltimore: Johns Hopkins University Press. ———. b. “The Social Agenda and the Politics of Reform in Latin America.” In Social Development in Latin America: The Politics of Reform, edited by Joseph S. Tulchin and Allison M. Garland. Boulder, Colo.: Lynne Rienner. Grugel, Jean, and María Pía Riggirozzi. . “The Return of the State in Argentina.” International Affairs (): –. Gutiérrez, Leandro H., and Luis Alberto Romero. . Sectores populares, cultura y política: Buenos Aires en la entreguerra. Buenos Aires: Editorial Sudamericana. Guzmán, Eugenio. . “Reflexiones sobre el sistema binomial.” Estudios públicos :–. Haber, Paul L. . Power from Experience: Urban Popular Movements in Late-Twentieth Century Mexico. University Park: Penn State University Press, . Habermas, Jürgen. . Between Facts and Norms: Contributions to a Discourse Theory of Law and Democracy. Cambridge: MIT Press. Hagopian, Frances. . “Democracy and Political Representation in Latin America in the s: Pause, Reorganization, or Decline?” In Fault Lines of Democracy in PostTransition Latin America, edited by Felipe Agüero and Jeffrey Stark. Boulder, Colo.: Lynne Rienner. Hall, Peter. . The Bootstrap and Edgeworth Expansion. New York: Springer-Verlag. Handlin, Samuel. . “Reinventing Class Mobilization: Class Cleavages and Participatory Populism in Hugo Chávez’s Venezuela.” Paper presented at the annual meeting of the American Political Science Association, Boston. Handlin, Samuel, and Ruth Berins Collier. Forthcoming. “Party Change and the Left in South America: Subjective Representation, Direct Contact, and Organizational Linkages.” In Latin America’s Left Turn, edited by Steven Levitsky and Kenneth Roberts. Johns Hopkins University Press. Hansen, John Mark. . “The Political Economy of Group Membership.” American Political Science Review (): –. Harnecker, Marta. . Gobiernos comunitarios: Transformando el estado desde abajo. Caracas: Monte Avila. Hawkins, Kirk A., and David R. Hansen. . “Dependent Civil Society: The Círculos Bolivarianos in Venezuela.” Latin American Research Review (): –. Hawkins, Kirk A., et al. . “Political Culture of Democracy in Venezuela: .” Latin American Public Opinion Project Series. Hellman, Judith Adler. . “The Riddle of New Social Movements: Who They Are and What They Do.” In Capital, Power, and Inequality in Latin America, edited by Sandor Halebsky and Richard L. Harris. Boulder, Colo.: Westview.
14 References.qxd
11/9/2009
11:53 AM
Page 369
,
———. . “Social Movements: Revolution, Reform, and Reaction.” Report on the Americas (): –. Hipsher, Patricia. . “Democratic Transitions and Social Movement Outcomes.” In From Contention to Democracy, edited by Marco G. Guigni, Doug McAdam, and Charles Tilly. Boston: Rowman & Littlefield. Hirschman, Albert O. . “The Political Economy of Latin American Development: Seven Exercises in Retrospection.” Latin American Research Review (): –. Hiskey, Jon, and Mitchell Seligson. . “Pitfalls of Power to the People: Decentralization, Local Government Performance, and System Support in Bolivia.” Studies in Comparative International Development (): –. Howes, Mick. . “ and the Institutional Development of Membership Organizations: Evidence from Six Cases.” Journal of International Development (): –. Huckfeldt, Robert, and John Sprague. . “Political Parties and Electoral Mobilization: Political Structure, Social Structure, and the Party Canvass.” American Political Science Review (): –. Ibarra Guëll, Pedro. . “Social Movements: From Promoters to Protagonists of Democracy.” In Social Movements and Democracy, edited by Pedro Ibarra Guëll. New York: Palgrave-Macmillan. . . Panorama laboral . Geneva. ———. . Panorama laboral . Geneva. ———. . Panorama laboral . Geneva. Inglehart, Ronald. . The Silent Revolution: Changing Values and Political Styles Among Western Publics. Princeton: Princeton University Press. ———. . “The Changing Structure of Political Cleavages in Western Society.” In Electoral Change in Advanced Industrial Democracies: Realignment or Dealignment? edited by Russell J. Dalton et al. Princeton: Princeton University Press. Inglehart, Ronald, and Jacques-René Rabier. .“Political Realignment in Advanced Industrial Society: From Class-Based Politics to Quality-of-Life Politics.” Government and Opposition (): –. Jaquette, Jane, ed. . The Women’s Movement in Latin America: Participation and Democracy. nd ed. Boulder, Colo.: Westview. Jelin, Elizabeth. . Los nuevos movimientos sociales. Buenos Aires: Centro Editor de América Latina. ———. . “Emergent Citizenship or Exclusion? Social Movements and NonGovernmental Organizations in the s.” In Politics, Social Change, and Economic Restructuring in Latin America, edited by William Smith and Roberto Korzeniewicz. Miami: North/South Center Press. ———. . “Toward a Culture of Participation and Citizenship: Challenges for a More Equitable World.” In Cultures of Politics/Politics of Cultures: Re-Visioning Latin American Social Movements, edited by Sonia E. Alvarez et al. Boulder, Colo.: Westview. Johnston, Richard, et al. . Letting the People Decide: Dynamics of a Canadian Election. Stanford: Stanford University Press. Jones, Mark P., and Wonjae Hwang. . “Provincial Party Bosses: Keystone of the Argentine Congress.” In Argentine Democracy: The Politics of Institutional Weakness, edited by Steven Levitsky and María Victoria Murillo. University Park: Pennsylvania State University Press.
14 References.qxd
11/9/2009
11:53 AM
Page 370
.
Katz, Richard, and Peter Mair. . “Changing Models of Party Organization and Party Democracy.” Party Politics (): –. Katznelson, Ira, and Aristide Zolberg, eds. . Working-Class Formation: NineteenthCentury Patterns in Western Europe and the United States. Princeton: Princeton University Press. Keck, Margaret E. . The Workers’ Party and Democratization in Brazil. New Haven: Yale University Press. Keck, Margaret E., and Kathryn Sikkink. . Activists Beyond Borders: Advocacy Networks in International Politics. Ithaca: Cornell University Press. Kersting, Norbert, and Sperberg Fuentealba. . “Pobreza urbana, sociedad civil y ciudadanía en Chile y Brasil.” Nueva Sociedad :–. Kessler, Timothy P. . “Political Capital: Mexican Financial Policy Under Salinas.” World Politics (): –. Key, V. O., Jr. . Public Opinion and American Democracy. John Wiley: New York. Key, V. O., Jr., and Milton C. Cummings Jr. . The Responsible Electorate: Rationality in Presidential Voting, –. Cambridge: Harvard University Press. Kinder, Donald. . “Opinion and Action in the Realm of Politics.” In The Handbook of Social Psychology, edited by Daniel Todd Gilbert et al. New York: McGraw-Hill. Kirchheimer, Otto. . “The Catch-All Party.” In The West European Party System, edited by Peter Mair. Oxford: Oxford University Press. Kish, Leslie, and Martin R. Frankel. . “Inference from Complex Samples (with Discussion).” Journal of the Royal Statistical Society. Series B (Methodological) (): –. Kitschelt, Herbert. . The Transformation of European Social Democracy. New York: Cambridge University Press. ———. . “Linkages Between Citizens and Politicians in Democratic Polities.” Comparative Political Studies (–): –. Kitschelt, Herbert, and Steven Wilkinson. . Patrons, Clients, and Policies: Patterns of Democratic Accountability and Political Competition. New York: Cambridge University Press. Klaiber, Jeffrey L. . The Church, Dictatorships, and Democracy in Latin America. Maryknoll, N.Y.: Orbis Books. Klandermans, Bert, and Dirk Oegema. . “Potentials, Networks, Motivations, and Barriers: Steps Toward Participation in Social Movements.” American Sociological Review (): –. Klarén, Peter F. . Peru: Society and Nationhood in the Andes. New York: Oxford University Press. Klesner, Joseph L. . “Social Capital and Political Participation in Latin America: Evidence from Argentina, Chile, Mexico, and Peru.” Latin American Research Review (): –. Klingemann, Hans-Dieter, and Bernhard Wessels. . “Voter Rationalities in Different Electoral Systems.” Paper presented at the International Political Science Association, Quebec. Korzeniewicz, Roberto Patricio, and William C. Smith. . “Poverty, Inequality, and Growth in Latin America: Searching for the High Road to Globalization.” Latin American Research Review (): –. ———. .“Protest and Collaboration: Transnational Civil Society Networks and the Politics of Summitry and Free Trade in the Americas.” North-South Agenda Papers .
14 References.qxd
11/9/2009
11:53 AM
Page 371
,
Kriesi, Hanspeter. . “The Organizational Structure of New Social Movements in a Polit ical Context.” In Comparative Perspectives on Social Movements: Political Opportunities, Mobilizing Structures, and Cultural Framings, edited by Doug McAdam, John D. McCarthy, and Mayer N. Zald. New York: Cambridge University Press. Kriesi, Hanspeter, et al., eds. . New Social Movements in Western Europe: A Comparative Analysis. Minneapolis: University of Minnesota Press. Krishna, Anirudh. . Active Social Capital: Tracing the Roots of Development and Democracy. New York: Columbia University Press. Kuenzi, Michelle T. . “Nonformal Education, Political Participation, and Democracy: Findings from Senegal.” Political Behavior (): –. Kurtz, Marcus J. . “Free Markets and Democratic Consolidation in Chile: The National Politics of Rural Transformation.” Politics & Society (): –. ———. a. “The Dilemmas of Democracy in the Open Economy: Lessons from Latin America.” World Politics (January): –. ———. b. Free Market Democracy and the Chilean and Mexican Countryside. New York: Cambridge University Press. Laclau, Ernesto, and Chantal Mouffe. . Hegemony and Socialist Strategy: Toward a Radical Democratic Politics. London: Verso Books. La Due Lake, Ronald, and Robert Huckfeldt. . “Social Capital, Social Networks, and Political Participation.” Political Psychology (): –. Lawson, Kay. . Political Parties and Linkage: A Comparative Perspective. New Haven: Yale University Press. Lawson, Chappell, and James McCann. . “Television Coverage, Media Effects, and Mexico’s Elections.” British Journal of Political Science (): –. Lehmbruch, Gerhard. . “Liberal Corporatism and Party Government.” Comparative Political Studies :–. ———. . “Consociational Democracy, Class Conflict and the New Corporatism.” In Trends Toward Corporatist Intermediation, edited by Philippe Schmitter and Gerhard Lehmbruch. London: Sage. Lehmbruch, Gerhard, and Philippe C. Schmitter. . Patterns of Corporatist Policy-Making. London: Sage Publications. Levine, Daniel, and Catalina Romero. . “Urban Citizen Movements and Disempowerment in Peru and Venezuela.” In The Crisis of Democratic Representation in the Andes, edited by Scott Mainwaring, Ana María Bejarano, and Eduardo Pizarro Leongómez. Stanford: Stanford University Press. Levitsky, Steven. . Transforming Labor-Based Parties in Latin America: Argentine Peronism in Comparative Perspective. New York: Cambridge University Press. Levitsky, Steven, and María Victoria Murillo. . “Argentina Weathers the Storm.” Journal of Democracy (): –. Levitsky, Steven, and Lucan A. Way. . “Between a Shock and a Hard Place: The Dynamics of Labor-Backed Adjustment in Poland and Argentina.” Comparative Politics (): –. Levy, Jonah D. . Tocqueville’s Revenge: State, Society, and Economy in Contemporary France. Cambridge: Harvard University Press. ———. . The State After Statism: New State Activities in the Age of Liberalization. Cambridge: Harvard University Press.
14 References.qxd
11/9/2009
11:53 AM
Page 372
.
Lindblom, Charles. . Politics and Markets: The World’s Political Economic Systems. New York: Basic Books. ———. . “The Market as Prison.” Journal of Politics (): –. Linz, Juan, and Alfred Stepan. . The Breakdown of Democratic Regimes: Crisis, Breakdown and Reequilibration. Baltimore: Johns Hopkins University Press. Lipietz, Alain. . Mirages and Miracles: The Crises of Global Fordism. London: Verso. Lipset, Seymour Martin, and Stein Rokkan. . Party Systems and Voter Alignments: CrossNational Perspectives. New York: Free Press. Lipset, Seymour Martin, Martin A. Trow, and James S. Coleman. . Union Democracy: The Internal Politics of the International Typographical Union. Glencoe, Ill.: Free Press. Lipsky, Michael. . “Protest as Political Resource.” American Political Science Review (): –. Londoño, Juan Luis, and Miguel Székely. . “Persistent Poverty and Excess Inequality: Latin America, –.” Journal of Applied Economics (): –. López Maya, Margarita. . Del Viernes Negro al Referendo Revocatorio. Caracas: Alfadil. ———. . “The State, Popular Participation, and How to Make Things Work.” Paper presented at the conference, “The Popular Sectors and the State in Chávez’s Venezuela,” Yale University. López Maya, Margarita, and Luis Lander. . “Popular Protest in Venezuela: Novelties and Continuities.” Latin American Perspectives (): –. Lowi, Theodore. . “American Business, Public Policy, Case-Studies, and Political Theory.” World Politics (): –. Luna, Juan Pablo. .“Shaky Foundations? The Transformation of Party-Society Linkages in Post-Transitional Chile and Its Implications for the Quality of Representation.” Paper presented at the Annual Meeting of the Midwest Political Science Association, Chicago. Lustig, Nora. . Coping with Austerity: Poverty and Inequality in Latin America. Washington, D.C.: Brookings Institution. Maccoby, Herbert. . “The Differential Political Activity of Participants in a Voluntary Association.” American Sociological Review (): –. Maceria, Daniel, and María Victoria Murillo. . “Social Sector Reform and the Role of Labor Unions.” Working Paper , . Machado, Jesús. . Estudio de los Consejos Comunales en Venezuela. Caracas: Fundación Centro Gumilla. Magaloni, Beatriz, Alberto Diaz-Cayeros, and Federico Estévez. . “Clientelism and Portfolio Diversification: A Model of Electoral Investment with Applications to Mexico.” In Patrons, Clients, and Policies: Patterns of Democratic Accountability and Political Competition, edited by Herbert Kitschelt and Steven Wilkinson. New York: Cambridge University Press. Mainwaring, Scott. . Rethinking Party Systems in the Third Wave of Democratization: The Case of Brazil. Stanford: Stanford University Press. Mainwaring, Scott, and Timothy Scully. . Christian Democracy in Latin America: Electoral Competition and Regime Conflicts. Stanford: Stanford University Press. Mair, Peter. . Party System Change: Approaches and Interpretations. Oxford: Oxford University Press, Clarendon Press. Manin, Bernard. . The Principles of Representative Government. Cambridge: Cambridge University Press.
14 References.qxd
11/9/2009
11:53 AM
Page 373
,
Manor, James. . The Political Economy of Democratic Decentralization. Washington, D.C.: World Bank. Mansbridge, Jane. . “Rethinking Representation.” American Political Science Review (): –. Manz, Thomas, and Moira Zuazo, eds. . Partidos políticos y representación en América Latina. Caracas: Nueva Sociedad. Manza, Jeff, Michael Hout, and Clem Brooks. . “Class Voting in Capitalist Democracies Since World War II: Dealignment, Realignment, or Trendless Fluctuation?” Annual Review of Sociology :–. Markowski, Radoslaw, and Joshua A. Tucker. . “Political Representation and Euroskepticism: Evidence from Poland.” Paper presented at the European Election Study Conference, Lisbon. McAdam, Doug. . Political Process and the Development of Black Insurgency, –. Chicago: University of Chicago Press. McAdam, Doug, John McCarthy, and Mayer Zald, eds. . “Social Movements.” In The Handbook of Sociology, edited by Neil J. Smelser. Newbury Park, Calif.: Sage Publications. ———. a. Comparative Perspectives on Social Movements: Political Opportunities, Mobilizing Structures, and Cultural Framings. New York: Cambridge University Press. ———. b. “Opportunities, Mobilizing Structures, and Framing Processes: Toward a Synthetic, Comparative Perspective on Social Movements.” In Comparative Perspectives on Social Movements: Political Opportunities, Mobilizing Structures, and Cultural Framings, edited by Doug McAdam, John D. McCarthy, and Mayer N. Zald. New York: Cambridge University Press. ———.c. “Introduction: Opportunities, Mobilizing Structures, and Framing Processes—Toward a Synthetic, Comparative Perspective on Social Movements.” In Comparative Perspectives on Social Movements: Political Opportunities, Mobilizing Structures, and Cultural Framings, edited by Doug McAdam, John D. McCarthy, and Mayer N. Zald. New York: Cambridge University Press. McAdam, Doug, Sidney Tarrow, and Charles Tilly. . The Dynamics of Contention. New York: Cambridge University Press. McCarthy, John D., and Mayer N. Zald. . “Resource Mobilization and Social Movements: A Partial Theory.” The American Journal of Sociology (): –. McConnell, Grant. . Private Power and American Democracy. New York: Knopf. McCoy, Jennifer. . “The Referendum in Venezuela: One Act in an Unfinished Drama.” Journal of Democracy (): –. McGuire, James W. . Peronism Without Perón: Unions, Parties, and Democracy in Argentina. Stanford: Stanford University Press. ———. . “Empowering Civil Society? Decentralizing the State and Increasing Local Participation in Peru.” PhD diss., George Washington University, Washington, D.C. Melucci, Alberto. . “The New Social Movements: A Theoretical Approach.” Social Science Information :–. Melucci, Alberto. . Challenging Codes: Collective Action in the Information Age. Cambridge: Cambridge University Press. Mericle, Kenneth. . “Corporatist Control of the Working Class: Authoritarian Brazil Since .” In Authoritarianism and Corporatism in Latin America, edited by James M. Malloy. Pittsburgh: University of Pittsburgh Press.
14 References.qxd
11/9/2009
11:53 AM
Page 374
.
Mesa-Lago, Carmelo. . “Social Welfare Reform in the Context of Economic-Political Liberalization: Latin American Cases.” World Development (): –. Meyer, David S., and Sidney Tarrow. . “A Movement Society: Contentious Politics for a New Century.” In The Social Movement Society: Contentious Politics for a New Century, edited by David Meyer and Sidney Tarrow. Lanham, Md.: Rowman and Littlefield. Michels, Robert. . Political Parties: A Sociological Study of the Oligarchical Tendencies of Modern Democracy. New York: Hearst’s International Library Co. Mignone, Emilio F. . Witness to the Truth: The Complicity of Church and Dictatorship in Argentina, –. Maryknoll, N.Y.: Orbis Books. Miller, Warren, and J. Merrill Shanks. . The New American Voter. Cambridge: Harvard University Press. Montero, Alfred P., and David J. Samuels, eds. a. Decentralization and Democracy in Latin America. Notre Dame: University of Notre Dame Press. ———. b. “The Political Determinants of Decentralization in Latin America.” In Decentralization and Democracy in Latin America, edited by Alfred P. Montero and David J. Samuels. Notre Dame: University of Notre Dame Press. Moreno, José Luis. . “La caída del estado de bienestar.” In De las cofradías a las organizaciones de la sociedad civil: Historia de la iniciativa asociativa en Argentina, – , edited by Roberto Di Stefano. Buenos Aires: Grupo de Análisis y Desarrollo Institucional y Social . Murillo, María Victoria. . Labor Unions, Partisan Coalitions, and Market Reforms in Latin America. New York: Cambridge University Press. ———. . “Partisanship Amidst Convergence: Labor Market Reforms in Latin America.” Comparative Politics (): –. Murillo, María Victoria, and Lucas Ronconi. . “Teachers’ Strikes in Argentina: Partisan Alignments and Public-Sector Labor Relations.” Studies in Comparative International Development (): –. Murillo, María Victoria, and Andrew Schrank. . “With a Little Help from My Friends: External and Domestic Allies and Labor Rights in Latin America.” Comparative Political Studies (): –. Myers, David. J., and Henry. A. Dietz. . Capital City Politics in Latin America: Democratization and Empowerment. Boulder, Colo.: Lynne Rienner Publishers. Nadaraya, E. A. . “On Estimating Regression.” Theory of Probability and Its Applications :–. Navia, Patricio. . “You Select the Rules of the Game and Lose? Advantages and Constraints When Choosing Electoral Rules: The Case of Chile.” PhD diss., New York University, New York. Nickson, R. Andrew. . Local Government in Latin America. Boulder, Colo.: Lynne Rienner. Nie, Norman H., Jane Junn, and Kenneth Stehlik-Barry. . Education and Democratic Citizenship in America. Chicago: University of Chicago Press. Novaro, Marcos. . Pilotos de tormentas: Crisis de representación y personalización en Argentina (–). Buenos Aires: Ediciones Letra Buena S.A. O’Donnell, Guillermo. . Modernization and Bureaucratic-Authoritarianism: Studies in South American Politics. Berkeley: Institute of International Studies, University of California, Berkeley.
14 References.qxd
11/9/2009
11:53 AM
Page 375
,
———. . “Corporatism and the Question of the State.” In Authoritarianism and Corporatism in Latin America, edited by James Malloy. Pittsburgh: University of Pittsburgh Press. ———. . “Delegative Democracy.” Journal of Democracy (): –. O’Donnell, Guillermo, and Philippe C. Schmitter. . Transitions from Authoritarian Rule: Tentative Conclusions About Uncertain Democracies. Baltimore: Johns Hopkins University Press. Offe, Claus. . “New Social Movements: Challenging the Boundaries of Institutional Politics.” Social Research (): –. Offe, Claus, and Helmut Wiesenthal. . “Two Logics of Collective Action: Theoretical Notes on Social Class and Organizational Form.” Political Power and Social Theory :–. ———. . “Two Logics of Collective Action.” In Disorganized Capitalism: Contemporary Transformation of Work and Politics, edited by Claus Offe and John Keane. Cambridge: MIT Press. Olavarría, Margot. . “Protected Neoliberalism: Perverse Institutionalization and the Crisis of Representation in Post-Dictatorship Chile.” Latin American Perspectives (): –. Olson, Mancur. . The Logic of Collective Action: Public Goods and the Theory of Groups. Cambridge: Harvard University Press. Opp, Karl-Dieter, and Christiane Gern. . “Dissident Groups, Personal Networks, and Spontaneous Collaboration: The East German Revolution of .” American Sociological Review (October): –. Oxhorn, Philip. a. “Understanding Political Change After Authoritarian Rule: The Popular Sectors and Chile’s New Democratic Regime.” Journal of Latin American Studies (): –. ———. b. “Where Did All the Protesters Go? Popular Mobilization and the Transition to Democracy in Chile.” Latin American Perspectives (): –. ———. . Organizing Civil Society: The Popular Sectors and the Struggle for Democracy in Chile. University Park: Pennsylvania State University Press. ———. a. “Is the Century of Corporatism Over? Neoliberalism and the Rise of Neopluralism.” In What Kind of Democracy? What Kind of Market? edited by Philip Oxhorn and Graciela Ducatenzeiler. University Park: Pennsylvania State University Press. ———. b. “The Social Foundations of Latin America’s Recurrent Populism: Problems of Popular Sector Class Formation and Collective Action.” Journal of Historical Sociology (): –. ———. . “Unraveling the Puzzle of Decentralization.” In Decentralization, Democratic Governance, and Civil Society in Comparative Perspective: Africa, Asia, and Latin America, edited by Philip Oxhorn et al. Washington, D.C.: Woodrow Wilson Center Press. Oxhorn, Philip, and Graciela Ducatenzeiler. a. “Economic Reform and Democratization in Latin America.” In What Kind of Democracy? What Kind of Market?: Latin America in the Age of Neoliberalism, edited by Philip Oxhorn and Graciela Ducatenzeiler. University Park: Pennsylvania State University Press. ———. b. What Kind of Democracy? What Kind of Market?: Latin America in the Age of Neoliberalism. University Park: Pennsylvania State University Press.
14 References.qxd
11/9/2009
11:53 AM
Page 376
.
Pakulski, Jan. . “Foundations of a Post-Class Analysis.” In Approaches to Class Analysis, edited by Erik Olin Wright. New York: Cambridge University Press. Palermo, Vicente, and Inés González Bombal. . “La política local.” In Los nuevos movimientos sociales, edited by Elizabeth Jelin. Buenos Aires: Centro Editor de América Latina. Panebianco, Angelo. . Political Parties: Organization and Power. New York: Cambridge University Press. Panfinchi, Aldo. . “Sociedad civil y democracia en los Andes y el Cono Sur a inicios del siglo XXI.” In Sociedad civil, esfera pública y democratización en América Latina: Andes y Cono Sur, edited by Aldo Panfinchi. Lima: Pontífica Universidad Católica del Perú. Parry, Jeraint, George Moyser, and Neil Day. . Political Participation and Democracy in Britain. Cambridge: Cambridge University Press. Pásara, Luis, N. Delpino, R. Valdeavellano, and A. Zarzar. . La otra cara de la luna: Nuevos actores sociales en el Perú. Buenos Aires: Manantial. Patterson, Eric. . “Religious Activity and Political Participation: The Brazilian and Chilean Cases.” Latin American Politics and Society (): –. Peattie, Lisa. . “An Idea in Good Currency and How It Grew: The Informal Sector.” World Development (): –. Penfold-Becerra, Michael. . “Clientelism and Social Funds: Evidence from Chávez’s Misiones.” Latin American Politics and Society (): –. Peruzzotti, Enrique. .“Emergencia, desarollo, crisis y reconstrucción de la sociedad civil argentina.” In Sociedad civil, esfera pública y democratización en América Latina: Andes y Cono Sur, edited by Aldo Panifinchi. Lima: Pontífica Universidad Católica del Perú. ———. a. “Redefiniendo la representación política: La sociedad civil argentina y el sistema representiva en los noventa.” Política y Gobierno (): –. ———. b. “Representation, Accountability and Civil Society.” Paper presented at the Annual Meetings of the Latin American Studies Association, Las Vegas. Peruzzotti, Enrique, and Catalina Smulovitz. . “Societal Accountability in Latin America.” Journal of Democracy (): –. ———, eds. . Enforcing the Rule of Law: Social Accountability in the New Latin American Democracies. Pittsburgh: University of Pittsburgh Press. Petras, James. . “The Political and Social Basis of Regional Variation in Land Occupations in Brazil.” Journal of Peasant Studies (): –. Piester, Kerianne. .“Targeting the Poor: The Politics of Social Policy Reforms in Mexico.” In The New Politics of Inequality in Latin America: Rethinking Participation and Representation, edited by Douglas. A. Chalmers, Carlos M. Vilas, Katherine Hite, Scott B. Martin, Kerianne Piester, and Monique Segarra New York: Oxford University Press. Pitkin, Hanna. . The Concept of Representation. Berkeley and Los Angeles: University of California Press. Piven, Frances Fox, and Richard A. Cloward. . Poor People’s Movements: Why They Succeed, How They Fail. New York: Pantheon Books. Pollock, Philip H. . “Organizations as Agents of Mobilization: How Does Group Activity Affect Political Participation?” American Journal of Political Science :–. Portes, Alejandro, and Kelly Hoffman. . “Latin American Class Structures: Their Composition and Change During the Neoliberal Era.” Latin American Research Review (): –.
14 References.qxd
11/9/2009
11:53 AM
Page 377
,
Portes, Alejandro, and Patricia Landolt, . “Social Capital: Promise and Pitfalls of Its Role in Development.” Journal of Latin American Studies (): –. Posner, Paul W. . “Popular Representation and Political Dissatisfaction in Chile’s New Democracy.” Journal of Interamerican Studies and World Affairs (): –. ———. . “Local Democracy and the Transformation of Popular Participation in Chile.” Latin American Politics and Society (): –. Powers, Nancy R. . Grassroots Expectations of Democracy and Economy: Argentina in Comparative Perspective. Pittsburgh: University of Pittsburgh Press. Pratt, Raymond. . “Parties, Neighborhood Associations, and the Politicization of the Urban Poor in Latin America: An Exploratory Analysis.” Midwest Journal of Political Science (): –. Putnam, Robert D., Robert Leonardi, and Raffaella Nanetti. . Making Democracy Work: Civic Traditions in Modern Italy. Princeton: Princeton University Press. Rabkin, Rhoda. . “Redemocratization, Electoral Engineering, and Party Strategies in Chile, –.” Comparative Political Studies (): –. Raczynski, Dagmar. . “The Crisis of Old Models of Social Protection in Latin America: New Alternatives for Dealing with Poverty.” In Poverty and Inequality in Latin America: Issues and New Challenges, edited by Víctor E. Tokman and Guillermo O’Donnell. Notre Dame: University of Notre Dame Press. Reilly, Charles A., ed. a. New Paths to Democratic Development in Latin America: The Rise of -Municipal Collaboration. Boulder, Colo.: Lynne Rienner. ———. b. “Topocrats, Technocrats, and .” In New Paths to Democratic Development in Latin America: The Rise of -Municipal Collaboration, edited by Charles A. Reilly. Boulder, Colo.: Lynne Rienner. Renaud, Coromoto, et al. . La participación de la sociedad civil en la constitución de . Caracas: . Richards, Patricia, and Bryan R. Roberts. . “Social Networks, Social Capital, Popular Organizations, and Urban Poverty: A Research Note.” Paper presented at the Seminar on Urban Poverty Sponsored by and the World Bank, Rio de Janeiro. Riker, William H. . Liberalism Against Populism: A Confrontation Between the Theory of Democracy and the Theory of Social Choice. San Francisco: W. H. Freeman. Roberts, Bryan R., and Alejandro Portes. . “Coping with the Free Market City: Collective Action in Six Latin American Cities at the End of the Twentieth Century.” Latin American Research Review (): –. Roberts, Kenneth. . “Neoliberalism and the Transformation of Populism in Latin America.” World Politics (): –. ———. . “Rethinking Economic Alternatives: Left Parties and the Articulation of Popular Demands in Chile and Peru.” In The New Politics of Inequality in Latin America: Rethinking Participation and Representation, edited by Douglas A. Chalmers, Carlos M. Vilas, Katherine Hite, Scott B. Martin, Kerianne Piester, and Monique Segarra. Oxford: Oxford University Press. ———. . Deepening Democracy? The Modern Left and Social Movements in Chile and Peru. Stanford: Stanford University Press. ———. . “Social Inequalities Without Class Cleavages in Latin America’s Neoliberal Era.” Studies in Comparative International Development (): –. ———. . “Populism, Political Conflict, and Grass-Roots Organization in Latin America.” Comparative Politics (): .
14 References.qxd
11/9/2009
11:53 AM
Page 378
.
———. Forthcoming. Party System Change in Latin America’s Neoliberal Era. New York: Cambridge University Press. Roever, Sally. . “Negotiating Formality: Informal Sector, Market, and State in Peru.” PhD diss., University of California, Berkeley. Rogowski, Ronald. . “Social Class and Partisanship in European Electorates: A ReAssessment.” World Politics (): –. Roitter, Mario, Regina List, and Lester M. Salamon. . “Argentina.” In Global Civil Society: Dimensions of the Nonprofit Sector, edited by Lester M. Salamon et al. Baltimore: Johns Hopkins Center for Civil Society Studies. Romero, Catalina. . “La democracia en el Perú: Una terca voluntad.” América Latina Hoy :–. Romero, Luis Alberto. . “El estado y las corporaciones.” In De las cofradías a las organizaciones de la sociedad civil: historia de la iniciativa asociativa en Argentina, – , edited by Roberto Di Stefano. Buenos Aires: Grupo de Análisis y Desarrollo Institucional y Social . Rose, Richard, and Derek W. Urwin. . “Persistence and Change in Western Party Systems Since .” Political Studies (): –. Rosenstone, Steven J., and John Mark Hansen. . Mobilization, Participation, and Democracy in America. New York: Macmillan. Rubin, Jeffrey W. . “In the Streets or in the Institutions?” Forum (): –. Sábato, Hilda. . The Many and the Few: Political Participation in Republican Buenos Aires. Stanford: Stanford University Press. ———. . “Estado y sociedad civil.” In De las cofradías a las organizaciones de la sociedad civil: historia de la iniciativa asociativa en Argentina, –, edited by Roberto Di Stefano. Buenos Aires: Grupo de Análisis y Desarrollo Institucional y Social . Salamon, Lester M., et al., eds. . Global Civil Society: Dimensions of the Nonprofit Sector. Baltimore: Johns Hopkins Center for Civil Society Studies. Sanborn, Cynthia, et al. . “Peru.” In Global Civil Society: Dimensions of the Nonprofit Sector, edited by Lester M. Salamon et al. Baltimore: Johns Hopkins Center for Civil Society Studies. Schattan Coelho, Vera, and Marcos Nobre. . Participação e deliberação: Teoria democrática e experiências institucionais no Brasil contemporâneo. São Paulo: Editora . Schattschneider, E. E. . The Semisovereign People: A Realist’s View of Democracy in America. New York: Holt, Rinehart and Winston. Schlozman, Kay Lehman. . “Interest Groups.” In The International Encyclopedia of the Social and Behavioral Sciences, edited by Neil Smelser and Paul Baltes. Amsterdam: Pergamon Press. Schlozman, Kay Lehman, and Sidney Verba. . The Private Roots of Public Action: Gender, Equality, and Political Participation. Cambridge: Harvard University Press. Schmitter, Philippe C. . Interest Conflict and Political Change in Brazil. Stanford: Stanford University Press. ———. . “Still the Century of Corporatism?” World Politics (): –. ———. . “Corporatism and Policy-Making in Contemporary Western Europe.” Comparative Political Studies (): –. ———. . “Interest Systems and the Consolidation of Democracies.” In Reexamining
14 References.qxd
11/9/2009
11:53 AM
Page 379
,
Democracy: Essays in Honor of Seymour Martin Lipset, edited by Gary Marks and Larry Diamond. Thousand Oaks, Calif.: Sage. Schneider, Ben Ross. . Politics Within the State: Elite Bureaucrats and Industrial Policy in Authoritarian Brazil. Pittsburgh: University of Pittsburgh Press. ———. . Business Politics and the State in Latin America. New York: Cambridge University Press. Schneider, Cathy Lisa. . Shantytown Protest in Pinochet’s Chile. Philadelphia: Temple University Press. Schönwälder, Gerd. . Linking Civil Society and the State: Urban Popular Movements, the Left, and Local Government in Peru, –. University Park: Pennsylvania State University Press. ———. . “Metropolitan Lima: A New Way of Making Politics.” In The Left in the City: Participatory Local Governments in Latin America, edited by Daniel Chávez and Benjamin Goldfrank. London: Latin America Bureau. Scott, John. . Social Network Analysis: A Handbook. London: Sage Publications. Seawright, Jason W. . “Crisis of Representation: Voters, Party Organizations, and PartySystem Collapse in South America.” PhD diss., University of California, Berkeley. Seidman, Gay W. . Manufacturing Militance. Berkeley and Los Angeles: University of California Press, . Seligson, Amber L. . “Civic Association and Democratic Participation in Central America: A Test of the Putnam Thesis.” Comparative Political Studies :–. Seligson, Mitchell. . “The Rise of Populism and the Left in Latin America.” Journal of Democracy :. Sen, Amartya. . Development as Freedom. New York: Knopf. Shadlen, Kenneth. . “Orphaned by Democracy: Small Industry in Contemporary Mexico.” Comparative Politics (): –. Sheahan, John. . Patterns of Development in Latin America: Poverty, Repression, and Economic Strategy. Princeton: Princeton University Press. Shi, Tianjian. . Political Participation in Beijing. Cambridge: Harvard University Press. Silver, Beverly J. . Forces of Labor: Workers’ Movements and Globalization Since . New York: Cambridge University Press. Simonoff, Jeffrey S. . Smoothing Methods in Statistics. New York: Springer. Smith, Brian H. . The Church and Politics in Chile: Challenges to Modern Catholicism. Princeton: Princeton University Press. Snyder, Richard, and David J. Samuels. . “Legislative Malapportionment in Latin America: Historical and Comparative Perspectives.” In Federalism and Democracy in Latin America, edited by Edward Gibson. Baltimore: Johns Hopkins University Press. Stallings, Barbara. . “International Influence on Economic Policy: Debt, Stabilization, and Structural Reform.” In The Politics of Economic Adjustment: International Constraints, Distributive Conflicts, and the State, edited by Stephan Haggard and Robert Kaufman. Princeton: Princeton University Press. Stanley, Harold W., William T. Bianco, and Richard G. Niemi. .“Partisanship and Group Support over Time: A Multivariate Analysis.” American Political Science Review (): –. Stepan, Alfred. . The State and Society: Peru in Comparative Perspective. Princeton: Princeton University Press.
14 References.qxd
11/9/2009
11:53 AM
Page 380
.
———. . “Religion, Democracy, and the ‘Twin Tolerations.’” Journal of Democracy (): –. Stirrat, Roderick L., and Heiko Henkel. . “The Development Gift: The Problem of Reciprocity in the World.” Annals of the American Academy of Political and Social Science :–. Stokes, Susan C. . Cultures in Conflict: Social Movements and the State in Peru. Berkeley and Los Angeles: University of California Press. ———. . Mandates and Democracy: Neoliberalism by Surprise in Latin America. New York: Cambridge University Press. ———. . “Perverse Accountability: A Formal Model of Machine Politics with Evidence from Argentina.” American Political Science Review (): –. Streeck, Wolfgang. . Social Institutions and Economic Performance: Studies of Industrial Relations in Advanced Capitalist Economies. London: Sage. Streeck, Wolfgang, and Philippe C. Schmitter. . “Community, Market, State—and Association?” In Private Interest Government and Public Policy, edited by Wolfgang Streeck and Philippe C. Schmitter. London: Sage. Streeten, Paul. . “Nongovernmental Organizations and Development.” Annals of the American Academy of Political and Social Science :–. Svampa, Maristella, and Sebástian Pereyra. . “Entre la ruta y el barro.” Biblos-Universidad Nacional de General. Buenos Aires: Biblos. Székely, Miguel, and Marianne Hilgert. . The s in Latin America: Another Decade of Persistent Inequality. Washington, D.C.: Inter-American Development Bank. Szwarcberg, Mariela. . “Making Local Democracy: Political Machines, Clientelism, and Social Networks in Argentina.” PhD diss., University of Chicago. Tabatabai, Hamid. . Statistics on Poverty and Income Distribution: An Compendium of Data. Geneva: International Labour Office. Tanaka, Martín. . Los espejismos de la democracia: El colapso del sistema de partidos en el Perú, -, en Perspectiva Comparada. Lima: Instituto de Estudios Peruanos. Tarrow, Sidney G. . “Social Movements and National Politics.” Annual Review of Sociology :–. ———. . Power in Movement: Social Movements and Contentious Politics. nd ed. New York: Cambridge University Press. ———. . The New Transnational Activism. New York: Cambridge University Press. Tendler, Judith. . Good Government in the Tropics. Baltimore: Johns Hopkins University Press. Thayer, Millicent S. . “Negotiating the Global: Northeast Brazilian Women’s Movements and the Transnational Feminist Public.” PhD diss., University of California, Berkeley. Thompson, Andrés, ed. . Público y privado. Buenos Aires: Losada. Thompson, E. P. . The Making of the English Working Class. London: V. Gollancz. Thornton, Douglas. . “Political Attitudes and Participation of Informal and Formal Sector Workers in Mexico.” Comparative Political Studies (): –. Thorp, Rosemary. . Progress, Poverty, and Exclusion: An Economic History of Latin America in the Twentieth Century. Washington, D.C.: Inter-American Development Bank. Tilly, Charles. . From Mobilization to Revolution. Reading: Addison-Wesley. ———. . “Afterword: An Agenda for Students of Social Movements.” In States, Parties,
14 References.qxd
11/9/2009
11:53 AM
Page 381
,
and Social Movements, edited by Jack A. Goldstone. Cambridge: Cambridge University Press. ———. . Regimes and Repertoires. . Chicago: University of Chicago Press. Torcal, Mariano, and Scott Mainwaring. . “The Political Recrafting of Social Bases of Party Competition: Chile, –.” British Journal of Political Science (): –. Touraine, Alain. . The May Movement: Revolt and Reform. New York: Random House. ———. . The Voice and the Eye: An Analysis of Social Movements. New York: Cambridge University Press; Paris: Maison des Sciences de l’Homme. Tova Samanez, Teresa. . “Vecinos y pobladores en la crisis (–).” In Movimientos sociales y crisis: El caso peruano, edited by Eduardo Ballón. Lima: Centro de Estudios y Promoción del Desarrollo. Truman, David. . The Governmental Process: Political Interests and Public Opinion. New York: Knopf. Uhlaner, Carole J. . “The Impact of Perceived Representation on Latino Political Participation.” Working Paper :, Center for the Study of Democracy, University of California, Irvine. Uquillas, Jorge, and Martien Van Nieuwkoop. . “Social Capital as a Factor in Indigenous Peoples Development in Ecuador.” Sustainable Development Working Paper , World Bank. Veltmeyer, Henry. . “Civil Society and Social Movements: The Dynamics of Intersectoral Alliances and Urban-Rural Linkages in Latin America.” Paper presented at the United Nations Research Institute for Social Development, Geneva. Verba, Sidney, Kay Lehman Schlozman, and Henry E. Brady. . Voice and Equality: Civic Voluntarism in American Politics. Cambridge: Harvard University Press. Verba, Sidney, and Norman Nie. . Participation in America: Political Democracy and Social Equality. New York: Harper and Row. Verba, Sidney, Norman H. Nie, and Jae-on Kim. . Participation and Political Equality: A Seven-Nation Comparison. New York: Cambridge University Press. Walker, Jack L. . Mobilizing Interest Groups in America: Patrons, Professions, and Social Movements. Ann Arbor: University of Michigan Press. Watson, G. S. . “Smooth Regression Analysis.” Sankhaya: The Indian Journal of Statistics A :–. Weber, Max. . The Theory of Social and Economic Organization. Translated by A. M. Henderson and Talcott Parsons. New York: Oxford University Press. Weyland, Kurt. a. Democracy Without Equity: Failures of Reform in Brazil. Pittsburgh: University of Pittsburgh Press. ———. b. “Neopopulism and Neoliberalism in Latin America: Unexpected Affinities.” Studies in Comparative International Development (): –. ———. . “Neoliberal Populism in Latin America and Eastern Europe.” Comparative Politics (): –. ———. . “The Growing Sustainability of Brazil’s Low-Quality Democracy.” In The Third Wave of Democratization in Latin America: Advances and Setbacks. New York: Cambridge University Press. ———. . “Assessing Latin American Neoliberalism: Introduction to a Debate.” Latin American Research Review (): –.
14 References.qxd
11/9/2009
11:53 AM
Page 382
.
Wielhouwer, Peter W., and Brad Lockerbie. . “Party Contacting and Political Participation: –.” American Journal of Political Science (): –. Wolfinger, Raymond, and Steven Rosenstone. . Who Votes? New Haven: Yale University Press. Woodruff, David M. . “Boom, Gloom, and Doom: Balance Sheets, Monetary Fragmentation, and the Politics of Financial Crisis in Argentina and Russia.” Politics and Society (): –. World Bank. . Together We Stand, Divided We Fall: Levels and Determinants of Social Capital in Argentina. Washington, D.C.: World Bank. Wright, Erik Olin. . “Working-Class Power, Capitalist-Class Interests, and Class Compromise.” American Journal of Sociology (): –. Yashar, Deborah. . Contesting Citizenship in Latin America: The Rise of Indigenous Movements and the Polistliberal Challenge. New York: Cambridge University Press. Yates, Douglas. . Neighborhood Democracy: The Politics and Impacts of Decentralization. Lexington, Mass.: Lexington Books. Zucco, Cesar, Jr. . “Where’s the Bias? A Reassessment of the Chilean Electoral System.” Electoral Studies (): –.
15 Index.qxd
11/9/2009
11:53 AM
Page 383
INDEX
. access A-Net and, , –, , –, –, –, , , –, –, , in Argentina, –, – autonomy and, , , , , , –, –, , –, –, in Chile, –, , as interest regime dimension, , –, – labor unions and, , , , , , , in Peru, –, – political parties and, –, –, , – -Hub and, , –, , –, in Venezuela, –, –, accessing strategies. See also claim making claim making and, , , – class and, – embeddedness and, Flexible Fronts and, at local level, –, –, , –, –, , , , –, at national level, –, , , , –, –, , , , , –, , , , , , –, , , , s and, patterns of, – protest and, scaling and, , state ties and, , –, , –, –, –, –, A-Net. See also associations accessing strategies and, , –, , –, –, –, , , –, –, , in Argentina, –, –, –, –, – autonomy and, , , –, –, –, , –, –, , – in Chile, –, –, –, –, –, – class and, –, –, – collective action and, –, –, – composition of, –, –, –, –, –, – demands and, , –
economic model and, , –, , , , , –, – embeddedness and, , – Flexible Fronts and, –, interest convergence and, , – labor unions and, , , –, –, –, –, –, –, , , , Nodal s and, –, participation and, –, –, –, –, –, –, –, , –, –, –, party activism and, – party electoral mobilization and, – party identification and, – in Peru, –, –, –, –, –, –, – political parties and, , –, –, –, –, –, –, , , , , – resources and, –, rise of, –, –, –, –, – scaling and, , , , , , , –, –, , , , , –, scope of, , , –, , –, –, –, –, , –, –, –, shared-member linkages and, , – state and, , –, –, –, –, –, –, –, –, –, – state policy and, – in Venezuela, –, –, –, –, – voting and, – Argentina access in, –, – A-Net in, –, –, –, –, – anticrime associations in, –, – associations in, , –, –, –, –, , –, –, –, –, –, –, –, –, –, –, –, –, –, –, –, – autonomy in, , –, –
15 Index.qxd
11/9/2009
11:53 AM
Page 384
.
Argentina (continued) capacity building in, –, –, –, –, – claim making in, –, , –, –, –, –, – class in, , –, –, –, , , –, –, – cooperatives in, –, – direct action in, –, –, electoral arena in, , , –, –, –, – embeddedness in, –, –, –, –, Flexible Fronts in, –, –, – focus districts in, – food-distribution associations in, –, – labor unions in, , , , –, , –, , –, , –, –, –, –, party electoral mobilization in, –, –, – neighborhood associations in, –, – Nodal s in, –, –, –, parents associations in, –, – participation in, –, , , –, –, –, –, – party activism in, –, – party identification in, –, –, – place-of-origin associations in, –, – political parties in, –, –, , , , –, , –, , , –, , –, –, –, –, –, –, problem solving in, –, –, –, –, –, –, – protest in, –, –, –, – provisioning in, –, , –, –, –, –, – recreational associations in, –, – religion-based associations in, –, – representational distortion in, , , –, –, – scaling in, , –, –, –, –, –, –, scope in, –, –, – shared-member linkages in, – state ties in, –, –, –, –, –, –, –, – voting in, –, –, –
associations. See also A-Net; interest organizations accessing strategies and, –, , –, –, –, , , –, –, , , –, –, anticrime associations, –, –, – in Argentina, , –, –, –, –, , –, –, –, –, –, –, –, –, –, –, –, –, –, –, –, – atomized, –, , –, – autonomy and, , , –, , , –, , , –, –, , –, –, , – bank accounts of, , , , –, –, budgets of, , , , –, –, –, , – capacity building by, , –, –, –, –, –, –, –, – in Chile, , –, –, –, –, , –, –, –, –, –, –, –, –, –, –, –, –, –, –, –, –, claim making by, , , , , , , , , , –, , –, –, –, –, –, , –, –, –, –, , , –, –, –, , –, – class and, –, –, –, –, , –, – clientelism and, –, –, –, collective action and, –, –, –, –, –, , confidence in, , , cooperatives, –, –, – demands and, , – electoral arena and, , –, –, , –, –, – embeddedness and, –, –, , , –, –, –, –, –, –, –, – Flexible Fronts, –, , –, food-distribution associations, –, –, – funding, –, –, –, , –, –, – gender and, – interest convergence and, , –
15 Index.qxd
11/9/2009
11:53 AM
Page 385
internal democracy and, , –, –, , – international actors and, , –, –, – labor unions and, , –, –, –, –, –, –, –, –, –, land invasions and, – legal processes and, , , –, –, – membership in, –, , –, –, , – neighborhood associations, –, –, –, –, –, – Nodal s and, –, , – organizational traits of, – parents associations, –, –, – participation in, –, –, , –, – , –, –, –, –, –, , , , , , –, –, –, , –, –, –, party activism and, – party electoral mobilization and, – party identification and, – in Peru, , –, –, –, –, , –, –, –, –, –, –, –, –, –, –, –, –, –, –, –, –, –, place-of-origin associations, –, , –, – policy and, –, , –, –, , – political parties and, , , –, , –, –, –, –, –, –, , –, , –, –, – problem-solving strategies of, –, , –, –, –, –, –, –, –, –, –, , , –, – protest by, , , –, –, –, –, –, –, , , , –, –, –, – provisioning by, , , , , , –, –, , , –, –, –, –, –, , , –, , – recreational associations, –, , –, – religion-based associations, –, , –, – representation and, , , –, –, –, –, , ,
,
resources and, –, , –, –, , , –, –, –, –, , –, –, , – scaling and, , , –, , –, , –, –, , , –, –, –, –, –, –, –, –, –, , , , , – scope and, , –, , –, –, –, –, , –, –, – shared-member linkages and, , –, social policy and, , – staff of, –, , –, –, state and, , , , , , , –, –, , –, –, –, –, , , –, –, –, –, –, , –, –, , –, –, –, –, –, –, – state recognition of, –, , –, –, unemployment associations, – in -Hub, –, – in Venezuela, , –, –, –, –, , –, –, –, –, –, –, –, –, –, –, –, –, –, –, –, voting and, – autonomy. See also dependence access and, , , , , , –, –, , –, –, in A-Net, , , –, –, –, , –, –, , – in Argentina, , –, – in Chile, , –, claim making and, , corporatism and, Flexible Fronts and, , , –, funding and, , –, – as interest regime dimension , , –, – labor unions and, , , , , new social movements and, Nodal s and, , , –, in Peru, , –, – political parties and, –, postcorporatism and, protest and, – provisioning and, – resources and, scaling and, , , –,
15 Index.qxd
11/9/2009
11:53 AM
Page 386
.
autonomy (continued) state and, , , –, , , –, , –, –, –, , –, –, , – -Hub and, , , – in Venezuela, , –, –, Bachelet, Michele, Barrantes, Alfonso, – Belaúnde, Fernando, capacity building in Argentina, –, –, –, –, – by associations, –, –, –, –, –, –, autonomy and, in Chile, –, –, –, –, – definition of, embeddedness and, –, , –, – Flexible Fronts and, –, – incidence of, –, – Nodal s and, –, – in Peru, –, –, –, –, – protest compared with, – resources and, – scaling and, , –, – state ties and, – in Venezuela, –, –, –, –, – Chávez, Hugo, n., , , –, , , – Chile access in, –, , A-Net in, –, –, –, –, –, – anticrime associations in, –, – associations in, , –, –, –, –, , –, –, –, –, –, –, –, –, –, –, –, –, –, –, –, –, autonomy in, , –, capacity building in, –, –, –, –, – claim making in, –, , –, –, –, –, – class in, , –, –, –, , , –, –, –
cooperatives in, –, – direct action in, –, –, electoral arena in, , , –, –, –, – embeddedness in, –, –, –, –, – Flexible Fronts in, –, –, – focus districts in, – food-distribution associations in, –, – labor unions in, , , , , –, , –, –, , –, –, –, –, neighborhood associations in, –, – Nodal s in, –, –, – parents associations in, –, – participation in, –, , , –, –, –, –, party activism in, –, – party electoral mobilization in, –, –, – party identification in, –, –, – place-of-origin associations in, –, – political parties in, –, , , , , –, , –, , , –, –, –, , –, –, –, –, , problem solving in, –, –, –, –, –, –, – protest in, –, –, –, – provisioning in, –, , –, –, –, –, – recreational associations in, –, – religion-based associations in, –, – representational distortion in, , , –, –, – scaling in, , –, –, –, –, –, –, scope in, –, –, –, shared-member linkages in, – state ties in, –, –, –, –, –, –, –, – -Hub in, voting in, –, –, – civil society. See associations; postcorporatism claim making. See also demands; direct action; intermediaries accessing strategies and, , , –
15 Index.qxd
11/9/2009
11:53 AM
Page 387
in Argentina, –, , –, –, –, –, – associational participation and, , , –, – by associations, , , , , , , , , , , –, –, –, –, –, , –, –, –, –, , , –, –, –, , –, – autonomy and, , in Chile, –, , –, –, –, –, – clientelism and, – corporatism and, embeddedness and, –, , –, –, – Flexible Fronts and, –, –, – incidence of, –, – labor unions and, , , , – at local level, , –, –, , , , –, , , , – at national level, , , , –, , , –, –, –, –, , –, –, , , , , , , , –, –, , – Nodal s and, –, –, – in Peru, –, , –, –, –, –, – political parties and, – protest compared with, , – representation and, – resources and, , – scaling and, , –, –, – in Venezuela, –, , –, –, –, – class. See also popular sectors; representational distortion; working classes A-Net and, –, –, – anticrime associations and, –, – in Argentina, , –, –, –, , , –, –, – associations and, –, –, –, –, , –, – in Chile, , –, –, –, , , –, –, – collective action and, –, –, – cooperatives and, –, – direct action and, , , –, –, , – education and, –, , –, –, – electoral arena and, –, –
,
food-distribution associations and, –, – gender and, , income and, , , , –, – interest arena and, –, – labor unions and, , –, – materialist dimensions of, – middle class, n., , neighborhood associations and, –, – parents associations and, –, – participation and, –, , –, –, –, –, , –, – party activism and, –, – party electoral mobilization and, –, –, – party identification and, –, –, – in Peru, , –, –, –, , , –, –, – place-of-origin associations and, –, – political activity and, political parties and, –, – recreational associations and, –, – religion-based associations and, –, – representation and, –, –, , –, – resources and, , rights movements and, – shared-member linkages and, , – in Venezuela, , –, –, –, , , –, –, – voting and, –, –, – clientelism associations and, –, –, –, direct action and, –, , , political parties and, –, – in Venezuela, collective action in interest regimes, –, –, – organizational traits and, – scaling and, , , –, , , –, scope and, , , –, state traits and, , , –, – contacting. See claim making contention. See protest coordinating associations. See also Flexible Fronts; Nodal s in A-Net generally, , , , –, , claim making and, , , –
15 Index.qxd
11/9/2009
11:53 AM
Page 388
.
coordinating associations (continued) individual collective action contrasted, –, –, – limitations on, provisioning and, , scaling and, –, , , , , types of, , , , union-linked associations as, , – in -Hub, , corporatism, , –, , de la Rúa, Fernando, , dependence. See autonomy; state ties descriptive representation, –, direct action in Argentina, –, –, atomized, –, , –, –, – in Chile, –, –, class and, , , –, –, , – clientelism and, –, , collective, –, , –, –, – contacting government, , –, –, , , contacting influential individuals, –, –, , , , contacting political parties, –, –, , , , definition of, , , – efficacy of, –, –, gender and, – incidence of, –, intermediaries and, –, –, , land invasions and, – legal processes as, –, –, in Peru, –, –, petitioning as, –, –, , political parties and, –, –, protest as, –, –, , , recreational associations and, – religion-based associations and, – resources and, –, , , –, , union membership and, – in Venezuela, –, –, Duhalde, Eduardo, , electoral arena in Argentina, , , –, –, –, – associations and, , –, –, , –, –, – in Chile, , , –, –, –, –
class and, –, – definition of, – equality in, – interest arena compared with, –, participation in, , , –, , – party activism, –, – party electoral mobilization, –, –, – party identification, –, –, – in Peru, , , –, –, –, – problem solving and, , –, – representation and, , , , –, , –, shared-member linkages and, , – in Venezuela, , , –, –, –, – voting in, –, –, – embeddedness. See also scaling accessing strategies and, A-Net and, , – in Argentina, –, –, –, –, – capacity building and, –, , –, – in Chile, –, –, –, –, – claim making and, –, , –, –, – definition of, , – electoral arena and, , , – incidence of, , –, – labor unions and, –, –, –, – legal processes and, – in Peru, –, –, –, –, – political parties and, , , –, – problem solving and, –, –, –, –, , –, –, – protest and, –, , –, –, – provisioning and, – resources and, –, – scaling, –, –, –, – state ties and, , –, – in Venezuela, –, –, –, – Flexible Fronts in Argentina, –, –, – capacity building by, –, –
15 Index.qxd
11/9/2009
11:53 AM
Page 389
in Chile, –, –, – claim making by, –, –, – definition of, , – dependence and, , , –, incidence of, –, – internal democracy and, , –, –, , – international actors and, , –, –, – participation in, –, –, , – in Peru, –, –, –, protest by, , –, –, – provisioning by, – state recognition of, –, –, resources and, –, –, , –, – scaling and, –, , –, state ties and, –, –, –, –, – in Venezuela, –, –, – focus districts, – Fujimori, Alberto. , n., nn., , , , , funding. See also resources; state ties of associations, –, –, –, , –, –, – dependence and, , –, – of Flexible Fronts, – by international actors, –, – of Nodal s, – by political parties, , – by state, , –, , –, –, – García, Alan, , gender, , –, – interest arena. See also direct action; interest regime associations in, , class and, –, –, – definition of, n., – electoral arena compared with, –, equality in, – individual participation in, –, –, –, –, – political activity in, – representation and, –, –, –, – interest organizations, –, , , , – –. See also associations; labor unions interest regime. See also A-Net; interest arena; -Hub
,
access of, , , – autonomy of, , , , – class and, – corporatism and, , – definition of, dimensions of, , – pluralism and, –, – political parties and, , postcorporatism and, , – representation and, –, –, , , , – scaling of, , – scope of, , – shift in, –, –, – intermediaries, , –, –, , . See also claim making; direct action international actors, , –, –, – Kirchner, Néstor, , –, – labor-based parties. See also political parties in A-Net, , , , –, , collective action and, economy and, , government and, in focus districts, , , , –, –, , – in labor movement, , – in -Hub, , , –, , , , , , , , , labor unions. See also -Hub accessing strategies and, , , , , , , A-Net and, , , –, –, –, –, –, –, , , , in Argentina, , , , –, , –, , –, , –, –, –, –, associations and, , –, –, –, –, –, –, –, –, –, dependence and, , , , , in Chile, , , , , –, , –, –, , –, –, –, –, claim making and, , , , – collective action and, , , –, – confidence in, corporatism, – demands of, , – direct action and, – electoral arena and, –
15 Index.qxd
11/9/2009
11:53 AM
Page 390
.
labor unions (continued) embeddedness and, –, –, –, – interest convergence and, , – organizational traits of, , – participation and, , –, , , –, –, –, –, – in Peru, , , , –, , –, –, , –, –, –, –, political parties and, , , –, –, –, , , –, , – protest and, – public sector unions, –, – resources and, –, , , , , – scaling and, , –, , , –, , –, , , , –, –, –, –, –, –, –, scope and, , –, social movement unions, –, – state policy and, –, , -Hub and, –, , –, –, –, –, , in Venezuela, , , –, , –, , –, , –, –, –, – Lagos, Ricardo, , , land invasions, –, – legal processes associations and, , , –, –, – as direct action, –, –, embeddedness and, – resources and, – lower classes. See popular sectors materialist issues, –, Menem, Carlos, , , , middle classes, , . See also class; representational distortion Nodal s in Argentina, –, –, –, capacity building by, –, – in Chile, –, –, – claim making by, –, –, – definition of, n., , – dependence and, , , –, incidence of, –, – internal democracy and, , –, –, , –
international actors and, , –, –, – participation in, –, –, , – in Peru, –, –, –, – protest by, , –, –, – provisioning by, – resources and, –, –, , –, – scaling and, –, , –, state recognition of, –, –, state ties and, –, –, –, –, – in Venezuela, –, –, – nongovernmental organizations. See Nodal s organizational infrastructure. See interest regime participation. See also direct action; problemsolving strategies; shared-member linkage; scope A-Net and, –, –, –, –, –, –, –, , –, –, –, in anticrime associations, –, – in Argentina, –, , , –, –, –, –, – in associations, –, –, , –, – , –, –, –, –, –, , , , , , –, –, –, , –, –, –, breadth of, , – in Chile, –, , , –, –, –, –, claim making and, , , , –, –, , – class and, , –, –, –, –, , –, – collective action and, , –, , , –, –, – in cooperatives, –, –, – demands and, density of, – direct action, –, –, –, – duration of, , in electoral arena, , , –, , – in Flexible Fronts, –, –, in food-distribution associations, –, –, – gender and, , –, – incidence of, , –, –, , –
15 Index.qxd
11/9/2009
11:53 AM
Page 391
intensity of, , in labor unions, , –, , , –, –, –, –, – land invasions and, –, – in neighborhood associations and, –, –, – in Nodal s, –, –, in parents associations and, –, –, – participant personality, – party activism and, –, – party electoral mobilization and, –, –, – party identification and, –, –, – in Peru, –, , , –, –, –, –, – in place-of-origin associations, –, , –, – political parties and, , , –, –, popular sectors and, –, , – problem solving and, , –, –, – provisioning and, –, , – in recreational associations, –, –, , –, – in religion-based associations, –, –, , –, – representation and, , , – representational distortion of, , , , , , –, – resources and, –, –, –, –, , , scope, , –, –, –, –, , , –, –, – shared-member linkages, , – state policy and, – in unemployment associations and, – -Hub and, –, – in Venezuela, –, , , –, –, –, –, –, – voting, –, –, – party activism, –, – party electoral mobilization, –, –, –, – party identification, –, –, – patronage. See clientelism Peru access in, –, – A-Net in, –, –, –, –, –, –, – anticrime associations in, –, –
,
associations in, , –, –, –, –, , –, –, –, –, –, –, –, –, –, –, –, –, –, –, –, –, –, autonomy in, , –, – capacity building in, –, –, –, –, – claim making in, –, , –, –, –, –, – class in, , –, –, –, , , –, –, – cooperatives in, –, – direct action in, –, –, electoral arena in, , , –, –, –, – embeddedness in, –, –, –, –, – Flexible Fronts in, –, –, –, focus districts in, – food-distribution associations in, –, – labor unions in, , , , –, , –, –, , –, –, –, –, neighborhood associations in, –, – Nodal s in, –, –, –, – parents associations in, –, – participation in, –, , , –, –, –, –, – party activism in, –, – party electoral mobilization in, –, –, – party identification in, –, –, – place-of-origin associations in, –, – political parties in, –, , , , –, , –, , , –, –, –, –, –, –, –, –, – problem solving in, –, –, –, –, –, –, – protest in, –, –, –, – provisioning in, –, , –, –, –, –, – recreational associations in, –, – religion-based associations in, –, – representational distortion in, , , –, –, –
15 Index.qxd
11/9/2009
11:53 AM
Page 392
.
Peru (continued) scaling in, , –, –, –, –, –, –, scope in, –, –, – shared-member linkages in, – state ties in, –, –, –, –, –, –, –, – voting in, –, –, – petitioning, –, –, , Pinochet, Augusto, , , , pluralism, –, – political activity, –, . See also participation political parties access and, –, –, , – A-Net and, , –, –, –, –, –, –, , , , , –, – associations and, , , –, , –, –, –, –, –, –, , –, , –, –, – candidate-based linkages, –, – class and, –, – clientelism and, –, – confidence in, , , –, dependence and, –, direct action and, –, –, electoral mobilization by, –, –, – embeddedness and, , , –, – funding of associations, , – interest convergence and, – issue-based linkages, –, – labor-based parties, , , , , , , , , , , , , , , , , , , , , , , , –, –, , – labor unions and, , , –, –, –, , , –, multifaceted linkages, –, – participation and, , , –, –, party activism, –, – party identification, –, –, – problem solving and, , , , –, –, , –, – rallies held by, , , –, – representation and, , –, , –, scaling and, , –, –, – shared-member linkages and, , – -Hub and, , –, –, –, –, , –, ,
popular organizations. See associations; labor unions popular sectors. See also class; representational distortion associations, class and, – definition of, n., –, education and, –, income and, representational distortion and, –, , , , , , , –, –, – postcorporatism, , – postmaterialist issues, –, problem-solving strategies. See also capacity building; claim making; direct action; participation; protest; provisioning by associations, –, – by individuals, –, – protest accessing strategies and, in Argentina, –, –, –, – by associations, , , –, –, –, –, –, –, , , , –, –, –, – capacity building compared with, – in Chile, –, –, –, – claim making compared with, , – dependence and, – embeddedness and, –, , –, –, – Flexible Fronts and, , –, –, – incidence of, , –, – by individuals, –, –, , , labor unions and, – Nodal s and, , –, –, – in Peru, –, –, –, – provisioning compared with, – resources and, – scaling and, , –, –, – in Venezuela, –, –, – provisioning in A-Net, – in Argentina, –, , –, –, –, –, – associational participation and, –, – by associations, , , , , , –, – , , , –, –, –, –, –, , , –, , – in Chile, –, , –, –, –, –, – dependence and, –
15 Index.qxd
11/9/2009
11:53 AM
Page 393
embeddedness and, – Flexible Fronts and, – incidence of, –, –, –, – Nodal s and, – outsourcing of state programs and, , –, , , , , , , in Peru, –, , –, –, –, –, – protest compared with, – resources and, , –, scaling and, – state ties and, –, , in Venezuela, –, , –, –, –, –, –, – representation. See also representational distortion concepts of, – confidence in institutions of, , descriptive, –, in electoral arena and, , , , –, , –, in interest arena and, –, –, –, – objective, –, – subjective, –, – representational distortion, –, , , , , , , –, –. See also class; representation resources. See also class; funding; provisioning of associations, –, , , –, , –, –, , –, –, , – collective action and, – dependence and, of individuals, –, –, , –, –, , , of labor unions, –, , , , , – scaling. See also embeddedness; Flexible Fronts; Nodal s A-Net and, , –, , –, –, –, , , , , –, in Argentina, , –, –, –, –, –, –, associational strategies and, , –, –, – associational traits and, –, –, , –, in Chile, , –, –, –, –, –, –,
,
collective action and, , , , , definition of, , embeddedness, , –, –, , , –, –, –, –, –, –, –, – Flexible Fronts and, –, , –, as interest regime dimension, –, –, – labor unions and, , , , , , , –, –, –, –, –, –, –, Nodal s and, –, , –, in Peru, , –, –, –, –, –, –, political parties and, , –, –, – -Hub and, –, in Venezuela, –, –, –, –, , – scope. See also participation, in associations. of A-Net, –, , –, –, –, –, , , –, –, – in Argentina, –, –, – in Chile, –, –, –, collective action and, , , –, definition of, as interest regime dimension, –, –, – in Peru, –, –, – representation and, of -Hub, –, in Venezuela, –, –, –, – shared-member linkages, , –, social movements, , , –, , –, , , society-targeted strategies. See capacity building; provisioning; state-targeted strategies state policy. See also claim making; provisioning; state ties A-Net and, –, – collective action and, , – outsourcing of state programs and, , – , , , –, –, , , -Hub and, – state-targeted strategies. See accessing; claim making; protest; society-targeted strategies state ties. See also autonomy; funding; state policy associational strategies and, – of associations, , –, –, –, , – of Flexible Fronts, –
15 Index.qxd
11/9/2009
11:53 AM
Page 394
.
state ties (continued) labor unions and, –, – of Nodal s, – Toledo, Alejandro, –, unions. See labor unions -Hub. See also labor unions; political parties access and, , –, , –, associations in, –, – autonomy and, , , – collective action and, –, –, – decline of, –, –, –, –, –, demands and, , – economic model and, , –, , – interest convergence and, , – labor unions in, –, , –, –, –, –, , political parties in, , –, –, –, –, , –, , resources and, – rise of, – scaling and, , , –, , –, scope of, –, , , –, –, , state and, –, , –, – state policy and, – urban popular interest regime. See interest regime Venezuela access in, –, –, A-Net in, –, –, –, –, – anticrime associations in, –, – associations in, , –, –, –, –, , –, –, –, –, –, –, –, –, –, –, –, –, –, –, –, autonomy in, , –, –, capacity building in, –, –, –, –, – claim making in, –, , –, –, –, – class in, , –, –, –, , , –, –, – cooperatives in, –, – direct action in, –, –,
electoral arena in, , , –, –, –, – embeddedness in, –, –, –, – Flexible Fronts in, –, –, – focus districts in, – food-distribution associations in, –, – labor unions in, , , –, , –, , –, , –, –, –, – neighborhood associations in, –, – Nodal s in, –, –, – parents associations in, –, – participation in, –, , , –, –, –, –, –, – party activism in, –, – party electoral mobilization in, –, –, – party identification in, –, –, – place-of-origin associations in, –, – political parties in, –, , , , –, , –, , , –, , –, –, –, –, –, –, – problem solving in, –, –, –, –, –, – protest in, –, –, – provisioning in, –, , –, –, –, –, –, – recreational associations in, –, – religion-based associations in, –, – representational distortion in, , , –, – scaling in, –, –, –, –, , – scope in, –, –, –, – shared-member linkages in, – state ties in, –, –, –, –, –, –, – -Hub in, voting in, –, –, – voting, –, –, –. See also electoral arena working classes, , , , –, –. See also class; popular sectors